- Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS v2] perf/core: Fix perf_mmap fail when CONFIG_PERF_USE_VMALLOC enabled
by Yang Jihong 29 Nov '23

29 Nov '23

From: Zhipeng Xie <xiezhipeng1(a)huawei.com> mainline inclusion from mainline-v5.18-rc4 commit 60490e7966659b26d74bf1fa4aa8693d9a94ca88 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8KFB3 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- This problem can be reproduced with CONFIG_PERF_USE_VMALLOC enabled on both x86_64 and aarch64 arch when using sysdig -B(using ebpf)[1]. sysdig -B works fine after rebuilding the kernel with CONFIG_PERF_USE_VMALLOC disabled. I tracked it down to the if condition event->rb->nr_pages != nr_pages in perf_mmap is true when CONFIG_PERF_USE_VMALLOC is enabled where event->rb->nr_pages = 1 and nr_pages = 2048 resulting perf_mmap to return -EINVAL. This is because when CONFIG_PERF_USE_VMALLOC is enabled, rb->nr_pages is always equal to 1. Arch with CONFIG_PERF_USE_VMALLOC enabled by default: arc/arm/csky/mips/sh/sparc/xtensa Arch with CONFIG_PERF_USE_VMALLOC disabled by default: x86_64/aarch64/... Fix this problem by using data_page_nr() [1] https://github.com/draios/sysdig Fixes: 906010b2134e ("perf_event: Provide vmalloc() based mmap() backing") Signed-off-by: Zhipeng Xie <xiezhipeng1(a)huawei.com> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Link: https://lkml.kernel.org/r/20220209145417.6495-1-xiezhipeng1@huawei.com conflict: kernel/events/core.c kernel/events/internal.h kernel/events/ring_buffer.c Signed-off-by: Yang Jihong <yangjihong1(a)huawei.com> --- kernel/events/core.c | 2 +- kernel/events/internal.h | 5 +++++ kernel/events/ring_buffer.c | 4 ---- 3 files changed, 6 insertions(+), 5 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index a49860879777..a1f12009e5ef 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -5760,7 +5760,7 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) again: mutex_lock(&event->mmap_mutex); if (event->rb) { - if (event->rb->nr_pages != nr_pages) { + if (data_page_nr(event->rb) != nr_pages) { ret = -EINVAL; goto unlock; } diff --git a/kernel/events/internal.h b/kernel/events/internal.h index 8fc0ddc38cb6..578de8c2694b 100644 --- a/kernel/events/internal.h +++ b/kernel/events/internal.h @@ -114,6 +114,11 @@ static inline int page_order(struct ring_buffer *rb) } #endif +static inline int data_page_nr(struct ring_buffer *rb) +{ + return rb->nr_pages << page_order(rb); +} + static inline unsigned long perf_data_size(struct ring_buffer *rb) { return rb->nr_pages << (PAGE_SHIFT + page_order(rb)); diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c index 34e1667e4254..a69378d901fa 100644 --- a/kernel/events/ring_buffer.c +++ b/kernel/events/ring_buffer.c @@ -803,10 +803,6 @@ void rb_free(struct ring_buffer *rb) } #else -static int data_page_nr(struct ring_buffer *rb) -{ - return rb->nr_pages << page_order(rb); -} static struct page * __perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff) -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS] perf/core: Fix perf_mmap fail when CONFIG_PERF_USE_VMALLOC enabled
by Yang Jihong 29 Nov '23

29 Nov '23

From: Zhipeng Xie <xiezhipeng1(a)huawei.com> mainline inclusion from mainline-v5.18-rc4 commit 60490e7966659b26d74bf1fa4aa category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8KFB3 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- This problem can be reproduced with CONFIG_PERF_USE_VMALLOC enabled on both x86_64 and aarch64 arch when using sysdig -B(using ebpf)[1]. sysdig -B works fine after rebuilding the kernel with CONFIG_PERF_USE_VMALLOC disabled. I tracked it down to the if condition event->rb->nr_pages != nr_pages in perf_mmap is true when CONFIG_PERF_USE_VMALLOC is enabled where event->rb->nr_pages = 1 and nr_pages = 2048 resulting perf_mmap to return -EINVAL. This is because when CONFIG_PERF_USE_VMALLOC is enabled, rb->nr_pages is always equal to 1. Arch with CONFIG_PERF_USE_VMALLOC enabled by default: arc/arm/csky/mips/sh/sparc/xtensa Arch with CONFIG_PERF_USE_VMALLOC disabled by default: x86_64/aarch64/... Fix this problem by using data_page_nr() [1] https://github.com/draios/sysdig Fixes: 906010b2134e ("perf_event: Provide vmalloc() based mmap() backing") Signed-off-by: Zhipeng Xie <xiezhipeng1(a)huawei.com> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Link: https://lkml.kernel.org/r/20220209145417.6495-1-xiezhipeng1@huawei.com conflict: kernel/events/core.c kernel/events/internal.h kernel/events/ring_buffer.c Signed-off-by: Yang Jihong <yangjihong1(a)huawei.com> --- kernel/events/core.c | 2 +- kernel/events/internal.h | 5 +++++ kernel/events/ring_buffer.c | 4 ---- 3 files changed, 6 insertions(+), 5 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index a49860879777..a1f12009e5ef 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -5760,7 +5760,7 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) again: mutex_lock(&event->mmap_mutex); if (event->rb) { - if (event->rb->nr_pages != nr_pages) { + if (data_page_nr(event->rb) != nr_pages) { ret = -EINVAL; goto unlock; } diff --git a/kernel/events/internal.h b/kernel/events/internal.h index 8fc0ddc38cb6..578de8c2694b 100644 --- a/kernel/events/internal.h +++ b/kernel/events/internal.h @@ -114,6 +114,11 @@ static inline int page_order(struct ring_buffer *rb) } #endif +static inline int data_page_nr(struct ring_buffer *rb) +{ + return rb->nr_pages << page_order(rb); +} + static inline unsigned long perf_data_size(struct ring_buffer *rb) { return rb->nr_pages << (PAGE_SHIFT + page_order(rb)); diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c index 34e1667e4254..a69378d901fa 100644 --- a/kernel/events/ring_buffer.c +++ b/kernel/events/ring_buffer.c @@ -803,10 +803,6 @@ void rb_free(struct ring_buffer *rb) } #else -static int data_page_nr(struct ring_buffer *rb) -{ - return rb->nr_pages << page_order(rb); -} static struct page * __perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff) -- 2.34.1

2 1

[PATCH v2 OLK-5.10] drivers: perf: Not enabled ARM64_BRBE by default
by Junhao He 29 Nov '23

29 Nov '23

driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8EC9K CVE: NA -------------------------------- The BRBE feature depend on ARMV8 PMU hardware support for branch stack sampling. Not enabled ARM64_BRBE by default. Signed-off-by: Junhao He <hejunhao3(a)huawei.com> --- arch/arm64/configs/openeuler_defconfig | 2 +- drivers/perf/Kconfig | 1 - 2 files changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 2c6aa6404190..f9371ebf21e7 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -6169,7 +6169,7 @@ CONFIG_QCOM_L3_PMU=y CONFIG_THUNDERX2_PMU=m CONFIG_XGENE_PMU=y CONFIG_ARM_SPE_PMU=y -CONFIG_ARM64_BRBE=y +# CONFIG_ARM64_BRBE is not set CONFIG_HISI_PMU=m CONFIG_HISI_PCIE_PMU=m CONFIG_HNS3_PMU=m diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig index 25a577530da4..67ad53cde11f 100644 --- a/drivers/perf/Kconfig +++ b/drivers/perf/Kconfig @@ -133,7 +133,6 @@ config ARM_SPE_PMU config ARM64_BRBE bool "Enable support for Branch Record Buffer Extension (BRBE)" depends on PERF_EVENTS && ARM64 && ARM_PMU - default y help Enable perf support for Branch Record Buffer Extension (BRBE) which records all branches taken in an execution path. This supports some -- 2.33.0

2 1

[PACH OLK-5.10 00/11] KVM: arm64: Add host/guest KVM-PTP support
by Dongxu Sun 29 Nov '23

29 Nov '23

Jianyong Wu (4): ptp: Reorganize ptp_kvm.c to make it arch-independent clocksource: Add clocksource id for arm arch counter KVM: arm64: Add support for the KVM PTP service ptp: arm/arm64: Enable ptp_kvm for arm/arm64 Jon Hunter (1): ptp: Don't print an error if ptp_kvm is not supported Kele Huang (1): ptp: fix error print of ptp_kvm on X86_64 platform Thomas Gleixner (1): time: Add mechanism to recognize clocksource in time_get_snapshot Will Deacon (2): arm/arm64: Probe for the presence of KVM hypervisor KVM: arm64: Advertise KVM UID to guests via SMCCC Yanan Wang (1): KVM: arm64: fix compile error/warning Zenghui Yu (1): KVM: arm64: Fix Function ID typo for PTP_KVM service Documentation/virt/kvm/api.rst | 9 ++ Documentation/virt/kvm/arm/index.rst | 1 + Documentation/virt/kvm/arm/ptp_kvm.rst | 25 ++++++ arch/arm/include/asm/hypervisor.h | 3 + arch/arm64/include/asm/hypervisor.h | 3 + arch/arm64/include/asm/kvm_para.h | 2 +- arch/arm64/include/uapi/asm/Kbuild | 2 +- arch/arm64/include/uapi/asm/kvm_para.h | 7 ++ arch/arm64/kvm/arm.c | 1 + arch/arm64/kvm/hypercalls.c | 94 ++++++++++++++++---- drivers/clocksource/arm_arch_timer.c | 36 ++++++++ drivers/firmware/psci/psci.c | 2 + drivers/firmware/smccc/Makefile | 2 +- drivers/firmware/smccc/kvm_guest.c | 50 +++++++++++ drivers/firmware/smccc/smccc.c | 1 + drivers/ptp/Kconfig | 2 +- drivers/ptp/Makefile | 2 + drivers/ptp/ptp_kvm_arm.c | 28 ++++++ drivers/ptp/{ptp_kvm.c => ptp_kvm_common.c} | 85 +++++------------- drivers/ptp/ptp_kvm_x86.c | 97 +++++++++++++++++++++ include/linux/arm-smccc.h | 41 +++++++++ include/linux/clocksource.h | 6 ++ include/linux/clocksource_ids.h | 12 +++ include/linux/ptp_kvm.h | 19 ++++ include/linux/timekeeping.h | 12 +-- include/uapi/linux/kvm.h | 1 + kernel/time/clocksource.c | 2 + kernel/time/timekeeping.c | 1 + 28 files changed, 459 insertions(+), 87 deletions(-) create mode 100644 Documentation/virt/kvm/arm/ptp_kvm.rst create mode 100644 arch/arm64/include/uapi/asm/kvm_para.h create mode 100644 drivers/firmware/smccc/kvm_guest.c create mode 100644 drivers/ptp/ptp_kvm_arm.c rename drivers/ptp/{ptp_kvm.c => ptp_kvm_common.c} (60%) create mode 100644 drivers/ptp/ptp_kvm_x86.c create mode 100644 include/linux/clocksource_ids.h create mode 100644 include/linux/ptp_kvm.h -- 2.33.0

1 11

[PATCH V3 OLK-5.10] vhost-vdpa: add reset state params to indicate reset level
by Jiang Dongxu 29 Nov '23

29 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- Add the interface parameter state to the vdpa reset interface, which respectively identifies the reset when the device is turned on/off and the virtio reset issued by the virtual machine. Signed-off-by: jiangdongxu <jiangdongxu1(a)huawei.com> --- drivers/vdpa/ifcvf/ifcvf_main.c | 2 +- drivers/vdpa/mlx5/net/mlx5_vnet.c | 2 +- drivers/vdpa/vdpa_sim/vdpa_sim.c | 2 +- drivers/vdpa/virtio_pci/vp_vdpa.c | 2 +- drivers/vhost/vdpa.c | 10 +++++----- drivers/virtio/virtio_vdpa.c | 2 +- include/linux/vdpa.h | 16 +++++++++++++--- 7 files changed, 23 insertions(+), 13 deletions(-) diff --git a/drivers/vdpa/ifcvf/ifcvf_main.c b/drivers/vdpa/ifcvf/ifcvf_main.c index 862f4591fff2..2a4991464cfd 100644 --- a/drivers/vdpa/ifcvf/ifcvf_main.c +++ b/drivers/vdpa/ifcvf/ifcvf_main.c @@ -233,7 +233,7 @@ static void ifcvf_vdpa_set_status(struct vdpa_device *vdpa_dev, u8 status) ifcvf_set_status(vf, status); } -static int ifcvf_vdpa_reset(struct vdpa_device *vdpa_dev) +static int ifcvf_vdpa_reset(struct vdpa_device *vdpa_dev, int state) { struct ifcvf_adapter *adapter; struct ifcvf_hw *vf; diff --git a/drivers/vdpa/mlx5/net/mlx5_vnet.c b/drivers/vdpa/mlx5/net/mlx5_vnet.c index 76e7b4bab9d3..c676bb70a453 100644 --- a/drivers/vdpa/mlx5/net/mlx5_vnet.c +++ b/drivers/vdpa/mlx5/net/mlx5_vnet.c @@ -1800,7 +1800,7 @@ static void mlx5_vdpa_set_status(struct vdpa_device *vdev, u8 status) ndev->mvdev.status |= VIRTIO_CONFIG_S_FAILED; } -static int mlx5_vdpa_reset(struct vdpa_device *vdev) +static int mlx5_vdpa_reset(struct vdpa_device *vdev, int state) { struct mlx5_vdpa_dev *mvdev = to_mvdev(vdev); struct mlx5_vdpa_net *ndev = to_mlx5_vdpa_ndev(mvdev); diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdpa_sim.c index c141f08b8274..be925462922d 100644 --- a/drivers/vdpa/vdpa_sim/vdpa_sim.c +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c @@ -454,7 +454,7 @@ static void vdpasim_set_status(struct vdpa_device *vdpa, u8 status) spin_unlock(&vdpasim->lock); } -static int vdpasim_reset(struct vdpa_device *vdpa) +static int vdpasim_reset(struct vdpa_device *vdpa, int state) { struct vdpasim *vdpasim = vdpa_to_sim(vdpa); diff --git a/drivers/vdpa/virtio_pci/vp_vdpa.c b/drivers/vdpa/virtio_pci/vp_vdpa.c index f4e375b1d903..bc0bdcc7bcea 100644 --- a/drivers/vdpa/virtio_pci/vp_vdpa.c +++ b/drivers/vdpa/virtio_pci/vp_vdpa.c @@ -220,7 +220,7 @@ static void vp_vdpa_set_status(struct vdpa_device *vdpa, u8 status) vp_modern_set_status(mdev, status); } -static int vp_vdpa_reset(struct vdpa_device *vdpa) +static int vp_vdpa_reset(struct vdpa_device *vdpa, int state) { struct vp_vdpa *vp_vdpa = vdpa_to_vp(vdpa); struct virtio_pci_modern_device *mdev = vp_vdpa_to_mdev(vp_vdpa); diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index eec8027dfc4f..eed51d004531 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -212,13 +212,13 @@ static void vhost_vdpa_unsetup_vq_irq(struct vhost_vdpa *v, u16 qid) irq_bypass_unregister_producer(&vq->call_ctx.producer); } -static int vhost_vdpa_reset(struct vhost_vdpa *v) +static int vhost_vdpa_reset(struct vhost_vdpa *v, int state) { struct vdpa_device *vdpa = v->vdpa; v->in_batch = 0; - return vdpa_reset(vdpa); + return vdpa_reset(vdpa, state); } static long vhost_vdpa_bind_mm(struct vhost_vdpa *v) @@ -297,7 +297,7 @@ static long vhost_vdpa_set_status(struct vhost_vdpa *v, u8 __user *statusp) vhost_vdpa_unsetup_vq_irq(v, i); if (status == 0) { - ret = vdpa_reset(vdpa); + ret = vdpa_reset(vdpa, VDPA_DEV_RESET_VIRTIO); if (ret) return ret; } else @@ -1496,7 +1496,7 @@ static int vhost_vdpa_open(struct inode *inode, struct file *filep) return r; nvqs = v->nvqs; - r = vhost_vdpa_reset(v); + r = vhost_vdpa_reset(v, VDPA_DEV_RESET_OPEN); if (r) goto err; @@ -1542,7 +1542,7 @@ static int vhost_vdpa_release(struct inode *inode, struct file *filep) mutex_lock(&d->mutex); filep->private_data = NULL; vhost_vdpa_clean_irq(v); - vhost_vdpa_reset(v); + vhost_vdpa_reset(v, VDPA_DEV_RESET_CLOSE); vhost_dev_stop(&v->vdev); vhost_vdpa_unbind_mm(v); vhost_vdpa_config_put(v); diff --git a/drivers/virtio/virtio_vdpa.c b/drivers/virtio/virtio_vdpa.c index 46c71653f508..8126a9be22cd 100644 --- a/drivers/virtio/virtio_vdpa.c +++ b/drivers/virtio/virtio_vdpa.c @@ -99,7 +99,7 @@ static void virtio_vdpa_reset(struct virtio_device *vdev) { struct vdpa_device *vdpa = vd_get_vdpa(vdev); - vdpa_reset(vdpa); + vdpa_reset(vdpa, VDPA_DEV_RESET_VIRTIO); } static bool virtio_vdpa_notify(struct virtqueue *vq) diff --git a/include/linux/vdpa.h b/include/linux/vdpa.h index 83df2380eeb9..ef53829c7a7a 100644 --- a/include/linux/vdpa.h +++ b/include/linux/vdpa.h @@ -120,6 +120,12 @@ struct vdpa_map_file { u64 offset; }; +enum vdpa_reset_state { + VDPA_DEV_RESET_VIRTIO = 0, + VDPA_DEV_RESET_OPEN = 1, + VDPA_DEV_RESET_CLOSE = 2, +}; + /** * struct vdpa_config_ops - operations for configuring a vDPA device. * Note: vDPA device drivers are required to implement all of the @@ -218,6 +224,10 @@ struct vdpa_map_file { * @status: virtio device status * @reset: Reset device * @vdev: vdpa device + * @state: state for reset + * VDPA_DEV_RESET_VIRTIO for virtio reset + * VDPA_DEV_RESET_OPEN for vhost-vdpa device open + * VDPA_DEV_RESET_CLOSE for vhost-vdpa device close * Returns integer: success (0) or error (< 0) * @suspend: Suspend the device (optional) * @vdev: vdpa device @@ -359,7 +369,7 @@ struct vdpa_config_ops { u32 (*get_vendor_id)(struct vdpa_device *vdev); u8 (*get_status)(struct vdpa_device *vdev); void (*set_status)(struct vdpa_device *vdev, u8 status); - int (*reset)(struct vdpa_device *vdev); + int (*reset)(struct vdpa_device *vdev, int state); int (*suspend)(struct vdpa_device *vdev); int (*resume)(struct vdpa_device *vdev); size_t (*get_config_size)(struct vdpa_device *vdev); @@ -482,14 +492,14 @@ static inline struct device *vdpa_get_dma_dev(struct vdpa_device *vdev) return vdev->dma_dev; } -static inline int vdpa_reset(struct vdpa_device *vdev) +static inline int vdpa_reset(struct vdpa_device *vdev, int state) { const struct vdpa_config_ops *ops = vdev->config; int ret; down_write(&vdev->cf_lock); vdev->features_valid = false; - ret = ops->reset(vdev); + ret = ops->reset(vdev, state); up_write(&vdev->cf_lock); return ret; } -- 2.27.0

2 1

[PATCH OLK-5.10 v3 00/10] rcu: Add RCU stall diagnosis information
by Zhen Lei 29 Nov '23

29 Nov '23

v2 --> v3: Add "# CONFIG_RCU_CPU_STALL_CPUTIME is not set" into openeuler_defconfig. v1 --> v2: Remove patch 1 in v1: rcu: Prevent lockdep-RCU splats on lock acquisition/release It had arleady been merged. v1: In some extreme cases, such as the I/O pressure test, the CPU usage may be 100%, causing RCU stall. In this case, the printed information about current is not useful. Displays the number and usage of hard interrupts, soft interrupts, and context switches that are generated within half of the CPU stall timeout, can help us make a general judgment. In other cases, we can preliminarily determine whether an infinite loop occurs when local_irq, local_bh or preempt is disabled. Neeraj Upadhyay (1): rcu: Check and report missed fqs timer wakeup on RCU stall Paul E. McKenney (2): rcu: For RCU grace-period kthread starvation, dump last CPU it ran on rcu: Do not NMI offline CPUs Zhen Lei (7): sched/debug: Try trigger_single_cpu_backtrace(cpu) in dump_cpu_task() sched/debug: Show the registers of 'current' in dump_cpu_task() sched: Add helper kstat_cpu_softirqs_sum() sched: Add helper nr_context_switches_cpu() rcu: Add RCU stall diagnosis information rcu: Align the output of RCU CPU stall warning messages config: update openeuler_defconfig for arm64 and x86 Documentation/RCU/stallwarn.rst | 23 ++++- .../admin-guide/kernel-parameters.txt | 6 ++ arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/kernel_stat.h | 12 +++ kernel/rcu/Kconfig.debug | 13 +++ kernel/rcu/rcu.h | 1 + kernel/rcu/tree.c | 43 +++++++--- kernel/rcu/tree.h | 19 ++++ kernel/rcu/tree_stall.h | 86 +++++++++++++++++-- kernel/rcu/update.c | 2 + kernel/sched/core.c | 19 ++++ kernel/smp.c | 3 +- 13 files changed, 211 insertions(+), 18 deletions(-) -- 2.25.1

2 11

[PATCH v3 OLK-5.10 0/3] support page eject
by Wupeng Ma 29 Nov '23

29 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> Page ejection is achieved by soft_offline_page and unpoison_memory. A linked list is maintained to log the pfns which are offlined by this module. Only the pfns present in the list is allowed to go online. The follow sysfs is introduced to online/offline page via this module, since there is a internal linked list to record all offlined pages, remove_page is introduced when a user no longer wishes to online this particular page. To reduce the length of the internal list, this function removes the specified page from the list. - /sys/kernel/page_eject/online_page - /sys/kernel/page_eject/offline_page - /sys/kernel/page_eject/remove_page Changelog since v2: - bugfix for page_eject_clear_list_locked Changelog since v1: - remove mce related patches temporarily. Ma Wupeng (3): mm/hwpoison: Export symbol soft_offline_page mm: page_eject: Introuduce page ejection config: update defconfig for PAGE_EJECT arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/ras/Kconfig | 10 + drivers/ras/Makefile | 1 + drivers/ras/page_eject.c | 248 +++++++++++++++++++++++++ mm/memory-failure.c | 1 + 6 files changed, 262 insertions(+) create mode 100644 drivers/ras/page_eject.c -- 2.25.1

2 4

[PATCH OLK-6.6 0/9] mm: PCP high auto-tuning
by Ze Zuo 29 Nov '23

29 Nov '23

The page allocation performance requirements of different workloads are often different. So, we need to tune the PCP (Per-CPU Pageset) high on each CPU automatically to optimize the page allocation performance. The list of patches in series is as follows, [1/9] mm, pcp: avoid to drain PCP when process exit [2/9] cacheinfo: calculate per-CPU data cache size [3/9] mm, pcp: reduce lock contention for draining high-order pages [4/9] mm: restrict the pcp batch scale factor to avoid too long latency [5/9] mm, page_alloc: scale the number of pages that are batch allocated [6/9] mm: add framework for PCP high auto-tuning [7/9] mm: tune PCP high automatically [8/9] mm, pcp: decrease PCP high if free pages < high watermark [9/9] mm, pcp: reduce detecting time of consecutive high order page freeing Patch [1/9], [2/9], [3/9] optimize the PCP draining for consecutive high-order pages freeing. Patch [4/9], [5/9] optimize batch freeing and allocating. Patch [6/9], [7/9], [8/9] implement and optimize a PCP high auto-tuning method. Patch [9/9] optimize the PCP draining for consecutive high order page freeing based on PCP high auto-tuning. The test results for patches with performance impact are as follows, kbuild ====== On a 2-socket Intel server with 224 logical CPU, we run 8 kbuild instances in parallel (each with `make -j 28`) in 8 cgroup. This simulates the kbuild server that is used by 0-Day kbuild service. build time lock contend% free_high alloc_zone ---------- ---------- --------- ---------- base 100.0 14.0 100.0 100.0 patch1 99.5 12.8 19.5 95.6 patch3 99.4 12.6 7.1 95.6 patch5 98.6 11.0 8.1 97.1 patch7 95.1 0.5 2.8 15.6 patch9 95.0 1.0 8.8 20.0 The PCP draining optimization (patch [1/9], [3/9]) and PCP batch allocation optimization (patch [5/9]) reduces zone lock contention a little. The PCP high auto-tuning (patch [7/9], [9/9]) reduces build time visibly. Where the tuning target: the number of pages allocated from zone reduces greatly. So, the zone contention cycles% reduces greatly. With PCP tuning patches (patch [7/9], [9/9]), the average used memory during test increases up to 18.4% because more pages are cached in PCP. But at the end of the test, the number of the used memory decreases to the same level as that of the base patch. That is, the pages cached in PCP will be released to zone after not being used actively. netperf SCTP_STREAM_MANY ======================== On a 2-socket Intel server with 128 logical CPU, we tested SCTP_STREAM_MANY test case of netperf test suite with 64-pair processes. score lock contend% free_high alloc_zone cache miss rate% ----- ---------- --------- ---------- ---------------- base 100.0 2.1 100.0 100.0 1.3 patch1 99.4 2.1 99.4 99.4 1.3 patch3 106.4 1.3 13.3 106.3 1.3 patch5 106.0 1.2 13.2 105.9 1.3 patch7 103.4 1.9 6.7 90.3 7.6 patch9 108.6 1.3 13.7 108.6 1.3 The PCP draining optimization (patch [1/9]+[3/9]) improves performance. The PCP high auto-tuning (patch [7/9]) reduces performance a little because PCP draining cannot be triggered in time sometimes. So, the cache miss rate% increases. The further PCP draining optimization (patch [9/9]) based on PCP tuning restore the performance. lmbench3 UNIX (AF_UNIX) ======================= On a 2-socket Intel server with 128 logical CPU, we tested UNIX (AF_UNIX socket) test case of lmbench3 test suite with 16-pair processes. score lock contend% free_high alloc_zone cache miss rate% ----- ---------- --------- ---------- ---------------- base 100.0 51.4 100.0 100.0 0.2 patch1 116.8 46.1 69.5 104.3 0.2 patch3 199.1 21.3 7.0 104.9 0.2 patch5 200.0 20.8 7.1 106.9 0.3 patch7 191.6 19.9 6.8 103.8 2.8 patch9 193.4 21.7 7.0 104.7 2.1 The PCP draining optimization (patch [1/9], [3/9]) improves performance much. The PCP tuning (patch [7/9]) reduces performance a little because PCP draining cannot be triggered in time sometimes. The further PCP draining optimization (patch [9/9]) based on PCP tuning restores the performance partly. The patchset adds several fields in struct per_cpu_pages. The struct layout before/after the patchset is as follows, base ==== struct per_cpu_pages { spinlock_t lock; /* 0 4 */ int count; /* 4 4 */ int high; /* 8 4 */ int batch; /* 12 4 */ short int free_factor; /* 16 2 */ short int expire; /* 18 2 */ /* XXX 4 bytes hole, try to pack */ struct list_head lists[13]; /* 24 208 */ /* size: 256, cachelines: 4, members: 7 */ /* sum members: 228, holes: 1, sum holes: 4 */ /* padding: 24 */ } __attribute__((__aligned__(64))); patched ======= struct per_cpu_pages { spinlock_t lock; /* 0 4 */ int count; /* 4 4 */ int high; /* 8 4 */ int high_min; /* 12 4 */ int high_max; /* 16 4 */ int batch; /* 20 4 */ u8 flags; /* 24 1 */ u8 alloc_factor; /* 25 1 */ u8 expire; /* 26 1 */ /* XXX 1 byte hole, try to pack */ short int free_count; /* 28 2 */ /* XXX 2 bytes hole, try to pack */ struct list_head lists[13]; /* 32 208 */ /* size: 256, cachelines: 4, members: 11 */ /* sum members: 237, holes: 2, sum holes: 3 */ /* padding: 16 */ } __attribute__((__aligned__(64))); The size of the struct doesn't changed with the patchset. Huang Ying (9): mm, pcp: avoid to drain PCP when process exit cacheinfo: calculate size of per-CPU data cache slice mm, pcp: reduce lock contention for draining high-order pages mm: restrict the pcp batch scale factor to avoid too long latency mm, page_alloc: scale the number of pages that are batch allocated mm: add framework for PCP high auto-tuning mm: tune PCP high automatically mm, pcp: decrease PCP high if free pages < high watermark mm, pcp: reduce detecting time of consecutive high order page freeing drivers/base/cacheinfo.c | 51 ++++++- include/linux/cacheinfo.h | 1 + include/linux/gfp.h | 2 + include/linux/mmzone.h | 27 +++- mm/Kconfig | 11 ++ mm/page_alloc.c | 310 +++++++++++++++++++++++++++++++------- mm/vmstat.c | 8 +- 7 files changed, 345 insertions(+), 65 deletions(-) -- 2.25.1

2 10

[PATCH OLK-6.6 0/9] mm: PCP high auto-tuning
by Ze Zuo 29 Nov '23

29 Nov '23

The page allocation performance requirements of different workloads are often different. So, we need to tune the PCP (Per-CPU Pageset) high on each CPU automatically to optimize the page allocation performance. The list of patches in series is as follows, [1/9] mm, pcp: avoid to drain PCP when process exit [2/9] cacheinfo: calculate per-CPU data cache size [3/9] mm, pcp: reduce lock contention for draining high-order pages [4/9] mm: restrict the pcp batch scale factor to avoid too long latency [5/9] mm, page_alloc: scale the number of pages that are batch allocated [6/9] mm: add framework for PCP high auto-tuning [7/9] mm: tune PCP high automatically [8/9] mm, pcp: decrease PCP high if free pages < high watermark [9/9] mm, pcp: reduce detecting time of consecutive high order page freeing Patch [1/9], [2/9], [3/9] optimize the PCP draining for consecutive high-order pages freeing. Patch [4/9], [5/9] optimize batch freeing and allocating. Patch [6/9], [7/9], [8/9] implement and optimize a PCP high auto-tuning method. Patch [9/9] optimize the PCP draining for consecutive high order page freeing based on PCP high auto-tuning. The test results for patches with performance impact are as follows, kbuild ====== On a 2-socket Intel server with 224 logical CPU, we run 8 kbuild instances in parallel (each with `make -j 28`) in 8 cgroup. This simulates the kbuild server that is used by 0-Day kbuild service. build time lock contend% free_high alloc_zone ---------- ---------- --------- ---------- base 100.0 14.0 100.0 100.0 patch1 99.5 12.8 19.5 95.6 patch3 99.4 12.6 7.1 95.6 patch5 98.6 11.0 8.1 97.1 patch7 95.1 0.5 2.8 15.6 patch9 95.0 1.0 8.8 20.0 The PCP draining optimization (patch [1/9], [3/9]) and PCP batch allocation optimization (patch [5/9]) reduces zone lock contention a little. The PCP high auto-tuning (patch [7/9], [9/9]) reduces build time visibly. Where the tuning target: the number of pages allocated from zone reduces greatly. So, the zone contention cycles% reduces greatly. With PCP tuning patches (patch [7/9], [9/9]), the average used memory during test increases up to 18.4% because more pages are cached in PCP. But at the end of the test, the number of the used memory decreases to the same level as that of the base patch. That is, the pages cached in PCP will be released to zone after not being used actively. netperf SCTP_STREAM_MANY ======================== On a 2-socket Intel server with 128 logical CPU, we tested SCTP_STREAM_MANY test case of netperf test suite with 64-pair processes. score lock contend% free_high alloc_zone cache miss rate% ----- ---------- --------- ---------- ---------------- base 100.0 2.1 100.0 100.0 1.3 patch1 99.4 2.1 99.4 99.4 1.3 patch3 106.4 1.3 13.3 106.3 1.3 patch5 106.0 1.2 13.2 105.9 1.3 patch7 103.4 1.9 6.7 90.3 7.6 patch9 108.6 1.3 13.7 108.6 1.3 The PCP draining optimization (patch [1/9]+[3/9]) improves performance. The PCP high auto-tuning (patch [7/9]) reduces performance a little because PCP draining cannot be triggered in time sometimes. So, the cache miss rate% increases. The further PCP draining optimization (patch [9/9]) based on PCP tuning restore the performance. lmbench3 UNIX (AF_UNIX) ======================= On a 2-socket Intel server with 128 logical CPU, we tested UNIX (AF_UNIX socket) test case of lmbench3 test suite with 16-pair processes. score lock contend% free_high alloc_zone cache miss rate% ----- ---------- --------- ---------- ---------------- base 100.0 51.4 100.0 100.0 0.2 patch1 116.8 46.1 69.5 104.3 0.2 patch3 199.1 21.3 7.0 104.9 0.2 patch5 200.0 20.8 7.1 106.9 0.3 patch7 191.6 19.9 6.8 103.8 2.8 patch9 193.4 21.7 7.0 104.7 2.1 The PCP draining optimization (patch [1/9], [3/9]) improves performance much. The PCP tuning (patch [7/9]) reduces performance a little because PCP draining cannot be triggered in time sometimes. The further PCP draining optimization (patch [9/9]) based on PCP tuning restores the performance partly. The patchset adds several fields in struct per_cpu_pages. The struct layout before/after the patchset is as follows, base ==== struct per_cpu_pages { spinlock_t lock; /* 0 4 */ int count; /* 4 4 */ int high; /* 8 4 */ int batch; /* 12 4 */ short int free_factor; /* 16 2 */ short int expire; /* 18 2 */ /* XXX 4 bytes hole, try to pack */ struct list_head lists[13]; /* 24 208 */ /* size: 256, cachelines: 4, members: 7 */ /* sum members: 228, holes: 1, sum holes: 4 */ /* padding: 24 */ } __attribute__((__aligned__(64))); patched ======= struct per_cpu_pages { spinlock_t lock; /* 0 4 */ int count; /* 4 4 */ int high; /* 8 4 */ int high_min; /* 12 4 */ int high_max; /* 16 4 */ int batch; /* 20 4 */ u8 flags; /* 24 1 */ u8 alloc_factor; /* 25 1 */ u8 expire; /* 26 1 */ /* XXX 1 byte hole, try to pack */ short int free_count; /* 28 2 */ /* XXX 2 bytes hole, try to pack */ struct list_head lists[13]; /* 32 208 */ /* size: 256, cachelines: 4, members: 11 */ /* sum members: 237, holes: 2, sum holes: 3 */ /* padding: 16 */ } __attribute__((__aligned__(64))); The size of the struct doesn't changed with the patchset. Huang Ying (9): mm, pcp: avoid to drain PCP when process exit cacheinfo: calculate size of per-CPU data cache slice mm, pcp: reduce lock contention for draining high-order pages mm: restrict the pcp batch scale factor to avoid too long latency mm, page_alloc: scale the number of pages that are batch allocated mm: add framework for PCP high auto-tuning mm: tune PCP high automatically mm, pcp: decrease PCP high if free pages < high watermark mm, pcp: reduce detecting time of consecutive high order page freeing drivers/base/cacheinfo.c | 51 ++++++- include/linux/cacheinfo.h | 1 + include/linux/gfp.h | 2 + include/linux/mmzone.h | 27 +++- mm/Kconfig | 11 ++ mm/page_alloc.c | 310 +++++++++++++++++++++++++++++++------- mm/vmstat.c | 8 +- 7 files changed, 345 insertions(+), 65 deletions(-) -- 2.25.1

1 9

[PACH OLK-5.10 00/11] KVM: arm64: Add host/guest KVM-PTP support
by Dongxu Sun 29 Nov '23

29 Nov '23

Jianyong Wu (4): ptp: Reorganize ptp_kvm.c to make it arch-independent clocksource: Add clocksource id for arm arch counter KVM: arm64: Add support for the KVM PTP service ptp: arm/arm64: Enable ptp_kvm for arm/arm64 Jon Hunter (1): ptp: Don't print an error if ptp_kvm is not supported Kele Huang (1): ptp: fix error print of ptp_kvm on X86_64 platform Thomas Gleixner (1): time: Add mechanism to recognize clocksource in time_get_snapshot Will Deacon (2): arm/arm64: Probe for the presence of KVM hypervisor KVM: arm64: Advertise KVM UID to guests via SMCCC Yanan Wang (1): KVM: arm64: fix compile error/warning Zenghui Yu (1): KVM: arm64: Fix Function ID typo for PTP_KVM service Documentation/virt/kvm/api.rst | 9 ++ Documentation/virt/kvm/arm/index.rst | 1 + Documentation/virt/kvm/arm/ptp_kvm.rst | 25 ++++++ arch/arm/include/asm/hypervisor.h | 3 + arch/arm64/include/asm/hypervisor.h | 3 + arch/arm64/include/asm/kvm_para.h | 2 +- arch/arm64/include/uapi/asm/Kbuild | 2 +- arch/arm64/include/uapi/asm/kvm_para.h | 7 ++ arch/arm64/kvm/arm.c | 1 + arch/arm64/kvm/hypercalls.c | 94 ++++++++++++++++---- drivers/clocksource/arm_arch_timer.c | 36 ++++++++ drivers/firmware/psci/psci.c | 2 + drivers/firmware/smccc/Makefile | 2 +- drivers/firmware/smccc/kvm_guest.c | 50 +++++++++++ drivers/firmware/smccc/smccc.c | 1 + drivers/ptp/Kconfig | 2 +- drivers/ptp/Makefile | 2 + drivers/ptp/ptp_kvm_arm.c | 28 ++++++ drivers/ptp/{ptp_kvm.c => ptp_kvm_common.c} | 85 +++++------------- drivers/ptp/ptp_kvm_x86.c | 97 +++++++++++++++++++++ include/linux/arm-smccc.h | 41 +++++++++ include/linux/clocksource.h | 6 ++ include/linux/clocksource_ids.h | 12 +++ include/linux/ptp_kvm.h | 19 ++++ include/linux/timekeeping.h | 12 +-- include/uapi/linux/kvm.h | 1 + kernel/time/clocksource.c | 2 + kernel/time/timekeeping.c | 1 + 28 files changed, 459 insertions(+), 87 deletions(-) create mode 100644 Documentation/virt/kvm/arm/ptp_kvm.rst create mode 100644 arch/arm64/include/uapi/asm/kvm_para.h create mode 100644 drivers/firmware/smccc/kvm_guest.c create mode 100644 drivers/ptp/ptp_kvm_arm.c rename drivers/ptp/{ptp_kvm.c => ptp_kvm_common.c} (60%) create mode 100644 drivers/ptp/ptp_kvm_x86.c create mode 100644 include/linux/clocksource_ids.h create mode 100644 include/linux/ptp_kvm.h -- 2.33.0

1 11

[PACH OLK-5.10 00/11] KVM: arm64: Add host/guest KVM-PTP support
by Dongxu Sun 29 Nov '23

29 Nov '23

Jianyong Wu (4): ptp: Reorganize ptp_kvm.c to make it arch-independent clocksource: Add clocksource id for arm arch counter KVM: arm64: Add support for the KVM PTP service ptp: arm/arm64: Enable ptp_kvm for arm/arm64 Jon Hunter (1): ptp: Don't print an error if ptp_kvm is not supported Kele Huang (1): ptp: fix error print of ptp_kvm on X86_64 platform Thomas Gleixner (1): time: Add mechanism to recognize clocksource in time_get_snapshot Will Deacon (2): arm/arm64: Probe for the presence of KVM hypervisor KVM: arm64: Advertise KVM UID to guests via SMCCC Yanan Wang (1): KVM: arm64: fix compile error/warning Zenghui Yu (1): KVM: arm64: Fix Function ID typo for PTP_KVM service Documentation/virt/kvm/api.rst | 9 ++ Documentation/virt/kvm/arm/index.rst | 1 + Documentation/virt/kvm/arm/ptp_kvm.rst | 25 ++++++ arch/arm/include/asm/hypervisor.h | 3 + arch/arm64/include/asm/hypervisor.h | 3 + arch/arm64/include/asm/kvm_para.h | 2 +- arch/arm64/include/uapi/asm/Kbuild | 2 +- arch/arm64/include/uapi/asm/kvm_para.h | 7 ++ arch/arm64/kvm/arm.c | 1 + arch/arm64/kvm/hypercalls.c | 94 ++++++++++++++++---- drivers/clocksource/arm_arch_timer.c | 36 ++++++++ drivers/firmware/psci/psci.c | 2 + drivers/firmware/smccc/Makefile | 2 +- drivers/firmware/smccc/kvm_guest.c | 50 +++++++++++ drivers/firmware/smccc/smccc.c | 1 + drivers/ptp/Kconfig | 2 +- drivers/ptp/Makefile | 2 + drivers/ptp/ptp_kvm_arm.c | 28 ++++++ drivers/ptp/{ptp_kvm.c => ptp_kvm_common.c} | 85 +++++------------- drivers/ptp/ptp_kvm_x86.c | 97 +++++++++++++++++++++ include/linux/arm-smccc.h | 41 +++++++++ include/linux/clocksource.h | 6 ++ include/linux/clocksource_ids.h | 12 +++ include/linux/ptp_kvm.h | 19 ++++ include/linux/timekeeping.h | 12 +-- include/uapi/linux/kvm.h | 1 + kernel/time/clocksource.c | 2 + kernel/time/timekeeping.c | 1 + 28 files changed, 459 insertions(+), 87 deletions(-) create mode 100644 Documentation/virt/kvm/arm/ptp_kvm.rst create mode 100644 arch/arm64/include/uapi/asm/kvm_para.h create mode 100644 drivers/firmware/smccc/kvm_guest.c create mode 100644 drivers/ptp/ptp_kvm_arm.c rename drivers/ptp/{ptp_kvm.c => ptp_kvm_common.c} (60%) create mode 100644 drivers/ptp/ptp_kvm_x86.c create mode 100644 include/linux/clocksource_ids.h create mode 100644 include/linux/ptp_kvm.h -- 2.33.0

1 4

[PATCH OLK-5.10] drivers: perf: Add feature flag check in armpmu_add/del
by Junhao He 29 Nov '23

29 Nov '23

driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8EC9K CVE: NA -------------------------------- Function armpmu_add will call armv8pmu_branch_reset() to reset BRBE. The BRBE feature depend on ARMV8 PMU hardware support for branch stack sampling. On older platforms the kernel panics. Internal error: Oops - Undefined instruction: 0000000002000000 [#1] SMP Call trace: armv8pmu_branch_reset+0xc/0x20 event_sched_in+0xc8/0x1a0 merge_sched_in+0x16c/0x41c Fixes: 2690e65ac9f9 ("drivers: perf: arm_pmuv3: Enable branch stack sampling via FEAT_BRBE") Signed-off-by: Junhao He <hejunhao3(a)huawei.com> --- drivers/perf/arm_pmu.c | 34 +++++++++++++++++++--------------- 1 file changed, 19 insertions(+), 15 deletions(-) diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c index d02a85c0bedc..4ef8aee84eea 100644 --- a/drivers/perf/arm_pmu.c +++ b/drivers/perf/arm_pmu.c @@ -313,10 +313,12 @@ armpmu_del(struct perf_event *event, int flags) struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; - WARN_ON_ONCE(!hw_events->brbe_users); - hw_events->brbe_users--; - if (!hw_events->brbe_users) - hw_events->brbe_context = NULL; + if (has_branch_stack(event)) { + WARN_ON_ONCE(!hw_events->brbe_users); + hw_events->brbe_users--; + if (!hw_events->brbe_users) + hw_events->brbe_context = NULL; + } armpmu_stop(event, PERF_EF_UPDATE); hw_events->events[idx] = NULL; @@ -334,18 +336,20 @@ armpmu_add(struct perf_event *event, int flags) struct hw_perf_event *hwc = &event->hw; int idx; - /* - * Reset branch records buffer if a new task event gets - * scheduled on a PMU which might have existing records. - * Otherwise older branch records present in the buffer - * might leak into the new task event. - */ - if (event->ctx->task && hw_events->brbe_context != event->ctx) { - hw_events->brbe_context = event->ctx; - if (armpmu->branch_reset) - armpmu->branch_reset(); + if (has_branch_stack(event)) { + /* + * Reset branch records buffer if a new task event gets + * scheduled on a PMU which might have existing records. + * Otherwise older branch records present in the buffer + * might leak into the new task event. + */ + if (event->ctx->task && hw_events->brbe_context != event->ctx) { + hw_events->brbe_context = event->ctx; + if (armpmu->branch_reset) + armpmu->branch_reset(); + } + hw_events->brbe_users++; } - hw_events->brbe_users++; /* An event following a process won't be stopped earlier */ if (!cpumask_test_cpu(smp_processor_id(), &armpmu->supported_cpus)) -- 2.33.0

2 1

[PATCH OLK-6.6 v2 0/6] Handle more faults under the VMA lock
by Tong Tiangen 29 Nov '23

29 Nov '23

At this point, we're handling the majority of file-backed page faults under the VMA lock, using the ->map_pages entry point. This patch set attempts to expand that for the following siutations: - We have to do a read. This could be because we've hit the point in the readahead window where we need to kick off the next readahead, or because the page is simply not present in cache. - We're handling a write fault. Most applications don't do I/O by writes to shared mmaps for very good reasons, but some do, and it'd be nice to not make that slow unnecessarily. - We're doing a COW of a private mapping (both PTE already present and PTE not-present). These are two different codepaths and I handle both of them in this patch set. There is no support in this patch set for drivers to mark themselves as being VMA lock friendly; they could implement the ->map_pages vm_operation, but if they do, they would be the first. This is probably something we want to change at some point in the future, and I've marked where to make that change in the code. There is very little performance change in the benchmarks we've run; mostly because the vast majority of page faults are handled through the other paths. I still think this patch series is useful for workloads that may take these paths more often, and just for cleaning up the fault path in general (it's now clearer why we have to retry in these cases). Matthew Wilcox (Oracle) (6): mm: make lock_folio_maybe_drop_mmap() VMA lock aware mm: call wp_page_copy() under the VMA lock mm: handle shared faults under the VMA lock mm: handle COW faults under the VMA lock mm: handle read faults under the VMA lock mm: handle write faults to RO pages under the VMA lock mm/filemap.c | 13 ++++---- mm/memory.c | 93 ++++++++++++++++++++++++++++++++-------------------- 2 files changed, 65 insertions(+), 41 deletions(-) -- 2.25.1

2 7

[PATCH OLK-6.6] tcp/dccp: Add another way to allocate local ports in connect()
by Zhengchao Shao 29 Nov '23

29 Nov '23

From: Lu Wei <luwei32(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8KBYH -------------------------------- Commit 07f4c90062f8 ("tcp/dccp: try to not exhaust ip_local_port_range in connect()") allocates even ports for connect() first while leaving odd ports for bind() and this works well in busy servers. But this strategy causes severe performance degradation in busy clients. when a client has used more than half of the local ports setted in proc/sys/net/ipv4/ip_local_port_range, if this client trys to connect to a server again, the connect time increases rapidly since it will traverse all the even ports though they are exhausted. So this path provides another strategy by introducing a system option: local_port_allocation. If it is a busy client, users should set it to 1 to use sequential allocation while it should be set to 0 in other situations. Its default value is 0. Signed-off-by: Lu Wei <luwei32(a)huawei.com> Signed-off-by: Liu Jian <liujian56(a)huawei.com> Conflicts: include/net/tcp.h net/ipv4/inet_hashtables.c net/ipv4/sysctl_net_ipv4.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- include/net/tcp.h | 1 + net/ipv4/inet_hashtables.c | 11 ++++++++--- net/ipv4/sysctl_net_ipv4.c | 8 ++++++++ 3 files changed, 17 insertions(+), 3 deletions(-) diff --git a/include/net/tcp.h b/include/net/tcp.h index 0239e815edf7..e9d387fffe22 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -258,6 +258,7 @@ DECLARE_PER_CPU(int, tcp_memory_per_cpu_fw_alloc); extern struct percpu_counter tcp_sockets_allocated; extern unsigned long tcp_memory_pressure; +extern int sysctl_local_port_allocation; /* optimized version of sk_under_memory_pressure() for TCP sockets */ static inline bool tcp_under_memory_pressure(const struct sock *sk) diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c index 598c1b114d2c..919f0f869118 100644 --- a/net/ipv4/inet_hashtables.c +++ b/net/ipv4/inet_hashtables.c @@ -1011,7 +1011,7 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row, struct inet_bind_bucket *tb; bool tb_created = false; u32 remaining, offset; - int ret, i, low, high; + int ret, i, low, high, span_size; int l3mdev; u32 index; @@ -1021,6 +1021,11 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row, local_bh_enable(); return ret; } + /* local_port_allocation 0 means even and odd port allocation strategy + * will be applied, so span size is 2; otherwise sequential allocation + * will be used and span size is 1. Default value is 0. + */ + span_size = sysctl_local_port_allocation ? 1 : 2; l3mdev = inet_sk_bound_l3mdev(sk); @@ -1043,7 +1048,7 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row, offset &= ~1U; other_parity_scan: port = low + offset; - for (i = 0; i < remaining; i += 2, port += 2) { + for (i = 0; i < remaining; i += span_size, port += span_size) { if (unlikely(port >= high)) port -= remaining; if (inet_is_local_reserved_port(net, port)) @@ -1084,7 +1089,7 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row, } offset++; - if ((offset & 1) && remaining > 1) + if ((offset & 1) && remaining > 1 && span_size == 2) goto other_parity_scan; return -EADDRNOTAVAIL; diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index 6ac890b4073f..b17eb28a9690 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -39,6 +39,7 @@ static unsigned long ip_ping_group_range_min[] = { 0, 0 }; static unsigned long ip_ping_group_range_max[] = { GID_T_MAX, GID_T_MAX }; static u32 u32_max_div_HZ = UINT_MAX / HZ; static int one_day_secs = 24 * 3600; +int sysctl_local_port_allocation; static u32 fib_multipath_hash_fields_all_mask __maybe_unused = FIB_MULTIPATH_HASH_FIELD_ALL_MASK; static unsigned int tcp_child_ehash_entries_max = 16 * 1024 * 1024; @@ -579,6 +580,13 @@ static struct ctl_table ipv4_table[] = { .extra1 = &sysctl_fib_sync_mem_min, .extra2 = &sysctl_fib_sync_mem_max, }, + { + .procname = "local_port_allocation", + .data = &sysctl_local_port_allocation, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, { } }; -- 2.34.1

2 1

[PACH OLK-5.10 v5 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 29 Nov '23

29 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 691 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 873 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH OLK-5.10] livepatch/powerpc: Fix issue that miss one layer on stack checking
by Zheng Yejian 29 Nov '23

29 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7BMC9 CVE: NA -------------------------------- Suppose task 'T' is running in function 'B', and user is doing livepatch for function 'A', and function 'A' is calling 'B'. Then task 'T' will be interrupted by stop_machine, and suppose that 'B' has not saved the link register (which point to address in 'A') into stack memory, and data in link register position may be invalid, then 'A' would be missed on stack checking. To find a solution, we list following cases (suppose calling chain: P -> A -> B, A runs `bl B`, 'nip' means instruction pointer found in interrupt frame, 'lr' means link register found in interrupt frame, 'frame->pc' means link register saved in first stack frame): 1. 'nip' is in 'A', `bl B` is not executed, then: 'lr' is in 'P', 'frame->pc' should be ignored; 2. 'nip' is in some ppc64_stub or ppc32_plt which also not save link register, then: 'lr' is in 'A', 'frame->pc' should be ignored; 3. 'nip' is in 'B', but 'B' has not saved link register, then: 'lr' is in 'A', 'frame->pc' should be ignored; 4. 'nip' is in 'B', 'B' has saved link register, then: 'lr' is in 'A', 'frame->pc' is also in 'A'; 5. 'nip' is in 'A', 'B' has returned but its stack pointer is not moved, then 'lr' is in 'A', 'frame->pc' is also in 'A'; 6. 'nip' is in 'A', 'B' has returned and its stack pointer is moved, then 'lr' is in 'A', 'frame->pc' is in 'P'. As a conclusion, we need to: 1. check 'nip' or 'lr' or both if they are not in same function; 2. ignore 'frame->pc' if 'nip' and 'lr' are not in same function. Signed-off-by: Zheng Yejian <zhengyejian1(a)huawei.com> --- arch/powerpc/include/asm/livepatch.h | 11 ++++++++++ arch/powerpc/kernel/livepatch.c | 28 +++++++++++++++++++++++-- arch/powerpc/kernel/livepatch_32.c | 31 +++++++++++++++++++--------- arch/powerpc/kernel/livepatch_64.c | 31 +++++++++++++++++++--------- 4 files changed, 79 insertions(+), 22 deletions(-) diff --git a/arch/powerpc/include/asm/livepatch.h b/arch/powerpc/include/asm/livepatch.h index ae674ea59ab3..285602e637f1 100644 --- a/arch/powerpc/include/asm/livepatch.h +++ b/arch/powerpc/include/asm/livepatch.h @@ -121,9 +121,20 @@ struct arch_klp_data { #define KLP_MAX_REPLACE_SIZE sizeof_field(struct arch_klp_data, old_insns) struct stackframe { + /* stack frame to be unwinded */ unsigned long sp; + /* link register saved in last stack frame */ unsigned long pc; + /* instruction register saved in pt_regs */ unsigned long nip; + /* link register saved in pt_regs */ + unsigned long link; + /* stack frame pointer (r1 register) saved in pt_regs */ + unsigned long sfp; + /* check if nip and link are in same function */ + unsigned int nip_link_in_same_func; + /* check if it is top frame before interrupt */ + unsigned int is_top_frame; }; #ifdef PPC64_ELF_ABI_v1 diff --git a/arch/powerpc/kernel/livepatch.c b/arch/powerpc/kernel/livepatch.c index d568e8c8b16b..5ba38c2c7c5c 100644 --- a/arch/powerpc/kernel/livepatch.c +++ b/arch/powerpc/kernel/livepatch.c @@ -23,6 +23,7 @@ #include <linux/module.h> #include <linux/ftrace.h> #include <linux/livepatch.h> +#include <linux/kallsyms.h> #include <asm/probes.h> #include <asm/livepatch.h> #include <asm/code-patching.h> @@ -67,6 +68,22 @@ int klp_brk_handler(struct pt_regs *regs) return 1; } +static int check_addr_in_same_func(unsigned long addr1, unsigned long addr2) +{ + unsigned long size = 0; + unsigned long offset = 0; + unsigned long start; + + if (addr1 == 0 || addr2 == 0) + return 0; + if (addr1 == addr2) + return 1; + if (!kallsyms_lookup_size_offset(addr1, &size, &offset)) + return 0; + start = addr1 - offset; + return (addr2 >= start) && (addr2 - start < size); +} + int klp_unwind_frame(struct task_struct *tsk, struct stackframe *frame) { unsigned long *stack; @@ -79,7 +96,10 @@ int klp_unwind_frame(struct task_struct *tsk, struct stackframe *frame) if (frame->nip != 0) frame->nip = 0; + if (frame->link != 0) + frame->link = 0; + frame->is_top_frame = (frame->sfp == frame->sp); stack = (unsigned long *)frame->sp; /* @@ -94,10 +114,14 @@ int klp_unwind_frame(struct task_struct *tsk, struct stackframe *frame) struct pt_regs *regs = (struct pt_regs *) (frame->sp + STACK_FRAME_OVERHEAD); frame->nip = regs->nip; - pr_debug("--- interrupt: task = %d/%s, trap %lx at NIP=x%lx/%pS, LR=0x%lx/%pS\n", + frame->link = regs->link; + frame->sfp = regs->gpr[PT_R1]; + frame->nip_link_in_same_func = check_addr_in_same_func(frame->nip, frame->link); + pr_debug("--- interrupt: task = %d/%s, trap %lx at NIP=0x%lx/%pS, LR=0x%lx/%pS, SFP=0x%lx, nip_link_in_same_func=%u\n", tsk->pid, tsk->comm, regs->trap, regs->nip, (void *)regs->nip, - regs->link, (void *)regs->link); + regs->link, (void *)regs->link, + frame->sfp, frame->nip_link_in_same_func); } frame->sp = stack[0]; diff --git a/arch/powerpc/kernel/livepatch_32.c b/arch/powerpc/kernel/livepatch_32.c index 3fe4f3c5790b..134b08e12e74 100644 --- a/arch/powerpc/kernel/livepatch_32.c +++ b/arch/powerpc/kernel/livepatch_32.c @@ -272,13 +272,21 @@ static int klp_check_jump_func(struct stackframe *frame, void *data) struct walk_stackframe_args *args = data; struct klp_func_list *check_funcs = args->check_funcs; - /* check the PC first */ - if (!check_func_list(check_funcs, &args->ret, frame->pc)) - return args->ret; - /* check NIP when the exception stack switching */ if (frame->nip && !check_func_list(check_funcs, &args->ret, frame->nip)) return args->ret; + if (frame->link && !frame->nip_link_in_same_func && + !check_func_list(check_funcs, &args->ret, frame->link)) + return args->ret; + /* + * There are two cases that frame->pc is reliable: + * 1. frame->pc is not in top frame before interrupt; + * 2. nip and link are in same function; + */ + if (!frame->is_top_frame || frame->nip_link_in_same_func) { + if (!check_func_list(check_funcs, &args->ret, frame->pc)) + return args->ret; + } return 0; } @@ -298,10 +306,11 @@ static int do_check_calltrace(struct walk_stackframe_args *args, int (*fn)(struct stackframe *, void *)) { struct task_struct *g, *t; - struct stackframe frame; unsigned long *stack; for_each_process_thread(g, t) { + struct stackframe frame = { 0 }; + if (t == current) { /* * Handle the current carefully on each CPUs, we shouldn't @@ -329,7 +338,6 @@ static int do_check_calltrace(struct walk_stackframe_args *args, frame.sp = (unsigned long)stack; frame.pc = stack[STACK_FRAME_LR_SAVE]; - frame.nip = 0; klp_walk_stackframe(&frame, fn, t, args); if (args->ret) { pr_info("PID: %d Comm: %.20s\n", t->pid, t->comm); @@ -368,13 +376,16 @@ static int check_module_calltrace(struct stackframe *frame, void *data) { struct walk_stackframe_args *args = data; - /* check the PC first */ - if (within_module_core(frame->pc, args->mod)) - goto err_out; - /* check NIP when the exception stack switching */ if (frame->nip && within_module_core(frame->nip, args->mod)) goto err_out; + if (frame->link && !frame->nip_link_in_same_func && + within_module_core(frame->link, args->mod)) + goto err_out; + if (!frame->is_top_frame || frame->nip_link_in_same_func) { + if (within_module_core(frame->pc, args->mod)) + goto err_out; + } return 0; diff --git a/arch/powerpc/kernel/livepatch_64.c b/arch/powerpc/kernel/livepatch_64.c index a2ec7c8c1bad..b33839b5916a 100644 --- a/arch/powerpc/kernel/livepatch_64.c +++ b/arch/powerpc/kernel/livepatch_64.c @@ -273,13 +273,21 @@ static int klp_check_jump_func(struct stackframe *frame, void *data) struct walk_stackframe_args *args = data; struct klp_func_list *check_funcs = args->check_funcs; - /* check the PC first */ - if (!check_func_list(check_funcs, &args->ret, frame->pc)) - return args->ret; - /* check NIP when the exception stack switching */ if (frame->nip && !check_func_list(check_funcs, &args->ret, frame->nip)) return args->ret; + if (frame->link && !frame->nip_link_in_same_func && + !check_func_list(check_funcs, &args->ret, frame->link)) + return args->ret; + /* + * There are two cases that frame->pc is reliable: + * 1. frame->pc is not in top frame before interrupt; + * 2. nip and link are in same function; + */ + if (!frame->is_top_frame || frame->nip_link_in_same_func) { + if (!check_func_list(check_funcs, &args->ret, frame->pc)) + return args->ret; + } return 0; } @@ -299,10 +307,11 @@ static int do_check_calltrace(struct walk_stackframe_args *args, int (*fn)(struct stackframe *, void *)) { struct task_struct *g, *t; - struct stackframe frame; unsigned long *stack; for_each_process_thread(g, t) { + struct stackframe frame = { 0 }; + if (t == current) { /* * Handle the current carefully on each CPUs, @@ -332,7 +341,6 @@ static int do_check_calltrace(struct walk_stackframe_args *args, frame.sp = (unsigned long)stack; frame.pc = stack[STACK_FRAME_LR_SAVE]; - frame.nip = 0; klp_walk_stackframe(&frame, fn, t, args); if (args->ret) { pr_debug("%s FAILED when %s\n", __func__, @@ -373,13 +381,16 @@ static int check_module_calltrace(struct stackframe *frame, void *data) { struct walk_stackframe_args *args = data; - /* check the PC first */ - if (within_module_core(frame->pc, args->mod)) - goto err_out; - /* check NIP when the exception stack switching */ if (frame->nip && within_module_core(frame->nip, args->mod)) goto err_out; + if (frame->link && !frame->nip_link_in_same_func && + within_module_core(frame->link, args->mod)) + goto err_out; + if (!frame->is_top_frame || frame->nip_link_in_same_func) { + if (within_module_core(frame->pc, args->mod)) + goto err_out; + } return 0; -- 2.25.1

2 1

[PACH OLK-5.10 v5 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 29 Nov '23

29 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 691 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 873 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH v2 OLK-5.10 0/3] support page eject
by Wupeng Ma 29 Nov '23

29 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> Page ejection is achieved by soft_offline_page and unpoison_memory. A linked list is maintained to log the pfns which are offlined by this module. Only the pfns present in the list is allowed to go online. The follow sysfs is introduced to online/offline page via this module, since there is a internal linked list to record all offlined pages, remove_page is introduce to remove this page from list. - /sys/kernel/page_eject/online_page - /sys/kernel/page_eject/offline_page - /sys/kernel/page_eject/remove_page Changelog since v1: - remove mce related patches temporarily. Ma Wupeng (3): mm/hwpoison: Export symbol soft_offline_page mm: page_eject: Introuduce page ejection config: update defconfig for PAGE_EJECT arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/ras/Kconfig | 10 + drivers/ras/Makefile | 1 + drivers/ras/page_eject.c | 248 +++++++++++++++++++++++++ mm/memory-failure.c | 1 + 6 files changed, 262 insertions(+) create mode 100644 drivers/ras/page_eject.c -- 2.25.1

2 4

[PATCH v2 OLK-5.10 0/3] support page eject
by Wupeng Ma 29 Nov '23

29 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> Page ejection is achieved by soft_offline_page and unpoison_memory. A linked list is maintained to log the pfns which are offlined by this module. Only the pfns present in the list is allowed to go online. The follow sysfs is introduced to online/offline page via this module, since there is a internal linked list to record all offlined pages, remove_page is introduce to remove this page from list. - /sys/kernel/page_eject/online_page - /sys/kernel/page_eject/offline_page - /sys/kernel/page_eject/remove_page Besides pag eject, machine check safe is enhanced as follow: During soft offline page, page is copied to new page in kernel. If the origin page has UCE, there will lead to kernel panic. In order to solve this problem, use machine check safe to catch this error which can be achieved by using copy_mc_to_kernel to replace copy_page. Signal SIGBUS will be send to user task if this UCE is consumed by this situation to avoid kernel panic. Ma Wupeng (3): mm/hwpoison: Export symbol soft_offline_page mm: page_eject: Introuduce page ejection config: update defconfig for PAGE_EJECT arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/ras/Kconfig | 10 + drivers/ras/Makefile | 1 + drivers/ras/page_eject.c | 248 +++++++++++++++++++++++++ mm/memory-failure.c | 1 + 6 files changed, 262 insertions(+) create mode 100644 drivers/ras/page_eject.c -- 2.25.1

2 4

[PATCH OLK-5.10] driver/perf: Not enabled ARM64_BRBE by default
by Junhao He 29 Nov '23

29 Nov '23

driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8EC9K CVE: NA -------------------------------- The BRBE feature depend on ARMV8 PMU hardware support for branch stack sampling. Not enabled ARM64_BRBE by default. Signed-off-by: Junhao He <hejunhao3(a)huawei.com> --- drivers/perf/Kconfig | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/perf/Kconfig b/drivers/perf/Kconfig index 25a577530da4..67ad53cde11f 100644 --- a/drivers/perf/Kconfig +++ b/drivers/perf/Kconfig @@ -133,7 +133,6 @@ config ARM_SPE_PMU config ARM64_BRBE bool "Enable support for Branch Record Buffer Extension (BRBE)" depends on PERF_EVENTS && ARM64 && ARM_PMU - default y help Enable perf support for Branch Record Buffer Extension (BRBE) which records all branches taken in an execution path. This supports some -- 2.33.0

2 1

[PACH OLK-5.10 v5 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 29 Nov '23

29 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 691 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 873 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH OLK-6.6 v2 00/31] Backport ARM64-ILP32 patches
by Jinjie Ruan 29 Nov '23

29 Nov '23

Backport ILP32 for ARM64 patches to OLK6.6 form openEuler-23.09. Andrew Pinski (3): arm64: rename COMPAT to AARCH32_EL0 arm64: uapi: set __BITS_PER_LONG correctly for ILP32 and LP64 arm64:ilp32: add ARM64_ILP32 to Kconfig Chen Jiahao (4): arm64: fix AUDIT_ARCH_AARCH64ILP32 bug on audit subsystem arm64: fix address limit problem with TASK_SIZE_MAX arm64: set 32-bit compatible TASK_SIZE_MAX to fix U32 libc_write_01 error arm64: fix image size inflation with CONFIG_COMPAT_TASK_SIZE Dave Martin (1): arm64: signal: Make parse_user_sigframe() independent of rt_sigframe layout James Morse (1): ptrace: Add compat PTRACE_{G,S}ETSIGMASK handlers Philipp Tomsich (1): arm64:ilp32: add vdso-ilp32 and use for signal return Xiongfeng Wang (6): arm64: rename functions that reference compat term arm64: secomp: fix the secure computing mode 1 syscall check for ilp32 ilp32: avoid clearing upper 32 bits of syscall return value for ilp32 ilp32: skip ARM erratum 1418040 for ilp32 application arm64: fix abi change caused by ILP32 ilp32: fix compile problem when ARM64_ILP32 and UBSAN are both enabled Yury Norov (14): thread: move thread bits accessors to separated file arm64: ilp32: add documentation on the ILP32 ABI for ARM64 arm64: introduce is_a32_compat_{task,thread} for AArch32 compat arm64: ilp32: add is_ilp32_compat_{task,thread} and TIF_32BIT_AARCH64 arm64: introduce AUDIT_ARCH_AARCH64ILP32 for ilp32 arm64: introduce binfmt_elf32.c arm64: change compat_elf_hwcap and compat_elf_hwcap2 prefix to a32 arm64: ilp32: introduce binfmt_ilp32.c arm64: ilp32: share aarch32 syscall handlers arm64: ilp32: introduce syscall table for ILP32 arm64: signal: share lp64 signal structures and routines to ilp32 arm64: signal32: move ilp32 and aarch32 common code to separated file arm64: ilp32: introduce ilp32-specific sigframe and ucontext arm64: ptrace: handle ptrace_request differently for aarch32 and ilp32 Zhen Lei (1): arm64: replace is_compat_task() with is_ilp32_compat_task() in TASK_SIZE_MAX Documentation/arm64/ilp32.txt | 52 +++ arch/arm64/Kconfig | 26 +- arch/arm64/Makefile | 3 + arch/arm64/include/asm/arch_timer.h | 4 +- arch/arm64/include/asm/compat.h | 18 +- arch/arm64/include/asm/elf.h | 32 +- arch/arm64/include/asm/fpsimd.h | 2 +- arch/arm64/include/asm/ftrace.h | 2 +- arch/arm64/include/asm/hwcap.h | 8 +- arch/arm64/include/asm/is_compat.h | 83 ++++ arch/arm64/include/asm/processor.h | 24 +- arch/arm64/include/asm/ptrace.h | 16 +- arch/arm64/include/asm/seccomp.h | 32 +- arch/arm64/include/asm/signal32.h | 31 +- arch/arm64/include/asm/signal32_common.h | 13 + arch/arm64/include/asm/signal_common.h | 391 +++++++++++++++++ arch/arm64/include/asm/signal_ilp32.h | 23 + arch/arm64/include/asm/syscall.h | 14 +- arch/arm64/include/asm/thread_info.h | 4 +- arch/arm64/include/asm/unistd.h | 5 + arch/arm64/include/asm/vdso.h | 9 + arch/arm64/include/uapi/asm/bitsperlong.h | 9 +- arch/arm64/include/uapi/asm/unistd.h | 15 +- arch/arm64/kernel/Makefile | 8 +- arch/arm64/kernel/armv8_deprecated.c | 14 +- arch/arm64/kernel/asm-offsets.c | 13 +- arch/arm64/kernel/binfmt_elf32.c | 27 ++ arch/arm64/kernel/binfmt_ilp32.c | 89 ++++ arch/arm64/kernel/compat_alignment.c | 2 +- arch/arm64/kernel/cpufeature.c | 36 +- arch/arm64/kernel/cpuinfo.c | 18 +- arch/arm64/kernel/debug-monitors.c | 4 +- arch/arm64/kernel/entry-common.c | 6 +- arch/arm64/kernel/hw_breakpoint.c | 8 +- arch/arm64/kernel/perf_callchain.c | 28 +- arch/arm64/kernel/perf_regs.c | 4 +- arch/arm64/kernel/process.c | 15 +- arch/arm64/kernel/proton-pack.c | 2 +- arch/arm64/kernel/ptrace.c | 38 +- arch/arm64/kernel/signal.c | 397 +++--------------- arch/arm64/kernel/signal32.c | 97 ++--- arch/arm64/kernel/signal32_common.c | 37 ++ arch/arm64/kernel/signal_ilp32.c | 67 +++ arch/arm64/kernel/sys32.c | 104 +---- arch/arm64/kernel/sys32_common.c | 106 +++++ arch/arm64/kernel/sys_compat.c | 12 +- arch/arm64/kernel/sys_ilp32.c | 82 ++++ arch/arm64/kernel/syscall.c | 37 +- arch/arm64/kernel/traps.c | 5 +- arch/arm64/kernel/vdso-ilp32/.gitignore | 2 + arch/arm64/kernel/vdso-ilp32/Makefile | 111 +++++ arch/arm64/kernel/vdso-ilp32/vdso-ilp32.S | 22 + arch/arm64/kernel/vdso-ilp32/vdso-ilp32.lds.S | 88 ++++ arch/arm64/kernel/vdso.c | 46 +- arch/arm64/mm/fault.c | 2 +- include/linux/sched.h | 1 + include/linux/thread_bits.h | 87 ++++ include/linux/thread_info.h | 73 +--- kernel/ptrace.c | 52 ++- 59 files changed, 1790 insertions(+), 766 deletions(-) create mode 100644 Documentation/arm64/ilp32.txt create mode 100644 arch/arm64/include/asm/is_compat.h create mode 100644 arch/arm64/include/asm/signal32_common.h create mode 100644 arch/arm64/include/asm/signal_common.h create mode 100644 arch/arm64/include/asm/signal_ilp32.h create mode 100644 arch/arm64/kernel/binfmt_elf32.c create mode 100644 arch/arm64/kernel/binfmt_ilp32.c create mode 100644 arch/arm64/kernel/signal32_common.c create mode 100644 arch/arm64/kernel/signal_ilp32.c create mode 100644 arch/arm64/kernel/sys32_common.c create mode 100644 arch/arm64/kernel/sys_ilp32.c create mode 100644 arch/arm64/kernel/vdso-ilp32/.gitignore create mode 100644 arch/arm64/kernel/vdso-ilp32/Makefile create mode 100644 arch/arm64/kernel/vdso-ilp32/vdso-ilp32.S create mode 100644 arch/arm64/kernel/vdso-ilp32/vdso-ilp32.lds.S create mode 100644 include/linux/thread_bits.h -- 2.34.1

1 31

[openEuler-20.03-LTS-SP4 PATCH] arm64: cpufeature: Extract capped perfmon fields
by Kunkun Jiang 29 Nov '23

29 Nov '23

From: zhangyuan <zhangyuan162(a)huawei.com> KVM: arm64: limit PMU version to PMUv3 for ARMv8.1 Backport patches from upstream to fix the issue of PMU. Signed-off-by: Yuan Zhang <zhangyuan162(a)huawei.com> --- kernel.spec | 5 +- ...eature-Extract-capped-perfmon-fields.patch | 80 +++++++++++++++++ ...mit-PMU-version-to-PMUv3-for-ARMv8.1.patch | 90 +++++++++++++++++++ series.conf | 2 + 4 files changed, 176 insertions(+), 1 deletion(-) create mode 100644 patches/0758-arm64-cpufeature-Extract-capped-perfmon-fields.patch create mode 100644 patches/0759-KVM-arm64-limit-PMU-version-to-PMUv3-for-ARMv8.1.patch diff --git a/kernel.spec b/kernel.spec index 0ddbd93..7a06c2c 100644 --- a/kernel.spec +++ b/kernel.spec @@ -32,7 +32,7 @@ Name: kernel Version: 4.19.90 -Release: %{hulkrelease}.0252 +Release: %{hulkrelease}.0253 Summary: Linux Kernel License: GPLv2 URL: http://www.kernel.org/ @@ -849,6 +849,9 @@ fi %endif %changelog +* Tue Nov 28 2023 Yuan Zhang <zhangyuan162(a)huawei.com> - 4.19.90-2311.5.0.0253 +- arm64: cpufeature: Extract capped perfmon fields +- KVM: arm64: limit PMU version to PMUv3 for ARMv8.1 * Tue Nov 28 2023 Li Xiaodong <lixiaodong67(a)huawei.com> - 4.19.90-2311.5.0.0252 - !3000 [openEuler-1.0-LTS] add Phytium drivers CONFIG diff --git a/patches/0758-arm64-cpufeature-Extract-capped-perfmon-fields.patch b/patches/0758-arm64-cpufeature-Extract-capped-perfmon-fields.patch new file mode 100644 index 0000000..de1db70 --- /dev/null +++ b/patches/0758-arm64-cpufeature-Extract-capped-perfmon-fields.patch @@ -0,0 +1,80 @@ +From a207eaeac98757a86cce704be6c4e097439503c5 Mon Sep 17 00:00:00 2001 +From: Andrew Murray <andrew.murray(a)arm.com> +Date: Tue, 28 Nov 2023 15:46:32 +0800 +Subject: [PATCH 1/2] arm64: cpufeature: Extract capped perfmon fields + +mainline inclusion +from mainline-v5.7-rc1 +commit 8e35aa642ee4dab01b16cc4b2df59d1936f3b3c2 +category: bugfix +bugzilla: https://gitee.com/openeuler/kernel/issues/I8K8XV +CVE: NA + +Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… + +----------------------------------- + +commit 8e35aa642ee4dab01b16cc4b2df59d1936f3b3c2 upstream. + +When emulating ID registers there is often a need to cap the version +bits of a feature such that the guest will not use features that the +host is not aware of. For example, when KVM mediates access to the PMU +by emulating register accesses. + +Let's add a helper that extracts a performance monitors ID field and +caps the version to a given value. + +Fields that identify the version of the Performance Monitors Extension +do not follow the standard ID scheme, and instead follow the scheme +described in ARM DDI 0487E.a page D13-2825 "Alternative ID scheme used +for the Performance Monitors Extension version". The value 0xF means an +IMPLEMENTATION DEFINED PMU is present, and values 0x0-OxE can be treated +the same as an unsigned field with 0x0 meaning no PMU is present. + +Signed-off-by: Andrew Murray <andrew.murray(a)arm.com> +Reviewed-by: Suzuki K Poulose <suzuki.poulose(a)arm.com> +[Mark: rework to handle perfmon fields] +Signed-off-by: Mark Rutland <mark.rutland(a)arm.com> +Signed-off-by: Will Deacon <will(a)kernel.org> +Signed-off-by: zhangyuan <zhangyuan162(a)huawei.com> +--- + arch/arm64/include/asm/cpufeature.h | 23 +++++++++++++++++++++++ + 1 file changed, 23 insertions(+) + +diff --git a/arch/arm64/include/asm/cpufeature.h b/arch/arm64/include/asm/cpufeature.h +index eef5a9c9b823..e2ba573644d5 100644 +--- a/arch/arm64/include/asm/cpufeature.h ++++ b/arch/arm64/include/asm/cpufeature.h +@@ -449,6 +449,29 @@ cpuid_feature_extract_unsigned_field(u64 features, int field) + return cpuid_feature_extract_unsigned_field_width(features, field, 4); + } + ++/* ++ * Fields that identify the version of the Performance Monitors Extension do ++ * not follow the standard ID scheme. See ARM DDI 0487E.a page D13-2825, ++ * "Alternative ID scheme used for the Performance Monitors Extension version". ++ */ ++static inline u64 __attribute_const__ ++cpuid_feature_cap_perfmon_field(u64 features, int field, u64 cap) ++{ ++ u64 val = cpuid_feature_extract_unsigned_field(features, field); ++ u64 mask = GENMASK_ULL(field + 3, field); ++ ++ /* Treat IMPLEMENTATION DEFINED functionality as unimplemented */ ++ if (val == 0xf) ++ val = 0; ++ ++ if (val > cap) { ++ features &= ~mask; ++ features |= (cap << field) & mask; ++ } ++ ++ return features; ++} ++ + static inline u64 arm64_ftr_mask(const struct arm64_ftr_bits *ftrp) + { + return (u64)GENMASK(ftrp->shift + ftrp->width - 1, ftrp->shift); +-- +2.33.0 + diff --git a/patches/0759-KVM-arm64-limit-PMU-version-to-PMUv3-for-ARMv8.1.patch b/patches/0759-KVM-arm64-limit-PMU-version-to-PMUv3-for-ARMv8.1.patch new file mode 100644 index 0000000..1f72f92 --- /dev/null +++ b/patches/0759-KVM-arm64-limit-PMU-version-to-PMUv3-for-ARMv8.1.patch @@ -0,0 +1,90 @@ +From 6a416c5efa3ce99f5f0f7e9c603b42b0e047481c Mon Sep 17 00:00:00 2001 +From: Andrew Murray <andrew.murray(a)arm.com> +Date: Tue, 28 Nov 2023 15:46:33 +0800 +Subject: [PATCH 2/2] KVM: arm64: limit PMU version to PMUv3 for ARMv8.1 + +mainline inclusion +from mainline-v5.7-rc1 +commit c854188ea01062f5a5fd7f05658feb1863774eaa +category: bugfix +bugzilla: https://gitee.com/openeuler/kernel/issues/I8K8XV +CVE: NA + +Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… + +----------------------------------- + +commit c854188ea01062f5a5fd7f05658feb1863774eaa upstream. + +We currently expose the PMU version of the host to the guest via +emulation of the DFR0_EL1 and AA64DFR0_EL1 debug feature registers. +However many of the features offered beyond PMUv3 for 8.1 are not +supported in KVM. Examples of this include support for the PMMIR +registers (added in PMUv3 for ARMv8.4) and 64-bit event counters +added in (PMUv3 for ARMv8.5). + +Let's trap the Debug Feature Registers in order to limit +PMUVer/PerfMon in the Debug Feature Registers to PMUv3 for ARMv8.1 +to avoid unexpected behaviour. + +Both ID_AA64DFR0.PMUVer and ID_DFR0.PerfMon follow the "Alternative ID +scheme used for the Performance Monitors Extension version" where 0xF +means an IMPLEMENTATION DEFINED PMU is implemented, and values 0x0-0xE +are treated as with an unsigned field (with 0x0 meaning no PMU is +present). As we don't expect to expose an IMPLEMENTATION DEFINED PMU, +and our cap is below 0xF, we can treat these fields as unsigned when +applying the cap. + +Signed-off-by: Andrew Murray <andrew.murray(a)arm.com> +Reviewed-by: Suzuki K Poulose <suzuki.poulose(a)arm.com> +[Mark: make field names consistent, use perfmon cap] +Signed-off-by: Mark Rutland <mark.rutland(a)arm.com> +Signed-off-by: Will Deacon <will(a)kernel.org> +[zhangyuan162(a)huawei.com: adjust the context in read_id_reg()] +Signed-off-by: zhangyuan <zhangyuan162(a)huawei.com> +--- + arch/arm64/include/asm/sysreg.h | 6 ++++++ + arch/arm64/kvm/sys_regs.c | 10 ++++++++++ + 2 files changed, 16 insertions(+) + +diff --git a/arch/arm64/include/asm/sysreg.h b/arch/arm64/include/asm/sysreg.h +index de4c31f073bb..041100c28cab 100644 +--- a/arch/arm64/include/asm/sysreg.h ++++ b/arch/arm64/include/asm/sysreg.h +@@ -637,6 +637,12 @@ + #define ID_AA64DFR0_TRACEVER_SHIFT 4 + #define ID_AA64DFR0_DEBUGVER_SHIFT 0 + ++#define ID_AA64DFR0_PMUVER_8_1 0x4 ++ ++#define ID_DFR0_PERFMON_SHIFT 24 ++ ++#define ID_DFR0_PERFMON_8_1 0x4 ++ + #define ID_AA64DFR0_PMSVER_8_2 0x1 + #define ID_AA64DFR0_PMSVER_8_3 0x2 + +diff --git a/arch/arm64/kvm/sys_regs.c b/arch/arm64/kvm/sys_regs.c +index 25caa9ec95e4..cc824a208684 100644 +--- a/arch/arm64/kvm/sys_regs.c ++++ b/arch/arm64/kvm/sys_regs.c +@@ -1082,6 +1082,16 @@ static u64 read_id_reg(struct kvm_vcpu *vcpu, + kvm_debug("LORegions unsupported for guests, suppressing\n"); + + val &= ~(0xfUL << ID_AA64MMFR1_LOR_SHIFT); ++ } else if (id == SYS_ID_AA64DFR0_EL1) { ++ /* Limit guests to PMUv3 for ARMv8.1 */ ++ val = cpuid_feature_cap_perfmon_field(val, ++ ID_AA64DFR0_PMUVER_SHIFT, ++ ID_AA64DFR0_PMUVER_8_1); ++ } else if (id == SYS_ID_DFR0_EL1) { ++ /* Limit guests to PMUv3 for ARMv8.1 */ ++ val = cpuid_feature_cap_perfmon_field(val, ++ ID_DFR0_PERFMON_SHIFT, ++ ID_DFR0_PERFMON_8_1); + } + + return val; +-- +2.33.0 + diff --git a/series.conf b/series.conf index f5f2522..67b9be0 100644 --- a/series.conf +++ b/series.conf @@ -758,3 +758,5 @@ patches/0754-scsi-hisi_sas_v3_hw-Remove-extra-function-calls-for-.patch patches/0755-config-arm64-Enable-dubugfs-config-of-hisi-sas.patch patches/0756-crypto-hisilicon-Add-value-profile-support-for-kerne.patch patches/0757-Revert-genirq-Increase-the-number-of-IRQ-descriptors.patch +patches/arm64-cpufeature-Extract-capped-perfmon-fields.patch +patches/KVM-arm64-limit-PMU-version-to-PMUv3-for-ARMv8.1.patch -- 2.33.0

1 0

[PATCH OLK-5.10 0/5] SME bugfixes
by Wang ShaoBo 28 Nov '23

28 Nov '23

SME bugfixes Catalin Marinas (1): arm64: Remove the __user annotation for the restore_za_context() argument Christophe JAILLET (1): arm64/ptrace: Clean up error handling path in sve_set_common() Mark Brown (3): arm64: cpufeature: Add missing .field_width for GIC system registers arm64/signal: Flush FPSIMD register state when disabling streaming mode arm64/fpsimd: Sync and zero pad FPSIMD state for streaming SVE arch/arm64/kernel/cpufeature.c | 1 + arch/arm64/kernel/fpsimd.c | 3 ++- arch/arm64/kernel/ptrace.c | 3 ++- arch/arm64/kernel/signal.c | 12 +++++++++++- 4 files changed, 16 insertions(+), 3 deletions(-) -- 2.25.1

2 6

[PATCH OLK-5.10 0/7] support page eject
by Wupeng Ma 28 Nov '23

28 Nov '23

From: Ma Wupeng <mawupeng1(a)huawei.com> Page ejection is achieved by soft_offline_page and unpoison_memory. A linked list is maintained to log the pfns which are offlined by this module. Only the pfns present in the list is allowed to go online. The follow sysfs is introduced to online/offline page via this module, since there is a internal linked list to record all offlined pages, remove_page is introduce to remove this page from list. - /sys/kernel/page_eject/online_page - /sys/kernel/page_eject/offline_page - /sys/kernel/page_eject/remove_page Besides pag eject, machine check safe is enhanced as follow: During soft offline page, page is copied to new page in kernel. If the origin page has UCE, there will lead to kernel panic. In order to solve this problem, use machine check safe to catch this error which can be achieved by using copy_mc_to_kernel to replace copy_page. Signal SIGBUS will be send to user task if this UCE is consumed by this situation to avoid kernel panic. Jiaqi Yan (1): mm/hwpoison: introduce copy_mc_highpage Ma Wupeng (6): mm: introduce copy_mc_highpages mm: arm64: Add migrate_page_copy_mc() mm: Update PF_COREDUMP_MCS to PF_MCS mm/hwpoison: Export symbol soft_offline_page mm: page_eject: Introuduce page ejection mm: page_eject: Add mc support during offline page drivers/ras/Kconfig | 10 ++ drivers/ras/Makefile | 1 + drivers/ras/page_eject.c | 249 +++++++++++++++++++++++++++++++++++++++ fs/coredump.c | 4 +- include/linux/highmem.h | 49 ++++++++ include/linux/sched.h | 2 +- lib/iov_iter.c | 2 +- mm/memory-failure.c | 1 + mm/migrate.c | 63 ++++++++-- 9 files changed, 365 insertions(+), 16 deletions(-) create mode 100644 drivers/ras/page_eject.c -- 2.25.1

2 8

[PATCH OLK-5.10] mpam: Fix uninitialized value
by Yu Liao 28 Nov '23

28 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7ZRRD -------------------------------- Fix the following smatch warning: drivers/acpi/arm64/mpam.c:219 acpi_mpam_parse_version() error: uninitialized symbol 'ret'. Fixes: b1e2f3d84839 ("[Backport] ACPI/MPAM: Adapt to Arm's MPAM ACPI table version 2") Signed-off-by: Yu Liao <liaoyu15(a)huawei.com> --- drivers/acpi/arm64/mpam.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/acpi/arm64/mpam.c b/drivers/acpi/arm64/mpam.c index f625f539d50a..9d59652e49f8 100644 --- a/drivers/acpi/arm64/mpam.c +++ b/drivers/acpi/arm64/mpam.c @@ -219,7 +219,7 @@ int __init acpi_mpam_parse_version(void) { struct acpi_table_header *mpam, *pptt; acpi_status status; - int ret; + int ret = -EINVAL; if (!cpus_have_const_cap(ARM64_HAS_MPAM)) return 0; -- 2.25.1

2 1

[PATCH OLK-5.10] Revert "arm64/mpam: Fix mpam corrupt when cpu online"
by Yu Liao 28 Nov '23

28 Nov '23

From: Wang ShaoBo <bobo.shaobowang(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7R0WK ------------------------------------------------- BUG 'sleeping function called from invalid context' reported when setup MPAM driver, it was blamed to bc9e3f9895ef2 ("arm64/mpam: Fix mpam corrupt when cpu online"), which reported a 'Bad PC' BUG, but missing the right conclusion, finally disabling irqs before calling cpuhp_setup_state() may only affect the probability of reproduction. The reason why triggerring 'Bad PC' BUG report is because mpam_enable() is __init type function, and may schedule out after calling __cpuhp_setup_state()->__might_sleep(), so the space of mpam_enable() might be freed after scheduling back. As we have changed mpam_enable() to non-init type function, we can revert commit bc9e3f9895ef2 directly, to solve these both two problems. Fixes: bc9e3f9895ef2 ("arm64/mpam: Fix mpam corrupt when cpu online") Signed-off-by: Wang ShaoBo <bobo.shaobowang(a)huawei.com> Signed-off-by: Yu Liao <liaoyu15(a)huawei.com> --- arch/arm64/kernel/mpam/mpam_device.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/arch/arm64/kernel/mpam/mpam_device.c b/arch/arm64/kernel/mpam/mpam_device.c index 96288f66f619..b3de7f518423 100644 --- a/arch/arm64/kernel/mpam/mpam_device.c +++ b/arch/arm64/kernel/mpam/mpam_device.c @@ -621,11 +621,9 @@ static void mpam_enable(struct work_struct *work) pr_err("Failed to setup/init resctrl\n"); mutex_unlock(&mpam_devices_lock); - local_irq_disable(); mpam_cpuhp_state = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "mpam:online", mpam_cpu_online, mpam_cpu_offline); - local_irq_enable(); if (mpam_cpuhp_state <= 0) pr_err("Failed to re-register 'dyn' cpuhp callbacks"); mutex_unlock(&mpam_cpuhp_lock); -- 2.25.1

2 1

[PATCH OLK-5.10] posix-cpu-timers: Implement timer_wait_running callback
by Xiongfeng Wang 28 Nov '23

28 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7EQ8B -------------------------------- For posix CPU timers, the .timer_wait_running() callback is missing, which cause WARN_ON() in timer_wait_running(). Commit f7abf14f0001 ("posix-cpu-timers: Implement the missing timer_wait_running callback") solved the problem. But it introduce a kabi broken which is hard to fix. The commit also describe a livelock in preemptible kernel. To fix the livelock, we add cond_resched() in the retry loop. *** below is from origin commit merge tag The reason is that for CONFIG_POSIX_CPU_TIMERS_TASK_WORK enabled systems there is a livelock issue independent of RT. CONFIG_POSIX_CPU_TIMERS_TASK_WORK=y moves the expiry of POSIX CPU timers out from hard interrupt context to task work, which is handled before returning to user space or to a VM. The expiry mechanism moves the expired timers to a stack local list head with sighand lock held. Once sighand is dropped the task can be preempted and a task which wants to delete a timer will spin-wait until the expiry task is scheduled back in. In the worst case this will end up in a livelock when the preempting task and the expiry task are pinned on the same CPU. Signed-off-by: Xiongfeng Wang <wangxiongfeng2(a)huawei.com> --- kernel/time/posix-cpu-timers.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c index c97bc3e3b210..578d0ebedb67 100644 --- a/kernel/time/posix-cpu-timers.c +++ b/kernel/time/posix-cpu-timers.c @@ -1446,6 +1446,9 @@ static int do_cpu_nanosleep(const clockid_t which_clock, int flags, spin_lock_irq(&timer.it_lock); error = posix_cpu_timer_del(&timer); spin_unlock_irq(&timer.it_lock); +#ifdef CONFIG_POSIX_CPU_TIMERS_TASK_WORK + cond_resched(); +#endif } if ((it.it_value.tv_sec | it.it_value.tv_nsec) == 0) { @@ -1546,6 +1549,18 @@ static int thread_cpu_timer_create(struct k_itimer *timer) return posix_cpu_timer_create(timer); } +static void posix_cpu_timer_wait_running(struct k_itimer *timr) +{ +#ifdef CONFIG_POSIX_CPU_TIMERS_TASK_WORK + rcu_read_unlock(); + cond_resched(); + rcu_read_lock(); +#else + cpu_relax(); +#endif +} + + const struct k_clock clock_posix_cpu = { .clock_getres = posix_cpu_clock_getres, .clock_set = posix_cpu_clock_set, @@ -1556,6 +1571,7 @@ const struct k_clock clock_posix_cpu = { .timer_del = posix_cpu_timer_del, .timer_get = posix_cpu_timer_get, .timer_rearm = posix_cpu_timer_rearm, + .timer_wait_running = posix_cpu_timer_wait_running, }; const struct k_clock clock_process = { -- 2.20.1

2 1

[PATCH OLK-6.6 0/6] Handle more faults under the VMA lock
by Tong Tiangen 28 Nov '23

28 Nov '23

At this point, we're handling the majority of file-backed page faults under the VMA lock, using the ->map_pages entry point. This patch set attempts to expand that for the following siutations: - We have to do a read. This could be because we've hit the point in the readahead window where we need to kick off the next readahead, or because the page is simply not present in cache. - We're handling a write fault. Most applications don't do I/O by writes to shared mmaps for very good reasons, but some do, and it'd be nice to not make that slow unnecessarily. - We're doing a COW of a private mapping (both PTE already present and PTE not-present). These are two different codepaths and I handle both of them in this patch set. There is no support in this patch set for drivers to mark themselves as being VMA lock friendly; they could implement the ->map_pages vm_operation, but if they do, they would be the first. This is probably something we want to change at some point in the future, and I've marked where to make that change in the code. There is very little performance change in the benchmarks we've run; mostly because the vast majority of page faults are handled through the other paths. I still think this patch series is useful for workloads that may take these paths more often, and just for cleaning up the fault path in general (it's now clearer why we have to retry in these cases). Matthew Wilcox (Oracle) (6): mm: make lock_folio_maybe_drop_mmap() VMA lock aware mm: call wp_page_copy() under the VMA lock mm: handle shared faults under the VMA lock mm: handle COW faults under the VMA lock mm: handle read faults under the VMA lock mm: handle write faults to RO pages under the VMA lock mm/filemap.c | 13 ++++---- mm/memory.c | 93 ++++++++++++++++++++++++++++++++-------------------- 2 files changed, 65 insertions(+), 41 deletions(-) -- 2.25.1

2 7

[PATCH OLK-5.10 00/80] sme backport from mainline
by Wang ShaoBo 28 Nov '23

28 Nov '23

sme backport from mainline Alexandru Elisei (1): arm64: Do not trap PMSNEVFR_EL1 Marc Zyngier (10): KVM: arm64: Let vcpu_sve_pffr() handle HYP VAs KVM: arm64: Provide KVM's own save/restore SVE primitives KVM: arm64: Use {read,write}_sysreg_el1 to access ZCR_EL1 KVM: arm64: Introduce vcpu_sve_vq() helper arm64: sve: Provide a conditional update accessor for ZCR_ELx KVM: arm64: Map SVE context at EL2 when available KVM: arm64: Rework SVE host-save/guest-restore KVM: arm64: Save guest's ZCR_EL1 before saving the FPSIMD state KVM: arm64: Save/restore SVE state for nVHE KVM: arm64: Always start with clearing SME flag on load Mark Brown (65): arm64/sve: Remove redundant system_supports_sve() tests arm64/sve: Add compile time checks for SVE hooks in generic functions arm64/sve: Rework SVE access trap to convert state in registers arm64/sve: Split _sve_flush macro into separate Z and predicate flushes arm64/sve: Skip flushing Z registers with 128 bit vectors arm64/sve: Use the sve_flush macros in sve_load_from_fpsimd_state() arm64/sve: Remove sve_load_from_fpsimd_state() arm64/sve: Better handle failure to allocate SVE register storage arm64/fp: Reindent fpsimd_save() arm64/sve: Make access to FFR optional arm64/sve: Rename find_supported_vector_length() arm64/sve: Use accessor functions for vector lengths in thread_struct arm64/sve: Put system wide vector length information into structs arm64/sve: Explicitly load vector length when restoring SVE state arm64/sve: Track vector lengths for tasks in an array arm64/sve: Make sysctl interface for SVE reusable by SME arm64/sve: Generalise vector length configuration prctl() for SME arm64/sve: Minor clarification of ABI documentation arm64: cpufeature: Always specify and use a field width for capabilities arm64/sme: Provide ABI documentation for SME arm64/sme: System register and exception syndrome definitions arm64/sme: Manually encode SME instructions arm64: Disable fine grained traps on boot arm64/sme: Early CPU setup for SME arm64/sme: Basic enumeration support arm64/sme: Identify supported SME vector lengths at boot arm64/sme: Implement sysctl to set the default vector length arm64/sme: Implement vector length configuration prctl()s arm64/sme: Implement support for TPIDR2 arm64/sme: Implement SVCR context switching arm64/sme: Implement streaming SVE context switching arm64/sme: Implement ZA context switching arm64/sme: Implement traps and syscall handling for SME arm64/sme: Disable ZA and streaming mode when handling signals arm64/sme: Implement streaming SVE signal handling arm64/sme: Implement ZA signal handling arm64/sme: Implement ptrace support for streaming mode SVE registers arm64/sme: Add ptrace support for ZA arm64/sme: Disable streaming mode and ZA when flushing CPU state arm64/sme: Save and restore streaming mode over EFI runtime calls arm64/sme: Provide Kconfig for SME arm64/sme: Add ID_AA64SMFR0_EL1 to __read_sysreg_by_encoding() arm64/sme: More sensibly define the size for the ZA register set KVM: arm64: Hide SME system registers from guests KVM: arm64: Trap SME usage in guest KVM: arm64: Handle SME host state when running guests arm64/fp: Make SVE and SME length register definition match architecture arm64/fp: Rename SVE and SME LEN field name to _WIDTH arm64/sme: Drop SYS_ from SMIDR_EL1 defines arm64/sme: Standardise bitfield names for SVCR arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h arm64/sme: Fix tests for 0b1111 value ID registers arm64/sme: Fix SVE/SME typo in ABI documentation arm64/sme: Fix EFI save/restore arm64/sme: Don't flush SVE register state when allocating SME storage arm64/sme: Don't flush SVE register state when handling SME traps arm64/ptrace: Don't clear calling process' TIF_SME on OOM arm64/signal: Always allocate SVE signal frames on SME only systems arm64/sme: Don't use streaming mode to probe the maximum SME VL arm64/sme: Optimise SME exit on syscall entry arm64/fpsimd: Ensure SME storage is allocated after SVE VL changes arm64/sme: Set new vector length before reallocating arm64/fpsimd: Clear SME state in the target task when setting the VL arm64/fpsimd: Sync FPSIMD state with SVE for SME only systems arm64/ptrace: Ensure that SME is set up for target when writing SSVE state Wan Jiabing (1): arm64/sme: Fix NULL check after kzalloc Will Deacon (1): arm64: sme: Use STR P to clear FFR context field in streaming SVE mode Xiaofei Tan (1): arm64: sve: Provide sve_cond_update_zcr_vq fallback when !ARM64_SVE Zenghui Yu (1): arm64: ptrace: Use ARM64_SME to guard the SME register enumerations Documentation/arm64/elf_hwcaps.rst | 34 + Documentation/arm64/index.rst | 1 + Documentation/arm64/sme.rst | 428 ++++++++++ Documentation/arm64/sve.rst | 72 +- arch/arm64/Kconfig | 11 + arch/arm64/include/asm/cpu.h | 4 + arch/arm64/include/asm/cpucaps.h | 2 + arch/arm64/include/asm/cpufeature.h | 25 + arch/arm64/include/asm/esr.h | 13 +- arch/arm64/include/asm/exception.h | 1 + arch/arm64/include/asm/fpsimd.h | 266 ++++++- arch/arm64/include/asm/fpsimdmacros.h | 108 ++- arch/arm64/include/asm/hwcap.h | 8 + arch/arm64/include/asm/kvm_arm.h | 1 + arch/arm64/include/asm/kvm_host.h | 14 +- arch/arm64/include/asm/kvm_hyp.h | 2 + arch/arm64/include/asm/processor.h | 78 +- arch/arm64/include/asm/signal_common.h | 19 +- arch/arm64/include/asm/sysreg.h | 79 +- arch/arm64/include/asm/thread_info.h | 4 +- arch/arm64/include/uapi/asm/hwcap.h | 8 + arch/arm64/include/uapi/asm/ptrace.h | 69 +- arch/arm64/include/uapi/asm/sigcontext.h | 55 +- arch/arm64/kernel/cpufeature.c | 292 +++++-- arch/arm64/kernel/cpuinfo.c | 13 + arch/arm64/kernel/entry-common.c | 11 + arch/arm64/kernel/entry-fpsimd.S | 72 +- arch/arm64/kernel/fpsimd.c | 961 ++++++++++++++++++----- arch/arm64/kernel/head.S | 87 ++ arch/arm64/kernel/process.c | 45 +- arch/arm64/kernel/ptrace.c | 380 +++++++-- arch/arm64/kernel/signal.c | 179 ++++- arch/arm64/kernel/syscall.c | 25 +- arch/arm64/kernel/traps.c | 1 + arch/arm64/kvm/fpsimd.c | 70 +- arch/arm64/kvm/guest.c | 6 +- arch/arm64/kvm/hyp/fpsimd.S | 12 + arch/arm64/kvm/hyp/include/hyp/switch.h | 75 +- arch/arm64/kvm/hyp/nvhe/switch.c | 38 +- arch/arm64/kvm/hyp/vhe/switch.c | 10 +- arch/arm64/kvm/reset.c | 14 +- arch/arm64/kvm/sys_regs.c | 8 +- include/uapi/linux/elf.h | 2 + include/uapi/linux/prctl.h | 9 + kernel/sys.c | 12 + 45 files changed, 3148 insertions(+), 476 deletions(-) create mode 100644 Documentation/arm64/sme.rst -- 2.25.1

1 80

[PATCH OLK-6.6 0/8] memcg_qos and memcg_async_reclaim
by Lu Jialin 28 Nov '23

28 Nov '23

Lu Jialin (8): cgroup: Export cgroup.kill from cgroupv2 to cgroupv1 memcg: Export memcg.{min/low} from cgroupv2 to cgroupv1 memcg: Export memcg.high from cgroupv2 to cgroupv1 memcg: Export memory.events and memory.events.local from cgroupv2 to cgroupv1 memcg: Fix inconsistent oom event behavior for OOM_MEMCG_KILL memcg: enable memcg async reclaim memcg: export high_async_ratio to userland memcg: Introduce CONFIG_MEMCG_V1_THRESHOLD_QOS arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/memcontrol.h | 24 +- init/Kconfig | 6 + kernel/cgroup/cgroup-internal.h | 3 + kernel/cgroup/cgroup-v1.c | 5 + kernel/cgroup/cgroup.c | 4 +- mm/memcontrol.c | 378 ++++++++++++++++++------- 8 files changed, 310 insertions(+), 112 deletions(-) -- 2.34.1

2 9

[PATCH OLK-5.10 00/80] sme backport from mainline
by Wang ShaoBo 28 Nov '23

28 Nov '23

sme backport from mainline Alexandru Elisei (1): arm64: Do not trap PMSNEVFR_EL1 Marc Zyngier (10): KVM: arm64: Let vcpu_sve_pffr() handle HYP VAs KVM: arm64: Provide KVM's own save/restore SVE primitives KVM: arm64: Use {read,write}_sysreg_el1 to access ZCR_EL1 KVM: arm64: Introduce vcpu_sve_vq() helper arm64: sve: Provide a conditional update accessor for ZCR_ELx KVM: arm64: Map SVE context at EL2 when available KVM: arm64: Rework SVE host-save/guest-restore KVM: arm64: Save guest's ZCR_EL1 before saving the FPSIMD state KVM: arm64: Save/restore SVE state for nVHE KVM: arm64: Always start with clearing SME flag on load Mark Brown (65): arm64/sve: Remove redundant system_supports_sve() tests arm64/sve: Add compile time checks for SVE hooks in generic functions arm64/sve: Rework SVE access trap to convert state in registers arm64/sve: Split _sve_flush macro into separate Z and predicate flushes arm64/sve: Skip flushing Z registers with 128 bit vectors arm64/sve: Use the sve_flush macros in sve_load_from_fpsimd_state() arm64/sve: Remove sve_load_from_fpsimd_state() arm64/sve: Better handle failure to allocate SVE register storage arm64/fp: Reindent fpsimd_save() arm64/sve: Make access to FFR optional arm64/sve: Rename find_supported_vector_length() arm64/sve: Use accessor functions for vector lengths in thread_struct arm64/sve: Put system wide vector length information into structs arm64/sve: Explicitly load vector length when restoring SVE state arm64/sve: Track vector lengths for tasks in an array arm64/sve: Make sysctl interface for SVE reusable by SME arm64/sve: Generalise vector length configuration prctl() for SME arm64/sve: Minor clarification of ABI documentation arm64: cpufeature: Always specify and use a field width for capabilities arm64/sme: Provide ABI documentation for SME arm64/sme: System register and exception syndrome definitions arm64/sme: Manually encode SME instructions arm64: Disable fine grained traps on boot arm64/sme: Early CPU setup for SME arm64/sme: Basic enumeration support arm64/sme: Identify supported SME vector lengths at boot arm64/sme: Implement sysctl to set the default vector length arm64/sme: Implement vector length configuration prctl()s arm64/sme: Implement support for TPIDR2 arm64/sme: Implement SVCR context switching arm64/sme: Implement streaming SVE context switching arm64/sme: Implement ZA context switching arm64/sme: Implement traps and syscall handling for SME arm64/sme: Disable ZA and streaming mode when handling signals arm64/sme: Implement streaming SVE signal handling arm64/sme: Implement ZA signal handling arm64/sme: Implement ptrace support for streaming mode SVE registers arm64/sme: Add ptrace support for ZA arm64/sme: Disable streaming mode and ZA when flushing CPU state arm64/sme: Save and restore streaming mode over EFI runtime calls arm64/sme: Provide Kconfig for SME arm64/sme: Add ID_AA64SMFR0_EL1 to __read_sysreg_by_encoding() arm64/sme: More sensibly define the size for the ZA register set KVM: arm64: Hide SME system registers from guests KVM: arm64: Trap SME usage in guest KVM: arm64: Handle SME host state when running guests arm64/fp: Make SVE and SME length register definition match architecture arm64/fp: Rename SVE and SME LEN field name to _WIDTH arm64/sme: Drop SYS_ from SMIDR_EL1 defines arm64/sme: Standardise bitfield names for SVCR arm64/sme: Remove _EL0 from name of SVCR - FIXME sysreg.h arm64/sme: Fix tests for 0b1111 value ID registers arm64/sme: Fix SVE/SME typo in ABI documentation arm64/sme: Fix EFI save/restore arm64/sme: Don't flush SVE register state when allocating SME storage arm64/sme: Don't flush SVE register state when handling SME traps arm64/ptrace: Don't clear calling process' TIF_SME on OOM arm64/signal: Always allocate SVE signal frames on SME only systems arm64/sme: Don't use streaming mode to probe the maximum SME VL arm64/sme: Optimise SME exit on syscall entry arm64/fpsimd: Ensure SME storage is allocated after SVE VL changes arm64/sme: Set new vector length before reallocating arm64/fpsimd: Clear SME state in the target task when setting the VL arm64/fpsimd: Sync FPSIMD state with SVE for SME only systems arm64/ptrace: Ensure that SME is set up for target when writing SSVE state Wan Jiabing (1): arm64/sme: Fix NULL check after kzalloc Will Deacon (1): arm64: sme: Use STR P to clear FFR context field in streaming SVE mode Xiaofei Tan (1): arm64: sve: Provide sve_cond_update_zcr_vq fallback when !ARM64_SVE Zenghui Yu (1): arm64: ptrace: Use ARM64_SME to guard the SME register enumerations Documentation/arm64/elf_hwcaps.rst | 34 + Documentation/arm64/index.rst | 1 + Documentation/arm64/sme.rst | 428 ++++++++++ Documentation/arm64/sve.rst | 72 +- arch/arm64/Kconfig | 11 + arch/arm64/include/asm/cpu.h | 4 + arch/arm64/include/asm/cpucaps.h | 2 + arch/arm64/include/asm/cpufeature.h | 25 + arch/arm64/include/asm/esr.h | 13 +- arch/arm64/include/asm/exception.h | 1 + arch/arm64/include/asm/fpsimd.h | 266 ++++++- arch/arm64/include/asm/fpsimdmacros.h | 108 ++- arch/arm64/include/asm/hwcap.h | 8 + arch/arm64/include/asm/kvm_arm.h | 1 + arch/arm64/include/asm/kvm_host.h | 14 +- arch/arm64/include/asm/kvm_hyp.h | 2 + arch/arm64/include/asm/processor.h | 78 +- arch/arm64/include/asm/signal_common.h | 19 +- arch/arm64/include/asm/sysreg.h | 79 +- arch/arm64/include/asm/thread_info.h | 4 +- arch/arm64/include/uapi/asm/hwcap.h | 8 + arch/arm64/include/uapi/asm/ptrace.h | 69 +- arch/arm64/include/uapi/asm/sigcontext.h | 55 +- arch/arm64/kernel/cpufeature.c | 292 +++++-- arch/arm64/kernel/cpuinfo.c | 13 + arch/arm64/kernel/entry-common.c | 11 + arch/arm64/kernel/entry-fpsimd.S | 72 +- arch/arm64/kernel/fpsimd.c | 961 ++++++++++++++++++----- arch/arm64/kernel/head.S | 87 ++ arch/arm64/kernel/process.c | 45 +- arch/arm64/kernel/ptrace.c | 380 +++++++-- arch/arm64/kernel/signal.c | 179 ++++- arch/arm64/kernel/syscall.c | 25 +- arch/arm64/kernel/traps.c | 1 + arch/arm64/kvm/fpsimd.c | 70 +- arch/arm64/kvm/guest.c | 6 +- arch/arm64/kvm/hyp/fpsimd.S | 12 + arch/arm64/kvm/hyp/include/hyp/switch.h | 75 +- arch/arm64/kvm/hyp/nvhe/switch.c | 38 +- arch/arm64/kvm/hyp/vhe/switch.c | 10 +- arch/arm64/kvm/reset.c | 14 +- arch/arm64/kvm/sys_regs.c | 8 +- include/uapi/linux/elf.h | 2 + include/uapi/linux/prctl.h | 9 + kernel/sys.c | 12 + 45 files changed, 3148 insertions(+), 476 deletions(-) create mode 100644 Documentation/arm64/sme.rst -- 2.25.1

1 80

[PATCH 0/4] mm: hugetlb: Skip initialization of gigantic tail struct pages if freed by HVO
by Jinjiang Tu 28 Nov '23

28 Nov '23

This series moves the boot time initialization of tail struct pages of a gigantic page to later on in the boot. Only the HUGETLB_VMEMMAP_RESERVE_SIZE / sizeof(struct page) - 1 tail struct pages are initialized at the start. If HVO is successful, then no more tail struct pages need to be initialized. For a 1G hugepage, this series avoid initialization of 262144 - 63 = 262081 struct pages per hugepage. When tested on a 512G system (allocating 500 1G hugepages), the kexec-boot times with DEFERRED_STRUCT_PAGE_INIT enabled are: - with patches, HVO enabled: 1.32 seconds - with patches, HVO disabled: 2.15 seconds - without patches, HVO enabled: 3.90 seconds - without patches, HVO disabled: 3.58 seconds This represents an approximately 70% reduction in boot time and will significantly reduce server downtime when using a large number of gigantic pages. Usama Arif (4): mm: hugetlb_vmemmap: use nid of the head page to reallocate it memblock: pass memblock_type to memblock_setclr_flag memblock: introduce MEMBLOCK_RSRV_NOINIT flag mm: hugetlb: skip initialization of gigantic tail struct pages if freed by HVO include/linux/memblock.h | 9 ++++++ mm/hugetlb.c | 68 ++++++++++++++++++++++++++++++++++------ mm/hugetlb_vmemmap.c | 4 +-- mm/hugetlb_vmemmap.h | 9 +++--- mm/internal.h | 3 ++ mm/memblock.c | 49 +++++++++++++++++++++-------- mm/mm_init.c | 2 +- 7 files changed, 115 insertions(+), 29 deletions(-) -- 2.25.1

1 4

[PATCH OLK-6.6 0/4] mm: hugetlb: Skip initialization of gigantic tail struct pages if freed by HVO
by Jinjiang Tu 28 Nov '23

28 Nov '23

This series moves the boot time initialization of tail struct pages of a gigantic page to later on in the boot. Only the HUGETLB_VMEMMAP_RESERVE_SIZE / sizeof(struct page) - 1 tail struct pages are initialized at the start. If HVO is successful, then no more tail struct pages need to be initialized. For a 1G hugepage, this series avoid initialization of 262144 - 63 = 262081 struct pages per hugepage. When tested on a 512G system (allocating 500 1G hugepages), the kexec-boot times with DEFERRED_STRUCT_PAGE_INIT enabled are: - with patches, HVO enabled: 1.32 seconds - with patches, HVO disabled: 2.15 seconds - without patches, HVO enabled: 3.90 seconds - without patches, HVO disabled: 3.58 seconds This represents an approximately 70% reduction in boot time and will significantly reduce server downtime when using a large number of gigantic pages. Usama Arif (4): mm: hugetlb_vmemmap: use nid of the head page to reallocate it memblock: pass memblock_type to memblock_setclr_flag memblock: introduce MEMBLOCK_RSRV_NOINIT flag mm: hugetlb: skip initialization of gigantic tail struct pages if freed by HVO include/linux/memblock.h | 9 ++++++ mm/hugetlb.c | 68 ++++++++++++++++++++++++++++++++++------ mm/hugetlb_vmemmap.c | 4 +-- mm/hugetlb_vmemmap.h | 9 +++--- mm/internal.h | 3 ++ mm/memblock.c | 49 +++++++++++++++++++++-------- mm/mm_init.c | 2 +- 7 files changed, 115 insertions(+), 29 deletions(-) -- 2.25.1

2 5

[PATCH v3 OLK-6.6 00/17] Support Qos scheduler
by Wenyu Huang 28 Nov '23

28 Nov '23

This patch adds the Qos scheduler. Hui Tang (1): sched: Fix null-ptr-deref in free_fair_sched_group Xia Fukun (1): sched/qos: Fix warning in CPU hotplug scenarios Zhang Qiao (15): sched: Introduce qos scheduler for co-location sched: Throttle qos cfs_rq when current cpu is running online task sched: Fix offline task can't be killed in a timely sched: Unthrottle the throttled cfs rq when offline rq sched: Unthrottle qos cfs rq when free a task group sched: bugfix setscheduler unlock cpuset_rwsem sched: Introduce handle priority reversion mechanism sched: enable CONFIG_QOS_SCHED on arm64 sched: Fix sleeping in atomic context at cpu_qos_write() sched/fair: Add qos_throttle_list node in struct cfs_rq sched: Throttle offline task at tracehook_notify_resume() sched/qos: Add qos_tg_{throttle,unthrottle}_{up,down} sched/fair: Update rq clock before unthrottle a qos cfs_rq sched/qos: Don't unthrottle cfs_rq when cfs_rq is throttled by qos sched: Enable qos scheduler config arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/cgroup.h | 4 + include/linux/resume_user_mode.h | 5 + include/linux/sched.h | 23 ++ include/linux/sched/sysctl.h | 8 + init/Kconfig | 12 + kernel/cgroup/cgroup.c | 22 ++ kernel/sched/core.c | 155 +++++++++++ kernel/sched/fair.c | 357 ++++++++++++++++++++++++- kernel/sched/sched.h | 12 + kernel/signal.c | 3 + kernel/sysctl.c | 24 ++ 13 files changed, 626 insertions(+), 1 deletion(-) -- 2.34.1

2 18

[PATCH OLK-6.6 0/3] Introduce NEON implementation of CRCT10DIF algorithm
by GUO Zihua 28 Nov '23

28 Nov '23

This patchset introduces ARM64 NEON implementation of CRC T10 DIF algorithm. This algorithm could be enabled by enabling CONFIG_CRYPTO_CRCT10DIF_ARM64_NEON, and the compiled KO is named crct10dif-neon.ko GUO Zihua (2): arm64/crypto: Introduce individual config for CRCT10DIF-NEON algorithm crypto: crct10dif-neon - fix use via crypto_shash_digest() YueHaibing (1): arm64/crypto: Accelerated CRC T10 DIF computation arch/arm64/crypto/Kconfig | 10 + arch/arm64/crypto/Makefile | 4 + arch/arm64/crypto/crct10dif-neon-asm_64.S | 752 ++++++++++++++++++++++ arch/arm64/crypto/crct10dif-neon_glue.c | 114 ++++ 4 files changed, 880 insertions(+) create mode 100644 arch/arm64/crypto/crct10dif-neon-asm_64.S create mode 100644 arch/arm64/crypto/crct10dif-neon_glue.c -- 2.34.1

2 4

[PATCH OLK-6.6 0/1] Random boot-time optimization
by GUO Zihua 28 Nov '23

28 Nov '23

This patch introduced random boot-time initialization optimization. The optimization works by disabling the deferring of interrupt randomness work until random is fully initialized. Xingjun Liu (1): alinux: random: speed up the initialization of module drivers/char/random.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) -- 2.34.1

2 2

[PATCH 03/15] crypto: hisilicon/zip - add zip comp high perf mode configuration
by Zhiqi Song 28 Nov '23

28 Nov '23

From: Chenghai Huang <huangchenghai2(a)huawei.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8IVG6 CVE: NA ---------------------------------------------------------------------- To meet specific application scenarios, the function of switching between the high performance mode and the high compression mode is added. Use the perf_mode=0/1 configuration to set the compression high perf mode, 0(default, high compression mode), 1(high performance mode). These two modes only apply to the compression direction and are compatible with software algorithm in both directions. Signed-off-by: Chenghai Huang <huangchenghai2(a)huawei.com> Signed-off-by: JiangShui Yang <yangjiangshui(a)h-partners.com> --- drivers/crypto/hisilicon/zip/zip_main.c | 65 +++++++++++++++++++++++++ 1 file changed, 65 insertions(+) diff --git a/drivers/crypto/hisilicon/zip/zip_main.c b/drivers/crypto/hisilicon/zip/zip_main.c index 3c00bb1474ee..998ddffb662a 100644 --- a/drivers/crypto/hisilicon/zip/zip_main.c +++ b/drivers/crypto/hisilicon/zip/zip_main.c @@ -106,6 +106,14 @@ #define HZIP_CLOCK_GATED_EN (HZIP_CORE_GATED_EN | \ HZIP_CORE_GATED_OOO_EN) +/* zip comp high performance */ +#define HZIP_HIGH_PERF_OFFSET 0x301208 + +enum { + HZIP_HIGH_COMP_RATE, + HZIP_HIGH_COMP_PERF, +}; + static const char hisi_zip_name[] = "hisi_zip"; static struct dentry *hzip_debugfs_root; @@ -366,6 +374,37 @@ static int hzip_diff_regs_show(struct seq_file *s, void *unused) return 0; } DEFINE_SHOW_ATTRIBUTE(hzip_diff_regs); + +static int perf_mode_set(const char *val, const struct kernel_param *kp) +{ + int ret; + u32 n; + + if (!val) + return -EINVAL; + + ret = kstrtou32(val, 10, &n); + if (ret != 0 || (n != HZIP_HIGH_COMP_PERF && + n != HZIP_HIGH_COMP_RATE)) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops zip_com_perf_ops = { + .set = perf_mode_set, + .get = param_get_int, +}; + +/* + * perf_mode = 0 means enable high compression rate mode, + * perf_mode = 1 means enable high compression performance mode. + * These two modes only apply to the compression direction. + */ +static u32 perf_mode = HZIP_HIGH_COMP_RATE; +module_param_cb(perf_mode, &zip_com_perf_ops, &perf_mode, 0444); +MODULE_PARM_DESC(perf_mode, "ZIP high perf mode 0(default), 1(enable)"); + static const struct kernel_param_ops zip_uacce_mode_ops = { .set = uacce_mode_set, .get = param_get_int, @@ -431,6 +470,28 @@ bool hisi_zip_alg_support(struct hisi_qm *qm, u32 alg) return false; } +static int hisi_zip_set_high_perf(struct hisi_qm *qm) +{ + u32 val; + int ret; + + val = readl_relaxed(qm->io_base + HZIP_HIGH_PERF_OFFSET); + if (perf_mode == HZIP_HIGH_COMP_PERF) + val |= HZIP_HIGH_COMP_PERF; + else + val &= ~HZIP_HIGH_COMP_PERF; + + /* Set perf mode */ + writel(val, qm->io_base + HZIP_HIGH_PERF_OFFSET); + ret = readl_relaxed_poll_timeout(qm->io_base + HZIP_HIGH_PERF_OFFSET, + val, val == perf_mode, HZIP_DELAY_1_US, + HZIP_POLL_TIMEOUT_US); + if (ret) + pci_err(qm->pdev, "failed to set perf mode\n"); + + return ret; +} + static void hisi_zip_open_sva_prefetch(struct hisi_qm *qm) { u32 val; @@ -1113,6 +1174,10 @@ static int hisi_zip_pf_probe_init(struct hisi_zip *hisi_zip) if (ret) return ret; + ret = hisi_zip_set_high_perf(qm); + if (ret) + return ret; + hisi_zip_open_sva_prefetch(qm); hisi_qm_dev_err_init(qm); hisi_zip_debug_regs_clear(qm); -- 2.33.0

1 0

[PATCH 1/3] crypto: hisilicon/sgl - small cleanups for sgl.c
by Weili Qian 28 Nov '23

28 Nov '23

driver inclusion category: cleanup bugzilla: https://gitee.com/openeuler/kernel/issues/I8IVG6 CVE: NA ---------------------------------------------------------------------- 1. Remove unnecessary brackets in function hisi_acc_create_sgl_pool(). 2. Modify local variable type, ensure that the variable type is consistent with the variable type to be compared. 3. Because the function clear_hw_sgl_sge() is in the task process, obtain the value of le16_to_cpu(hw_sgl->entry_sum_in_sgl) before loop execting to shorten the loop execution time. Signed-off-by: Weili Qian <qianweili(a)huawei.com> Signed-off-by: JiangShui Yang <yangjiangshui(a)h-partners.com> --- drivers/crypto/hisilicon/sgl.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/drivers/crypto/hisilicon/sgl.c b/drivers/crypto/hisilicon/sgl.c index 09586a837b1e..6f281d808a0f 100644 --- a/drivers/crypto/hisilicon/sgl.c +++ b/drivers/crypto/hisilicon/sgl.c @@ -121,10 +121,10 @@ struct hisi_acc_sgl_pool *hisi_acc_create_sgl_pool(struct device *dev, return pool; err_free_mem: - for (j = 0; j < i; j++) { + for (j = 0; j < i; j++) dma_free_coherent(dev, block_size, block[j].sgl, block[j].sgl_dma); - } + kfree_sensitive(pool); return ERR_PTR(-ENOMEM); } @@ -140,7 +140,7 @@ EXPORT_SYMBOL_GPL(hisi_acc_create_sgl_pool); void hisi_acc_free_sgl_pool(struct device *dev, struct hisi_acc_sgl_pool *pool) { struct mem_block *block; - int i; + u32 i; if (!dev || !pool) return; @@ -196,9 +196,10 @@ static void update_hw_sgl_sum_sge(struct hisi_acc_hw_sgl *hw_sgl, u16 sum) static void clear_hw_sgl_sge(struct hisi_acc_hw_sgl *hw_sgl) { struct acc_hw_sge *hw_sge = hw_sgl->sge_entries; + u16 entry_sum = le16_to_cpu(hw_sgl->entry_sum_in_sgl); int i; - for (i = 0; i < le16_to_cpu(hw_sgl->entry_sum_in_sgl); i++) { + for (i = 0; i < entry_sum; i++) { hw_sge[i].page_ctrl = NULL; hw_sge[i].buf = 0; hw_sge[i].len = 0; @@ -223,10 +224,11 @@ hisi_acc_sg_buf_map_to_hw_sgl(struct device *dev, u32 index, dma_addr_t *hw_sgl_dma) { struct hisi_acc_hw_sgl *curr_hw_sgl; + unsigned int i, sg_n_mapped; dma_addr_t curr_sgl_dma = 0; struct acc_hw_sge *curr_hw_sge; struct scatterlist *sg; - int i, sg_n, sg_n_mapped; + int sg_n; if (!dev || !sgl || !pool || !hw_sgl_dma) return ERR_PTR(-EINVAL); -- 2.33.0

1 2

[PATCH v2 OLK-6.6 00/17] Support Qos scheduler
by Wenyu Huang 28 Nov '23

28 Nov '23

Hui Tang (1): sched: Fix null-ptr-deref in free_fair_sched_group Xia Fukun (1): sched/qos: Fix warning in CPU hotplug scenarios Zhang Qiao (15): sched: Introduce qos scheduler for co-location sched: Throttle qos cfs_rq when current cpu is running online task sched: Fix offline task can't be killed in a timely sched: Unthrottle the throttled cfs rq when offline rq sched: Unthrottle qos cfs rq when free a task group sched: bugfix setscheduler unlock cpuset_rwsem sched: Introduce handle priority reversion mechanism sched: enable CONFIG_QOS_SCHED on arm64 sched: Fix sleeping in atomic context at cpu_qos_write() sched/fair: Add qos_throttle_list node in struct cfs_rq sched: Throttle offline task at tracehook_notify_resume() sched/qos: Add qos_tg_{throttle,unthrottle}_{up,down} sched/fair: Update rq clock before unthrottle a qos cfs_rq sched/qos: Don't unthrottle cfs_rq when cfs_rq is throttled by qos sched: Enable qos scheduler config arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/cgroup.h | 4 + include/linux/resume_user_mode.h | 5 + include/linux/sched.h | 23 ++ include/linux/sched/sysctl.h | 8 + init/Kconfig | 12 + kernel/cgroup/cgroup.c | 22 ++ kernel/sched/core.c | 155 +++++++++++ kernel/sched/fair.c | 357 ++++++++++++++++++++++++- kernel/sched/sched.h | 12 + kernel/signal.c | 3 + kernel/sysctl.c | 24 ++ 13 files changed, 626 insertions(+), 1 deletion(-) -- 2.34.1

1 17

[PATCH OLK-5.10] ipv4: igmp: fix refcnt uaf issue when receiving igmp query packet
by Zhengchao Shao 28 Nov '23

28 Nov '23

mainline inclusion from mainline-v6.7-rc2 commit e2b706c691905fe78468c361aaabc719d0a496f1 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8K0N2 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When I perform the following test operations: 1.ip link add br0 type bridge 2.brctl addif br0 eth0 3.ip addr add 239.0.0.1/32 dev eth0 4.ip addr add 239.0.0.1/32 dev br0 5.ip addr add 224.0.0.1/32 dev br0 6.while ((1)) do ifconfig br0 up ifconfig br0 down done 7.send IGMPv2 query packets to port eth0 continuously. For example, ./mausezahn ethX -c 0 "01 00 5e 00 00 01 00 72 19 88 aa 02 08 00 45 00 00 1c 00 01 00 00 01 02 0e 7f c0 a8 0a b7 e0 00 00 01 11 64 ee 9b 00 00 00 00" The preceding tests may trigger the refcnt uaf issue of the mc list. The stack is as follows: refcount_t: addition on 0; use-after-free. WARNING: CPU: 21 PID: 144 at lib/refcount.c:25 refcount_warn_saturate (lib/refcount.c:25) CPU: 21 PID: 144 Comm: ksoftirqd/21 Kdump: loaded Not tainted 6.7.0-rc1-next-20231117-dirty #80 Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011 RIP: 0010:refcount_warn_saturate (lib/refcount.c:25) RSP: 0018:ffffb68f00657910 EFLAGS: 00010286 RAX: 0000000000000000 RBX: ffff8a00c3bf96c0 RCX: ffff8a07b6160908 RDX: 00000000ffffffd8 RSI: 0000000000000027 RDI: ffff8a07b6160900 RBP: ffff8a00cba36862 R08: 0000000000000000 R09: 00000000ffff7fff R10: ffffb68f006577c0 R11: ffffffffb0fdcdc8 R12: ffff8a00c3bf9680 R13: ffff8a00c3bf96f0 R14: 0000000000000000 R15: ffff8a00d8766e00 FS: 0000000000000000(0000) GS:ffff8a07b6140000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055f10b520b28 CR3: 000000039741a000 CR4: 00000000000006f0 Call Trace: <TASK> igmp_heard_query (net/ipv4/igmp.c:1068) igmp_rcv (net/ipv4/igmp.c:1132) ip_protocol_deliver_rcu (net/ipv4/ip_input.c:205) ip_local_deliver_finish (net/ipv4/ip_input.c:234) __netif_receive_skb_one_core (net/core/dev.c:5529) netif_receive_skb_internal (net/core/dev.c:5729) netif_receive_skb (net/core/dev.c:5788) br_handle_frame_finish (net/bridge/br_input.c:216) nf_hook_bridge_pre (net/bridge/br_input.c:294) __netif_receive_skb_core (net/core/dev.c:5423) __netif_receive_skb_list_core (net/core/dev.c:5606) __netif_receive_skb_list (net/core/dev.c:5674) netif_receive_skb_list_internal (net/core/dev.c:5764) napi_gro_receive (net/core/gro.c:609) e1000_clean_rx_irq (drivers/net/ethernet/intel/e1000/e1000_main.c:4467) e1000_clean (drivers/net/ethernet/intel/e1000/e1000_main.c:3805) __napi_poll (net/core/dev.c:6533) net_rx_action (net/core/dev.c:6735) __do_softirq (kernel/softirq.c:554) run_ksoftirqd (kernel/softirq.c:913) smpboot_thread_fn (kernel/smpboot.c:164) kthread (kernel/kthread.c:388) ret_from_fork (arch/x86/kernel/process.c:153) ret_from_fork_asm (arch/x86/entry/entry_64.S:250) </TASK> The root causes are as follows: Thread A Thread B ... netif_receive_skb br_dev_stop ... br_multicast_leave_snoopers ... __ip_mc_dec_group ... __igmp_group_dropped igmp_rcv igmp_stop_timer igmp_heard_query //ref = 1 ip_ma_put igmp_mod_timer refcount_dec_and_test igmp_start_timer //ref = 0 ... refcount_inc //ref increases from 0 When the device receives an IGMPv2 Query message, it starts the timer immediately, regardless of whether the device is running. If the device is down and has left the multicast group, it will cause the mc list refcount uaf issue. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Reviewed-by: Hangbin Liu <liuhangbin(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Conflicts: net/ipv4/igmp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/ipv4/igmp.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index c71b863093ac..73c24ce011f0 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -216,8 +216,10 @@ static void igmp_start_timer(struct ip_mc_list *im, int max_delay) int tv = prandom_u32() % max_delay; im->tm_running = 1; - if (!mod_timer(&im->timer, jiffies+tv+2)) - refcount_inc(&im->refcnt); + if (refcount_inc_not_zero(&im->refcnt)) { + if (mod_timer(&im->timer, jiffies + tv + 2)) + ip_ma_put(im); + } } static void igmp_gq_start_timer(struct in_device *in_dev) -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS] ipv4: igmp: fix refcnt uaf issue when receiving igmp query packet
by Zhengchao Shao 28 Nov '23

28 Nov '23

mainline inclusion from mainline-v6.7-rc2 commit e2b706c691905fe78468c361aaabc719d0a496f1 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8K0N2 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When I perform the following test operations: 1.ip link add br0 type bridge 2.brctl addif br0 eth0 3.ip addr add 239.0.0.1/32 dev eth0 4.ip addr add 239.0.0.1/32 dev br0 5.ip addr add 224.0.0.1/32 dev br0 6.while ((1)) do ifconfig br0 up ifconfig br0 down done 7.send IGMPv2 query packets to port eth0 continuously. For example, ./mausezahn ethX -c 0 "01 00 5e 00 00 01 00 72 19 88 aa 02 08 00 45 00 00 1c 00 01 00 00 01 02 0e 7f c0 a8 0a b7 e0 00 00 01 11 64 ee 9b 00 00 00 00" The preceding tests may trigger the refcnt uaf issue of the mc list. The stack is as follows: refcount_t: addition on 0; use-after-free. WARNING: CPU: 21 PID: 144 at lib/refcount.c:25 refcount_warn_saturate (lib/refcount.c:25) CPU: 21 PID: 144 Comm: ksoftirqd/21 Kdump: loaded Not tainted 6.7.0-rc1-next-20231117-dirty #80 Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011 RIP: 0010:refcount_warn_saturate (lib/refcount.c:25) RSP: 0018:ffffb68f00657910 EFLAGS: 00010286 RAX: 0000000000000000 RBX: ffff8a00c3bf96c0 RCX: ffff8a07b6160908 RDX: 00000000ffffffd8 RSI: 0000000000000027 RDI: ffff8a07b6160900 RBP: ffff8a00cba36862 R08: 0000000000000000 R09: 00000000ffff7fff R10: ffffb68f006577c0 R11: ffffffffb0fdcdc8 R12: ffff8a00c3bf9680 R13: ffff8a00c3bf96f0 R14: 0000000000000000 R15: ffff8a00d8766e00 FS: 0000000000000000(0000) GS:ffff8a07b6140000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055f10b520b28 CR3: 000000039741a000 CR4: 00000000000006f0 Call Trace: <TASK> igmp_heard_query (net/ipv4/igmp.c:1068) igmp_rcv (net/ipv4/igmp.c:1132) ip_protocol_deliver_rcu (net/ipv4/ip_input.c:205) ip_local_deliver_finish (net/ipv4/ip_input.c:234) __netif_receive_skb_one_core (net/core/dev.c:5529) netif_receive_skb_internal (net/core/dev.c:5729) netif_receive_skb (net/core/dev.c:5788) br_handle_frame_finish (net/bridge/br_input.c:216) nf_hook_bridge_pre (net/bridge/br_input.c:294) __netif_receive_skb_core (net/core/dev.c:5423) __netif_receive_skb_list_core (net/core/dev.c:5606) __netif_receive_skb_list (net/core/dev.c:5674) netif_receive_skb_list_internal (net/core/dev.c:5764) napi_gro_receive (net/core/gro.c:609) e1000_clean_rx_irq (drivers/net/ethernet/intel/e1000/e1000_main.c:4467) e1000_clean (drivers/net/ethernet/intel/e1000/e1000_main.c:3805) __napi_poll (net/core/dev.c:6533) net_rx_action (net/core/dev.c:6735) __do_softirq (kernel/softirq.c:554) run_ksoftirqd (kernel/softirq.c:913) smpboot_thread_fn (kernel/smpboot.c:164) kthread (kernel/kthread.c:388) ret_from_fork (arch/x86/kernel/process.c:153) ret_from_fork_asm (arch/x86/entry/entry_64.S:250) </TASK> The root causes are as follows: Thread A Thread B ... netif_receive_skb br_dev_stop ... br_multicast_leave_snoopers ... __ip_mc_dec_group ... __igmp_group_dropped igmp_rcv igmp_stop_timer igmp_heard_query //ref = 1 ip_ma_put igmp_mod_timer refcount_dec_and_test igmp_start_timer //ref = 0 ... refcount_inc //ref increases from 0 When the device receives an IGMPv2 Query message, it starts the timer immediately, regardless of whether the device is running. If the device is down and has left the multicast group, it will cause the mc list refcount uaf issue. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Reviewed-by: Hangbin Liu <liuhangbin(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Conflicts: net/ipv4/igmp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/ipv4/igmp.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index 834780a8baf6..c4b396968acd 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -221,8 +221,10 @@ static void igmp_start_timer(struct ip_mc_list *im, int max_delay) int tv = prandom_u32() % max_delay; im->tm_running = 1; - if (!mod_timer(&im->timer, jiffies+tv+2)) - refcount_inc(&im->refcnt); + if (refcount_inc_not_zero(&im->refcnt)) { + if (mod_timer(&im->timer, jiffies + tv + 2)) + ip_ma_put(im); + } } static void igmp_gq_start_timer(struct in_device *in_dev) -- 2.34.1

2 1

[PATCH OLK-5.10 v8 00/20] introduce smart_grid zone
by Yipeng Zou 28 Nov '23

28 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v7: 1. Update x86 openeuler_defconfig Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 22 files changed, 1606 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 21

[PATCH OLK-6.6 v2 0/1] Added SM3 as module signing algorithm
by GUO Zihua 28 Nov '23

28 Nov '23

This patch adds SM3 as one of the module signing algorithm. v2: Updated related issue. GUO Zihua (1): keys: Allow automatic module signature with SM3 kernel/module/Kconfig | 5 +++++ 1 file changed, 5 insertions(+) -- 2.34.1

2 2

[PATCH OLK-5.10 0/2] fix ksmbd to release the ones allocated for async work
by Zizhi Wo 28 Nov '23

28 Nov '23

fix typo and add release_async_work() to delete asynchronous work. Dawei Li (1): ksmbd: fix typo, syncronous->synchronous Namjae Jeon (1): ksmbd: delete asynchronous work from list fs/ksmbd/connection.c | 12 +++++------- fs/ksmbd/ksmbd_work.h | 2 +- fs/ksmbd/smb2pdu.c | 28 +++++++++++++++++++++------- fs/ksmbd/smb2pdu.h | 1 + 4 files changed, 28 insertions(+), 15 deletions(-) -- 2.39.2

2 3

[PATCH OLK-6.6 00/17] Support Qos scheduler
by Wenyu Huang 28 Nov '23

28 Nov '23

Hui Tang (1): sched: Fix null-ptr-deref in free_fair_sched_group Xia Fukun (1): sched/qos: Fix warning in CPU hotplug scenarios Zhang Qiao (15): sched: Introduce qos scheduler for co-location sched: Throttle qos cfs_rq when current cpu is running online task sched: Fix offline task can't be killed in a timely sched: Unthrottle the throttled cfs rq when offline rq sched: Unthrottle qos cfs rq when free a task group sched: bugfix setscheduler unlock cpuset_rwsem sched: Introduce handle priority reversion mechanism sched: enable CONFIG_QOS_SCHED on arm64 sched: Fix sleeping in atomic context at cpu_qos_write() sched/fair: Add qos_throttle_list node in struct cfs_rq sched: Throttle offline task at tracehook_notify_resume() sched/qos: Add qos_tg_{throttle,unthrottle}_{up,down} sched/fair: Update rq clock before unthrottle a qos cfs_rq sched/qos: Don't unthrottle cfs_rq when cfs_rq is throttled by qos sched: Enable qos scheduler config arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/cgroup.h | 4 + include/linux/resume_user_mode.h | 5 + include/linux/sched.h | 23 ++ include/linux/sched/sysctl.h | 8 + init/Kconfig | 12 + kernel/cgroup/cgroup.c | 22 ++ kernel/sched/core.c | 155 +++++++++++ kernel/sched/fair.c | 357 ++++++++++++++++++++++++- kernel/sched/sched.h | 12 + kernel/signal.c | 3 + kernel/sysctl.c | 24 ++ 13 files changed, 626 insertions(+), 1 deletion(-) -- 2.34.1

1 17

[PATCH OLK-5.10 v4 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 28 Nov '23

28 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 691 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 873 insertions(+), 171 deletions(-) -- 2.34.1

2 21

[PATCH OLK-5.10] ipmi_si: fix a memleak in try_smi_init()
by Yi Yang 28 Nov '23

28 Nov '23

stable inclusion from stable-v5.10.195 commit 5c5f02e16b919c8cb6024dc3778c8d8f1fb1f26b category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7GMMZ Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 6cf1a126de2992b4efe1c3c4d398f8de4aed6e3f upstream. Kmemleak reported the following leak info in try_smi_init(): unreferenced object 0xffff00018ecf9400 (size 1024): comm "modprobe", pid 2707763, jiffies 4300851415 (age 773.308s) backtrace: [<000000004ca5b312>] __kmalloc+0x4b8/0x7b0 [<00000000953b1072>] try_smi_init+0x148/0x5dc [ipmi_si] [<000000006460d325>] 0xffff800081b10148 [<0000000039206ea5>] do_one_initcall+0x64/0x2a4 [<00000000601399ce>] do_init_module+0x50/0x300 [<000000003c12ba3c>] load_module+0x7a8/0x9e0 [<00000000c246fffe>] __se_sys_init_module+0x104/0x180 [<00000000eea99093>] __arm64_sys_init_module+0x24/0x30 [<0000000021b1ef87>] el0_svc_common.constprop.0+0x94/0x250 [<0000000070f4f8b7>] do_el0_svc+0x48/0xe0 [<000000005a05337f>] el0_svc+0x24/0x3c [<000000005eb248d6>] el0_sync_handler+0x160/0x164 [<0000000030a59039>] el0_sync+0x160/0x180 The problem was that when an error occurred before handlers registration and after allocating `new_smi->si_sm`, the variable wouldn't be freed in the error handling afterwards since `shutdown_smi()` hadn't been registered yet. Fix it by adding a `kfree()` in the error handling path in `try_smi_init()`. Cc: stable(a)vger.kernel.org # 4.19+ Fixes: 7960f18a5647 ("ipmi_si: Convert over to a shutdown handler") Signed-off-by: Yi Yang <yiyang13(a)huawei.com> Co-developed-by: GONG, Ruiqi <gongruiqi(a)huaweicloud.com> Signed-off-by: GONG, Ruiqi <gongruiqi(a)huaweicloud.com> Message-Id: <20230629123328.2402075-1-gongruiqi(a)huaweicloud.com> Signed-off-by: Corey Minyard <minyard(a)acm.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yi Yang <yiyang13(a)huawei.com> --- drivers/char/ipmi/ipmi_si_intf.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 9e3e24c43168..227eb9e8e47d 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -2089,6 +2089,11 @@ static int try_smi_init(struct smi_info *new_smi) new_smi->io.io_cleanup = NULL; } + if (rv && new_smi->si_sm) { + kfree(new_smi->si_sm); + new_smi->si_sm = NULL; + } + return rv; } -- 2.25.1

2 1

[PATCH OLK-6.6 0/1] Added SM3 as module signing algorithm
by GUO Zihua 28 Nov '23

28 Nov '23

This patch adds SM3 as one of the module signing algorithm. GUO Zihua (1): keys: Allow automatic module signature with SM3 kernel/module/Kconfig | 5 +++++ 1 file changed, 5 insertions(+) -- 2.34.1

2 2

[PATCH openEuler-1.0-LTS v2] x86/mce/amd: Publish the bank pointer only after setup has succeeded
by liwei 28 Nov '23

28 Nov '23

From: Borislav Petkov <bp(a)suse.de> mainline inclusion from mainline-v5.6-rc3 commit 6e5cf31fbe651bed7ba1df768f2e123531132417 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JXM2 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- threshold_create_bank() creates a bank descriptor per MCA error thresholding counter which can be controlled over sysfs. It publishes the pointer to that bank in a per-CPU variable and then goes on to create additional thresholding blocks if the bank has such. However, that creation of additional blocks in allocate_threshold_blocks() can fail, leading to a use-after-free through the per-CPU pointer. Therefore, publish that pointer only after all blocks have been setup successfully. Fixes: 019f34fccfd5 ("x86, MCE, AMD: Move shared bank to node descriptor") Reported-by: Saar Amar <Saar.Amar(a)microsoft.com> Reported-by: Dan Carpenter <dan.carpenter(a)oracle.com> Signed-off-by: Borislav Petkov <bp(a)suse.de> Cc: <stable(a)vger.kernel.org> Link: http://lkml.kernel.org/r/20200128140846.phctkvx5btiexvbx@kili.mountain Signed-off-by: liwei <liwei728(a)huawei.com> --- arch/x86/kernel/cpu/mce/amd.c | 33 ++++++++++++++++----------------- 1 file changed, 16 insertions(+), 17 deletions(-) diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c index 38b18fa055ef..af4a83695970 100644 --- a/arch/x86/kernel/cpu/mce/amd.c +++ b/arch/x86/kernel/cpu/mce/amd.c @@ -1200,8 +1200,9 @@ static const char *get_name(unsigned int bank, struct threshold_block *b) return buf_mcatype; } -static int allocate_threshold_blocks(unsigned int cpu, unsigned int bank, - unsigned int block, u32 address) +static int allocate_threshold_blocks(unsigned int cpu, struct threshold_bank *tb, + unsigned int bank, unsigned int block, + u32 address) { struct threshold_block *b = NULL; u32 low, high; @@ -1245,16 +1246,12 @@ static int allocate_threshold_blocks(unsigned int cpu, unsigned int bank, INIT_LIST_HEAD(&b->miscj); - if (per_cpu(threshold_banks, cpu)[bank]->blocks) { - list_add(&b->miscj, - &per_cpu(threshold_banks, cpu)[bank]->blocks->miscj); - } else { - per_cpu(threshold_banks, cpu)[bank]->blocks = b; - } + if (tb->blocks) + list_add(&b->miscj, &tb->blocks->miscj); + else + tb->blocks = b; - err = kobject_init_and_add(&b->kobj, &threshold_ktype, - per_cpu(threshold_banks, cpu)[bank]->kobj, - get_name(bank, b)); + err = kobject_init_and_add(&b->kobj, &threshold_ktype, tb->kobj, get_name(bank, b)); if (err) goto out_free; recurse: @@ -1262,7 +1259,7 @@ static int allocate_threshold_blocks(unsigned int cpu, unsigned int bank, if (!address) return 0; - err = allocate_threshold_blocks(cpu, bank, block, address); + err = allocate_threshold_blocks(cpu, tb, bank, block, address); if (err) goto out_free; @@ -1347,8 +1344,6 @@ static int threshold_create_bank(unsigned int cpu, unsigned int bank) goto out_free; } - per_cpu(threshold_banks, cpu)[bank] = b; - if (is_shared_bank(bank)) { refcount_set(&b->cpus, 1); @@ -1359,9 +1354,13 @@ static int threshold_create_bank(unsigned int cpu, unsigned int bank) } } - err = allocate_threshold_blocks(cpu, bank, 0, msr_ops.misc(bank)); - if (!err) - goto out; + err = allocate_threshold_blocks(cpu, b, bank, 0, msr_ops.misc(bank)); + if (err) + goto out_free; + + per_cpu(threshold_banks, cpu)[bank] = b; + + return 0; out_free: kfree(b); -- 2.25.1

2 3

[PATCH OLK-5.10 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 28 Nov '23

28 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

2 21

[PATCH OLK-6.6 V2 00/12] Add SDEI Watchdog Support
by Xiongfeng Wang 28 Nov '23

28 Nov '23

Xiongfeng Wang (12): firmware: arm_sdei: add interrupt binding api firmware: arm_sdei: make 'sdei_api_event_disable/enable' public lockup_detector: init lockup detector after all the init_calls watchdog: add nmi_watchdog support for arm64 based on SDEI sdei_watchdog: clear EOI of the secure timer before kdump sdei_watchdog: set secure timer period base on 'watchdog_thresh' sdei_watchdog: avoid possible false hardlockup init: only move down lockup_detector_init() when sdei_watchdog is enabled kprobes/arm64: Blacklist sdei watchdog callback functions openeuler_defconfig: Enable SDEI Watchdog stop_machine: mask sdei before running the callback arm64: kexec: only clear EOI for SDEI in NMI context arch/arm64/configs/openeuler_defconfig | 3 +- arch/arm64/kernel/Makefile | 1 + arch/arm64/kernel/machine_kexec.c | 11 ++ arch/arm64/kernel/watchdog_sdei.c | 151 +++++++++++++++++++++++++ drivers/firmware/arm_sdei.c | 26 ++++- include/linux/arm_sdei.h | 5 + include/linux/nmi.h | 8 ++ include/uapi/linux/arm_sdei.h | 2 + init/main.c | 7 +- kernel/stop_machine.c | 10 ++ kernel/watchdog.c | 3 + lib/Kconfig.debug | 9 ++ 12 files changed, 232 insertions(+), 4 deletions(-) create mode 100644 arch/arm64/kernel/watchdog_sdei.c -- 2.20.1

2 13

[PATCH OLK-5.10 v3 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 28 Nov '23

28 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 691 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 873 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH OLK-5.10 v3 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 27 Nov '23

27 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 691 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 873 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH OLK-5.10 v2 0/9] rcu: Add RCU stall diagnosis information
by Zhen Lei 27 Nov '23

27 Nov '23

v1 --> v2: Remove patch 1 in v1: rcu: Prevent lockdep-RCU splats on lock acquisition/release It had arleady been merged. v1: In some extreme cases, such as the I/O pressure test, the CPU usage may be 100%, causing RCU stall. In this case, the printed information about current is not useful. Displays the number and usage of hard interrupts, soft interrupts, and context switches that are generated within half of the CPU stall timeout, can help us make a general judgment. In other cases, we can preliminarily determine whether an infinite loop occurs when local_irq, local_bh or preempt is disabled. Neeraj Upadhyay (1): rcu: Check and report missed fqs timer wakeup on RCU stall Paul E. McKenney (2): rcu: For RCU grace-period kthread starvation, dump last CPU it ran on rcu: Do not NMI offline CPUs Zhen Lei (6): sched/debug: Try trigger_single_cpu_backtrace(cpu) in dump_cpu_task() sched/debug: Show the registers of 'current' in dump_cpu_task() sched: Add helper kstat_cpu_softirqs_sum() sched: Add helper nr_context_switches_cpu() rcu: Add RCU stall diagnosis information rcu: Align the output of RCU CPU stall warning messages Documentation/RCU/stallwarn.rst | 23 ++++- .../admin-guide/kernel-parameters.txt | 6 ++ include/linux/kernel_stat.h | 12 +++ kernel/rcu/Kconfig.debug | 13 +++ kernel/rcu/rcu.h | 1 + kernel/rcu/tree.c | 43 +++++++--- kernel/rcu/tree.h | 19 ++++ kernel/rcu/tree_stall.h | 86 +++++++++++++++++-- kernel/rcu/update.c | 2 + kernel/sched/core.c | 19 ++++ kernel/smp.c | 3 +- 11 files changed, 209 insertions(+), 18 deletions(-) -- 2.25.1

2 10

[PATCH OLK-5.10 v2 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 27 Nov '23

27 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 691 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 873 insertions(+), 171 deletions(-) -- 2.34.1

1 0

[PATCH OLK-5.10 v2 0/9] rcu: Add RCU stall diagnosis information
by Zhen Lei 27 Nov '23

27 Nov '23

v1 --> v2: Remove patch 1 in v1: rcu: Prevent lockdep-RCU splats on lock acquisition/release It had arleady been merged. v1: In some extreme cases, such as the I/O pressure test, the CPU usage may be 100%, causing RCU stall. In this case, the printed information about current is not useful. Displays the number and usage of hard interrupts, soft interrupts, and context switches that are generated within half of the CPU stall timeout, can help us make a general judgment. In other cases, we can preliminarily determine whether an infinite loop occurs when local_irq, local_bh or preempt is disabled. Neeraj Upadhyay (1): rcu: Check and report missed fqs timer wakeup on RCU stall Paul E. McKenney (2): rcu: For RCU grace-period kthread starvation, dump last CPU it ran on rcu: Do not NMI offline CPUs Zhen Lei (6): sched/debug: Try trigger_single_cpu_backtrace(cpu) in dump_cpu_task() sched/debug: Show the registers of 'current' in dump_cpu_task() sched: Add helper kstat_cpu_softirqs_sum() sched: Add helper nr_context_switches_cpu() rcu: Add RCU stall diagnosis information rcu: Align the output of RCU CPU stall warning messages Documentation/RCU/stallwarn.rst | 23 ++++- .../admin-guide/kernel-parameters.txt | 6 ++ include/linux/kernel_stat.h | 12 +++ kernel/rcu/Kconfig.debug | 13 +++ kernel/rcu/rcu.h | 1 + kernel/rcu/tree.c | 43 +++++++--- kernel/rcu/tree.h | 19 ++++ kernel/rcu/tree_stall.h | 86 +++++++++++++++++-- kernel/rcu/update.c | 2 + kernel/sched/core.c | 19 ++++ kernel/smp.c | 3 +- 11 files changed, 209 insertions(+), 18 deletions(-) -- 2.25.1

1 9

[PATCH OLK-6.6 00/31] Backport ARM64-ILP32 patches
by Jinjie Ruan 27 Nov '23

27 Nov '23

Backport ILP32 for ARM64 patches to OLK6.6 form openEuler-23.09. Andrew Pinski (3): arm64: rename COMPAT to AARCH32_EL0 arm64: uapi: set __BITS_PER_LONG correctly for ILP32 and LP64 arm64:ilp32: add ARM64_ILP32 to Kconfig Chen Jiahao (4): arm64: fix AUDIT_ARCH_AARCH64ILP32 bug on audit subsystem arm64: fix address limit problem with TASK_SIZE_MAX arm64: set 32-bit compatible TASK_SIZE_MAX to fix U32 libc_write_01 error arm64: fix image size inflation with CONFIG_COMPAT_TASK_SIZE Dave Martin (1): arm64: signal: Make parse_user_sigframe() independent of rt_sigframe layout James Morse (1): ptrace: Add compat PTRACE_{G,S}ETSIGMASK handlers Philipp Tomsich (1): arm64:ilp32: add vdso-ilp32 and use for signal return Xiongfeng Wang (6): arm64: rename functions that reference compat term arm64: secomp: fix the secure computing mode 1 syscall check for ilp32 ilp32: avoid clearing upper 32 bits of syscall return value for ilp32 ilp32: skip ARM erratum 1418040 for ilp32 application arm64: fix abi change caused by ILP32 ilp32: fix compile problem when ARM64_ILP32 and UBSAN are both enabled Yury Norov (14): thread: move thread bits accessors to separated file arm64: ilp32: add documentation on the ILP32 ABI for ARM64 arm64: introduce is_a32_compat_{task,thread} for AArch32 compat arm64: ilp32: add is_ilp32_compat_{task,thread} and TIF_32BIT_AARCH64 arm64: introduce AUDIT_ARCH_AARCH64ILP32 for ilp32 arm64: introduce binfmt_elf32.c arm64: change compat_elf_hwcap and compat_elf_hwcap2 prefix to a32 arm64: ilp32: introduce binfmt_ilp32.c arm64: ilp32: share aarch32 syscall handlers arm64: ilp32: introduce syscall table for ILP32 arm64: signal: share lp64 signal structures and routines to ilp32 arm64: signal32: move ilp32 and aarch32 common code to separated file arm64: ilp32: introduce ilp32-specific sigframe and ucontext arm64: ptrace: handle ptrace_request differently for aarch32 and ilp32 Zhen Lei (1): arm64: replace is_compat_task() with is_ilp32_compat_task() in TASK_SIZE_MAX Documentation/arm64/ilp32.txt | 52 +++ arch/arm64/Kconfig | 26 +- arch/arm64/Makefile | 3 + arch/arm64/include/asm/arch_timer.h | 4 +- arch/arm64/include/asm/compat.h | 18 +- arch/arm64/include/asm/elf.h | 32 +- arch/arm64/include/asm/fpsimd.h | 2 +- arch/arm64/include/asm/ftrace.h | 2 +- arch/arm64/include/asm/hwcap.h | 8 +- arch/arm64/include/asm/is_compat.h | 83 ++++ arch/arm64/include/asm/processor.h | 24 +- arch/arm64/include/asm/ptrace.h | 16 +- arch/arm64/include/asm/seccomp.h | 32 +- arch/arm64/include/asm/signal32.h | 31 +- arch/arm64/include/asm/signal32_common.h | 13 + arch/arm64/include/asm/signal_common.h | 391 +++++++++++++++++ arch/arm64/include/asm/signal_ilp32.h | 23 + arch/arm64/include/asm/syscall.h | 14 +- arch/arm64/include/asm/thread_info.h | 4 +- arch/arm64/include/asm/unistd.h | 5 + arch/arm64/include/asm/vdso.h | 9 + arch/arm64/include/uapi/asm/bitsperlong.h | 9 +- arch/arm64/include/uapi/asm/unistd.h | 15 +- arch/arm64/kernel/Makefile | 8 +- arch/arm64/kernel/armv8_deprecated.c | 14 +- arch/arm64/kernel/asm-offsets.c | 13 +- arch/arm64/kernel/binfmt_elf32.c | 27 ++ arch/arm64/kernel/binfmt_ilp32.c | 89 ++++ arch/arm64/kernel/compat_alignment.c | 2 +- arch/arm64/kernel/cpufeature.c | 36 +- arch/arm64/kernel/cpuinfo.c | 18 +- arch/arm64/kernel/debug-monitors.c | 4 +- arch/arm64/kernel/entry-common.c | 6 +- arch/arm64/kernel/hw_breakpoint.c | 8 +- arch/arm64/kernel/perf_callchain.c | 28 +- arch/arm64/kernel/perf_regs.c | 4 +- arch/arm64/kernel/process.c | 15 +- arch/arm64/kernel/proton-pack.c | 2 +- arch/arm64/kernel/ptrace.c | 38 +- arch/arm64/kernel/signal.c | 397 +++--------------- arch/arm64/kernel/signal32.c | 97 ++--- arch/arm64/kernel/signal32_common.c | 37 ++ arch/arm64/kernel/signal_ilp32.c | 67 +++ arch/arm64/kernel/sys32.c | 104 +---- arch/arm64/kernel/sys32_common.c | 106 +++++ arch/arm64/kernel/sys_compat.c | 12 +- arch/arm64/kernel/sys_ilp32.c | 82 ++++ arch/arm64/kernel/syscall.c | 37 +- arch/arm64/kernel/traps.c | 5 +- arch/arm64/kernel/vdso-ilp32/.gitignore | 2 + arch/arm64/kernel/vdso-ilp32/Makefile | 111 +++++ arch/arm64/kernel/vdso-ilp32/vdso-ilp32.S | 22 + arch/arm64/kernel/vdso-ilp32/vdso-ilp32.lds.S | 88 ++++ arch/arm64/kernel/vdso.c | 46 +- arch/arm64/mm/fault.c | 2 +- include/linux/sched.h | 1 + include/linux/thread_bits.h | 87 ++++ include/linux/thread_info.h | 73 +--- kernel/ptrace.c | 52 ++- 59 files changed, 1790 insertions(+), 766 deletions(-) create mode 100644 Documentation/arm64/ilp32.txt create mode 100644 arch/arm64/include/asm/is_compat.h create mode 100644 arch/arm64/include/asm/signal32_common.h create mode 100644 arch/arm64/include/asm/signal_common.h create mode 100644 arch/arm64/include/asm/signal_ilp32.h create mode 100644 arch/arm64/kernel/binfmt_elf32.c create mode 100644 arch/arm64/kernel/binfmt_ilp32.c create mode 100644 arch/arm64/kernel/signal32_common.c create mode 100644 arch/arm64/kernel/signal_ilp32.c create mode 100644 arch/arm64/kernel/sys32_common.c create mode 100644 arch/arm64/kernel/sys_ilp32.c create mode 100644 arch/arm64/kernel/vdso-ilp32/.gitignore create mode 100644 arch/arm64/kernel/vdso-ilp32/Makefile create mode 100644 arch/arm64/kernel/vdso-ilp32/vdso-ilp32.S create mode 100644 arch/arm64/kernel/vdso-ilp32/vdso-ilp32.lds.S create mode 100644 include/linux/thread_bits.h -- 2.34.1

2 32

[PATCH OLK-6.6 00/16] Support Qos scheduler
by Wenyu Huang 27 Nov '23

27 Nov '23

*** BLURB HERE *** Xia Fukun (1): sched/qos: Fix warning in CPU hotplug scenarios Zhang Qiao (15): sched: Introduce qos scheduler for co-location sched: Throttle qos cfs_rq when current cpu is running online task sched: Fix offline task can't be killed in a timely sched: Unthrottle the throttled cfs rq when offline rq sched: Unthrottle qos cfs rq when free a task group sched: bugfix setscheduler unlock cpuset_rwsem sched: Introduce handle priority reversion mechanism sched: enable CONFIG_QOS_SCHED on arm64 sched: Fix sleeping in atomic context at cpu_qos_write() sched/fair: Add qos_throttle_list node in struct cfs_rq sched: Throttle offline task at tracehook_notify_resume() sched/qos: Add qos_tg_{throttle,unthrottle}_{up,down} sched/fair: Update rq clock before unthrottle a qos cfs_rq sched/qos: Don't unthrottle cfs_rq when cfs_rq is throttled by qos sched: Enable qos scheduler config arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/cgroup.h | 4 + include/linux/resume_user_mode.h | 5 + include/linux/sched.h | 23 ++ include/linux/sched/sysctl.h | 8 + init/Kconfig | 12 + kernel/cgroup/cgroup.c | 22 ++ kernel/sched/core.c | 155 +++++++++++ kernel/sched/fair.c | 357 ++++++++++++++++++++++++- kernel/sched/sched.h | 12 + kernel/signal.c | 3 + kernel/sysctl.c | 24 ++ 13 files changed, 626 insertions(+), 1 deletion(-) -- 2.34.1

2 17

[PATCH OLK-6.6 00/16] Support Qos scheduler
by Wenyu Huang 27 Nov '23

27 Nov '23

Xia Fukun (1): sched/qos: Fix warning in CPU hotplug scenarios Zhang Qiao (15): sched: Introduce qos scheduler for co-location sched: Throttle qos cfs_rq when current cpu is running online task sched: Fix offline task can't be killed in a timely sched: Unthrottle the throttled cfs rq when offline rq sched: Unthrottle qos cfs rq when free a task group sched: bugfix setscheduler unlock cpuset_rwsem sched: Introduce handle priority reversion mechanism sched: enable CONFIG_QOS_SCHED on arm64 sched: Fix sleeping in atomic context at cpu_qos_write() sched/fair: Add qos_throttle_list node in struct cfs_rq sched: Throttle offline task at tracehook_notify_resume() sched/qos: Add qos_tg_{throttle,unthrottle}_{up,down} sched/fair: Update rq clock before unthrottle a qos cfs_rq sched/qos: Don't unthrottle cfs_rq when cfs_rq is throttled by qos sched: Enable qos scheduler config arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/cgroup.h | 4 + include/linux/resume_user_mode.h | 5 + include/linux/sched.h | 23 ++ include/linux/sched/sysctl.h | 8 + init/Kconfig | 12 + kernel/cgroup/cgroup.c | 22 ++ kernel/sched/core.c | 155 +++++++++++ kernel/sched/fair.c | 357 ++++++++++++++++++++++++- kernel/sched/sched.h | 12 + kernel/signal.c | 3 + kernel/sysctl.c | 24 ++ 13 files changed, 626 insertions(+), 1 deletion(-) -- 2.34.1

1 16

[PATCH OLK-6.6 00/31] Backport ARM64-ILP32 patches
by Jinjie Ruan 27 Nov '23

27 Nov '23

Backport ILP32 for ARM64 patches to OLK6.6 form openEuler-23.09. Andrew Pinski (3): arm64: rename COMPAT to AARCH32_EL0 arm64: uapi: set __BITS_PER_LONG correctly for ILP32 and LP64 arm64:ilp32: add ARM64_ILP32 to Kconfig Chen Jiahao (4): arm64: fix AUDIT_ARCH_AARCH64ILP32 bug on audit subsystem arm64: fix address limit problem with TASK_SIZE_MAX arm64: set 32-bit compatible TASK_SIZE_MAX to fix U32 libc_write_01 error arm64: fix image size inflation with CONFIG_COMPAT_TASK_SIZE Dave Martin (1): arm64: signal: Make parse_user_sigframe() independent of rt_sigframe layout James Morse (1): ptrace: Add compat PTRACE_{G,S}ETSIGMASK handlers Philipp Tomsich (1): arm64:ilp32: add vdso-ilp32 and use for signal return Xiongfeng Wang (6): arm64: rename functions that reference compat term arm64: secomp: fix the secure computing mode 1 syscall check for ilp32 ilp32: avoid clearing upper 32 bits of syscall return value for ilp32 ilp32: skip ARM erratum 1418040 for ilp32 application arm64: fix abi change caused by ILP32 ilp32: fix compile problem when ARM64_ILP32 and UBSAN are both enabled Yury Norov (14): thread: move thread bits accessors to separated file arm64: ilp32: add documentation on the ILP32 ABI for ARM64 arm64: introduce is_a32_compat_{task,thread} for AArch32 compat arm64: ilp32: add is_ilp32_compat_{task,thread} and TIF_32BIT_AARCH64 arm64: introduce AUDIT_ARCH_AARCH64ILP32 for ilp32 arm64: introduce binfmt_elf32.c arm64: change compat_elf_hwcap and compat_elf_hwcap2 prefix to a32 arm64: ilp32: introduce binfmt_ilp32.c arm64: ilp32: share aarch32 syscall handlers arm64: ilp32: introduce syscall table for ILP32 arm64: signal: share lp64 signal structures and routines to ilp32 arm64: signal32: move ilp32 and aarch32 common code to separated file arm64: ilp32: introduce ilp32-specific sigframe and ucontext arm64: ptrace: handle ptrace_request differently for aarch32 and ilp32 Zhen Lei (1): arm64: replace is_compat_task() with is_ilp32_compat_task() in TASK_SIZE_MAX Documentation/arm64/ilp32.txt | 52 +++ arch/arm64/Kconfig | 26 +- arch/arm64/Makefile | 3 + arch/arm64/include/asm/arch_timer.h | 4 +- arch/arm64/include/asm/compat.h | 18 +- arch/arm64/include/asm/elf.h | 32 +- arch/arm64/include/asm/fpsimd.h | 2 +- arch/arm64/include/asm/ftrace.h | 2 +- arch/arm64/include/asm/hwcap.h | 8 +- arch/arm64/include/asm/is_compat.h | 83 ++++ arch/arm64/include/asm/processor.h | 24 +- arch/arm64/include/asm/ptrace.h | 16 +- arch/arm64/include/asm/seccomp.h | 32 +- arch/arm64/include/asm/signal32.h | 31 +- arch/arm64/include/asm/signal32_common.h | 13 + arch/arm64/include/asm/signal_common.h | 391 +++++++++++++++++ arch/arm64/include/asm/signal_ilp32.h | 23 + arch/arm64/include/asm/syscall.h | 14 +- arch/arm64/include/asm/thread_info.h | 4 +- arch/arm64/include/asm/unistd.h | 5 + arch/arm64/include/asm/vdso.h | 9 + arch/arm64/include/uapi/asm/bitsperlong.h | 9 +- arch/arm64/include/uapi/asm/unistd.h | 15 +- arch/arm64/kernel/Makefile | 8 +- arch/arm64/kernel/armv8_deprecated.c | 14 +- arch/arm64/kernel/asm-offsets.c | 13 +- arch/arm64/kernel/binfmt_elf32.c | 27 ++ arch/arm64/kernel/binfmt_ilp32.c | 89 ++++ arch/arm64/kernel/compat_alignment.c | 2 +- arch/arm64/kernel/cpufeature.c | 36 +- arch/arm64/kernel/cpuinfo.c | 18 +- arch/arm64/kernel/debug-monitors.c | 4 +- arch/arm64/kernel/entry-common.c | 6 +- arch/arm64/kernel/hw_breakpoint.c | 8 +- arch/arm64/kernel/perf_callchain.c | 28 +- arch/arm64/kernel/perf_regs.c | 4 +- arch/arm64/kernel/process.c | 15 +- arch/arm64/kernel/proton-pack.c | 2 +- arch/arm64/kernel/ptrace.c | 38 +- arch/arm64/kernel/signal.c | 397 +++--------------- arch/arm64/kernel/signal32.c | 97 ++--- arch/arm64/kernel/signal32_common.c | 37 ++ arch/arm64/kernel/signal_ilp32.c | 67 +++ arch/arm64/kernel/sys32.c | 104 +---- arch/arm64/kernel/sys32_common.c | 106 +++++ arch/arm64/kernel/sys_compat.c | 12 +- arch/arm64/kernel/sys_ilp32.c | 82 ++++ arch/arm64/kernel/syscall.c | 37 +- arch/arm64/kernel/traps.c | 5 +- arch/arm64/kernel/vdso-ilp32/.gitignore | 2 + arch/arm64/kernel/vdso-ilp32/Makefile | 111 +++++ arch/arm64/kernel/vdso-ilp32/vdso-ilp32.S | 22 + arch/arm64/kernel/vdso-ilp32/vdso-ilp32.lds.S | 88 ++++ arch/arm64/kernel/vdso.c | 46 +- arch/arm64/mm/fault.c | 2 +- include/linux/sched.h | 1 + include/linux/thread_bits.h | 87 ++++ include/linux/thread_info.h | 73 +--- kernel/ptrace.c | 52 ++- 59 files changed, 1790 insertions(+), 766 deletions(-) create mode 100644 Documentation/arm64/ilp32.txt create mode 100644 arch/arm64/include/asm/is_compat.h create mode 100644 arch/arm64/include/asm/signal32_common.h create mode 100644 arch/arm64/include/asm/signal_common.h create mode 100644 arch/arm64/include/asm/signal_ilp32.h create mode 100644 arch/arm64/kernel/binfmt_elf32.c create mode 100644 arch/arm64/kernel/binfmt_ilp32.c create mode 100644 arch/arm64/kernel/signal32_common.c create mode 100644 arch/arm64/kernel/signal_ilp32.c create mode 100644 arch/arm64/kernel/sys32_common.c create mode 100644 arch/arm64/kernel/sys_ilp32.c create mode 100644 arch/arm64/kernel/vdso-ilp32/.gitignore create mode 100644 arch/arm64/kernel/vdso-ilp32/Makefile create mode 100644 arch/arm64/kernel/vdso-ilp32/vdso-ilp32.S create mode 100644 arch/arm64/kernel/vdso-ilp32/vdso-ilp32.lds.S create mode 100644 include/linux/thread_bits.h -- 2.34.1

1 6

[PATCH OLK-6.6 00/10] Add SDEI Watchdog Support
by Xiongfeng Wang 27 Nov '23

27 Nov '23

Xiongfeng Wang (10): firmware: arm_sdei: add interrupt binding api firmware: arm_sdei: make 'sdei_api_event_disable/enable' public lockup_detector: init lockup detector after all the init_calls watchdog: add nmi_watchdog support for arm64 based on SDEI sdei_watchdog: clear EOI of the secure timer before kdump sdei_watchdog: set secure timer period base on 'watchdog_thresh' sdei_watchdog: avoid possible false hardlockup init: only move down lockup_detector_init() when sdei_watchdog is enabled kprobes/arm64: Blacklist sdei watchdog callback functions openeuler_defconfig: Enable SDEI Watchdog arch/arm64/configs/openeuler_defconfig | 3 +- arch/arm64/kernel/Makefile | 1 + arch/arm64/kernel/machine_kexec.c | 10 ++ arch/arm64/kernel/watchdog_sdei.c | 151 +++++++++++++++++++++++++ drivers/firmware/arm_sdei.c | 26 ++++- include/linux/arm_sdei.h | 5 + include/linux/nmi.h | 8 ++ include/uapi/linux/arm_sdei.h | 2 + init/main.c | 7 +- kernel/watchdog.c | 3 + lib/Kconfig.debug | 9 ++ 11 files changed, 221 insertions(+), 4 deletions(-) create mode 100644 arch/arm64/kernel/watchdog_sdei.c -- 2.20.1

1 10

[PATCH OLK-5.5 00/10] Add SDEI Watchdog Support
by Xiongfeng Wang 27 Nov '23

27 Nov '23

Xiongfeng Wang (10): firmware: arm_sdei: add interrupt binding api firmware: arm_sdei: make 'sdei_api_event_disable/enable' public lockup_detector: init lockup detector after all the init_calls watchdog: add nmi_watchdog support for arm64 based on SDEI sdei_watchdog: clear EOI of the secure timer before kdump sdei_watchdog: set secure timer period base on 'watchdog_thresh' sdei_watchdog: avoid possible false hardlockup init: only move down lockup_detector_init() when sdei_watchdog is enabled kprobes/arm64: Blacklist sdei watchdog callback functions openeuler_defconfig: Enable SDEI Watchdog arch/arm64/configs/openeuler_defconfig | 3 +- arch/arm64/kernel/Makefile | 1 + arch/arm64/kernel/machine_kexec.c | 10 ++ arch/arm64/kernel/watchdog_sdei.c | 151 +++++++++++++++++++++++++ drivers/firmware/arm_sdei.c | 26 ++++- include/linux/arm_sdei.h | 5 + include/linux/nmi.h | 8 ++ include/uapi/linux/arm_sdei.h | 2 + init/main.c | 7 +- kernel/watchdog.c | 3 + lib/Kconfig.debug | 9 ++ 11 files changed, 221 insertions(+), 4 deletions(-) create mode 100644 arch/arm64/kernel/watchdog_sdei.c -- 2.20.1

1 10

[PATCH OLK-5.10] ipmi_si: fix a memleak in try_smi_init()
by Yi Yang 27 Nov '23

27 Nov '23

stable inclusion from stable-v5.10.195 commit 5c5f02e16b919c8cb6024dc3778c8d8f1fb1f26b category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7GMMZ Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 6cf1a126de2992b4efe1c3c4d398f8de4aed6e3f upstream. Kmemleak reported the following leak info in try_smi_init(): unreferenced object 0xffff00018ecf9400 (size 1024): comm "modprobe", pid 2707763, jiffies 4300851415 (age 773.308s) backtrace: [<000000004ca5b312>] __kmalloc+0x4b8/0x7b0 [<00000000953b1072>] try_smi_init+0x148/0x5dc [ipmi_si] [<000000006460d325>] 0xffff800081b10148 [<0000000039206ea5>] do_one_initcall+0x64/0x2a4 [<00000000601399ce>] do_init_module+0x50/0x300 [<000000003c12ba3c>] load_module+0x7a8/0x9e0 [<00000000c246fffe>] __se_sys_init_module+0x104/0x180 [<00000000eea99093>] __arm64_sys_init_module+0x24/0x30 [<0000000021b1ef87>] el0_svc_common.constprop.0+0x94/0x250 [<0000000070f4f8b7>] do_el0_svc+0x48/0xe0 [<000000005a05337f>] el0_svc+0x24/0x3c [<000000005eb248d6>] el0_sync_handler+0x160/0x164 [<0000000030a59039>] el0_sync+0x160/0x180 The problem was that when an error occurred before handlers registration and after allocating `new_smi->si_sm`, the variable wouldn't be freed in the error handling afterwards since `shutdown_smi()` hadn't been registered yet. Fix it by adding a `kfree()` in the error handling path in `try_smi_init()`. Cc: stable(a)vger.kernel.org # 4.19+ Fixes: 7960f18a5647 ("ipmi_si: Convert over to a shutdown handler") Signed-off-by: Yi Yang <yiyang13(a)huawei.com> Co-developed-by: GONG, Ruiqi <gongruiqi(a)huaweicloud.com> Signed-off-by: GONG, Ruiqi <gongruiqi(a)huaweicloud.com> Message-Id: <20230629123328.2402075-1-gongruiqi(a)huaweicloud.com> Signed-off-by: Corey Minyard <minyard(a)acm.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yi Yang <yiyang13(a)huawei.com> --- drivers/char/ipmi/ipmi_si_intf.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c index 9e3e24c43168..227eb9e8e47d 100644 --- a/drivers/char/ipmi/ipmi_si_intf.c +++ b/drivers/char/ipmi/ipmi_si_intf.c @@ -2089,6 +2089,11 @@ static int try_smi_init(struct smi_info *new_smi) new_smi->io.io_cleanup = NULL; } + if (rv && new_smi->si_sm) { + kfree(new_smi->si_sm); + new_smi->si_sm = NULL; + } + return rv; } -- 2.25.1

1 0

[PATCH OLK-5.10] fs/dirty_pages: fix some errors in seq_read_dirty()
by Zizhi Wo 27 Nov '23

27 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JUUC?from=project-issue -------------------------------- In seq_read_dirty(), the parameter passed into simple_read_from_buffer() will cause problem on a 32-bit system; The simple_read_from_buffer() function has a lot of redundancy checks inside. Replaced it with copy_to_user() and add the return value check. And the update of the "copied" character count in seq_read_dirty() is also incorrect, fixed it. Fixes: aeb96447becc ("fs/dirty_pages: dump the number of dirty pages for each inode") Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/dirty_pages.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/dirty_pages.c b/fs/dirty_pages.c index 21e2251b6619..0eb818b599ba 100644 --- a/fs/dirty_pages.c +++ b/fs/dirty_pages.c @@ -217,12 +217,12 @@ static ssize_t seq_read_dirty( } n = min(m->count - m->from, size); - err = simple_read_from_buffer(buf, n, - (loff_t *) &m->from, m->buf, m->count); - if (err < 0) { + n -= copy_to_user(buf, m->buf + m->from, n); + if (unlikely(!n)) { err = -EFAULT; goto done; } + m->from += n; copied += n; done: if (!copied) -- 2.39.2

2 1

[PATCH openEuler-1.0-LTS 0/2] fix CFS bandwidth vs. hrtimer self deadlock
by Yu Liao 27 Nov '23

27 Nov '23

Thomas Gleixner (1): hrtimers: Push pending hrtimers away from outgoing CPU earlier Yu Liao (1): cpu/hotplug: fix kabi breakage in enum cpuhp_state include/linux/hrtimer.h | 4 ++-- include/linux/smp.h | 1 + kernel/cpu.c | 17 +++++++++++++++-- kernel/smp.c | 8 ++++++++ kernel/time/hrtimer.c | 33 ++++++++++++--------------------- 5 files changed, 38 insertions(+), 25 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10] fs/dirty_pages: fix inode reference count leakage error in dump_dirtypages_sb()
by Zizhi Wo 27 Nov '23

27 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JUUC?from=project-issue -------------------------------- Inode reference count leakage exists in dump_dirtypages_sb() function. Executing iput(inode) in advance to avoid this problem. Fixes: aeb96447becc ("fs/dirty_pages: dump the number of dirty pages for each inode") Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/dirty_pages.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/dirty_pages.c b/fs/dirty_pages.c index 0eb818b599ba..b0cf442c5730 100644 --- a/fs/dirty_pages.c +++ b/fs/dirty_pages.c @@ -159,6 +159,7 @@ static void dump_dirtypages_sb(struct super_block *sb, struct seq_file *m) if (m->size <= m->count) { seq_set_overflow(m); strncpy(m->buf+m->count-12, "terminated\n\0", 12); + iput(inode); goto done; } seq_printf(m, "FSType: %s, Dev ID: %u(%u:%u) ino %lu, dirty pages %lu, path %s\n", -- 2.39.2

2 1

[PATCH OLK-6.6 00/33] folio conversions for numa balance
by Peng Zhang 27 Nov '23

27 Nov '23

From: ZhangPeng <zhangpeng362(a)huawei.com> Includes 3 patch series "mm: migrate: more folio conversion and unification", "mm: convert numa balancing functions to use a folio" and "mm: convert page cpupid functions to folios" to implement folio conversions of numa balance. Kefeng Wang (33): mm: migrate: remove PageTransHuge check in numamigrate_isolate_page() mm: migrate: remove THP mapcount check in numamigrate_isolate_page() mm: migrate: convert numamigrate_isolate_page() to numamigrate_isolate_folio() mm: migrate: convert migrate_misplaced_page() to migrate_misplaced_folio() mm: migrate: use __folio_test_movable() mm: migrate: use a folio in add_page_for_migration() mm: migrate: remove PageHead() check for HugeTLB in add_page_for_migration() mm: migrate: remove isolated variable in add_page_for_migration() mm: memory: add vm_normal_folio_pmd() mm: huge_memory: use a folio in do_huge_pmd_numa_page() mm: memory: use a folio in do_numa_page() mm: memory: make numa_migrate_prep() to take a folio mm: mempolicy: make mpol_misplaced() to take a folio sched/numa, mm: make numa migrate functions to take a folio mm_types: add virtual and _last_cpupid into struct folio mm: add folio_last_cpupid() mm: memory: use folio_last_cpupid() in do_numa_page() mm: huge_memory: use folio_last_cpupid() in do_huge_pmd_numa_page() mm: huge_memory: use folio_last_cpupid() in __split_huge_page_tail() mm: remove page_cpupid_last() mm: add folio_xchg_access_time() sched/fair: use folio_xchg_access_time() in numa_hint_fault_latency() mm: mprotect: use a folio in change_pte_range() mm: huge_memory: use a folio in change_huge_pmd() mm: remove xchg_page_access_time() mm: add folio_xchg_last_cpupid() sched/fair: use folio_xchg_last_cpupid() in should_numa_migrate_memory() mm: migrate: use folio_xchg_last_cpupid() in folio_migrate_flags() mm: huge_memory: use folio_xchg_last_cpupid() in __split_huge_page_tail() mm: make finish_mkwrite_fault() static mm: convert wp_page_reuse() and finish_mkwrite_fault() to take a folio mm: use folio_xchg_last_cpupid() in wp_page_reuse() mm: remove page_cpupid_xchg_last() include/linux/mempolicy.h | 5 +- include/linux/migrate.h | 4 +- include/linux/mm.h | 32 ++++--- include/linux/mm_types.h | 22 ++++- include/linux/sched/numa_balancing.h | 6 +- kernel/sched/fair.c | 12 +-- mm/huge_memory.c | 44 +++++---- mm/internal.h | 2 +- mm/memory.c | 85 +++++++++-------- mm/mempolicy.c | 22 +++-- mm/migrate.c | 134 +++++++++++++-------------- mm/mmzone.c | 6 +- mm/mprotect.c | 16 ++-- 13 files changed, 205 insertions(+), 185 deletions(-) -- 2.25.1

2 71

[PATCH OLK-5.10 v3 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 27 Nov '23

27 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH OLK-6.6 v2] anolis: bond: broadcast ARP or ND messages to all slaves
by Zhengchao Shao 27 Nov '23

27 Nov '23

From: Tony Lu <tonylu(a)linux.alibaba.com> anolis inclusion from devel-5.10-v5.10.134-12 commit b90e28f7170e1ae40c572f9f80a50bbdc8f8b99f category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8JTCU Reference: https://gitee.com/anolis/cloud-kernel/commit/b90e28f7170e1ae40c572f9f80a50b… --------------------------- OpenAnolis Bug Tracker:0000282 This is achieved by broadcasting ARP or ND packets to all of its slave devices on transmit side. The switch will take further actions based on proper configuration. A new sysctl knob "net.bonding.broadcast_arp_or_nd" is introduced which controls the behaviour of broadcasting. Signed-off-by: Tony Lu <tonylu(a)linux.alibaba.com> Acked-by: Dust Li <dust.li(a)linux.alibaba.com> Signed-off-by: Qiao Ma <mqaio(a)linux.alibaba.com> Reviewed-by: Shile Zhang <shile.zhang(a)linux.alibaba.com> Acked-by: Dust Li <dust.li(a)linux.alibaba.com> Signed-off-by: Wang Yufen <wangyufen(a)huawei.com> (cherry picked from commit 7c902772d8ee8840b6092b1de94ed0c17c9a8a50) Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- v2: modify email address error --- Documentation/networking/ip-sysctl.rst | 11 +++++++ drivers/net/bonding/Makefile | 2 +- drivers/net/bonding/bond_main.c | 40 ++++++++++++++++++++++++++ drivers/net/bonding/bond_sysctl.c | 32 +++++++++++++++++++++ include/net/bonding.h | 4 +++ 5 files changed, 88 insertions(+), 1 deletion(-) create mode 100644 drivers/net/bonding/bond_sysctl.c diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index a66054d0763a..12fa38639102 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -3206,3 +3206,14 @@ max_dgram_qlen - INTEGER Default: 10 + +``/proc/sys/net/bonding/*`` +======================== + +broadcast_arp_or_nd - INTEGER + Control broadcasting ARP or ND messages to all slaves + + 0: Not broadcasting + 1: Broadcasting + + Default: 0 diff --git a/drivers/net/bonding/Makefile b/drivers/net/bonding/Makefile index 30e8ae3da2da..9dd55d91185a 100644 --- a/drivers/net/bonding/Makefile +++ b/drivers/net/bonding/Makefile @@ -5,7 +5,7 @@ obj-$(CONFIG_BONDING) += bonding.o -bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_sysfs_slave.o bond_debugfs.o bond_netlink.o bond_options.o +bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_sysfs_slave.o bond_debugfs.o bond_netlink.o bond_options.o bond_sysctl.o proc-$(CONFIG_PROC_FS) += bond_procfs.o bonding-objs += $(proc-y) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 51d47eda1c87..842a138a574d 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -5127,6 +5127,39 @@ int bond_update_slave_arr(struct bonding *bond, struct slave *skipslave) return ret; } +/* Check whether the skb is arp or nd msg */ +static inline bool skb_is_arp_or_nd(struct sk_buff *skb) +{ + switch (ntohs(skb->protocol)) { + case ETH_P_ARP: + return true; + case ETH_P_IPV6: + if (pskb_may_pull(skb, sizeof(struct ipv6hdr) + + sizeof(struct nd_msg))) { + struct ipv6hdr *hdr = ipv6_hdr(skb); + u8 nexthdr = hdr->nexthdr; + struct icmp6hdr *icmp6; + + if (nexthdr == IPPROTO_ICMPV6) { + icmp6 = icmp6_hdr(skb); + + if ((icmp6->icmp6_type == + NDISC_NEIGHBOUR_SOLICITATION || + icmp6->icmp6_type == + NDISC_NEIGHBOUR_ADVERTISEMENT) && + icmp6->icmp6_code == 0) { + return true; + } + } + } + } + + return false; +} + +static netdev_tx_t bond_xmit_broadcast(struct sk_buff *skb, + struct net_device *bond_dev); + static struct slave *bond_xmit_3ad_xor_slave_get(struct bonding *bond, struct sk_buff *skb, struct bond_up_slave *slaves) @@ -5171,6 +5204,10 @@ static netdev_tx_t bond_3ad_xor_xmit(struct sk_buff *skb, struct bond_up_slave *slaves; struct slave *slave; + /* Broadcast to all slaves. */ + if (sysctl_bond_broadcast_arp_or_nd && skb_is_arp_or_nd(skb)) + return bond_xmit_broadcast(skb, dev); + slaves = rcu_dereference(bond->usable_slaves); slave = bond_xmit_3ad_xor_slave_get(bond, skb, slaves); if (likely(slave)) @@ -6488,6 +6525,7 @@ static int __init bonding_init(void) goto err_link; bond_create_debugfs(); + bond_create_sysctl(); for (i = 0; i < max_bonds; i++) { res = bond_create(&init_net, NULL); @@ -6504,6 +6542,7 @@ static int __init bonding_init(void) return res; err: bond_destroy_debugfs(); + bond_destroy_sysctl(); bond_netlink_fini(); err_link: unregister_pernet_subsys(&bond_net_ops); @@ -6516,6 +6555,7 @@ static void __exit bonding_exit(void) unregister_netdevice_notifier(&bond_netdev_notifier); bond_destroy_debugfs(); + bond_destroy_sysctl(); bond_netlink_fini(); unregister_pernet_subsys(&bond_net_ops); diff --git a/drivers/net/bonding/bond_sysctl.c b/drivers/net/bonding/bond_sysctl.c new file mode 100644 index 000000000000..17404d37a2fd --- /dev/null +++ b/drivers/net/bonding/bond_sysctl.c @@ -0,0 +1,32 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <net/net_namespace.h> +#include <linux/sysctl.h> +#include <net/bonding.h> + +int sysctl_bond_broadcast_arp_or_nd __read_mostly; +EXPORT_SYMBOL(sysctl_bond_broadcast_arp_or_nd); + +struct ctl_table_header *bond_broadcast_arp_or_nd_table_header; + +static struct ctl_table bond_broadcast_arp_or_nd_table[] = { + { + .procname = "broadcast_arp_or_nd", + .data = &sysctl_bond_broadcast_arp_or_nd, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + {} +}; + +void bond_create_sysctl(void) +{ + bond_broadcast_arp_or_nd_table_header = + register_net_sysctl(&init_net, "net/bonding", + bond_broadcast_arp_or_nd_table); +} + +void bond_destroy_sysctl(void) +{ + unregister_net_sysctl_table(bond_broadcast_arp_or_nd_table_header); +} diff --git a/include/net/bonding.h b/include/net/bonding.h index 5b8b1b644a2d..0f065f150b81 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -115,6 +115,8 @@ static inline int is_netpoll_tx_blocked(struct net_device *dev) #define is_netpoll_tx_blocked(dev) (0) #endif +extern int sysctl_bond_broadcast_arp_or_nd; + struct bond_params { int mode; int xmit_policy; @@ -684,6 +686,8 @@ struct bond_vlan_tag *bond_verify_device_path(struct net_device *start_dev, int bond_update_slave_arr(struct bonding *bond, struct slave *skipslave); void bond_slave_arr_work_rearm(struct bonding *bond, unsigned long delay); void bond_work_init_all(struct bonding *bond); +void bond_create_sysctl(void); +void bond_destroy_sysctl(void); #ifdef CONFIG_PROC_FS void bond_create_proc_entry(struct bonding *bond); -- 2.34.1

2 1

[PATCH OLK-6.6] anolis: bond: broadcast ARP or ND messages to all slaves
by Zhengchao Shao 27 Nov '23

27 Nov '23

From: Tony Lu <tonylu(a)linux.alibaba.com> anolis inclusion from devel-5.10-v5.10.134-12 commit b90e28f7170e1ae40c572f9f80a50bbdc8f8b99f category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8JTCU Reference: https://gitee.com/anolis/cloud-kernel/commit/b90e28f7170e1ae40c572f9f80a50b… --------------------------- OpenAnolis Bug Tracker:0000282 This is achieved by broadcasting ARP or ND packets to all of its slave devices on transmit side. The switch will take further actions based on proper configuration. A new sysctl knob "net.bonding.broadcast_arp_or_nd" is introduced which controls the behaviour of broadcasting. Signed-off-by: Tony Lu <tonylu(a)linux.alibaba.com> Acked-by: Dust Li <dust.li(a)linux.alibaba.com> Signed-off-by: Qiao Ma <mqaio(a)linux.alibaba.com> Reviewed-by: Shile Zhang <shile.zhang(a)linux.alibaba.com> Acked-by: Dust Li <dust.li(a)linux.alibaba.com> Signed-off-by: Wang Yufen <wangyufen(a)huawei.com> (cherry picked from commit 7c902772d8ee8840b6092b1de94ed0c17c9a8a50) Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- Documentation/networking/ip-sysctl.rst | 11 +++++++ drivers/net/bonding/Makefile | 2 +- drivers/net/bonding/bond_main.c | 40 ++++++++++++++++++++++++++ drivers/net/bonding/bond_sysctl.c | 32 +++++++++++++++++++++ include/net/bonding.h | 4 +++ 5 files changed, 88 insertions(+), 1 deletion(-) create mode 100644 drivers/net/bonding/bond_sysctl.c diff --git a/Documentation/networking/ip-sysctl.rst b/Documentation/networking/ip-sysctl.rst index a66054d0763a..12fa38639102 100644 --- a/Documentation/networking/ip-sysctl.rst +++ b/Documentation/networking/ip-sysctl.rst @@ -3206,3 +3206,14 @@ max_dgram_qlen - INTEGER Default: 10 + +``/proc/sys/net/bonding/*`` +======================== + +broadcast_arp_or_nd - INTEGER + Control broadcasting ARP or ND messages to all slaves + + 0: Not broadcasting + 1: Broadcasting + + Default: 0 diff --git a/drivers/net/bonding/Makefile b/drivers/net/bonding/Makefile index 30e8ae3da2da..9dd55d91185a 100644 --- a/drivers/net/bonding/Makefile +++ b/drivers/net/bonding/Makefile @@ -5,7 +5,7 @@ obj-$(CONFIG_BONDING) += bonding.o -bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_sysfs_slave.o bond_debugfs.o bond_netlink.o bond_options.o +bonding-objs := bond_main.o bond_3ad.o bond_alb.o bond_sysfs.o bond_sysfs_slave.o bond_debugfs.o bond_netlink.o bond_options.o bond_sysctl.o proc-$(CONFIG_PROC_FS) += bond_procfs.o bonding-objs += $(proc-y) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 51d47eda1c87..842a138a574d 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -5127,6 +5127,39 @@ int bond_update_slave_arr(struct bonding *bond, struct slave *skipslave) return ret; } +/* Check whether the skb is arp or nd msg */ +static inline bool skb_is_arp_or_nd(struct sk_buff *skb) +{ + switch (ntohs(skb->protocol)) { + case ETH_P_ARP: + return true; + case ETH_P_IPV6: + if (pskb_may_pull(skb, sizeof(struct ipv6hdr) + + sizeof(struct nd_msg))) { + struct ipv6hdr *hdr = ipv6_hdr(skb); + u8 nexthdr = hdr->nexthdr; + struct icmp6hdr *icmp6; + + if (nexthdr == IPPROTO_ICMPV6) { + icmp6 = icmp6_hdr(skb); + + if ((icmp6->icmp6_type == + NDISC_NEIGHBOUR_SOLICITATION || + icmp6->icmp6_type == + NDISC_NEIGHBOUR_ADVERTISEMENT) && + icmp6->icmp6_code == 0) { + return true; + } + } + } + } + + return false; +} + +static netdev_tx_t bond_xmit_broadcast(struct sk_buff *skb, + struct net_device *bond_dev); + static struct slave *bond_xmit_3ad_xor_slave_get(struct bonding *bond, struct sk_buff *skb, struct bond_up_slave *slaves) @@ -5171,6 +5204,10 @@ static netdev_tx_t bond_3ad_xor_xmit(struct sk_buff *skb, struct bond_up_slave *slaves; struct slave *slave; + /* Broadcast to all slaves. */ + if (sysctl_bond_broadcast_arp_or_nd && skb_is_arp_or_nd(skb)) + return bond_xmit_broadcast(skb, dev); + slaves = rcu_dereference(bond->usable_slaves); slave = bond_xmit_3ad_xor_slave_get(bond, skb, slaves); if (likely(slave)) @@ -6488,6 +6525,7 @@ static int __init bonding_init(void) goto err_link; bond_create_debugfs(); + bond_create_sysctl(); for (i = 0; i < max_bonds; i++) { res = bond_create(&init_net, NULL); @@ -6504,6 +6542,7 @@ static int __init bonding_init(void) return res; err: bond_destroy_debugfs(); + bond_destroy_sysctl(); bond_netlink_fini(); err_link: unregister_pernet_subsys(&bond_net_ops); @@ -6516,6 +6555,7 @@ static void __exit bonding_exit(void) unregister_netdevice_notifier(&bond_netdev_notifier); bond_destroy_debugfs(); + bond_destroy_sysctl(); bond_netlink_fini(); unregister_pernet_subsys(&bond_net_ops); diff --git a/drivers/net/bonding/bond_sysctl.c b/drivers/net/bonding/bond_sysctl.c new file mode 100644 index 000000000000..17404d37a2fd --- /dev/null +++ b/drivers/net/bonding/bond_sysctl.c @@ -0,0 +1,32 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <net/net_namespace.h> +#include <linux/sysctl.h> +#include <net/bonding.h> + +int sysctl_bond_broadcast_arp_or_nd __read_mostly; +EXPORT_SYMBOL(sysctl_bond_broadcast_arp_or_nd); + +struct ctl_table_header *bond_broadcast_arp_or_nd_table_header; + +static struct ctl_table bond_broadcast_arp_or_nd_table[] = { + { + .procname = "broadcast_arp_or_nd", + .data = &sysctl_bond_broadcast_arp_or_nd, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec, + }, + {} +}; + +void bond_create_sysctl(void) +{ + bond_broadcast_arp_or_nd_table_header = + register_net_sysctl(&init_net, "net/bonding", + bond_broadcast_arp_or_nd_table); +} + +void bond_destroy_sysctl(void) +{ + unregister_net_sysctl_table(bond_broadcast_arp_or_nd_table_header); +} diff --git a/include/net/bonding.h b/include/net/bonding.h index 5b8b1b644a2d..0f065f150b81 100644 --- a/include/net/bonding.h +++ b/include/net/bonding.h @@ -115,6 +115,8 @@ static inline int is_netpoll_tx_blocked(struct net_device *dev) #define is_netpoll_tx_blocked(dev) (0) #endif +extern int sysctl_bond_broadcast_arp_or_nd; + struct bond_params { int mode; int xmit_policy; @@ -684,6 +686,8 @@ struct bond_vlan_tag *bond_verify_device_path(struct net_device *start_dev, int bond_update_slave_arr(struct bonding *bond, struct slave *skipslave); void bond_slave_arr_work_rearm(struct bonding *bond, unsigned long delay); void bond_work_init_all(struct bonding *bond); +void bond_create_sysctl(void); +void bond_destroy_sysctl(void); #ifdef CONFIG_PROC_FS void bond_create_proc_entry(struct bonding *bond); -- 2.34.1

2 1

[PATCH OLK-5.10 v3 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 27 Nov '23

27 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[src-openEuler/rdma-core 0/5] Fix several bugs
by Chengchang Tang 27 Nov '23

27 Nov '23

From: Ran Zhou <zhouran10(a)h-partners.com> This series of patches focus on some minor bugfix. Chengchang Tang (3): libhns: Fix possible overflow in cq clean libhns: Fix uninitialized qp attr when flush cqe libhns: Fix unnecessary dca memory detach Luoyouming (1): libhns: Add input parameter check for hnsdv_query_device() Ran Zhou (1): libhns: Corrects several issues with output format and variable types. providers/hns/hns_roce_u.c | 2 +- providers/hns/hns_roce_u_hw_v2.c | 45 ++++++++++++++++---------------- providers/hns/hns_roce_u_verbs.c | 5 ++-- 3 files changed, 27 insertions(+), 25 deletions(-) -- 2.30.0

1 5

[OLK-5.10 v2 00/12] A set of patchs for openEuler
by Chengchang Tang 27 Nov '23

27 Nov '23

From: Juan Zhou <zhoujuan51(a)h-partners.com> This group of patches fix the issues found in the code review. Chengchang Tang (11): RDMA/hns: Use complete parentheses in macros RDMA/hns: Fix sleeping in setup_dca_buf_to_hw() RDMA/hns: Remove unnecessary checks for NULL in mtr_alloc_bufs() RDMA/hns: Improve the readability of free mr exception branch RDMA/hns: Fix incorrect print type in bond RDMA/hns: Fix a meaningless loop in free_buffer_pages_proc() RDMA/hns: Delete redundant assignment values RDMA/hns: Remove redundant address operations in print DCA logs RDMA/hns: Fix memory leak in free_mr_exit() and free_mr_init() RDMA/hns: Fix unhandled error code in dca uverbs RDMA/hns: Improve readabilties for dca uverbs Junxian Huang (1): RDMA/hns: Fix several memory issues in roce bonding drivers/infiniband/hw/hns/hns_roce_bond.c | 17 ++++--- drivers/infiniband/hw/hns/hns_roce_dca.c | 45 ++++++++++--------- drivers/infiniband/hw/hns/hns_roce_hem.h | 12 ++--- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 52 ++++++++++++++-------- drivers/infiniband/hw/hns/hns_roce_main.c | 12 ++++- drivers/infiniband/hw/hns/hns_roce_mr.c | 4 +- 6 files changed, 87 insertions(+), 55 deletions(-) -- 2.30.0

1 12

[PATCH openEuler-22.03-LTS-SP2] config: update openeuler_defconfig for x86 with gcc 10.3.1
by Jialin Zhang 27 Nov '23

27 Nov '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8JANU ----------------------------------------- Using the following commands introduced in commit ("kconfig: Add script to check & update openeuler_defconfig") to update openeuler_defconfig for x86: make update_oedefconfig Updating with gcc 10.3.1 is different from updating with gcc 8.4.0 in commit 0de6142922da ("config: update openeuler_defconfig for x86"). Fixes: 0de6142922da ("config: update openeuler_defconfig for x86") Signed-off-by: Jialin Zhang <zhangjialin11(a)huawei.com> --- arch/x86/configs/openeuler_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 9b65f44e62c5..ebe4fb56da52 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -8181,6 +8181,7 @@ CONFIG_DEBUG_SECTION_MISMATCH=y CONFIG_SECTION_MISMATCH_WARN_ONLY=y CONFIG_STACK_VALIDATION=y # CONFIG_DEBUG_FORCE_WEAK_PER_CPU is not set +# CONFIG_PGO_KERNEL is not set # end of Compile-time checks and compiler options # -- 2.25.1

2 1

[PATCH OLK-5.10] config: update openeuler_defconfig for x86 with gcc 10.3.1
by Jialin Zhang 27 Nov '23

27 Nov '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8JANU ----------------------------------------- Using the following commands introduced in commit ("kconfig: Add script to check & update openeuler_defconfig") to update openeuler_defconfig for x86: make update_oedefconfig Updating with gcc 10.3.1 is different from updating with gcc 8.4.0 in commit 0de6142922da ("config: update openeuler_defconfig for x86"). Fixes: 0de6142922da ("config: update openeuler_defconfig for x86") Signed-off-by: Jialin Zhang <zhangjialin11(a)huawei.com> --- arch/x86/configs/openeuler_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 225ef354637a..32236e4fa96d 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -8191,6 +8191,7 @@ CONFIG_DEBUG_SECTION_MISMATCH=y CONFIG_SECTION_MISMATCH_WARN_ONLY=y CONFIG_STACK_VALIDATION=y # CONFIG_DEBUG_FORCE_WEAK_PER_CPU is not set +# CONFIG_PGO_KERNEL is not set # end of Compile-time checks and compiler options # -- 2.25.1

2 1

[PATCH V2 OLK-5.10] vhost-vdpa: add reset state params to indicate reset level
by Jiang Dongxu 27 Nov '23

27 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- Add the interface parameter state to the vdpa reset interface, which respectively identifies the reset when the device is turned on/off and the virtio reset issued by the virtual machine. Signed-off-by: jiangdongxu <jiangdongxu1(a)huawei.com> --- drivers/vdpa/ifcvf/ifcvf_main.c | 2 +- drivers/vdpa/mlx5/net/mlx5_vnet.c | 2 +- drivers/vdpa/virtio_pci/vp_vdpa.c | 2 +- drivers/vhost/vdpa.c | 10 +++++----- drivers/virtio/virtio_vdpa.c | 2 +- include/linux/vdpa.h | 16 +++++++++++++--- 6 files changed, 22 insertions(+), 12 deletions(-) diff --git a/drivers/vdpa/ifcvf/ifcvf_main.c b/drivers/vdpa/ifcvf/ifcvf_main.c index 862f4591fff2..2a4991464cfd 100644 --- a/drivers/vdpa/ifcvf/ifcvf_main.c +++ b/drivers/vdpa/ifcvf/ifcvf_main.c @@ -233,7 +233,7 @@ static void ifcvf_vdpa_set_status(struct vdpa_device *vdpa_dev, u8 status) ifcvf_set_status(vf, status); } -static int ifcvf_vdpa_reset(struct vdpa_device *vdpa_dev) +static int ifcvf_vdpa_reset(struct vdpa_device *vdpa_dev, int state) { struct ifcvf_adapter *adapter; struct ifcvf_hw *vf; diff --git a/drivers/vdpa/mlx5/net/mlx5_vnet.c b/drivers/vdpa/mlx5/net/mlx5_vnet.c index 76e7b4bab9d3..c676bb70a453 100644 --- a/drivers/vdpa/mlx5/net/mlx5_vnet.c +++ b/drivers/vdpa/mlx5/net/mlx5_vnet.c @@ -1800,7 +1800,7 @@ static void mlx5_vdpa_set_status(struct vdpa_device *vdev, u8 status) ndev->mvdev.status |= VIRTIO_CONFIG_S_FAILED; } -static int mlx5_vdpa_reset(struct vdpa_device *vdev) +static int mlx5_vdpa_reset(struct vdpa_device *vdev, int state) { struct mlx5_vdpa_dev *mvdev = to_mvdev(vdev); struct mlx5_vdpa_net *ndev = to_mlx5_vdpa_ndev(mvdev); diff --git a/drivers/vdpa/virtio_pci/vp_vdpa.c b/drivers/vdpa/virtio_pci/vp_vdpa.c index f4e375b1d903..bc0bdcc7bcea 100644 --- a/drivers/vdpa/virtio_pci/vp_vdpa.c +++ b/drivers/vdpa/virtio_pci/vp_vdpa.c @@ -220,7 +220,7 @@ static void vp_vdpa_set_status(struct vdpa_device *vdpa, u8 status) vp_modern_set_status(mdev, status); } -static int vp_vdpa_reset(struct vdpa_device *vdpa) +static int vp_vdpa_reset(struct vdpa_device *vdpa, int state) { struct vp_vdpa *vp_vdpa = vdpa_to_vp(vdpa); struct virtio_pci_modern_device *mdev = vp_vdpa_to_mdev(vp_vdpa); diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 8e3bf64123ae..7c1c40a221c7 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -212,13 +212,13 @@ static void vhost_vdpa_unsetup_vq_irq(struct vhost_vdpa *v, u16 qid) irq_bypass_unregister_producer(&vq->call_ctx.producer); } -static int vhost_vdpa_reset(struct vhost_vdpa *v) +static int vhost_vdpa_reset(struct vhost_vdpa *v, int state) { struct vdpa_device *vdpa = v->vdpa; v->in_batch = 0; - return vdpa_reset(vdpa); + return vdpa_reset(vdpa, state); } static long vhost_vdpa_bind_mm(struct vhost_vdpa *v) @@ -297,7 +297,7 @@ static long vhost_vdpa_set_status(struct vhost_vdpa *v, u8 __user *statusp) vhost_vdpa_unsetup_vq_irq(v, i); if (status == 0) { - ret = vdpa_reset(vdpa); + ret = vdpa_reset(vdpa, VDPA_DEV_RESET_VIRTIO); if (ret) return ret; } else @@ -1493,7 +1493,7 @@ static int vhost_vdpa_open(struct inode *inode, struct file *filep) return r; nvqs = v->nvqs; - r = vhost_vdpa_reset(v); + r = vhost_vdpa_reset(v, VDPA_DEV_RESET_OPEN); if (r) goto err; @@ -1539,7 +1539,7 @@ static int vhost_vdpa_release(struct inode *inode, struct file *filep) mutex_lock(&d->mutex); filep->private_data = NULL; vhost_vdpa_clean_irq(v); - vhost_vdpa_reset(v); + vhost_vdpa_reset(v, VDPA_DEV_RESET_CLOSE); vhost_dev_stop(&v->vdev); vhost_vdpa_unbind_mm(v); vhost_vdpa_config_put(v); diff --git a/drivers/virtio/virtio_vdpa.c b/drivers/virtio/virtio_vdpa.c index 46c71653f508..8126a9be22cd 100644 --- a/drivers/virtio/virtio_vdpa.c +++ b/drivers/virtio/virtio_vdpa.c @@ -99,7 +99,7 @@ static void virtio_vdpa_reset(struct virtio_device *vdev) { struct vdpa_device *vdpa = vd_get_vdpa(vdev); - vdpa_reset(vdpa); + vdpa_reset(vdpa, VDPA_DEV_RESET_VIRTIO); } static bool virtio_vdpa_notify(struct virtqueue *vq) diff --git a/include/linux/vdpa.h b/include/linux/vdpa.h index 83df2380eeb9..ef53829c7a7a 100644 --- a/include/linux/vdpa.h +++ b/include/linux/vdpa.h @@ -120,6 +120,12 @@ struct vdpa_map_file { u64 offset; }; +enum vdpa_reset_state { + VDPA_DEV_RESET_VIRTIO = 0, + VDPA_DEV_RESET_OPEN = 1, + VDPA_DEV_RESET_CLOSE = 2, +}; + /** * struct vdpa_config_ops - operations for configuring a vDPA device. * Note: vDPA device drivers are required to implement all of the @@ -218,6 +224,10 @@ struct vdpa_map_file { * @status: virtio device status * @reset: Reset device * @vdev: vdpa device + * @state: state for reset + * VDPA_DEV_RESET_VIRTIO for virtio reset + * VDPA_DEV_RESET_OPEN for vhost-vdpa device open + * VDPA_DEV_RESET_CLOSE for vhost-vdpa device close * Returns integer: success (0) or error (< 0) * @suspend: Suspend the device (optional) * @vdev: vdpa device @@ -359,7 +369,7 @@ struct vdpa_config_ops { u32 (*get_vendor_id)(struct vdpa_device *vdev); u8 (*get_status)(struct vdpa_device *vdev); void (*set_status)(struct vdpa_device *vdev, u8 status); - int (*reset)(struct vdpa_device *vdev); + int (*reset)(struct vdpa_device *vdev, int state); int (*suspend)(struct vdpa_device *vdev); int (*resume)(struct vdpa_device *vdev); size_t (*get_config_size)(struct vdpa_device *vdev); @@ -482,14 +492,14 @@ static inline struct device *vdpa_get_dma_dev(struct vdpa_device *vdev) return vdev->dma_dev; } -static inline int vdpa_reset(struct vdpa_device *vdev) +static inline int vdpa_reset(struct vdpa_device *vdev, int state) { const struct vdpa_config_ops *ops = vdev->config; int ret; down_write(&vdev->cf_lock); vdev->features_valid = false; - ret = ops->reset(vdev); + ret = ops->reset(vdev, state); up_write(&vdev->cf_lock); return ret; } -- 2.27.0

2 1

[PATCH V2 OLK-5.10] vhost-vdpa: allow set feature VHOST_F_LOG_ALL when been negotiated.
by Jiang Dongxu 27 Nov '23

27 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- It's not allowed to change the features after vhost-vdpa devices have been negotiated. But log start/end is allowed. Add exception to feature VHOST_F_LOG_ALL. Signed-off-by: jiangdongxu <jiangdongxu1(a)huawei.com> --- drivers/vhost/vdpa.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 7c1c40a221c7..eed51d004531 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -425,16 +425,19 @@ static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep) u64 features; int i; + if (copy_from_user(&features, featurep, sizeof(features))) + return -EFAULT; + + actual_features = ops->get_driver_features(vdpa); + /* * It's not allowed to change the features after they have - * been negotiated. + * been negotiated. But log start/end is allowed. */ - if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) + if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) && + (features & ~(BIT_ULL(VHOST_F_LOG_ALL))) != actual_features) return -EBUSY; - if (copy_from_user(&features, featurep, sizeof(features))) - return -EFAULT; - if (vdpa_set_features(vdpa, features)) return -EINVAL; -- 2.27.0

2 1

[PATCH OLK-6.6] proc: fix ubsan warning in mem_lseek
by Zizhi Wo 27 Nov '23

27 Nov '23

From: yangerkun <yangerkun(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JE4E?from=project-issue --------------------------- UBSAN has reported a overflow with mem_lseek. And it's fine with mem_open set file mode with FMODE_UNSIGNED_OFFSET(memory_lseek). However, another file use mem_lseek do lseek can have not FMODE_UNSIGNED_OFFSET(proc_kpagecount_operations/proc_pagemap_operations), fix it by checking overflow and FMODE_UNSIGNED_OFFSET. Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> ================================================================== UBSAN: Undefined behaviour in ../fs/proc/base.c:941:15 signed integer overflow: 4611686018427387904 + 4611686018427387904 cannot be represented in type 'long long int' CPU: 4 PID: 4762 Comm: syz-executor.1 Not tainted 4.4.189 #3 Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015 Call trace: [<ffffff90080a5f28>] dump_backtrace+0x0/0x590 arch/arm64/kernel/traps.c:91 [<ffffff90080a64f0>] show_stack+0x38/0x60 arch/arm64/kernel/traps.c:234 [<ffffff9008986a34>] __dump_stack lib/dump_stack.c:15 [inline] [<ffffff9008986a34>] dump_stack+0x128/0x184 lib/dump_stack.c:51 [<ffffff9008a2d120>] ubsan_epilogue+0x34/0x9c lib/ubsan.c:166 [<ffffff9008a2d8b8>] handle_overflow+0x228/0x280 lib/ubsan.c:197 [<ffffff9008a2da2c>] __ubsan_handle_add_overflow+0x4c/0x68 lib/ubsan.c:204 [<ffffff900862b9f4>] mem_lseek+0x12c/0x130 fs/proc/base.c:941 [<ffffff90084ef78c>] vfs_llseek fs/read_write.c:260 [inline] [<ffffff90084ef78c>] SYSC_lseek fs/read_write.c:285 [inline] [<ffffff90084ef78c>] SyS_lseek+0x164/0x1f0 fs/read_write.c:276 [<ffffff9008093c80>] el0_svc_naked+0x30/0x34 ================================================================== Signed-off-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: zhangyi (F) <yi.zhang(a)huawei.com> (cherry picked from commit a422358aa04c53a08b215b8dcd6814d916ef5cf1) Conflicts: fs/read_write.c Signed-off-by: Li Ming <limingming.li(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/proc/base.c | 32 ++++++++++++++++++++++++-------- fs/read_write.c | 5 ----- include/linux/fs.h | 4 ++++ 3 files changed, 28 insertions(+), 13 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index ffd54617c354..d0ae6e4511aa 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -903,18 +903,34 @@ static ssize_t mem_write(struct file *file, const char __user *buf, loff_t mem_lseek(struct file *file, loff_t offset, int orig) { + loff_t ret = 0; + + spin_lock(&file->f_lock); switch (orig) { - case 0: - file->f_pos = offset; - break; - case 1: - file->f_pos += offset; + case SEEK_CUR: + offset += file->f_pos; + __attribute__((__fallthrough__)); + case SEEK_SET: + /* to avoid userland mistaking f_pos=-9 as -EBADF=-9 */ + if ((unsigned long long)offset >= -MAX_ERRNO) + ret = -EOVERFLOW; break; default: - return -EINVAL; + ret = -EINVAL; + } + + if (!ret) { + if (offset < 0 && !(unsigned_offsets(file))) { + ret = -EINVAL; + } else { + file->f_pos = offset; + ret = file->f_pos; + force_successful_syscall_return(); + } } - force_successful_syscall_return(); - return file->f_pos; + + spin_unlock(&file->f_lock); + return ret; } static int mem_release(struct inode *inode, struct file *file) diff --git a/fs/read_write.c b/fs/read_write.c index 4771701c896b..2f456d5a1df5 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -34,11 +34,6 @@ const struct file_operations generic_ro_fops = { EXPORT_SYMBOL(generic_ro_fops); -static inline bool unsigned_offsets(struct file *file) -{ - return file->f_mode & FMODE_UNSIGNED_OFFSET; -} - /** * vfs_setpos - update the file offset for lseek * @file: file structure in question diff --git a/include/linux/fs.h b/include/linux/fs.h index 4f6b15e960ec..59d4edee1b96 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2932,6 +2932,10 @@ extern long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, extern void file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping); extern loff_t noop_llseek(struct file *file, loff_t offset, int whence); +static inline bool unsigned_offsets(struct file *file) +{ + return file->f_mode & FMODE_UNSIGNED_OFFSET; +} #define no_llseek NULL extern loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize); extern loff_t generic_file_llseek(struct file *file, loff_t offset, int whence); -- 2.39.2

2 1

[PATCH OLK-5.10] open: make RESOLVE_CACHED correctly test for O_TMPFILE
by Zizhi Wo 27 Nov '23

27 Nov '23

From: Aleksa Sarai <cyphar(a)cyphar.com> stable inclusion from stable-v5.10.190 commit c9c78b91c7834de6e4ae49392c3177608bcd47ab category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JN4Q?from=project-issue Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- commit a0fc452a5d7fed986205539259df1d60546f536c upstream. O_TMPFILE is actually __O_TMPFILE|O_DIRECTORY. This means that the old fast-path check for RESOLVE_CACHED would reject all users passing O_DIRECTORY with -EAGAIN, when in fact the intended test was to check for __O_TMPFILE. Cc: stable(a)vger.kernel.org # v5.12+ Fixes: 99668f618062 ("fs: expose LOOKUP_CACHED through openat2() RESOLVE_CACHED") Signed-off-by: Aleksa Sarai <cyphar(a)cyphar.com> Message-Id: <20230806-resolve_cached-o_tmpfile-v1-1-7ba16308465e(a)cyphar.com> Signed-off-by: Christian Brauner <brauner(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/open.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/open.c b/fs/open.c index 8530ca1c0345..8092178ceab0 100644 --- a/fs/open.c +++ b/fs/open.c @@ -1101,7 +1101,7 @@ inline int build_open_flags(const struct open_how *how, struct open_flags *op) lookup_flags |= LOOKUP_IN_ROOT; if (how->resolve & RESOLVE_CACHED) { /* Don't bother even trying for create/truncate/tmpfile open */ - if (flags & (O_TRUNC | O_CREAT | O_TMPFILE)) + if (flags & (O_TRUNC | O_CREAT | __O_TMPFILE)) return -EAGAIN; lookup_flags |= LOOKUP_CACHED; } -- 2.39.2

2 1

[PATCH OLK-5.10] open: make RESOLVE_CACHED correctly test for O_TMPFILE
by Zizhi Wo 27 Nov '23

27 Nov '23

From: Aleksa Sarai <cyphar(a)cyphar.com> stable inclusion from stable-v6.5-rc5 commit a0fc452a5d7fed986205539259df1d60546f536c category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JN4Q?from=project-issue Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- O_TMPFILE is actually __O_TMPFILE|O_DIRECTORY. This means that the old fast-path check for RESOLVE_CACHED would reject all users passing O_DIRECTORY with -EAGAIN, when in fact the intended test was to check for __O_TMPFILE. Cc: stable(a)vger.kernel.org # v5.12+ Fixes: 99668f618062 ("fs: expose LOOKUP_CACHED through openat2() RESOLVE_CACHED") Signed-off-by: Aleksa Sarai <cyphar(a)cyphar.com> Message-Id: <20230806-resolve_cached-o_tmpfile-v1-1-7ba16308465e(a)cyphar.com> Signed-off-by: Christian Brauner <brauner(a)kernel.org> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/open.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/open.c b/fs/open.c index 8530ca1c0345..8092178ceab0 100644 --- a/fs/open.c +++ b/fs/open.c @@ -1101,7 +1101,7 @@ inline int build_open_flags(const struct open_how *how, struct open_flags *op) lookup_flags |= LOOKUP_IN_ROOT; if (how->resolve & RESOLVE_CACHED) { /* Don't bother even trying for create/truncate/tmpfile open */ - if (flags & (O_TRUNC | O_CREAT | O_TMPFILE)) + if (flags & (O_TRUNC | O_CREAT | __O_TMPFILE)) return -EAGAIN; lookup_flags |= LOOKUP_CACHED; } -- 2.39.2

2 1

[PATCH OLK-5.10 v2 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 27 Nov '23

27 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

2 21

[PATCH openEuler-1.0-LTS] umh: fix memory leak on execve failure
by Wenyu Huang 27 Nov '23

27 Nov '23

From: Vincent Minet <v.minet(a)criteo.com> mainline inclusion from mainline commit db803036ada7d61d096783726f9771b3fc540370 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J9MA CVE: NA ------------------------------------------------- If a UMH process created by fork_usermode_blob() fails to execute, a pair of struct file allocated by umh_pipe_setup() will leak. Under normal conditions, the caller (like bpfilter) needs to manage the lifetime of the UMH and its two pipes. But when fork_usermode_blob() fails, the caller doesn't really have a way to know what needs to be done. It seems better to do the cleanup ourselves in this case. Fixes: 449325b52b7a ("umh: introduce fork_usermode_blob() helper") Signed-off-by: Vincent Minet <v.minet(a)criteo.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Wenyu Huang <huangwenyu5(a)huawei.com> --- kernel/umh.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/umh.c b/kernel/umh.c index 53611efb10cb..715b4368f9ed 100644 --- a/kernel/umh.c +++ b/kernel/umh.c @@ -495,6 +495,12 @@ static void umh_clean_and_save_pid(struct subprocess_info *info) { struct umh_info *umh_info = info->data; + /* cleanup if umh_pipe_setup() was successful but exec failed */ + if (info->pid && info->retval) { + fput(umh_info->pipe_to_umh); + fput(umh_info->pipe_from_umh); + } + argv_free(info->argv); umh_info->pid = info->pid; } -- 2.34.1

2 1

[PATCH OLK-6.6] proc: fix ubsan warning in mem_lseek
by Zizhi Wo 27 Nov '23

27 Nov '23

From: yangerkun <yangerkun(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JE4E?from=project-issue --------------------------- UBSAN has reported a overflow with mem_lseek. And it's fine with mem_open set file mode with FMODE_UNSIGNED_OFFSET(memory_lseek). However, another file use mem_lseek do lseek can have not FMODE_UNSIGNED_OFFSET(proc_kpagecount_operations/proc_pagemap_operations), fix it by checking overflow and FMODE_UNSIGNED_OFFSET. Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> ================================================================== UBSAN: Undefined behaviour in ../fs/proc/base.c:941:15 signed integer overflow: 4611686018427387904 + 4611686018427387904 cannot be represented in type 'long long int' CPU: 4 PID: 4762 Comm: syz-executor.1 Not tainted 4.4.189 #3 Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015 Call trace: [<ffffff90080a5f28>] dump_backtrace+0x0/0x590 arch/arm64/kernel/traps.c:91 [<ffffff90080a64f0>] show_stack+0x38/0x60 arch/arm64/kernel/traps.c:234 [<ffffff9008986a34>] __dump_stack lib/dump_stack.c:15 [inline] [<ffffff9008986a34>] dump_stack+0x128/0x184 lib/dump_stack.c:51 [<ffffff9008a2d120>] ubsan_epilogue+0x34/0x9c lib/ubsan.c:166 [<ffffff9008a2d8b8>] handle_overflow+0x228/0x280 lib/ubsan.c:197 [<ffffff9008a2da2c>] __ubsan_handle_add_overflow+0x4c/0x68 lib/ubsan.c:204 [<ffffff900862b9f4>] mem_lseek+0x12c/0x130 fs/proc/base.c:941 [<ffffff90084ef78c>] vfs_llseek fs/read_write.c:260 [inline] [<ffffff90084ef78c>] SYSC_lseek fs/read_write.c:285 [inline] [<ffffff90084ef78c>] SyS_lseek+0x164/0x1f0 fs/read_write.c:276 [<ffffff9008093c80>] el0_svc_naked+0x30/0x34 ================================================================== Signed-off-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: zhangyi (F) <yi.zhang(a)huawei.com> (cherry picked from commit a422358aa04c53a08b215b8dcd6814d916ef5cf1) Conflicts: fs/read_write.c Signed-off-by: Li Ming <limingming.li(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/proc/base.c | 32 ++++++++++++++++++++++++-------- fs/read_write.c | 5 ----- include/linux/fs.h | 4 ++++ 3 files changed, 28 insertions(+), 13 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index ffd54617c354..d0ae6e4511aa 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -903,18 +903,34 @@ static ssize_t mem_write(struct file *file, const char __user *buf, loff_t mem_lseek(struct file *file, loff_t offset, int orig) { + loff_t ret = 0; + + spin_lock(&file->f_lock); switch (orig) { - case 0: - file->f_pos = offset; - break; - case 1: - file->f_pos += offset; + case SEEK_CUR: + offset += file->f_pos; + __attribute__((fallthrough)); + case SEEK_SET: + /* to avoid userland mistaking f_pos=-9 as -EBADF=-9 */ + if ((unsigned long long)offset >= -MAX_ERRNO) + ret = -EOVERFLOW; break; default: - return -EINVAL; + ret = -EINVAL; + } + + if (!ret) { + if (offset < 0 && !(unsigned_offsets(file))) { + ret = -EINVAL; + } else { + file->f_pos = offset; + ret = file->f_pos; + force_successful_syscall_return(); + } } - force_successful_syscall_return(); - return file->f_pos; + + spin_unlock(&file->f_lock); + return ret; } static int mem_release(struct inode *inode, struct file *file) diff --git a/fs/read_write.c b/fs/read_write.c index 4771701c896b..2f456d5a1df5 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -34,11 +34,6 @@ const struct file_operations generic_ro_fops = { EXPORT_SYMBOL(generic_ro_fops); -static inline bool unsigned_offsets(struct file *file) -{ - return file->f_mode & FMODE_UNSIGNED_OFFSET; -} - /** * vfs_setpos - update the file offset for lseek * @file: file structure in question diff --git a/include/linux/fs.h b/include/linux/fs.h index 4f6b15e960ec..59d4edee1b96 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2932,6 +2932,10 @@ extern long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, extern void file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping); extern loff_t noop_llseek(struct file *file, loff_t offset, int whence); +static inline bool unsigned_offsets(struct file *file) +{ + return file->f_mode & FMODE_UNSIGNED_OFFSET; +} #define no_llseek NULL extern loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize); extern loff_t generic_file_llseek(struct file *file, loff_t offset, int whence); -- 2.39.2

2 1

[PATCH openEuler-1.0-LTS] umh: fix memory leak on execve failure
by Wenyu Huang 27 Nov '23

27 Nov '23

From: Vincent Minet <v.minet(a)criteo.com> mainline inclusion from mainline commit db803036ada7d61d096783726f9771b3fc540370 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J9MA CVE: NA ------------------------------------------------- If a UMH process created by fork_usermode_blob() fails to execute, a pair of struct file allocated by umh_pipe_setup() will leak. Under normal conditions, the caller (like bpfilter) needs to manage the lifetime of the UMH and its two pipes. But when fork_usermode_blob() fails, the caller doesn't really have a way to know what needs to be done. It seems better to do the cleanup ourselves in this case. Fixes: 449325b52b7a ("umh: introduce fork_usermode_blob() helper") Signed-off-by: Vincent Minet <v.minet(a)criteo.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Wenyu Huang <huangwenyu5(a)huawei.com> --- kernel/umh.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/umh.c b/kernel/umh.c index 53611efb10cb..715b4368f9ed 100644 --- a/kernel/umh.c +++ b/kernel/umh.c @@ -495,6 +495,12 @@ static void umh_clean_and_save_pid(struct subprocess_info *info) { struct umh_info *umh_info = info->data; + /* cleanup if umh_pipe_setup() was successful but exec failed */ + if (info->pid && info->retval) { + fput(umh_info->pipe_to_umh); + fput(umh_info->pipe_from_umh); + } + argv_free(info->argv); umh_info->pid = info->pid; } -- 2.34.1

2 1

[PATCH OLK-6.6] proc: fix ubsan warning in mem_lseek
by Zizhi Wo 25 Nov '23

25 Nov '23

From: yangerkun <yangerkun(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JE4E?from=project-issue --------------------------- UBSAN has reported a overflow with mem_lseek. And it's fine with mem_open set file mode with FMODE_UNSIGNED_OFFSET(memory_lseek). However, another file use mem_lseek do lseek can have not FMODE_UNSIGNED_OFFSET(proc_kpagecount_operations/proc_pagemap_operations), fix it by checking overflow and FMODE_UNSIGNED_OFFSET. Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> ================================================================== UBSAN: Undefined behaviour in ../fs/proc/base.c:941:15 signed integer overflow: 4611686018427387904 + 4611686018427387904 cannot be represented in type 'long long int' CPU: 4 PID: 4762 Comm: syz-executor.1 Not tainted 4.4.189 #3 Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015 Call trace: [<ffffff90080a5f28>] dump_backtrace+0x0/0x590 arch/arm64/kernel/traps.c:91 [<ffffff90080a64f0>] show_stack+0x38/0x60 arch/arm64/kernel/traps.c:234 [<ffffff9008986a34>] __dump_stack lib/dump_stack.c:15 [inline] [<ffffff9008986a34>] dump_stack+0x128/0x184 lib/dump_stack.c:51 [<ffffff9008a2d120>] ubsan_epilogue+0x34/0x9c lib/ubsan.c:166 [<ffffff9008a2d8b8>] handle_overflow+0x228/0x280 lib/ubsan.c:197 [<ffffff9008a2da2c>] __ubsan_handle_add_overflow+0x4c/0x68 lib/ubsan.c:204 [<ffffff900862b9f4>] mem_lseek+0x12c/0x130 fs/proc/base.c:941 [<ffffff90084ef78c>] vfs_llseek fs/read_write.c:260 [inline] [<ffffff90084ef78c>] SYSC_lseek fs/read_write.c:285 [inline] [<ffffff90084ef78c>] SyS_lseek+0x164/0x1f0 fs/read_write.c:276 [<ffffff9008093c80>] el0_svc_naked+0x30/0x34 ================================================================== Signed-off-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: zhangyi (F) <yi.zhang(a)huawei.com> (cherry picked from commit a422358aa04c53a08b215b8dcd6814d916ef5cf1) Conflicts: fs/read_write.c Signed-off-by: Li Ming <limingming.li(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/proc/base.c | 32 ++++++++++++++++++++++++-------- fs/read_write.c | 5 ----- include/linux/fs.h | 4 ++++ 3 files changed, 28 insertions(+), 13 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index ffd54617c354..d0ae6e4511aa 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -903,18 +903,34 @@ static ssize_t mem_write(struct file *file, const char __user *buf, loff_t mem_lseek(struct file *file, loff_t offset, int orig) { + loff_t ret = 0; + + spin_lock(&file->f_lock); switch (orig) { - case 0: - file->f_pos = offset; - break; - case 1: - file->f_pos += offset; + case SEEK_CUR: + offset += file->f_pos; + /* fall through */ + case SEEK_SET: + /* to avoid userland mistaking f_pos=-9 as -EBADF=-9 */ + if ((unsigned long long)offset >= -MAX_ERRNO) + ret = -EOVERFLOW; break; default: - return -EINVAL; + ret = -EINVAL; + } + + if (!ret) { + if (offset < 0 && !(unsigned_offsets(file))) { + ret = -EINVAL; + } else { + file->f_pos = offset; + ret = file->f_pos; + force_successful_syscall_return(); + } } - force_successful_syscall_return(); - return file->f_pos; + + spin_unlock(&file->f_lock); + return ret; } static int mem_release(struct inode *inode, struct file *file) diff --git a/fs/read_write.c b/fs/read_write.c index 4771701c896b..2f456d5a1df5 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -34,11 +34,6 @@ const struct file_operations generic_ro_fops = { EXPORT_SYMBOL(generic_ro_fops); -static inline bool unsigned_offsets(struct file *file) -{ - return file->f_mode & FMODE_UNSIGNED_OFFSET; -} - /** * vfs_setpos - update the file offset for lseek * @file: file structure in question diff --git a/include/linux/fs.h b/include/linux/fs.h index 4f6b15e960ec..59d4edee1b96 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2932,6 +2932,10 @@ extern long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, extern void file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping); extern loff_t noop_llseek(struct file *file, loff_t offset, int whence); +static inline bool unsigned_offsets(struct file *file) +{ + return file->f_mode & FMODE_UNSIGNED_OFFSET; +} #define no_llseek NULL extern loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize); extern loff_t generic_file_llseek(struct file *file, loff_t offset, int whence); -- 2.39.2

2 1

[PATCH OLK-6.6] fs/dcache.c: avoid softlock since too many negative dentry
by Zizhi Wo 25 Nov '23

25 Nov '23

From: yangerkun <yangerkun(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JH5M?from=project-issue --------------------------- Parallel thread to add negative dentry under root dir. Sometimes later, 'systemctl daemon-reload' will report softlockup since __fsnotify_update_child_dentry_flags need update all child under root dentry without distinguish does it active or not. It will waste so long time with catching d_lock of root dentry. And other thread try to spin_lock d_lock will run overtime. Limit negative dentry under dir can avoid this. Signed-off-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: Miao Xie <miaoxie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Conflicts: fs/dcache.c include/linux/dcache.h Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/dcache.c | 43 ++++++++++++++++++++++++++++++++++++++++-- include/linux/dcache.h | 4 ++++ 2 files changed, 45 insertions(+), 2 deletions(-) diff --git a/fs/dcache.c b/fs/dcache.c index 7cdd8bbedd1c..c27aa8ff78f4 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -341,10 +341,18 @@ static inline void __d_set_inode_and_type(struct dentry *dentry, unsigned type_flags) { unsigned flags; + struct dentry *parent; + + parent = dentry->d_parent; + if ((dentry->d_flags & DCACHE_NEGATIVE_ACCOUNT) && parent) { + WARN_ON(!inode); + atomic_dec(&parent->d_neg_dnum); + } dentry->d_inode = inode; flags = READ_ONCE(dentry->d_flags); - flags &= ~(DCACHE_ENTRY_TYPE | DCACHE_FALLTHRU); + flags &= ~(DCACHE_ENTRY_TYPE | DCACHE_FALLTHRU | + DCACHE_NEGATIVE_ACCOUNT); flags |= type_flags; smp_store_release(&dentry->d_flags, flags); } @@ -363,6 +371,7 @@ static inline void __d_clear_type_and_inode(struct dentry *dentry) static void dentry_free(struct dentry *dentry) { WARN_ON(!hlist_unhashed(&dentry->d_u.d_alias)); + WARN_ON(dentry->d_flags & DCACHE_NEGATIVE_ACCOUNT); if (unlikely(dname_external(dentry))) { struct external_name *p = external_name(dentry); if (likely(atomic_dec_and_test(&p->u.count))) { @@ -601,8 +610,14 @@ static void __dentry_kill(struct dentry *dentry) /* if it was on the hash then remove it */ __d_drop(dentry); dentry_unlist(dentry, parent); - if (parent) + if (parent) { + if (dentry->d_flags & DCACHE_NEGATIVE_ACCOUNT) { + atomic_dec(&parent->d_neg_dnum); + dentry->d_flags &= ~DCACHE_NEGATIVE_ACCOUNT; + } + spin_unlock(&parent->d_lock); + } if (dentry->d_inode) dentry_unlink_inode(dentry); else @@ -662,6 +677,8 @@ static inline struct dentry *lock_parent(struct dentry *dentry) static inline bool retain_dentry(struct dentry *dentry) { + struct dentry *parent; + WARN_ON(d_in_lookup(dentry)); /* Unreachable? Get rid of it */ @@ -679,6 +696,27 @@ static inline bool retain_dentry(struct dentry *dentry) if (unlikely(dentry->d_flags & DCACHE_DONTCACHE)) return false; + if (unlikely(!dentry->d_parent)) + goto noparent; + + parent = dentry->d_parent; + /* Return false if it's negative */ + WARN_ON((atomic_read(&parent->d_neg_dnum) < 0)); + if (!dentry->d_inode) { + if (!(dentry->d_flags & DCACHE_NEGATIVE_ACCOUNT)) { + unsigned int flags = READ_ONCE(dentry->d_flags); + + flags |= DCACHE_NEGATIVE_ACCOUNT; + WRITE_ONCE(dentry->d_flags, flags); + atomic_inc(&parent->d_neg_dnum); + } + } + + if (!dentry->d_inode && + atomic_read(&parent->d_neg_dnum) >= NEG_DENTRY_LIMIT) + return false; + +noparent: /* retain; LRU fodder */ dentry->d_lockref.count--; if (unlikely(!(dentry->d_flags & DCACHE_LRU_LIST))) @@ -1809,6 +1847,7 @@ static struct dentry *__d_alloc(struct super_block *sb, const struct qstr *name) seqcount_spinlock_init(&dentry->d_seq, &dentry->d_lock); dentry->d_inode = NULL; dentry->d_parent = dentry; + atomic_set(&dentry->d_neg_dnum, 0); dentry->d_sb = sb; dentry->d_op = NULL; dentry->d_fsdata = NULL; diff --git a/include/linux/dcache.h b/include/linux/dcache.h index 6b351e009f59..1c5b08d804ce 100644 --- a/include/linux/dcache.h +++ b/include/linux/dcache.h @@ -77,6 +77,7 @@ extern const struct qstr dotdot_name; # endif #endif +#define NEG_DENTRY_LIMIT 16384 #define d_lock d_lockref.lock struct dentry { @@ -111,6 +112,8 @@ struct dentry { struct hlist_bl_node d_in_lookup_hash; /* only for in-lookup ones */ struct rcu_head d_rcu; } d_u; + /* negative dentry under this dentry, if it's dir */ + atomic_t d_neg_dnum; } __randomize_layout; /* @@ -212,6 +215,7 @@ struct dentry_operations { #define DCACHE_PAR_LOOKUP 0x10000000 /* being looked up (with parent locked shared) */ #define DCACHE_DENTRY_CURSOR 0x20000000 #define DCACHE_NORCU 0x40000000 /* No RCU delay for freeing */ +#define DCACHE_NEGATIVE_ACCOUNT 0x80000000 extern seqlock_t rename_lock; -- 2.39.2

2 1

[PATCH OKLK-6.6] proc: fix ubsan warning in mem_lseek
by Zizhi Wo 25 Nov '23

25 Nov '23

From: yangerkun <yangerkun(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JE4E?from=project-issue --------------------------- UBSAN has reported a overflow with mem_lseek. And it's fine with mem_open set file mode with FMODE_UNSIGNED_OFFSET(memory_lseek). However, another file use mem_lseek do lseek can have not FMODE_UNSIGNED_OFFSET(proc_kpagecount_operations/proc_pagemap_operations), fix it by checking overflow and FMODE_UNSIGNED_OFFSET. Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> ================================================================== UBSAN: Undefined behaviour in ../fs/proc/base.c:941:15 signed integer overflow: 4611686018427387904 + 4611686018427387904 cannot be represented in type 'long long int' CPU: 4 PID: 4762 Comm: syz-executor.1 Not tainted 4.4.189 #3 Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015 Call trace: [<ffffff90080a5f28>] dump_backtrace+0x0/0x590 arch/arm64/kernel/traps.c:91 [<ffffff90080a64f0>] show_stack+0x38/0x60 arch/arm64/kernel/traps.c:234 [<ffffff9008986a34>] __dump_stack lib/dump_stack.c:15 [inline] [<ffffff9008986a34>] dump_stack+0x128/0x184 lib/dump_stack.c:51 [<ffffff9008a2d120>] ubsan_epilogue+0x34/0x9c lib/ubsan.c:166 [<ffffff9008a2d8b8>] handle_overflow+0x228/0x280 lib/ubsan.c:197 [<ffffff9008a2da2c>] __ubsan_handle_add_overflow+0x4c/0x68 lib/ubsan.c:204 [<ffffff900862b9f4>] mem_lseek+0x12c/0x130 fs/proc/base.c:941 [<ffffff90084ef78c>] vfs_llseek fs/read_write.c:260 [inline] [<ffffff90084ef78c>] SYSC_lseek fs/read_write.c:285 [inline] [<ffffff90084ef78c>] SyS_lseek+0x164/0x1f0 fs/read_write.c:276 [<ffffff9008093c80>] el0_svc_naked+0x30/0x34 ================================================================== Signed-off-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: zhangyi (F) <yi.zhang(a)huawei.com> (cherry picked from commit a422358aa04c53a08b215b8dcd6814d916ef5cf1) Conflicts: fs/read_write.c Signed-off-by: Li Ming <limingming.li(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/proc/base.c | 32 ++++++++++++++++++++++++-------- fs/read_write.c | 5 ----- include/linux/fs.h | 4 ++++ 3 files changed, 28 insertions(+), 13 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index ffd54617c354..d0ae6e4511aa 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -903,18 +903,34 @@ static ssize_t mem_write(struct file *file, const char __user *buf, loff_t mem_lseek(struct file *file, loff_t offset, int orig) { + loff_t ret = 0; + + spin_lock(&file->f_lock); switch (orig) { - case 0: - file->f_pos = offset; - break; - case 1: - file->f_pos += offset; + case SEEK_CUR: + offset += file->f_pos; + /* fall through */ + case SEEK_SET: + /* to avoid userland mistaking f_pos=-9 as -EBADF=-9 */ + if ((unsigned long long)offset >= -MAX_ERRNO) + ret = -EOVERFLOW; break; default: - return -EINVAL; + ret = -EINVAL; + } + + if (!ret) { + if (offset < 0 && !(unsigned_offsets(file))) { + ret = -EINVAL; + } else { + file->f_pos = offset; + ret = file->f_pos; + force_successful_syscall_return(); + } } - force_successful_syscall_return(); - return file->f_pos; + + spin_unlock(&file->f_lock); + return ret; } static int mem_release(struct inode *inode, struct file *file) diff --git a/fs/read_write.c b/fs/read_write.c index 4771701c896b..2f456d5a1df5 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -34,11 +34,6 @@ const struct file_operations generic_ro_fops = { EXPORT_SYMBOL(generic_ro_fops); -static inline bool unsigned_offsets(struct file *file) -{ - return file->f_mode & FMODE_UNSIGNED_OFFSET; -} - /** * vfs_setpos - update the file offset for lseek * @file: file structure in question diff --git a/include/linux/fs.h b/include/linux/fs.h index 4f6b15e960ec..59d4edee1b96 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2932,6 +2932,10 @@ extern long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, extern void file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping); extern loff_t noop_llseek(struct file *file, loff_t offset, int whence); +static inline bool unsigned_offsets(struct file *file) +{ + return file->f_mode & FMODE_UNSIGNED_OFFSET; +} #define no_llseek NULL extern loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize); extern loff_t generic_file_llseek(struct file *file, loff_t offset, int whence); -- 2.39.2

1 0

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 21

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 21

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH OKLK-6.6] proc: fix ubsan warning in mem_lseek
by Zizhi Wo 25 Nov '23

25 Nov '23

From: yangerkun <yangerkun(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JE4E?from=project-issue --------------------------- UBSAN has reported a overflow with mem_lseek. And it's fine with mem_open set file mode with FMODE_UNSIGNED_OFFSET(memory_lseek). However, another file use mem_lseek do lseek can have not FMODE_UNSIGNED_OFFSET(proc_kpagecount_operations/proc_pagemap_operations), fix it by checking overflow and FMODE_UNSIGNED_OFFSET. Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> ================================================================== UBSAN: Undefined behaviour in ../fs/proc/base.c:941:15 signed integer overflow: 4611686018427387904 + 4611686018427387904 cannot be represented in type 'long long int' CPU: 4 PID: 4762 Comm: syz-executor.1 Not tainted 4.4.189 #3 Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015 Call trace: [<ffffff90080a5f28>] dump_backtrace+0x0/0x590 arch/arm64/kernel/traps.c:91 [<ffffff90080a64f0>] show_stack+0x38/0x60 arch/arm64/kernel/traps.c:234 [<ffffff9008986a34>] __dump_stack lib/dump_stack.c:15 [inline] [<ffffff9008986a34>] dump_stack+0x128/0x184 lib/dump_stack.c:51 [<ffffff9008a2d120>] ubsan_epilogue+0x34/0x9c lib/ubsan.c:166 [<ffffff9008a2d8b8>] handle_overflow+0x228/0x280 lib/ubsan.c:197 [<ffffff9008a2da2c>] __ubsan_handle_add_overflow+0x4c/0x68 lib/ubsan.c:204 [<ffffff900862b9f4>] mem_lseek+0x12c/0x130 fs/proc/base.c:941 [<ffffff90084ef78c>] vfs_llseek fs/read_write.c:260 [inline] [<ffffff90084ef78c>] SYSC_lseek fs/read_write.c:285 [inline] [<ffffff90084ef78c>] SyS_lseek+0x164/0x1f0 fs/read_write.c:276 [<ffffff9008093c80>] el0_svc_naked+0x30/0x34 ================================================================== Signed-off-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: zhangyi (F) <yi.zhang(a)huawei.com> (cherry picked from commit a422358aa04c53a08b215b8dcd6814d916ef5cf1) Conflicts: fs/read_write.c Signed-off-by: Li Ming <limingming.li(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/proc/base.c | 32 ++++++++++++++++++++++++-------- fs/read_write.c | 5 ----- include/linux/fs.h | 4 ++++ 3 files changed, 28 insertions(+), 13 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index ffd54617c354..d0ae6e4511aa 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -903,18 +903,34 @@ static ssize_t mem_write(struct file *file, const char __user *buf, loff_t mem_lseek(struct file *file, loff_t offset, int orig) { + loff_t ret = 0; + + spin_lock(&file->f_lock); switch (orig) { - case 0: - file->f_pos = offset; - break; - case 1: - file->f_pos += offset; + case SEEK_CUR: + offset += file->f_pos; + /* fall through */ + case SEEK_SET: + /* to avoid userland mistaking f_pos=-9 as -EBADF=-9 */ + if ((unsigned long long)offset >= -MAX_ERRNO) + ret = -EOVERFLOW; break; default: - return -EINVAL; + ret = -EINVAL; + } + + if (!ret) { + if (offset < 0 && !(unsigned_offsets(file))) { + ret = -EINVAL; + } else { + file->f_pos = offset; + ret = file->f_pos; + force_successful_syscall_return(); + } } - force_successful_syscall_return(); - return file->f_pos; + + spin_unlock(&file->f_lock); + return ret; } static int mem_release(struct inode *inode, struct file *file) diff --git a/fs/read_write.c b/fs/read_write.c index 4771701c896b..2f456d5a1df5 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -34,11 +34,6 @@ const struct file_operations generic_ro_fops = { EXPORT_SYMBOL(generic_ro_fops); -static inline bool unsigned_offsets(struct file *file) -{ - return file->f_mode & FMODE_UNSIGNED_OFFSET; -} - /** * vfs_setpos - update the file offset for lseek * @file: file structure in question diff --git a/include/linux/fs.h b/include/linux/fs.h index 4f6b15e960ec..59d4edee1b96 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2932,6 +2932,10 @@ extern long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, extern void file_ra_state_init(struct file_ra_state *ra, struct address_space *mapping); extern loff_t noop_llseek(struct file *file, loff_t offset, int whence); +static inline bool unsigned_offsets(struct file *file) +{ + return file->f_mode & FMODE_UNSIGNED_OFFSET; +} #define no_llseek NULL extern loff_t vfs_setpos(struct file *file, loff_t offset, loff_t maxsize); extern loff_t generic_file_llseek(struct file *file, loff_t offset, int whence); -- 2.39.2

1 0

[PATCH OLK-5.10 0/2] fix CFS bandwidth vs. hrtimer self deadlock
by Yu Liao 25 Nov '23

25 Nov '23

Fix the following issue: CPU1 CPU2 CPU3 T1 sets cfs_quota starts hrtimer cfs_bandwidth 'period_timer' T1 is migrated to CPU3 T2(worker thread) initiates offlining of CPU1 Hotplug operation starts ... 'period_timer' expires and is re-enqueued on CPU1 ... take_cpu_down() CPU1 shuts down and does not handle timers anymore. They have to be migrated in the post dead hotplug steps by the control task. T2(worker thread) runs the post dead offline operation T1 holds lockA T1 is scheduled out //throttled by CFS bandwidth control T1 waits for 'period_timer' to expire T2(worker thread) waits for lockA T1 waits there forever if it is scheduled out before it can execute the hrtimer offline callback hrtimers_dead_cpu(). Thus T2 waits for lockA forever. Thomas Gleixner (1): hrtimers: Push pending hrtimers away from outgoing CPU earlier Yu Liao (1): cpu/hotplug: fix kabi breakage in enum cpuhp_state include/linux/hrtimer.h | 4 ++-- include/linux/smp.h | 1 + kernel/cpu.c | 17 +++++++++++++++-- kernel/smp.c | 8 ++++++++ kernel/time/hrtimer.c | 33 ++++++++++++--------------------- 5 files changed, 38 insertions(+), 25 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH OLK-6.6] fs/dcache.c: avoid panic while lockref of dentry overflow
by Zizhi Wo 25 Nov '23

25 Nov '23

From: yangerkun <yangerkun(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JDP6?from=project-issue --------------------------- We use lockref for dentry reference without notice that so many negative dentry under one dir can lead to overflow of lockref. This can lead to system crash if we do this under root dir. Since there is not a perfect solution, we just limit max number of dentry count up to INT_MAX / 2. Also, it will cost a lot of time from INT_MAX / 2 to INT_MAX, so we no need to do this under protection of dentry lock. Also, we limit the FILES_MAX to INT_MAX / 2, since a lot open for same file can lead to overflow too. Changelog: v1->v2: add a function to do check / add a Macro to mean INT_MAX / 2 Signed-off-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: Miao Xie <miaoxie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Conflicts: fs/dcache.c Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/dcache.c | 32 ++++++++++++++++++++++++++++---- include/linux/fs.h | 3 +++ 2 files changed, 31 insertions(+), 4 deletions(-) diff --git a/fs/dcache.c b/fs/dcache.c index 25ac74d30bff..7cdd8bbedd1c 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -1834,6 +1834,18 @@ static struct dentry *__d_alloc(struct super_block *sb, const struct qstr *name) return dentry; } +static inline bool d_forbid_overflow(struct dentry *dentry) +{ + if (unlikely(d_count(dentry) >= D_COUNT_MAX)) { + shrink_dcache_parent(dentry); + + if (d_count(dentry) >= D_COUNT_MAX) + return false; + } + + return true; +} + /** * d_alloc - allocate a dcache entry * @parent: parent of entry to allocate @@ -1845,9 +1857,15 @@ static struct dentry *__d_alloc(struct super_block *sb, const struct qstr *name) */ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name) { - struct dentry *dentry = __d_alloc(parent->d_sb, name); + struct dentry *dentry = NULL; + + if (unlikely(!d_forbid_overflow(parent))) + goto out; + + dentry = __d_alloc(parent->d_sb, name); if (!dentry) - return NULL; + goto out; + spin_lock(&parent->d_lock); /* * don't need child lock because it is not subject @@ -1857,7 +1875,7 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name) dentry->d_parent = parent; list_add(&dentry->d_child, &parent->d_subdirs); spin_unlock(&parent->d_lock); - +out: return dentry; } EXPORT_SYMBOL(d_alloc); @@ -1870,11 +1888,17 @@ EXPORT_SYMBOL(d_alloc_anon); struct dentry *d_alloc_cursor(struct dentry * parent) { - struct dentry *dentry = d_alloc_anon(parent->d_sb); + struct dentry *dentry = NULL; + + if (unlikely(!d_forbid_overflow(parent))) + goto out; + + dentry = d_alloc_anon(parent->d_sb); if (dentry) { dentry->d_flags |= DCACHE_DENTRY_CURSOR; dentry->d_parent = dget(parent); } +out: return dentry; } diff --git a/include/linux/fs.h b/include/linux/fs.h index 4a40823c3c67..4f6b15e960ec 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -47,6 +47,9 @@ #include <asm/byteorder.h> #include <uapi/linux/fs.h> +#define D_COUNT_MAX (INT_MAX / 2) + + struct backing_dev_info; struct bdi_writeback; struct bio; -- 2.39.2

2 1

[PATCH OLK-6.6] hugetlbfs: avoid overflow in hugetlbfs_fallocate
by Zizhi Wo 25 Nov '23

25 Nov '23

From: yangerkun <yangerkun(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8JDQ7?from=project-issue --------------------------- luojiajun report a problem[1] two years ago which seems still exists in mainline. vfs_fallocate can avoid 'offset + len' trigger overflow, but 'offset + len + hpage_size - 1' may overflow too and will lead to a wrong 'end'. luojiajun give a solution which can fix the wrong 'end' but leave the overflow still happened. Fix it with DIV_ROUND_UP_ULL. [1] https://patchwork.kernel.org/project/linux-mm/patch/1554775226-67213-1-git-… Signed-off-by: yangerkun <yangerkun(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/hugetlbfs/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 316c4cebd3f3..cb6a5ea74cda 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -836,7 +836,7 @@ static long hugetlbfs_fallocate(struct file *file, int mode, loff_t offset, * as well as being converted to page offsets. */ start = offset >> hpage_shift; - end = (offset + len + hpage_size - 1) >> hpage_shift; + end = DIV_ROUND_UP_ULL(offset + len, hpage_size); inode_lock(inode); -- 2.39.2

2 1

[PATCH openEuler-22.03-LTS 0/4] Add script to check & update openeuler_defconfig
by Jialin Zhang 25 Nov '23

25 Nov '23

Jialin Zhang (3): Revert "kconfig: Add script to update openeuler_defconfig" config: update openeuler_defconfig for x86 config: update openeuler_defconfig for arm64 Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig arch/arm64/configs/openeuler_defconfig | 32 ++++++++++++++++++-------- arch/x86/configs/openeuler_defconfig | 12 ++++++---- scripts/kconfig/Makefile | 14 +++++++++++ scripts/kconfig/makeconfig.sh | 5 ++++ 4 files changed, 49 insertions(+), 14 deletions(-) -- 2.25.1

2 5

[PATCH openEuler-22.03-LTS-SP1 0/4] Add script to check & update openeuler_defconfig
by Jialin Zhang 25 Nov '23

25 Nov '23

Jialin Zhang (3): Revert "kconfig: Add script to update openeuler_defconfig" config: update openeuler_defconfig for x86 config: update openeuler_defconfig for arm64 Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig arch/arm64/configs/openeuler_defconfig | 26 ++++++++++++++++++-------- arch/x86/configs/openeuler_defconfig | 8 +++----- scripts/kconfig/Makefile | 14 ++++++++++++++ scripts/kconfig/makeconfig.sh | 5 +++++ 4 files changed, 40 insertions(+), 13 deletions(-) -- 2.25.1

2 5

[PATCH openEuler-22.03-LTS-SP2 0/4] Add script to check & update openeuler_defconfig
by Jialin Zhang 25 Nov '23

25 Nov '23

Jialin Zhang (3): Revert "kconfig: Add script to update openeuler_defconfig" config: update openeuler_defconfig for x86 config: update openeuler_defconfig for arm64 Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig arch/arm64/configs/openeuler_defconfig | 51 +++++++++++++++++--------- arch/x86/configs/openeuler_defconfig | 32 ++++++++++------ scripts/kconfig/Makefile | 14 +++++++ scripts/kconfig/makeconfig.sh | 5 +++ 4 files changed, 74 insertions(+), 28 deletions(-) -- 2.25.1

2 5

[PATCH openEuler-1.0-LTS] x86/mce/inject: Fix a wrong assignment of i_mce.status
by liwei 25 Nov '23

25 Nov '23

From: Zhenzhong Duan <zhenzhong.duan(a)gmail.com> mainline inclusion from mainline-v5.9-rc1 commit 5d7f7d1d5e01c22894dee7c9c9266500478dca99 category: bugfix bugzilla: 189383 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- The original code is a nop as i_mce.status is or'ed with part of itself, fix it. Fixes: a1300e505297 ("x86/ras/mce_amd_inj: Trigger deferred and thresholding errors interrupts") Signed-off-by: Zhenzhong Duan <zhenzhong.duan(a)gmail.com> Signed-off-by: Borislav Petkov <bp(a)suse.de> Acked-by: Yazen Ghannam <yazen.ghannam(a)amd.com> Link: https://lkml.kernel.org/r/20200611023238.3830-1-zhenzhong.duan@gmail.com Signed-off-by: liwei <liwei728(a)huawei.com> --- arch/x86/kernel/cpu/mce/inject.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kernel/cpu/mce/inject.c b/arch/x86/kernel/cpu/mce/inject.c index 3f82afd0f46f..f21f3f26c555 100644 --- a/arch/x86/kernel/cpu/mce/inject.c +++ b/arch/x86/kernel/cpu/mce/inject.c @@ -518,7 +518,7 @@ static void do_inject(void) */ if (inj_type == DFR_INT_INJ) { i_mce.status |= MCI_STATUS_DEFERRED; - i_mce.status |= (i_mce.status & ~MCI_STATUS_UC); + i_mce.status &= ~MCI_STATUS_UC; } /* -- 2.25.1

2 1

[PATCH OLK-6.6] io_uring: fix soft lockup in io_submit_sqes()
by Zizhi Wo 25 Nov '23

25 Nov '23

From: Guo Xuenan <guoxuenan(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I5930C ------------------- when set up sq ring size with IORING_MAX_ENTRIES, io_submit_sqes may looping ~32768 times which may trigger soft lockups. add cond_resched condition to avoid this bad situation. set sq ring size 32768 to perform stress test as follows: watchdog: BUG: soft lockup - CPU#3 stuck for 26s! [poc:691] Modules linked in: CPU: 3 PID: 691 Comm: poc Not tainted 5.18.0+ #9 Hardware name: linux,dummy-virt (DT) pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--) pc : arch_local_irq_enable+0xc/0x28 lr : io_issue_sqe+0x870/0x28e8 sp : ffff80000e0f7800 x29: ffff80000e0f7800 x28: ffff0000cf850dd0 x27: ffff0000cf2c2000 x26: ffff8000096f4b20 x25: ffff0000cd769c00 x24: ffff0000c12b6650 x23: ffff800009dad958 x22: 00000006fc23ac00 x21: ffff0000cd769c08 x20: 1ffff00001c1ef1a x19: ffff0000cd767e00 x18: 0000000000000000 x17: ffff800008032b74 x16: ffff800008636448 x15: 0000fffff7166568 x14: ffff80000861edf0 x13: ffff600019e58449 x12: 1fffe00019e58448 x11: 1fffe00019e58448 x10: ffff600019e58448 x9 : dfff800000000000 x8 : ffff0000cf2c2244 x7 : 0000000000000001 x6 : ffff600019e58449 x5 : ffff600019e58449 x4 : ffff600019e58449 x3 : ffff8000086306c0 x2 : 0000000000000001 x1 : ffff0000cf2c2244 x0 : 00000000000000e0 Call trace: arch_local_irq_enable+0xc/0x28 io_submit_sqes+0x530/0x29d8 __arm64_sys_io_uring_enter+0x380/0xd18 invoke_syscall+0x64/0x180 el0_svc_common.constprop.2+0x178/0x208 do_el0_svc+0x84/0xa0 el0_svc+0x48/0x1a0 el0t_64_sync_handler+0x90/0xb8 el0t_64_sync+0x180/0x184 Kernel panic - not syncing: softlockup: hung tasks CPU: 3 PID: 691 Comm: poc Tainted: G L 5.18.0+ #9 Hardware name: linux,dummy-virt (DT) Call trace: dump_backtrace+0x218/0x228 show_stack+0x20/0x68 dump_stack_lvl+0x68/0x84 dump_stack+0x1c/0x38 panic+0x1ec/0x3ec watchdog_timer_fn+0x28c/0x300 __hrtimer_run_queues+0x1d8/0x498 hrtimer_interrupt+0x238/0x558 arch_timer_handler_virt+0x48/0x60 handle_percpu_devid_irq+0xdc/0x270 generic_handle_domain_irq+0x50/0x70 gic_handle_irq+0x8c/0x4bc call_on_irq_stack+0x2c/0x38 do_interrupt_handler+0xc4/0xc8 el1_interrupt+0x48/0xb0 el1h_64_irq_handler+0x18/0x28 el1h_64_irq+0x74/0x78 arch_local_irq_enable+0xc/0x28 io_submit_sqes+0x530/0x29d8 __arm64_sys_io_uring_enter+0x380/0xd18 invoke_syscall+0x64/0x180 el0_svc_common.constprop.2+0x178/0x208 do_el0_svc+0x84/0xa0 el0_svc+0x48/0x1a0 el0t_64_sync_handler+0x90/0xb8 el0t_64_sync+0x180/0x184 SMP: stopping secondary CPUs Kernel Offset: disabled CPU features: 0x110,00008f09,00001006 Memory Limit: none ---[ end Kernel panic - not syncing: softlockup: hung tasks ]--- Link: https://lore.kernel.org/all/d4bc3afb-02d5-1793-cffa-e15b2bdb0028@huawei.com/ Signed-off-by: Guo Xuenan <guoxuenan(a)huawei.com> Conflict: io_uring/io_uring.c Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Reviewed-by: Wang Weiyang <wangweiyang2(a)huawei.com> Signed-off-by: Jialin Zhang <zhangjialin11(a)huawei.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- io_uring/io_uring.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c index 8d1bc6cdfe71..d79e78971ec5 100644 --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -2440,6 +2440,9 @@ int io_submit_sqes(struct io_ring_ctx *ctx, unsigned int nr) left--; break; } + /* to avoid doing too much in one submit round */ + if (left > IORING_MAX_ENTRIES / 2) + cond_resched(); } while (--left); if (unlikely(left)) { -- 2.39.2

2 3

[PATCH openEuler-1.0-LTS] x86/mce/amd: Publish the bank pointer only after setup has succeeded
by liwei 25 Nov '23

25 Nov '23

From: Borislav Petkov <bp(a)suse.de> mainline inclusion from mainline-v5.6-rc3 commit 6e5cf31fbe651bed7ba1df768f2e123531132417 category: bugfix bugzilla: 189381 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- threshold_create_bank() creates a bank descriptor per MCA error thresholding counter which can be controlled over sysfs. It publishes the pointer to that bank in a per-CPU variable and then goes on to create additional thresholding blocks if the bank has such. However, that creation of additional blocks in allocate_threshold_blocks() can fail, leading to a use-after-free through the per-CPU pointer. Therefore, publish that pointer only after all blocks have been setup successfully. Fixes: 019f34fccfd5 ("x86, MCE, AMD: Move shared bank to node descriptor") Reported-by: Saar Amar <Saar.Amar(a)microsoft.com> Reported-by: Dan Carpenter <dan.carpenter(a)oracle.com> Signed-off-by: Borislav Petkov <bp(a)suse.de> Cc: <stable(a)vger.kernel.org> Link: http://lkml.kernel.org/r/20200128140846.phctkvx5btiexvbx@kili.mountain Signed-off-by: liwei <liwei728(a)huawei.com> --- arch/x86/kernel/cpu/mce/amd.c | 33 ++++++++++++++++----------------- 1 file changed, 16 insertions(+), 17 deletions(-) diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c index 38b18fa055ef..af4a83695970 100644 --- a/arch/x86/kernel/cpu/mce/amd.c +++ b/arch/x86/kernel/cpu/mce/amd.c @@ -1200,8 +1200,9 @@ static const char *get_name(unsigned int bank, struct threshold_block *b) return buf_mcatype; } -static int allocate_threshold_blocks(unsigned int cpu, unsigned int bank, - unsigned int block, u32 address) +static int allocate_threshold_blocks(unsigned int cpu, struct threshold_bank *tb, + unsigned int bank, unsigned int block, + u32 address) { struct threshold_block *b = NULL; u32 low, high; @@ -1245,16 +1246,12 @@ static int allocate_threshold_blocks(unsigned int cpu, unsigned int bank, INIT_LIST_HEAD(&b->miscj); - if (per_cpu(threshold_banks, cpu)[bank]->blocks) { - list_add(&b->miscj, - &per_cpu(threshold_banks, cpu)[bank]->blocks->miscj); - } else { - per_cpu(threshold_banks, cpu)[bank]->blocks = b; - } + if (tb->blocks) + list_add(&b->miscj, &tb->blocks->miscj); + else + tb->blocks = b; - err = kobject_init_and_add(&b->kobj, &threshold_ktype, - per_cpu(threshold_banks, cpu)[bank]->kobj, - get_name(bank, b)); + err = kobject_init_and_add(&b->kobj, &threshold_ktype, tb->kobj, get_name(bank, b)); if (err) goto out_free; recurse: @@ -1262,7 +1259,7 @@ static int allocate_threshold_blocks(unsigned int cpu, unsigned int bank, if (!address) return 0; - err = allocate_threshold_blocks(cpu, bank, block, address); + err = allocate_threshold_blocks(cpu, tb, bank, block, address); if (err) goto out_free; @@ -1347,8 +1344,6 @@ static int threshold_create_bank(unsigned int cpu, unsigned int bank) goto out_free; } - per_cpu(threshold_banks, cpu)[bank] = b; - if (is_shared_bank(bank)) { refcount_set(&b->cpus, 1); @@ -1359,9 +1354,13 @@ static int threshold_create_bank(unsigned int cpu, unsigned int bank) } } - err = allocate_threshold_blocks(cpu, bank, 0, msr_ops.misc(bank)); - if (!err) - goto out; + err = allocate_threshold_blocks(cpu, b, bank, 0, msr_ops.misc(bank)); + if (err) + goto out_free; + + per_cpu(threshold_banks, cpu)[bank] = b; + + return 0; out_free: kfree(b); -- 2.25.1

2 1

[PATCH OLK-5.10 0/4] Add script to check & update openeuler_defconfig
by Jialin Zhang 25 Nov '23

25 Nov '23

Jialin Zhang (3): Revert "kconfig: Add script to update openeuler_defconfig" config: update openeuler_defconfig for x86 config: update openeuler_defconfig for arm64 Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig arch/arm64/configs/openeuler_defconfig | 60 +++++++++++++++++--------- arch/x86/configs/openeuler_defconfig | 39 +++++++++++------ scripts/kconfig/Makefile | 14 ++++++ scripts/kconfig/makeconfig.sh | 5 +++ 4 files changed, 86 insertions(+), 32 deletions(-) -- 2.25.1

2 5

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 25 Nov '23

25 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 v7 00/20] introduce smart_grid zone
by Yipeng Zou 25 Nov '23

25 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v6: 1. check is timer active in affinity timer Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: check is active in affinity timer arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 500 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1605 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH openEuler-1.0-LTS] umh: fix memory leak on execve failure
by Wenyu Huang 25 Nov '23

25 Nov '23

From: Vincent Minet <v.minet(a)criteo.com> mainline inclusion from mainline commit db803036ada7 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J9MA CVE: NA ------------------------------------------------- If a UMH process created by fork_usermode_blob() fails to execute, a pair of struct file allocated by umh_pipe_setup() will leak. Under normal conditions, the caller (like bpfilter) needs to manage the lifetime of the UMH and its two pipes. But when fork_usermode_blob() fails, the caller doesn't really have a way to know what needs to be done. It seems better to do the cleanup ourselves in this case. Fixes: 449325b52b7a ("umh: introduce fork_usermode_blob() helper") Signed-off-by: Vincent Minet <v.minet(a)criteo.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Wenyu Huang <huangwenyu5(a)huawei.com> --- kernel/umh.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/umh.c b/kernel/umh.c index 53611efb10cb..715b4368f9ed 100644 --- a/kernel/umh.c +++ b/kernel/umh.c @@ -495,6 +495,12 @@ static void umh_clean_and_save_pid(struct subprocess_info *info) { struct umh_info *umh_info = info->data; + /* cleanup if umh_pipe_setup() was successful but exec failed */ + if (info->pid && info->retval) { + fput(umh_info->pipe_to_umh); + fput(umh_info->pipe_from_umh); + } + argv_free(info->argv); umh_info->pid = info->pid; } -- 2.34.1

2 1

[PATCH OLK-6.6 v3] net: sched: sch_qfq: Use non-work-conserving warning handler
by Zhengchao Shao 25 Nov '23

25 Nov '23

From: Liu Jian <liujian56(a)huawei.com> mainline inclusion from mainline-v6.7-rc1 commit 6d25d1dc76bf5943a5c1f4bb74d66d5eac58eb77 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J4OL Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- A helper function for printing non-work-conserving alarms is added in commit b00355db3f88 ("pkt_sched: sch_hfsc: sch_htb: Add non-work-conserving warning handler."). In this commit, use qdisc_warn_nonwc() instead of WARN_ONCE() to handle the non-work-conserving warning in qfq Qdisc. Signed-off-by: Liu Jian <liujian56(a)huawei.com> Link: https://lore.kernel.org/r/20231023064729.370649-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/sched/sch_qfq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index 546c10adcacd..5598f8be18ae 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -1003,7 +1003,7 @@ static inline struct sk_buff *qfq_peek_skb(struct qfq_aggregate *agg, *cl = list_first_entry(&agg->active, struct qfq_class, alist); skb = (*cl)->qdisc->ops->peek((*cl)->qdisc); if (skb == NULL) - WARN_ONCE(1, "qfq_dequeue: non-workconserving leaf\n"); + qdisc_warn_nonwc("qfq_dequeue", (*cl)->qdisc); else *len = qdisc_pkt_len(skb); -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS 00/17] Add error handle for add_disk
by Li Nan 25 Nov '23

25 Nov '23

Christoph Hellwig (5): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference Luis Chamberlain (4): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling Tetsuo Handa (1): block: check minor range in device_add_disk() Yu Kuai (1): block: fix memory leak for elevator on add_disk failure Zhong Jinghua (6): block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path block/blk.h | 5 +- include/linux/genhd.h | 13 +++ block/blk-integrity.c | 12 ++- block/genhd.c | 246 +++++++++++++++++++++++++----------------- 4 files changed, 173 insertions(+), 103 deletions(-) -- 2.39.2

2 19

[PATCH openEuler-1.0-LTS] umh: fix memory leak on execve failure
by Wenyu Huang 25 Nov '23

25 Nov '23

From: Vincent Minet <v.minet(a)criteo.com> mainline inclusion from mainline-v5.10 commit db803036ada7 ("umh: fix memory leak on execve failure") category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J9MA CVE: NA ------------------------------------------------- If a UMH process created by fork_usermode_blob() fails to execute, a pair of struct file allocated by umh_pipe_setup() will leak. Under normal conditions, the caller (like bpfilter) needs to manage the lifetime of the UMH and its two pipes. But when fork_usermode_blob() fails, the caller doesn't really have a way to know what needs to be done. It seems better to do the cleanup ourselves in this case. Fixes: 449325b52b7a ("umh: introduce fork_usermode_blob() helper") Signed-off-by: Vincent Minet <v.minet(a)criteo.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Wenyu Huang <huangwenyu5(a)huawei.com> --- kernel/umh.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/umh.c b/kernel/umh.c index 53611efb10cb..715b4368f9ed 100644 --- a/kernel/umh.c +++ b/kernel/umh.c @@ -495,6 +495,12 @@ static void umh_clean_and_save_pid(struct subprocess_info *info) { struct umh_info *umh_info = info->data; + /* cleanup if umh_pipe_setup() was successful but exec failed */ + if (info->pid && info->retval) { + fput(umh_info->pipe_to_umh); + fput(umh_info->pipe_from_umh); + } + argv_free(info->argv); umh_info->pid = info->pid; } -- 2.34.1

2 1

[PATCH OLK-6.6 0/1] checkpatch: Update link tags to fix ci warning
by Zheng Zengkai 24 Nov '23

24 Nov '23

Update link tags in checkpatch.pl script to fix openEuler kernel ci warning: WARNING:COMMIT_LOG_USE_LINK: Unknown link reference 'bugzilla:', use 'Link:' or 'Closes:' instead. Wei Li (1): checkpatch: Update check of link tags scripts/checkpatch.pl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.20.1

2 2

[PATCH] crypto: hisilicon/qm - remove incorrect type cast
by Zhiqi Song 24 Nov '23

24 Nov '23

From: Weili Qian <qianweili(a)huawei.com> maillist inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8IVG6 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- The 'offset' type is unsigned long in 'struct debugfs_reg32', so type of values casts to unsigned long long is incorrect, and the values do not require type cast, remove them. Signed-off-by: Weili Qian <qianweili(a)huawei.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: JiangShui Yang <yangjiangshui(a)h-partners.com> --- drivers/crypto/hisilicon/debugfs.c | 50 +++++++++++------------ drivers/crypto/hisilicon/zip/zip_main.c | 54 ++++++++++++------------- 2 files changed, 52 insertions(+), 52 deletions(-) diff --git a/drivers/crypto/hisilicon/debugfs.c b/drivers/crypto/hisilicon/debugfs.c index 88885268851b..848195515de8 100644 --- a/drivers/crypto/hisilicon/debugfs.c +++ b/drivers/crypto/hisilicon/debugfs.c @@ -53,34 +53,34 @@ static struct qm_dfx_item qm_dfx_files[] = { #define CNT_CYC_REGS_NUM 10 static const struct debugfs_reg32 qm_dfx_regs[] = { /* XXX_CNT are reading clear register */ - {"QM_ECC_1BIT_CNT ", 0x104000ull}, - {"QM_ECC_MBIT_CNT ", 0x104008ull}, - {"QM_DFX_MB_CNT ", 0x104018ull}, - {"QM_DFX_DB_CNT ", 0x104028ull}, - {"QM_DFX_SQE_CNT ", 0x104038ull}, - {"QM_DFX_CQE_CNT ", 0x104048ull}, - {"QM_DFX_SEND_SQE_TO_ACC_CNT ", 0x104050ull}, - {"QM_DFX_WB_SQE_FROM_ACC_CNT ", 0x104058ull}, - {"QM_DFX_ACC_FINISH_CNT ", 0x104060ull}, - {"QM_DFX_CQE_ERR_CNT ", 0x1040b4ull}, - {"QM_DFX_FUNS_ACTIVE_ST ", 0x200ull}, - {"QM_ECC_1BIT_INF ", 0x104004ull}, - {"QM_ECC_MBIT_INF ", 0x10400cull}, - {"QM_DFX_ACC_RDY_VLD0 ", 0x1040a0ull}, - {"QM_DFX_ACC_RDY_VLD1 ", 0x1040a4ull}, - {"QM_DFX_AXI_RDY_VLD ", 0x1040a8ull}, - {"QM_DFX_FF_ST0 ", 0x1040c8ull}, - {"QM_DFX_FF_ST1 ", 0x1040ccull}, - {"QM_DFX_FF_ST2 ", 0x1040d0ull}, - {"QM_DFX_FF_ST3 ", 0x1040d4ull}, - {"QM_DFX_FF_ST4 ", 0x1040d8ull}, - {"QM_DFX_FF_ST5 ", 0x1040dcull}, - {"QM_DFX_FF_ST6 ", 0x1040e0ull}, - {"QM_IN_IDLE_ST ", 0x1040e4ull}, + {"QM_ECC_1BIT_CNT ", 0x104000}, + {"QM_ECC_MBIT_CNT ", 0x104008}, + {"QM_DFX_MB_CNT ", 0x104018}, + {"QM_DFX_DB_CNT ", 0x104028}, + {"QM_DFX_SQE_CNT ", 0x104038}, + {"QM_DFX_CQE_CNT ", 0x104048}, + {"QM_DFX_SEND_SQE_TO_ACC_CNT ", 0x104050}, + {"QM_DFX_WB_SQE_FROM_ACC_CNT ", 0x104058}, + {"QM_DFX_ACC_FINISH_CNT ", 0x104060}, + {"QM_DFX_CQE_ERR_CNT ", 0x1040b4}, + {"QM_DFX_FUNS_ACTIVE_ST ", 0x200}, + {"QM_ECC_1BIT_INF ", 0x104004}, + {"QM_ECC_MBIT_INF ", 0x10400c}, + {"QM_DFX_ACC_RDY_VLD0 ", 0x1040a0}, + {"QM_DFX_ACC_RDY_VLD1 ", 0x1040a4}, + {"QM_DFX_AXI_RDY_VLD ", 0x1040a8}, + {"QM_DFX_FF_ST0 ", 0x1040c8}, + {"QM_DFX_FF_ST1 ", 0x1040cc}, + {"QM_DFX_FF_ST2 ", 0x1040d0}, + {"QM_DFX_FF_ST3 ", 0x1040d4}, + {"QM_DFX_FF_ST4 ", 0x1040d8}, + {"QM_DFX_FF_ST5 ", 0x1040dc}, + {"QM_DFX_FF_ST6 ", 0x1040e0}, + {"QM_IN_IDLE_ST ", 0x1040e4}, }; static const struct debugfs_reg32 qm_vf_dfx_regs[] = { - {"QM_DFX_FUNS_ACTIVE_ST ", 0x200ull}, + {"QM_DFX_FUNS_ACTIVE_ST ", 0x200}, }; /* define the QM's dfx regs region and region length */ diff --git a/drivers/crypto/hisilicon/zip/zip_main.c b/drivers/crypto/hisilicon/zip/zip_main.c index e8dbf7759c80..944afeb9da73 100644 --- a/drivers/crypto/hisilicon/zip/zip_main.c +++ b/drivers/crypto/hisilicon/zip/zip_main.c @@ -285,28 +285,28 @@ static const u64 core_offsets[] = { }; static const struct debugfs_reg32 hzip_dfx_regs[] = { - {"HZIP_GET_BD_NUM ", 0x00ull}, - {"HZIP_GET_RIGHT_BD ", 0x04ull}, - {"HZIP_GET_ERROR_BD ", 0x08ull}, - {"HZIP_DONE_BD_NUM ", 0x0cull}, - {"HZIP_WORK_CYCLE ", 0x10ull}, - {"HZIP_IDLE_CYCLE ", 0x18ull}, - {"HZIP_MAX_DELAY ", 0x20ull}, - {"HZIP_MIN_DELAY ", 0x24ull}, - {"HZIP_AVG_DELAY ", 0x28ull}, - {"HZIP_MEM_VISIBLE_DATA ", 0x30ull}, - {"HZIP_MEM_VISIBLE_ADDR ", 0x34ull}, - {"HZIP_CONSUMED_BYTE ", 0x38ull}, - {"HZIP_PRODUCED_BYTE ", 0x40ull}, - {"HZIP_COMP_INF ", 0x70ull}, - {"HZIP_PRE_OUT ", 0x78ull}, - {"HZIP_BD_RD ", 0x7cull}, - {"HZIP_BD_WR ", 0x80ull}, - {"HZIP_GET_BD_AXI_ERR_NUM ", 0x84ull}, - {"HZIP_GET_BD_PARSE_ERR_NUM ", 0x88ull}, - {"HZIP_ADD_BD_AXI_ERR_NUM ", 0x8cull}, - {"HZIP_DECOMP_STF_RELOAD_CURR_ST ", 0x94ull}, - {"HZIP_DECOMP_LZ77_CURR_ST ", 0x9cull}, + {"HZIP_GET_BD_NUM ", 0x00}, + {"HZIP_GET_RIGHT_BD ", 0x04}, + {"HZIP_GET_ERROR_BD ", 0x08}, + {"HZIP_DONE_BD_NUM ", 0x0c}, + {"HZIP_WORK_CYCLE ", 0x10}, + {"HZIP_IDLE_CYCLE ", 0x18}, + {"HZIP_MAX_DELAY ", 0x20}, + {"HZIP_MIN_DELAY ", 0x24}, + {"HZIP_AVG_DELAY ", 0x28}, + {"HZIP_MEM_VISIBLE_DATA ", 0x30}, + {"HZIP_MEM_VISIBLE_ADDR ", 0x34}, + {"HZIP_CONSUMED_BYTE ", 0x38}, + {"HZIP_PRODUCED_BYTE ", 0x40}, + {"HZIP_COMP_INF ", 0x70}, + {"HZIP_PRE_OUT ", 0x78}, + {"HZIP_BD_RD ", 0x7c}, + {"HZIP_BD_WR ", 0x80}, + {"HZIP_GET_BD_AXI_ERR_NUM ", 0x84}, + {"HZIP_GET_BD_PARSE_ERR_NUM ", 0x88}, + {"HZIP_ADD_BD_AXI_ERR_NUM ", 0x8c}, + {"HZIP_DECOMP_STF_RELOAD_CURR_ST ", 0x94}, + {"HZIP_DECOMP_LZ77_CURR_ST ", 0x9c}, }; static const struct debugfs_reg32 hzip_com_dfx_regs[] = { @@ -318,11 +318,11 @@ static const struct debugfs_reg32 hzip_com_dfx_regs[] = { }; static const struct debugfs_reg32 hzip_dump_dfx_regs[] = { - {"HZIP_GET_BD_NUM ", 0x00ull}, - {"HZIP_GET_RIGHT_BD ", 0x04ull}, - {"HZIP_GET_ERROR_BD ", 0x08ull}, - {"HZIP_DONE_BD_NUM ", 0x0cull}, - {"HZIP_MAX_DELAY ", 0x20ull}, + {"HZIP_GET_BD_NUM ", 0x00}, + {"HZIP_GET_RIGHT_BD ", 0x04}, + {"HZIP_GET_ERROR_BD ", 0x08}, + {"HZIP_DONE_BD_NUM ", 0x0c}, + {"HZIP_MAX_DELAY ", 0x20}, }; /* define the ZIP's dfx regs region and region length */ -- 2.30.0

1 0

[PATCH] crypto: hisilicon/qm - print device abnormal information
by Zhiqi Song 24 Nov '23

24 Nov '23

From: Weili Qian <qianweili(a)huawei.com> maillist inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8IVG6 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- When device is abnormal and reports abnormal interrupt event to driver, the driver can print device information for error analysis. This patch adds some device error-related information output after the device reports an abnormal interrupt. Signed-off-by: Weili Qian <qianweili(a)huawei.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: JiangShui Yang <yangjiangshui(a)h-partners.com> --- drivers/crypto/hisilicon/qm.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c index 8f99683bcdb5..b7eab69318f5 100644 --- a/drivers/crypto/hisilicon/qm.c +++ b/drivers/crypto/hisilicon/qm.c @@ -131,16 +131,21 @@ #define QM_FIFO_OVERFLOW_TYPE 0xc0 #define QM_FIFO_OVERFLOW_TYPE_SHIFT 6 #define QM_FIFO_OVERFLOW_VF 0x3f +#define QM_FIFO_OVERFLOW_QP_SHIFT 16 #define QM_ABNORMAL_INF01 0x100014 #define QM_DB_TIMEOUT_TYPE 0xc0 #define QM_DB_TIMEOUT_TYPE_SHIFT 6 #define QM_DB_TIMEOUT_VF 0x3f +#define QM_DB_TIMEOUT_QP_SHIFT 16 +#define QM_ABNORMAL_INF02 0x100018 +#define QM_AXI_POISON_ERR BIT(22) #define QM_RAS_CE_ENABLE 0x1000ec #define QM_RAS_FE_ENABLE 0x1000f0 #define QM_RAS_NFE_ENABLE 0x1000f4 #define QM_RAS_CE_THRESHOLD 0x1000f8 #define QM_RAS_CE_TIMES_PER_IRQ 1 #define QM_OOO_SHUTDOWN_SEL 0x1040f8 +#define QM_AXI_RRESP_ERR BIT(0) #define QM_ECC_MBIT BIT(2) #define QM_DB_TIMEOUT BIT(10) #define QM_OF_FIFO_OF BIT(11) @@ -1538,7 +1543,7 @@ static void qm_log_hw_error(struct hisi_qm *qm, u32 error_status) { const struct hisi_qm_hw_error *err; struct device *dev = &qm->pdev->dev; - u32 reg_val, type, vf_num; + u32 reg_val, type, vf_num, qp_id; int i; for (i = 0; i < ARRAY_SIZE(qm_hw_error); i++) { @@ -1554,19 +1559,24 @@ static void qm_log_hw_error(struct hisi_qm *qm, u32 error_status) type = (reg_val & QM_DB_TIMEOUT_TYPE) >> QM_DB_TIMEOUT_TYPE_SHIFT; vf_num = reg_val & QM_DB_TIMEOUT_VF; - dev_err(dev, "qm %s doorbell timeout in function %u\n", - qm_db_timeout[type], vf_num); + qp_id = reg_val >> QM_DB_TIMEOUT_QP_SHIFT; + dev_err(dev, "qm %s doorbell timeout in function %u qp %u\n", + qm_db_timeout[type], vf_num, qp_id); } else if (err->int_msk & QM_OF_FIFO_OF) { reg_val = readl(qm->io_base + QM_ABNORMAL_INF00); type = (reg_val & QM_FIFO_OVERFLOW_TYPE) >> QM_FIFO_OVERFLOW_TYPE_SHIFT; vf_num = reg_val & QM_FIFO_OVERFLOW_VF; - + qp_id = reg_val >> QM_FIFO_OVERFLOW_QP_SHIFT; if (type < ARRAY_SIZE(qm_fifo_overflow)) - dev_err(dev, "qm %s fifo overflow in function %u\n", - qm_fifo_overflow[type], vf_num); + dev_err(dev, "qm %s fifo overflow in function %u qp %u\n", + qm_fifo_overflow[type], vf_num, qp_id); else dev_err(dev, "unknown error type\n"); + } else if (err->int_msk & QM_AXI_RRESP_ERR) { + reg_val = readl(qm->io_base + QM_ABNORMAL_INF02); + if (reg_val & QM_AXI_POISON_ERR) + dev_err(dev, "qm axi poison error happened\n"); } } } -- 2.30.0

1 0

[PATCH] crypto: hisilicon/qm - check function qp num before alg register
by Zhiqi Song 24 Nov '23

24 Nov '23

From: Weili Qian <qianweili(a)huawei.com> mainline inclusion from mainline-v6.7-rc1 commit b42ab1c61a77832040ad42ebf9adf237360e49f7 category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8IVG6 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ---------------------------------------------------------------------- When the Kunpeng accelerator executes tasks such as encryption and decryption have minimum requirements on the number of device queues. If the number of queues does not meet the requirement, the process initialization will fail. Therefore, the driver checks the number of queues on the device before registering the algorithm. If the number does not meet the requirements, the driver does not register the algorithm to crypto subsystem, the device is still added to the qm_list. Signed-off-by: Weili Qian <qianweili(a)huawei.com> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: JiangShui Yang <yangjiangshui(a)h-partners.com> --- drivers/crypto/hisilicon/hpre/hpre_crypto.c | 25 ++++++++++- drivers/crypto/hisilicon/hpre/hpre_main.c | 14 +++--- drivers/crypto/hisilicon/qm.c | 47 +++++++-------------- drivers/crypto/hisilicon/sec2/sec_crypto.c | 31 ++++++++++++-- drivers/crypto/hisilicon/sec2/sec_main.c | 24 +++++------ drivers/crypto/hisilicon/zip/zip_crypto.c | 23 +++++++++- drivers/crypto/hisilicon/zip/zip_main.c | 14 +++--- include/linux/hisi_acc_qm.h | 18 +++++++- 8 files changed, 133 insertions(+), 63 deletions(-) diff --git a/drivers/crypto/hisilicon/hpre/hpre_crypto.c b/drivers/crypto/hisilicon/hpre/hpre_crypto.c index b0a7eb3b6362..1f64e63b75cb 100644 --- a/drivers/crypto/hisilicon/hpre/hpre_crypto.c +++ b/drivers/crypto/hisilicon/hpre/hpre_crypto.c @@ -57,6 +57,9 @@ struct hpre_ctx; #define HPRE_DRV_ECDH_MASK_CAP BIT(2) #define HPRE_DRV_X25519_MASK_CAP BIT(5) +static DEFINE_MUTEX(hpre_algs_lock); +static unsigned int hpre_available_devs; + typedef void (*hpre_cb)(struct hpre_ctx *ctx, void *sqe); struct hpre_rsa_ctx { @@ -2192,11 +2195,17 @@ static void hpre_unregister_x25519(struct hisi_qm *qm) int hpre_algs_register(struct hisi_qm *qm) { - int ret; + int ret = 0; + + mutex_lock(&hpre_algs_lock); + if (hpre_available_devs) { + hpre_available_devs++; + goto unlock; + } ret = hpre_register_rsa(qm); if (ret) - return ret; + goto unlock; ret = hpre_register_dh(qm); if (ret) @@ -2210,6 +2219,9 @@ int hpre_algs_register(struct hisi_qm *qm) if (ret) goto unreg_ecdh; + hpre_available_devs++; + mutex_unlock(&hpre_algs_lock); + return ret; unreg_ecdh: @@ -2218,13 +2230,22 @@ int hpre_algs_register(struct hisi_qm *qm) hpre_unregister_dh(qm); unreg_rsa: hpre_unregister_rsa(qm); +unlock: + mutex_unlock(&hpre_algs_lock); return ret; } void hpre_algs_unregister(struct hisi_qm *qm) { + mutex_lock(&hpre_algs_lock); + if (--hpre_available_devs) + goto unlock; + hpre_unregister_x25519(qm); hpre_unregister_ecdh(qm); hpre_unregister_dh(qm); hpre_unregister_rsa(qm); + +unlock: + mutex_unlock(&hpre_algs_lock); } diff --git a/drivers/crypto/hisilicon/hpre/hpre_main.c b/drivers/crypto/hisilicon/hpre/hpre_main.c index 005e8c46ce35..580159cb6582 100644 --- a/drivers/crypto/hisilicon/hpre/hpre_main.c +++ b/drivers/crypto/hisilicon/hpre/hpre_main.c @@ -107,6 +107,7 @@ #define HPRE_VIA_MSI_DSM 1 #define HPRE_SQE_MASK_OFFSET 8 #define HPRE_SQE_MASK_LEN 24 +#define HPRE_CTX_Q_NUM_DEF 1 #define HPRE_DFX_BASE 0x301000 #define HPRE_DFX_COMMON1 0x301400 @@ -1382,10 +1383,11 @@ static int hpre_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) dev_warn(&pdev->dev, "init debugfs fail!\n"); - ret = hisi_qm_alg_register(qm, &hpre_devices); + hisi_qm_add_list(qm, &hpre_devices); + ret = hisi_qm_alg_register(qm, &hpre_devices, HPRE_CTX_Q_NUM_DEF); if (ret < 0) { pci_err(pdev, "fail to register algs to crypto!\n"); - goto err_with_qm_start; + goto err_qm_del_list; } ret = qm_register_uacce(qm); @@ -1405,9 +1407,10 @@ static int hpre_probe(struct pci_dev *pdev, const struct pci_device_id *id) return 0; err_with_alg_register: - hisi_qm_alg_unregister(qm, &hpre_devices); + hisi_qm_alg_unregister(qm, &hpre_devices, HPRE_CTX_Q_NUM_DEF); -err_with_qm_start: +err_qm_del_list: + hisi_qm_del_list(qm, &hpre_devices); hpre_debugfs_exit(qm); hisi_qm_stop(qm, QM_NORMAL); @@ -1427,7 +1430,8 @@ static void hpre_remove(struct pci_dev *pdev) hisi_qm_pm_uninit(qm); hisi_qm_wait_task_finish(qm, &hpre_devices); - hisi_qm_alg_unregister(qm, &hpre_devices); + hisi_qm_alg_unregister(qm, &hpre_devices, HPRE_CTX_Q_NUM_DEF); + hisi_qm_del_list(qm, &hpre_devices); if (qm->fun_type == QM_HW_PF && qm->vfs_num) hisi_qm_sriov_disable(pdev, true); diff --git a/drivers/crypto/hisilicon/qm.c b/drivers/crypto/hisilicon/qm.c index d67d2d83d590..2c45d5ecce69 100644 --- a/drivers/crypto/hisilicon/qm.c +++ b/drivers/crypto/hisilicon/qm.c @@ -5027,63 +5027,48 @@ static void qm_cmd_process(struct work_struct *cmd_process) } /** - * hisi_qm_alg_register() - Register alg to crypto and add qm to qm_list. + * hisi_qm_alg_register() - Register alg to crypto. * @qm: The qm needs add. * @qm_list: The qm list. + * @guard: Guard of qp_num. * - * This function adds qm to qm list, and will register algorithm to - * crypto when the qm list is empty. + * Register algorithm to crypto when the function is satisfy guard. */ -int hisi_qm_alg_register(struct hisi_qm *qm, struct hisi_qm_list *qm_list) +int hisi_qm_alg_register(struct hisi_qm *qm, struct hisi_qm_list *qm_list, int guard) { struct device *dev = &qm->pdev->dev; - int flag = 0; - int ret = 0; - - mutex_lock(&qm_list->lock); - if (list_empty(&qm_list->list)) - flag = 1; - list_add_tail(&qm->list, &qm_list->list); - mutex_unlock(&qm_list->lock); if (qm->ver <= QM_HW_V2 && qm->use_sva) { dev_info(dev, "HW V2 not both use uacce sva mode and hardware crypto algs.\n"); return 0; } - if (flag) { - ret = qm_list->register_to_crypto(qm); - if (ret) { - mutex_lock(&qm_list->lock); - list_del(&qm->list); - mutex_unlock(&qm_list->lock); - } + if (qm->qp_num < guard) { + dev_info(dev, "qp_num is less than task need.\n"); + return 0; } - return ret; + return qm_list->register_to_crypto(qm); } EXPORT_SYMBOL_GPL(hisi_qm_alg_register); /** - * hisi_qm_alg_unregister() - Unregister alg from crypto and delete qm from - * qm list. + * hisi_qm_alg_unregister() - Unregister alg from crypto. * @qm: The qm needs delete. * @qm_list: The qm list. + * @guard: Guard of qp_num. * - * This function deletes qm from qm list, and will unregister algorithm - * from crypto when the qm list is empty. + * Unregister algorithm from crypto when the last function is satisfy guard. */ -void hisi_qm_alg_unregister(struct hisi_qm *qm, struct hisi_qm_list *qm_list) +void hisi_qm_alg_unregister(struct hisi_qm *qm, struct hisi_qm_list *qm_list, int guard) { - mutex_lock(&qm_list->lock); - list_del(&qm->list); - mutex_unlock(&qm_list->lock); - if (qm->ver <= QM_HW_V2 && qm->use_sva) return; - if (list_empty(&qm_list->list)) - qm_list->unregister_from_crypto(qm); + if (qm->qp_num < guard) + return; + + qm_list->unregister_from_crypto(qm); } EXPORT_SYMBOL_GPL(hisi_qm_alg_unregister); diff --git a/drivers/crypto/hisilicon/sec2/sec_crypto.c b/drivers/crypto/hisilicon/sec2/sec_crypto.c index 27db9fa65ddf..4d30d49c605e 100644 --- a/drivers/crypto/hisilicon/sec2/sec_crypto.c +++ b/drivers/crypto/hisilicon/sec2/sec_crypto.c @@ -104,6 +104,9 @@ #define IV_CTR_INIT 0x1 #define IV_BYTE_OFFSET 0x8 +static DEFINE_MUTEX(sec_algs_lock); +static unsigned int sec_available_devs; + struct sec_skcipher { u64 alg_msk; struct skcipher_alg alg; @@ -2536,16 +2539,31 @@ static int sec_register_aead(u64 alg_mask) int sec_register_to_crypto(struct hisi_qm *qm) { u64 alg_mask = sec_get_alg_bitmap(qm, SEC_DRV_ALG_BITMAP_HIGH, SEC_DRV_ALG_BITMAP_LOW); - int ret; + int ret = 0; + + mutex_lock(&sec_algs_lock); + if (sec_available_devs) { + sec_available_devs++; + goto unlock; + } ret = sec_register_skcipher(alg_mask); if (ret) - return ret; + goto unlock; ret = sec_register_aead(alg_mask); if (ret) - sec_unregister_skcipher(alg_mask, ARRAY_SIZE(sec_skciphers)); + goto unreg_skcipher; + sec_available_devs++; + mutex_unlock(&sec_algs_lock); + + return 0; + +unreg_skcipher: + sec_unregister_skcipher(alg_mask, ARRAY_SIZE(sec_skciphers)); +unlock: + mutex_unlock(&sec_algs_lock); return ret; } @@ -2553,6 +2571,13 @@ void sec_unregister_from_crypto(struct hisi_qm *qm) { u64 alg_mask = sec_get_alg_bitmap(qm, SEC_DRV_ALG_BITMAP_HIGH, SEC_DRV_ALG_BITMAP_LOW); + mutex_lock(&sec_algs_lock); + if (--sec_available_devs) + goto unlock; + sec_unregister_aead(alg_mask, ARRAY_SIZE(sec_aeads)); sec_unregister_skcipher(alg_mask, ARRAY_SIZE(sec_skciphers)); + +unlock: + mutex_unlock(&sec_algs_lock); } diff --git a/drivers/crypto/hisilicon/sec2/sec_main.c b/drivers/crypto/hisilicon/sec2/sec_main.c index 09831a8c6fad..29461754490e 100644 --- a/drivers/crypto/hisilicon/sec2/sec_main.c +++ b/drivers/crypto/hisilicon/sec2/sec_main.c @@ -1205,15 +1205,11 @@ static int sec_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) pci_warn(pdev, "Failed to init debugfs!\n"); - if (qm->qp_num >= ctx_q_num) { - ret = hisi_qm_alg_register(qm, &sec_devices); - if (ret < 0) { - pr_err("Failed to register driver to crypto.\n"); - goto err_qm_stop; - } - } else { - pci_warn(qm->pdev, - "Failed to use kernel mode, qp not enough!\n"); + hisi_qm_add_list(qm, &sec_devices); + ret = hisi_qm_alg_register(qm, &sec_devices, ctx_q_num); + if (ret < 0) { + pr_err("Failed to register driver to crypto.\n"); + goto err_qm_del_list; } if (qm->uacce) { @@ -1235,9 +1231,9 @@ static int sec_probe(struct pci_dev *pdev, const struct pci_device_id *id) return 0; err_alg_unregister: - if (qm->qp_num >= ctx_q_num) - hisi_qm_alg_unregister(qm, &sec_devices); -err_qm_stop: + hisi_qm_alg_unregister(qm, &sec_devices, ctx_q_num); +err_qm_del_list: + hisi_qm_del_list(qm, &sec_devices); sec_debugfs_exit(qm); hisi_qm_stop(qm, QM_NORMAL); err_probe_uninit: @@ -1254,8 +1250,8 @@ static void sec_remove(struct pci_dev *pdev) hisi_qm_pm_uninit(qm); hisi_qm_wait_task_finish(qm, &sec_devices); - if (qm->qp_num >= ctx_q_num) - hisi_qm_alg_unregister(qm, &sec_devices); + hisi_qm_alg_unregister(qm, &sec_devices, ctx_q_num); + hisi_qm_del_list(qm, &sec_devices); if (qm->fun_type == QM_HW_PF && qm->vfs_num) hisi_qm_sriov_disable(pdev, true); diff --git a/drivers/crypto/hisilicon/zip/zip_crypto.c b/drivers/crypto/hisilicon/zip/zip_crypto.c index 09f60f786779..304411ac5a36 100644 --- a/drivers/crypto/hisilicon/zip/zip_crypto.c +++ b/drivers/crypto/hisilicon/zip/zip_crypto.c @@ -44,6 +44,9 @@ #define HZIP_ALG_GZIP GENMASK(3, 2) #define HZIP_ALG_DEFLATE GENMASK(5, 4) +static DEFINE_MUTEX(zip_algs_lock); +static unsigned int zip_available_devs; + static const u8 zlib_head[HZIP_ZLIB_HEAD_SIZE] = {0x78, 0x9c}; static const u8 gzip_head[HZIP_GZIP_HEAD_SIZE] = { 0x1f, 0x8b, 0x08, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x03 @@ -864,9 +867,15 @@ int hisi_zip_register_to_crypto(struct hisi_qm *qm) { int ret = 0; + mutex_lock(&zip_algs_lock); + if (zip_available_devs) { + zip_available_devs++; + goto err_unlock; + } + ret = hisi_zip_register_deflate(qm); if (ret) - return ret; + goto err_unlock; ret = hisi_zip_register_zlib(qm); if (ret) @@ -876,18 +885,30 @@ int hisi_zip_register_to_crypto(struct hisi_qm *qm) if (ret) goto err_unreg_zlib; + zip_available_devs++; + mutex_unlock(&zip_algs_lock); + return 0; err_unreg_zlib: hisi_zip_unregister_zlib(qm); err_unreg_deflate: hisi_zip_unregister_deflate(qm); +err_unlock: + mutex_unlock(&zip_algs_lock); return ret; } void hisi_zip_unregister_from_crypto(struct hisi_qm *qm) { + mutex_lock(&zip_algs_lock); + if (--zip_available_devs) + goto unlock; + hisi_zip_unregister_deflate(qm); hisi_zip_unregister_zlib(qm); hisi_zip_unregister_gzip(qm); + +unlock: + mutex_unlock(&zip_algs_lock); } diff --git a/drivers/crypto/hisilicon/zip/zip_main.c b/drivers/crypto/hisilicon/zip/zip_main.c index 2d56a546c76f..7bb512e3cb32 100644 --- a/drivers/crypto/hisilicon/zip/zip_main.c +++ b/drivers/crypto/hisilicon/zip/zip_main.c @@ -67,6 +67,7 @@ #define HZIP_SQE_SIZE 128 #define HZIP_PF_DEF_Q_NUM 64 #define HZIP_PF_DEF_Q_BASE 0 +#define HZIP_CTX_Q_NUM_DEF 2 #define HZIP_SOFT_CTRL_CNT_CLR_CE 0x301000 #define HZIP_SOFT_CTRL_CNT_CLR_CE_BIT BIT(0) @@ -1215,10 +1216,11 @@ static int hisi_zip_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (ret) pci_err(pdev, "failed to init debugfs (%d)!\n", ret); - ret = hisi_qm_alg_register(qm, &zip_devices); + hisi_qm_add_list(qm, &zip_devices); + ret = hisi_qm_alg_register(qm, &zip_devices, HZIP_CTX_Q_NUM_DEF); if (ret < 0) { pci_err(pdev, "failed to register driver to crypto!\n"); - goto err_qm_stop; + goto err_qm_del_list; } if (qm->uacce) { @@ -1240,9 +1242,10 @@ static int hisi_zip_probe(struct pci_dev *pdev, const struct pci_device_id *id) return 0; err_qm_alg_unregister: - hisi_qm_alg_unregister(qm, &zip_devices); + hisi_qm_alg_unregister(qm, &zip_devices, HZIP_CTX_Q_NUM_DEF); -err_qm_stop: +err_qm_del_list: + hisi_qm_del_list(qm, &zip_devices); hisi_zip_debugfs_exit(qm); hisi_qm_stop(qm, QM_NORMAL); @@ -1262,7 +1265,8 @@ static void hisi_zip_remove(struct pci_dev *pdev) hisi_qm_pm_uninit(qm); hisi_qm_wait_task_finish(qm, &zip_devices); - hisi_qm_alg_unregister(qm, &zip_devices); + hisi_qm_alg_unregister(qm, &zip_devices, HZIP_CTX_Q_NUM_DEF); + hisi_qm_del_list(qm, &zip_devices); if (qm->fun_type == QM_HW_PF && qm->vfs_num) hisi_qm_sriov_disable(pdev, true); diff --git a/include/linux/hisi_acc_qm.h b/include/linux/hisi_acc_qm.h index 03860e33f8eb..b0147be3a606 100644 --- a/include/linux/hisi_acc_qm.h +++ b/include/linux/hisi_acc_qm.h @@ -519,6 +519,20 @@ static inline void hisi_qm_init_list(struct hisi_qm_list *qm_list) mutex_init(&qm_list->lock); } +static inline void hisi_qm_add_list(struct hisi_qm *qm, struct hisi_qm_list *qm_list) +{ + mutex_lock(&qm_list->lock); + list_add_tail(&qm->list, &qm_list->list); + mutex_unlock(&qm_list->lock); +} + +static inline void hisi_qm_del_list(struct hisi_qm *qm, struct hisi_qm_list *qm_list) +{ + mutex_lock(&qm_list->lock); + list_del(&qm->list); + mutex_unlock(&qm_list->lock); +} + int qm_register_uacce(struct hisi_qm *qm); int hisi_qm_init(struct hisi_qm *qm); void hisi_qm_uninit(struct hisi_qm *qm); @@ -561,8 +575,8 @@ int hisi_qm_alloc_qps_node(struct hisi_qm_list *qm_list, int qp_num, void hisi_qm_free_qps(struct hisi_qp **qps, int qp_num); void hisi_qm_dev_shutdown(struct pci_dev *pdev); void hisi_qm_wait_task_finish(struct hisi_qm *qm, struct hisi_qm_list *qm_list); -int hisi_qm_alg_register(struct hisi_qm *qm, struct hisi_qm_list *qm_list); -void hisi_qm_alg_unregister(struct hisi_qm *qm, struct hisi_qm_list *qm_list); +int hisi_qm_alg_register(struct hisi_qm *qm, struct hisi_qm_list *qm_list, int guard); +void hisi_qm_alg_unregister(struct hisi_qm *qm, struct hisi_qm_list *qm_list, int guard); int hisi_qm_resume(struct device *dev); int hisi_qm_suspend(struct device *dev); void hisi_qm_pm_uninit(struct hisi_qm *qm); -- 2.30.0

1 0

[PATCH 0/5] crypto: hisilicon - fix the process to obtain capability register value
by Zhiqi Song 24 Nov '23

24 Nov '23

This seires patch modify the process to obtain the value of capablity registers. Pre-store the valid values of them. Wenkai Lin (1): crypto: hisilicon/qm - add a function to set qm algs Zhiqi Song (4): crypto: hisilicon/qm - save capability registers in qm init process crypto: hisilicon/hpre - save capability registers in probe process crypto: hisilicon/sec2 - save capability registers in probe process crypto: hisilicon/zip - save capability registers in probe process drivers/crypto/hisilicon/hpre/hpre_main.c | 122 +++++++++++---------- drivers/crypto/hisilicon/qm.c | 98 +++++++++++++++-- drivers/crypto/hisilicon/sec2/sec.h | 7 ++ drivers/crypto/hisilicon/sec2/sec_crypto.c | 8 +- drivers/crypto/hisilicon/sec2/sec_main.c | 78 ++++++------- drivers/crypto/hisilicon/zip/zip_main.c | 120 +++++++++++--------- include/linux/hisi_acc_qm.h | 20 +++- 7 files changed, 295 insertions(+), 158 deletions(-) -- 2.30.0

1 5

【邀请函】操作系统大会&openEuler Summit 2023
by openEuler 24 Nov '23

24 Nov '23

1 0

[PATCH openEuler-1.0-LTS] umh: fix memory leak on execve failure
by Wenyu Huang 24 Nov '23

24 Nov '23

From: Vincent Minet <v.minet(a)criteo.com> mainline inclusion from mainline-5.10 commit db803036ada7 ("umh: fix memory leak on execve failure") category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J9MA CVE: NA ------------------------------------------------- If a UMH process created by fork_usermode_blob() fails to execute, a pair of struct file allocated by umh_pipe_setup() will leak. Under normal conditions, the caller (like bpfilter) needs to manage the lifetime of the UMH and its two pipes. But when fork_usermode_blob() fails, the caller doesn't really have a way to know what needs to be done. It seems better to do the cleanup ourselves in this case. Fixes: 449325b52b7a ("umh: introduce fork_usermode_blob() helper") Signed-off-by: Vincent Minet <v.minet(a)criteo.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Wenyu Huang <huangwenyu5(a)huawei.com> --- kernel/umh.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/kernel/umh.c b/kernel/umh.c index 53611efb10cb..715b4368f9ed 100644 --- a/kernel/umh.c +++ b/kernel/umh.c @@ -495,6 +495,12 @@ static void umh_clean_and_save_pid(struct subprocess_info *info) { struct umh_info *umh_info = info->data; + /* cleanup if umh_pipe_setup() was successful but exec failed */ + if (info->pid && info->retval) { + fput(umh_info->pipe_to_umh); + fput(umh_info->pipe_from_umh); + } + argv_free(info->argv); umh_info->pid = info->pid; } -- 2.34.1

2 1

[PATCH OLK-6.6 v2] net: sched: sch_qfq: Use non-work-conserving warning handler
by Zhengchao Shao 24 Nov '23

24 Nov '23

From: Liu Jian <liujian56(a)huawei.com> mainline inclusion from mainline-v6.7-rc1 commit 6d25d1dc76bf5943a5c1f4bb74d66d5eac58eb77 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J4OL Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- A helper function for printing non-work-conserving alarms is added in commit b00355db3f88 ("pkt_sched: sch_hfsc: sch_htb: Add non-work-conserving warning handler."). In this commit, use qdisc_warn_nonwc() instead of WARN_ONCE() to handle the non-work-conserving warning in qfq Qdisc. Signed-off-by: Liu Jian <liujian56(a)huawei.com> Link: https://lore.kernel.org/r/20231023064729.370649-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/sched/sch_qfq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index 546c10adcacd..5598f8be18ae 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -1003,7 +1003,7 @@ static inline struct sk_buff *qfq_peek_skb(struct qfq_aggregate *agg, *cl = list_first_entry(&agg->active, struct qfq_class, alist); skb = (*cl)->qdisc->ops->peek((*cl)->qdisc); if (skb == NULL) - WARN_ONCE(1, "qfq_dequeue: non-workconserving leaf\n"); + qdisc_warn_nonwc("qfq_dequeue", (*cl)->qdisc); else *len = qdisc_pkt_len(skb); -- 2.34.1

2 1

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 5 +++++ 2 files changed, 32 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10] time/sched_clock: Remove clock_read_data's read mostly members
by Zeng Heng 24 Nov '23

24 Nov '23

hulk inclusion category: performance bugzilla: https://gitee.com/openeuler/kernel/issues/I8J8UH -------------------------------- In the clock_read_data structure, the epoch_ns and epoch_cyc members are often updated by the update_sched_clock(). In fact, the sched_clock_mask, read_sched_clock, mult, and shift will no longer change after they are initialized and registered in sched_clock_register(). Because all of them are in the same cache line, especially in multi-core scenarios, this will cause frequent cache miss. Here we move these almostly read-only members into the struct clock_data to avoid false share with epoch_* members, and separate them into different cache line with padding. Here are the detailed test results of unixbench. ./Run -i 3 -c 96 context1 Without Patch ------------------------------------------------------------------------ Pipe-based Context Switching 7423119.7 lps (10.0 s, 2 samples) System Benchmarks Partial Index BASELINE RESULT INDEX Pipe-based Context Switching 4000.0 7423119.7 18557.8 ======== System Benchmarks Index Score (Partial Only) 18557.8 With Patch ------------------------------------------------------------------------ Pipe-based Context Switching 15869071.9 lps (10.0 s, 2 samples) System Benchmarks Partial Index BASELINE RESULT INDEX Pipe-based Context Switching 4000.0 15869071.9 39672.7 ======== System Benchmarks Index Score (Partial Only) 39672.7 Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- arch/arm64/kernel/perf_event.c | 12 ++++---- include/linux/sched_clock.h | 42 ++++++++++++++++++++++---- kernel/time/sched_clock.c | 55 ++++++++++------------------------ 3 files changed, 59 insertions(+), 50 deletions(-) diff --git a/arch/arm64/kernel/perf_event.c b/arch/arm64/kernel/perf_event.c index cdb3d4549b3a..4894fb0668fa 100644 --- a/arch/arm64/kernel/perf_event.c +++ b/arch/arm64/kernel/perf_event.c @@ -1255,6 +1255,8 @@ static int __init armv8_pmu_driver_init(void) } device_initcall(armv8_pmu_driver_init) +extern struct clock_data cd; + void arch_perf_update_userpage(struct perf_event *event, struct perf_event_mmap_page *userpg, u64 now) { @@ -1269,21 +1271,21 @@ void arch_perf_update_userpage(struct perf_event *event, do { rd = sched_clock_read_begin(&seq); - if (rd->read_sched_clock != arch_timer_read_counter) + if (cd.read_sched_clock != arch_timer_read_counter) return; - userpg->time_mult = rd->mult; - userpg->time_shift = rd->shift; + userpg->time_mult = cd.mult; + userpg->time_shift = cd.shift; userpg->time_zero = rd->epoch_ns; userpg->time_cycles = rd->epoch_cyc; - userpg->time_mask = rd->sched_clock_mask; + userpg->time_mask = cd.sched_clock_mask; /* * Subtract the cycle base, such that software that * doesn't know about cap_user_time_short still 'works' * assuming no wraps. */ - ns = mul_u64_u32_shr(rd->epoch_cyc, rd->mult, rd->shift); + ns = mul_u64_u32_shr(rd->epoch_cyc, cd.mult, cd.shift); userpg->time_zero -= ns; } while (sched_clock_read_retry(seq)); diff --git a/include/linux/sched_clock.h b/include/linux/sched_clock.h index 835ee87ed792..579b6d02c80c 100644 --- a/include/linux/sched_clock.h +++ b/include/linux/sched_clock.h @@ -11,11 +11,6 @@ * * @epoch_ns: sched_clock() value at last update * @epoch_cyc: Clock cycle value at last update. - * @sched_clock_mask: Bitmask for two's complement subtraction of non 64bit - * clocks. - * @read_sched_clock: Current clock source (or dummy source when suspended). - * @mult: Multiplier for scaled math conversion. - * @shift: Shift value for scaled math conversion. * * Care must be taken when updating this structure; it is read by * some very hot code paths. It occupies <=40 bytes and, when combined @@ -25,10 +20,47 @@ struct clock_read_data { u64 epoch_ns; u64 epoch_cyc; +}; + +/** + * struct clock_data - all data needed for sched_clock() (including + * registration of a new clock source) + * + * @seq: Sequence counter for protecting updates. The lowest + * bit is the index for @read_data. + * @read_data: Data required to read from sched_clock. + * + * @sched_clock_mask: Bitmask for two's complement subtraction of non 64bit + * clocks. + * @read_sched_clock: Current clock source (or dummy source when suspended). + * @mult: Multiplier for scaled math conversion. + * @shift: Shift value for scaled math conversion. + * + * @wrap_kt: Duration for which clock can run before wrapping. + * @rate: Tick rate of the registered clock. + * @actual_read_sched_clock: Registered hardware level clock read function. + * + * The ordering of this structure has been chosen to optimize cache + * performance. In particular 'seq' and 'read_data[0]' (combined) should fit + * into a single 64-byte cache line. + */ +struct clock_data { + seqcount_latch_t seq; + struct clock_read_data read_data[2]; + + struct { + char padding[0]; + } ____cacheline_aligned; + u64 sched_clock_mask; u64 (*read_sched_clock)(void); u32 mult; u32 shift; + + ktime_t wrap_kt; + unsigned long rate; + + u64 (*actual_read_sched_clock)(void); }; extern struct clock_read_data *sched_clock_read_begin(unsigned int *seq); diff --git a/kernel/time/sched_clock.c b/kernel/time/sched_clock.c index b1b9b12899f5..de50a321f835 100644 --- a/kernel/time/sched_clock.c +++ b/kernel/time/sched_clock.c @@ -19,30 +19,6 @@ #include "timekeeping.h" -/** - * struct clock_data - all data needed for sched_clock() (including - * registration of a new clock source) - * - * @seq: Sequence counter for protecting updates. The lowest - * bit is the index for @read_data. - * @read_data: Data required to read from sched_clock. - * @wrap_kt: Duration for which clock can run before wrapping. - * @rate: Tick rate of the registered clock. - * @actual_read_sched_clock: Registered hardware level clock read function. - * - * The ordering of this structure has been chosen to optimize cache - * performance. In particular 'seq' and 'read_data[0]' (combined) should fit - * into a single 64-byte cache line. - */ -struct clock_data { - seqcount_latch_t seq; - struct clock_read_data read_data[2]; - ktime_t wrap_kt; - unsigned long rate; - - u64 (*actual_read_sched_clock)(void); -}; - static struct hrtimer sched_clock_timer; static int irqtime = -1; @@ -57,9 +33,9 @@ static u64 notrace jiffy_sched_clock_read(void) return (u64)(jiffies - INITIAL_JIFFIES); } -static struct clock_data cd ____cacheline_aligned = { - .read_data[0] = { .mult = NSEC_PER_SEC / HZ, - .read_sched_clock = jiffy_sched_clock_read, }, +struct clock_data cd ____cacheline_aligned = { + .mult = NSEC_PER_SEC / HZ, + .read_sched_clock = jiffy_sched_clock_read, .actual_read_sched_clock = jiffy_sched_clock_read, }; @@ -88,9 +64,9 @@ unsigned long long notrace sched_clock(void) do { rd = sched_clock_read_begin(&seq); - cyc = (rd->read_sched_clock() - rd->epoch_cyc) & - rd->sched_clock_mask; - res = rd->epoch_ns + cyc_to_ns(cyc, rd->mult, rd->shift); + cyc = (cd.read_sched_clock() - rd->epoch_cyc) & + cd.sched_clock_mask; + res = rd->epoch_ns + cyc_to_ns(cyc, cd.mult, cd.shift); } while (sched_clock_read_retry(seq)); return res; @@ -133,7 +109,7 @@ static void update_sched_clock(void) rd = cd.read_data[0]; cyc = cd.actual_read_sched_clock(); - ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); + ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & cd.sched_clock_mask, cd.mult, cd.shift); rd.epoch_ns = ns; rd.epoch_cyc = cyc; @@ -179,13 +155,14 @@ sched_clock_register(u64 (*read)(void), int bits, unsigned long rate) /* Update epoch for new counter and update 'epoch_ns' from old counter*/ new_epoch = read(); cyc = cd.actual_read_sched_clock(); - ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); + ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & cd.sched_clock_mask, cd.mult, cd.shift); + cd.actual_read_sched_clock = read; + cd.read_sched_clock = read; + cd.sched_clock_mask = new_mask; + cd.mult = new_mult; + cd.shift = new_shift; - rd.read_sched_clock = read; - rd.sched_clock_mask = new_mask; - rd.mult = new_mult; - rd.shift = new_shift; rd.epoch_cyc = new_epoch; rd.epoch_ns = ns; @@ -265,11 +242,9 @@ static u64 notrace suspended_sched_clock_read(void) int sched_clock_suspend(void) { - struct clock_read_data *rd = &cd.read_data[0]; - update_sched_clock(); hrtimer_cancel(&sched_clock_timer); - rd->read_sched_clock = suspended_sched_clock_read; + cd.read_sched_clock = suspended_sched_clock_read; return 0; } @@ -280,7 +255,7 @@ void sched_clock_resume(void) rd->epoch_cyc = cd.actual_read_sched_clock(); hrtimer_start(&sched_clock_timer, cd.wrap_kt, HRTIMER_MODE_REL_HARD); - rd->read_sched_clock = cd.actual_read_sched_clock; + cd.read_sched_clock = cd.actual_read_sched_clock; } static struct syscore_ops sched_clock_ops = { -- 2.25.1

2 1

[PATCH OLK-5.10] time/sched_clock: Remove clock_read_data's read mostly members
by Zeng Heng 24 Nov '23

24 Nov '23

In the clock_read_data structure, the epoch_ns and epoch_cyc members are often updated by the update_sched_clock(). In fact, the sched_clock_mask, read_sched_clock, mult, and shift will no longer change after they are initialized and registered in sched_clock_register(). Because all of them are in the same cache line, especially in multi-core scenarios, this will cause frequent cache miss. Here we move these almostly read-only members into the struct clock_data to avoid false share with epoch_* members, and separate them into different cache line with padding. Here are the detailed test results of unixbench. ./Run -i 3 -c 96 context1 Without Patch ------------------------------------------------------------------------ Pipe-based Context Switching 7423119.7 lps (10.0 s, 2 samples) System Benchmarks Partial Index BASELINE RESULT INDEX Pipe-based Context Switching 4000.0 7423119.7 18557.8 ======== System Benchmarks Index Score (Partial Only) 18557.8 With Patch ------------------------------------------------------------------------ Pipe-based Context Switching 15869071.9 lps (10.0 s, 2 samples) System Benchmarks Partial Index BASELINE RESULT INDEX Pipe-based Context Switching 4000.0 15869071.9 39672.7 ======== System Benchmarks Index Score (Partial Only) 39672.7 Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- arch/arm64/kernel/perf_event.c | 12 ++++---- include/linux/sched_clock.h | 42 ++++++++++++++++++++++---- kernel/time/sched_clock.c | 55 ++++++++++------------------------ 3 files changed, 59 insertions(+), 50 deletions(-) diff --git a/arch/arm64/kernel/perf_event.c b/arch/arm64/kernel/perf_event.c index cdb3d4549b3a..4894fb0668fa 100644 --- a/arch/arm64/kernel/perf_event.c +++ b/arch/arm64/kernel/perf_event.c @@ -1255,6 +1255,8 @@ static int __init armv8_pmu_driver_init(void) } device_initcall(armv8_pmu_driver_init) +extern struct clock_data cd; + void arch_perf_update_userpage(struct perf_event *event, struct perf_event_mmap_page *userpg, u64 now) { @@ -1269,21 +1271,21 @@ void arch_perf_update_userpage(struct perf_event *event, do { rd = sched_clock_read_begin(&seq); - if (rd->read_sched_clock != arch_timer_read_counter) + if (cd.read_sched_clock != arch_timer_read_counter) return; - userpg->time_mult = rd->mult; - userpg->time_shift = rd->shift; + userpg->time_mult = cd.mult; + userpg->time_shift = cd.shift; userpg->time_zero = rd->epoch_ns; userpg->time_cycles = rd->epoch_cyc; - userpg->time_mask = rd->sched_clock_mask; + userpg->time_mask = cd.sched_clock_mask; /* * Subtract the cycle base, such that software that * doesn't know about cap_user_time_short still 'works' * assuming no wraps. */ - ns = mul_u64_u32_shr(rd->epoch_cyc, rd->mult, rd->shift); + ns = mul_u64_u32_shr(rd->epoch_cyc, cd.mult, cd.shift); userpg->time_zero -= ns; } while (sched_clock_read_retry(seq)); diff --git a/include/linux/sched_clock.h b/include/linux/sched_clock.h index 835ee87ed792..579b6d02c80c 100644 --- a/include/linux/sched_clock.h +++ b/include/linux/sched_clock.h @@ -11,11 +11,6 @@ * * @epoch_ns: sched_clock() value at last update * @epoch_cyc: Clock cycle value at last update. - * @sched_clock_mask: Bitmask for two's complement subtraction of non 64bit - * clocks. - * @read_sched_clock: Current clock source (or dummy source when suspended). - * @mult: Multiplier for scaled math conversion. - * @shift: Shift value for scaled math conversion. * * Care must be taken when updating this structure; it is read by * some very hot code paths. It occupies <=40 bytes and, when combined @@ -25,10 +20,47 @@ struct clock_read_data { u64 epoch_ns; u64 epoch_cyc; +}; + +/** + * struct clock_data - all data needed for sched_clock() (including + * registration of a new clock source) + * + * @seq: Sequence counter for protecting updates. The lowest + * bit is the index for @read_data. + * @read_data: Data required to read from sched_clock. + * + * @sched_clock_mask: Bitmask for two's complement subtraction of non 64bit + * clocks. + * @read_sched_clock: Current clock source (or dummy source when suspended). + * @mult: Multiplier for scaled math conversion. + * @shift: Shift value for scaled math conversion. + * + * @wrap_kt: Duration for which clock can run before wrapping. + * @rate: Tick rate of the registered clock. + * @actual_read_sched_clock: Registered hardware level clock read function. + * + * The ordering of this structure has been chosen to optimize cache + * performance. In particular 'seq' and 'read_data[0]' (combined) should fit + * into a single 64-byte cache line. + */ +struct clock_data { + seqcount_latch_t seq; + struct clock_read_data read_data[2]; + + struct { + char padding[0]; + } ____cacheline_aligned; + u64 sched_clock_mask; u64 (*read_sched_clock)(void); u32 mult; u32 shift; + + ktime_t wrap_kt; + unsigned long rate; + + u64 (*actual_read_sched_clock)(void); }; extern struct clock_read_data *sched_clock_read_begin(unsigned int *seq); diff --git a/kernel/time/sched_clock.c b/kernel/time/sched_clock.c index b1b9b12899f5..de50a321f835 100644 --- a/kernel/time/sched_clock.c +++ b/kernel/time/sched_clock.c @@ -19,30 +19,6 @@ #include "timekeeping.h" -/** - * struct clock_data - all data needed for sched_clock() (including - * registration of a new clock source) - * - * @seq: Sequence counter for protecting updates. The lowest - * bit is the index for @read_data. - * @read_data: Data required to read from sched_clock. - * @wrap_kt: Duration for which clock can run before wrapping. - * @rate: Tick rate of the registered clock. - * @actual_read_sched_clock: Registered hardware level clock read function. - * - * The ordering of this structure has been chosen to optimize cache - * performance. In particular 'seq' and 'read_data[0]' (combined) should fit - * into a single 64-byte cache line. - */ -struct clock_data { - seqcount_latch_t seq; - struct clock_read_data read_data[2]; - ktime_t wrap_kt; - unsigned long rate; - - u64 (*actual_read_sched_clock)(void); -}; - static struct hrtimer sched_clock_timer; static int irqtime = -1; @@ -57,9 +33,9 @@ static u64 notrace jiffy_sched_clock_read(void) return (u64)(jiffies - INITIAL_JIFFIES); } -static struct clock_data cd ____cacheline_aligned = { - .read_data[0] = { .mult = NSEC_PER_SEC / HZ, - .read_sched_clock = jiffy_sched_clock_read, }, +struct clock_data cd ____cacheline_aligned = { + .mult = NSEC_PER_SEC / HZ, + .read_sched_clock = jiffy_sched_clock_read, .actual_read_sched_clock = jiffy_sched_clock_read, }; @@ -88,9 +64,9 @@ unsigned long long notrace sched_clock(void) do { rd = sched_clock_read_begin(&seq); - cyc = (rd->read_sched_clock() - rd->epoch_cyc) & - rd->sched_clock_mask; - res = rd->epoch_ns + cyc_to_ns(cyc, rd->mult, rd->shift); + cyc = (cd.read_sched_clock() - rd->epoch_cyc) & + cd.sched_clock_mask; + res = rd->epoch_ns + cyc_to_ns(cyc, cd.mult, cd.shift); } while (sched_clock_read_retry(seq)); return res; @@ -133,7 +109,7 @@ static void update_sched_clock(void) rd = cd.read_data[0]; cyc = cd.actual_read_sched_clock(); - ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); + ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & cd.sched_clock_mask, cd.mult, cd.shift); rd.epoch_ns = ns; rd.epoch_cyc = cyc; @@ -179,13 +155,14 @@ sched_clock_register(u64 (*read)(void), int bits, unsigned long rate) /* Update epoch for new counter and update 'epoch_ns' from old counter*/ new_epoch = read(); cyc = cd.actual_read_sched_clock(); - ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & rd.sched_clock_mask, rd.mult, rd.shift); + ns = rd.epoch_ns + cyc_to_ns((cyc - rd.epoch_cyc) & cd.sched_clock_mask, cd.mult, cd.shift); + cd.actual_read_sched_clock = read; + cd.read_sched_clock = read; + cd.sched_clock_mask = new_mask; + cd.mult = new_mult; + cd.shift = new_shift; - rd.read_sched_clock = read; - rd.sched_clock_mask = new_mask; - rd.mult = new_mult; - rd.shift = new_shift; rd.epoch_cyc = new_epoch; rd.epoch_ns = ns; @@ -265,11 +242,9 @@ static u64 notrace suspended_sched_clock_read(void) int sched_clock_suspend(void) { - struct clock_read_data *rd = &cd.read_data[0]; - update_sched_clock(); hrtimer_cancel(&sched_clock_timer); - rd->read_sched_clock = suspended_sched_clock_read; + cd.read_sched_clock = suspended_sched_clock_read; return 0; } @@ -280,7 +255,7 @@ void sched_clock_resume(void) rd->epoch_cyc = cd.actual_read_sched_clock(); hrtimer_start(&sched_clock_timer, cd.wrap_kt, HRTIMER_MODE_REL_HARD); - rd->read_sched_clock = cd.actual_read_sched_clock; + cd.read_sched_clock = cd.actual_read_sched_clock; } static struct syscore_ops sched_clock_ops = { -- 2.25.1

2 1

[PATCH OLK-5.10 0/3] can: raw: fix memory leak
by Ziyang Xuan 24 Nov '23

24 Nov '23

Patchset for memory leak fix. Eric Dumazet (1): can: raw: fix lockdep issue in raw_release() Oliver Hartkopp (1): can: raw: add missing refcount for memory leak fix Ziyang Xuan (1): can: raw: fix receiver memory leak net/can/raw.c | 80 +++++++++++++++++++++++++++------------------------ 1 file changed, 42 insertions(+), 38 deletions(-) -- 2.25.1

2 4

[PATCH openEuler-1.0-LTS 0/4] Add script to check & update openeuler_defconfig
by Zhang Changzhong 24 Nov '23

24 Nov '23

Masahiro Yamada (1): kbuild: ensure full rebuild when the compiler is updated Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zhang Changzhong (2): config: update openeuler_defconfig for x86 config: update openeuler_defconfig for arm64 Kconfig | 2 - arch/arm64/configs/openeuler_defconfig | 191 +++++++++++++++++-------- arch/x86/configs/openeuler_defconfig | 135 ++++++++--------- include/linux/kconfig.h | 2 + init/Kconfig | 17 +++ scripts/kconfig/Makefile | 22 +++ scripts/kconfig/makeconfig.sh | 24 ++++ 7 files changed, 264 insertions(+), 129 deletions(-) create mode 100644 scripts/kconfig/makeconfig.sh -- 2.34.1

2 5

[PATCH openEuler-1.0-LTS] x86/MCE/AMD: Clear DFR errors found in THR handler
by liwei 24 Nov '23

24 Nov '23

From: Yazen Ghannam <yazen.ghannam(a)amd.com> mainline inclusion from mainline-v6.2-rc1 commit bc1b705b0eee4c645ad8b3bbff3c8a66e9688362 category: bugfix bugzilla: 189379 CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- AMD's MCA Thresholding feature counts errors of all severity levels, not just correctable errors. If a deferred error causes the threshold limit to be reached (it was the error that caused the overflow), then both a deferred error interrupt and a thresholding interrupt will be triggered. The order of the interrupts is not guaranteed. If the threshold interrupt handler is executed first, then it will clear MCA_STATUS for the error. It will not check or clear MCA_DESTAT which also holds a copy of the deferred error. When the deferred error interrupt handler runs it will not find an error in MCA_STATUS, but it will find the error in MCA_DESTAT. This will cause two errors to be logged. Check for deferred errors when handling a threshold interrupt. If a bank contains a deferred error, then clear the bank's MCA_DESTAT register. Define a new helper function to do the deferred error check and clearing of MCA_DESTAT. [ bp: Simplify, convert comment to passive voice. ] Fixes: 37d43acfd79f ("x86/mce/AMD: Redo error logging from APIC LVT interrupt handlers") Signed-off-by: Yazen Ghannam <yazen.ghannam(a)amd.com> Signed-off-by: Borislav Petkov <bp(a)suse.de> Cc: stable(a)vger.kernel.org Link: https://lore.kernel.org/r/20220621155943.33623-1-yazen.ghannam@amd.com Conflicts: arch/x86/kernel/cpu/mce/amd.c Signed-off-by: liwei <liwei728(a)huawei.com> --- arch/x86/kernel/cpu/mce/amd.c | 33 ++++++++++++++++++++------------- 1 file changed, 20 insertions(+), 13 deletions(-) diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c index 38b18fa055ef..5af882dbd979 100644 --- a/arch/x86/kernel/cpu/mce/amd.c +++ b/arch/x86/kernel/cpu/mce/amd.c @@ -941,6 +941,24 @@ _log_error_bank(unsigned int bank, u32 msr_stat, u32 msr_addr, u64 misc) return status & MCI_STATUS_DEFERRED; } +static bool _log_error_deferred(unsigned int bank, u32 misc) +{ + if (!_log_error_bank(bank, mca_msr_reg(bank, MCA_STATUS), + mca_msr_reg(bank, MCA_ADDR), misc)) + return false; + + /* + * Non-SMCA systems don't have MCA_DESTAT/MCA_DEADDR registers. + * Return true here to avoid accessing these registers. + */ + if (!mce_flags.smca) + return true; + + /* Clear MCA_DESTAT if the deferred error was logged from MCA_STATUS. */ + wrmsrl(MSR_AMD64_SMCA_MCx_DESTAT(bank), 0); + return true; +} + /* * We have three scenarios for checking for Deferred errors: * @@ -952,19 +970,8 @@ _log_error_bank(unsigned int bank, u32 msr_stat, u32 msr_addr, u64 misc) */ static void log_error_deferred(unsigned int bank) { - bool defrd; - - defrd = _log_error_bank(bank, msr_ops.status(bank), - msr_ops.addr(bank), 0); - - if (!mce_flags.smca) - return; - - /* Clear MCA_DESTAT if we logged the deferred error from MCA_STATUS. */ - if (defrd) { - wrmsrl(MSR_AMD64_SMCA_MCx_DESTAT(bank), 0); + if (_log_error_deferred(bank, 0)) return; - } /* * Only deferred errors are logged in MCA_DE{STAT,ADDR} so just check @@ -985,7 +992,7 @@ static void amd_deferred_error_interrupt(void) static void log_error_thresholding(unsigned int bank, u64 misc) { - _log_error_bank(bank, msr_ops.status(bank), msr_ops.addr(bank), misc); + _log_error_deferred(bank, misc); } static void log_and_reset_block(struct threshold_block *block) -- 2.25.1

2 1

[PATCH OLK-5.10 v6 00/20] introduce smart_grid zone
by Yipeng Zou 24 Nov '23

24 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: cancel affinity timer in lock free arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 506 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1611 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 21

[PATCH OLK-6.6] net: sched: sch_qfq: Use non-work-conserving warning handler
by Zhengchao Shao 24 Nov '23

24 Nov '23

From: Liu Jian <liujian56(a)huawei.com> mainline inclusion from mainline-v6.7-rc1 commit 6d25d1dc76bf5943a5c1f4bb74d66d5eac58eb77 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8J4OL Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- A helper function for printing non-work-conserving alarms is added in commit b00355db3f88 ("pkt_sched: sch_hfsc: sch_htb: Add non-work-conserving warning handler."). In this commit, use qdisc_warn_nonwc() instead of WARN_ONCE() to handle the non-work-conserving warning in qfq Qdisc. Signed-off-by: Liu Jian <liujian56(a)huawei.com> Link: https://lore.kernel.org/r/20231023064729.370649-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/sched/sch_qfq.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/sched/sch_qfq.c b/net/sched/sch_qfq.c index 546c10adcacd..5598f8be18ae 100644 --- a/net/sched/sch_qfq.c +++ b/net/sched/sch_qfq.c @@ -1003,7 +1003,7 @@ static inline struct sk_buff *qfq_peek_skb(struct qfq_aggregate *agg, *cl = list_first_entry(&agg->active, struct qfq_class, alist); skb = (*cl)->qdisc->ops->peek((*cl)->qdisc); if (skb == NULL) - WARN_ONCE(1, "qfq_dequeue: non-workconserving leaf\n"); + qdisc_warn_nonwc("qfq_dequeue", (*cl)->qdisc); else *len = qdisc_pkt_len(skb); -- 2.34.1

2 1

[PATCH OLK-5.10 v6 00/20] introduce smart_grid zone
by Yipeng Zou 24 Nov '23

24 Nov '23

The patch sets include two parts: 1. patch 1~15,20: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v5: 1. Fix race condition in smart_grid_governor_enable. 2. Fix race condition in stop_auto_affinity. 3. Fix some str to int input in smart_grid_level_write. Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (5): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control sched: smart grid: cancel affinity timer in lock free arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 232 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 78 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 506 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 +++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1611 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 20

[PATCH OLK-5.10 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 5 +++++ 2 files changed, 32 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 24 Nov '23

24 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

2 21

[PATCH OLK-5.10 0/2] Cmdline for ARMv8 Pointer Authentication
by GONG, Ruiqi 24 Nov '23

24 Nov '23

RT. GONG, Ruiqi (1): [Huawei] arm64: add cmdline to control ptrauth feature at boot Srinivas Ramana (1): [Backport] arm64: Defer enabling pointer authentication on boot core arch/arm64/include/asm/pointer_auth.h | 10 +++++ arch/arm64/include/asm/stackprotector.h | 1 + arch/arm64/kernel/cpufeature.c | 54 ++++++++++++++++++++----- arch/arm64/kernel/head.S | 4 -- 4 files changed, 55 insertions(+), 14 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] Backport crypto bugfix
by Yi Yang 24 Nov '23

24 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] Cmdline for ARMv8 Pointer Authentication
by GONG, Ruiqi 24 Nov '23

24 Nov '23

RT. GONG, Ruiqi (1): [Huawei] arm64: add cmdline to control ptrauth feature at boot Srinivas Ramana (1): [Backport] arm64: Defer enabling pointer authentication on boot core arch/arm64/include/asm/pointer_auth.h | 10 +++++ arch/arm64/include/asm/stackprotector.h | 1 + arch/arm64/kernel/cpufeature.c | 54 ++++++++++++++++++++----- arch/arm64/kernel/head.S | 4 -- 4 files changed, 55 insertions(+), 14 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10] LoongArch: disable 40bit user space by default
by Hongchen Zhang 24 Nov '23

24 Nov '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8IRUM ------------------------------------------ Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/configs/loongson3_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/loongarch/configs/loongson3_defconfig b/arch/loongarch/configs/loongson3_defconfig index d2beed31ab74..2f917ae20f76 100644 --- a/arch/loongarch/configs/loongson3_defconfig +++ b/arch/loongarch/configs/loongson3_defconfig @@ -44,6 +44,7 @@ CONFIG_CPU_HAS_LSX=y CONFIG_CPU_HAS_LASX=y CONFIG_NR_CPUS=256 CONFIG_NUMA=y +# CONFIG_VA_BITS_40 is not set CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y CONFIG_PARAVIRT_TIME_ACCOUNTING=y -- 2.33.0

2 1

[PATCH OLK-5.10 0/2] ext4: mitigate cacheline false sharing in struct ext4_inode_info
by Zeng Heng 23 Nov '23

23 Nov '23

Mitigate cacheline false sharing in struct ext4_inode_info Yang Yingliang (2): ext4: mitigatin cacheline false sharing in struct ext4_inode_info enable MITIGATION_FALSE_SHARING by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/ext4/Kconfig | 9 +++++++++ fs/ext4/ext4.h | 5 +++++ fs/ext4/super.c | 4 ++++ 5 files changed, 20 insertions(+) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] ext4: mitigate cacheline false sharing in struct ext4_inode_info
by Zeng Heng 23 Nov '23

23 Nov '23

Yang Yingliang (2): ext4: mitigatin cacheline false sharing in struct ext4_inode_info enable MITIGATION_FALSE_SHARING by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/ext4/Kconfig | 9 +++++++++ fs/ext4/ext4.h | 5 +++++ fs/ext4/super.c | 4 ++++ 5 files changed, 20 insertions(+) -- 2.25.1

2 4

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 23 Nov '23

23 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 5 +++++ 2 files changed, 32 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH openEuler-1.0-LTS 0/2] Backport crypto bugfix
by Yi Yang 23 Nov '23

23 Nov '23

crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 5 +++++ 2 files changed, 32 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] Backport crypto bugfix
by Yi Yang 23 Nov '23

23 Nov '23

fix crypto bugfix and fix kabi broken. Herbert Xu (1): crypto: api - Use work queue in crypto_destroy_instance Yi Yang (1): crypto: fix kabi broken in struct crypto_instance crypto/algapi.c | 29 +++++++++++++++++++++++++++-- include/crypto/algapi.h | 6 ++++++ 2 files changed, 33 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] ext4: mitigatin cacheline false sharing in struct ext4_inode_info
by Zeng Heng 23 Nov '23

23 Nov '23

Yang Yingliang (2): ext4: mitigatin cacheline false sharing in struct ext4_inode_info enable MITIGATION_FALSE_SHARING by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/ext4/Kconfig | 9 +++++++++ fs/ext4/ext4.h | 5 +++++ fs/ext4/super.c | 4 ++++ 5 files changed, 20 insertions(+) -- 2.25.1

2 4

[PATCH OLK-6.6 000/602] Backport linux 6.6.2 LTS patches
by Zheng Zengkai 23 Nov '23

23 Nov '23

Backport linux-6.6.2 LTS patches from upstream. git cherry-pick v6.6.1..v6.6.2~1 -s No conflicts. Aananth V (1): tcp: call tcp_try_undo_recovery when an RTOd TFO SYNACK is ACKed Aaron Plattner (1): objtool: Propagate early errors Abel Vesa (2): clk: imx: Select MXC_CLK for CLK_IMX8QXP regulator: qcom-rpmh: Fix smps4 regulator for pm8550ve Adam Dunlap (1): x86/sev-es: Allow copy_from_kernel_nofault() in earlier boot Adam Ford (3): ARM: dts: am3517-evm: Fix LED3/4 pinmux arm64: dts: imx8mm: Add sound-dai-cells to micfil node arm64: dts: imx8mn: Add sound-dai-cells to micfil node Adam Guerin (1): crypto: qat - enable dc chaining service Adam Skladowski (1): arm64: dts: qcom: msm8976: Fix ipc bit shifts Aditya Gupta (1): powerpc/vmcore: Add MMU information to vmcoreinfo Aditya Kumar Singh (1): wifi: ath11k: fix Tx power value during active CAC Alex Deucher (2): drm/amdgpu/gfx10,11: use memcpy_to/fromio for MQDs drm/amdgpu: don't put MQDs in VRAM on ARM | ARM64 Alexander Aring (4): dlm: fix creating multiple node structures dlm: fix remove member after close call dlm: be sure we reset all nodes at forced shutdown dlm: fix no ack after final message Alexandre Ghiti (2): libbpf: Fix syscall access arguments on riscv drivers: perf: Do not broadcast to other cpus when starting a counter Alison Schofield (5): x86/numa: Introduce numa_fill_memblks() ACPI/NUMA: Apply SRAT proximity domain to entire CFMWS window cxl/region: Prepare the decoder match range helper for reuse cxl/region: Calculate a target position in a region interleave cxl/region: Use cxl_calc_interleave_pos() for auto-discovery Amit Kumar Mahapatra (1): spi: spi-zynq-qspi: add spi-mem to driver kconfig dependencies Andre Przywara (1): clocksource/drivers/arm_arch_timer: limit XGene-1 workaround Andrea Righi (2): module/decompress: use vmalloc() for gzip decompression workspace module/decompress: use kvmalloc() consistently Andrii Staikov (1): i40e: fix potential memory leaks in i40e_remove() Andy Shevchenko (2): ACPI: property: Allow _DSD buffer data only for byte accessors interconnect: qcom: osm-l3: Replace custom implementation of COUNT_ARGS() AngeloGioacchino Del Regno (1): drm: mediatek: mtk_dsi: Fix NO_EOT_PACKET settings/handling Aniruddha Paul (1): ice: Fix VF-VF filter rules in switchdev mode Anuj Gupta (1): nvme: fix error-handling for io_uring nvme-passthrough Anup Patel (2): irqchip/sifive-plic: Fix syscore registration for multi-socket systems RISC-V: Don't fail in riscv_of_parent_hartid() for disabled HARTs Aradhya Bhatia (1): arm64: dts: ti: Fix HDMI Audio overlay in Makefile Armin Wolf (4): platform/x86: wmi: Fix probe failure when failing to register WMI devices platform/x86: wmi: Fix opening of char device hwmon: (sch5627) Use bit macros when accessing the control register hwmon: (sch5627) Disallow write access if virtual registers are locked Arnaldo Carvalho de Melo (1): perf build: Add missing comment about NO_LIBTRACEEVENT=1 Arnd Bergmann (1): fbdev: fsl-diu-fb: mark wr_reg_wa() static Artem Savkov (1): selftests/bpf: Skip module_fentry_shadow test when bpf_testmod is not available Avraham Stern (2): wifi: iwlwifi: mvm: update station's MFP flag after association wifi: iwlwifi: mvm: fix removing pasn station for responder Baochen Qiang (2): wifi: ath12k: fix DMA unmap warning on NULL DMA address wifi: ath11k: fix boot failure with one MSI vector Bard Liao (1): ASoC: Intel: sof_sdw_rt_sdca_jack_common: add rt713 support Bartosz Golaszewski (1): gpio: sim: initialize a managed pointer when declaring it Basavaraj Natikar (1): xhci: Loosen RPM as default policy to cover for AMD xHC 1.1 Ben Wolsieffer (2): futex: Don't include process MM in futex key on no-MMU regmap: prevent noinc writes from clobbering cache Benjamin Gaignard (1): media: verisilicon: Fixes clock list for rk3588 av1 decoder Benjamin Gray (1): powerpc/xive: Fix endian conversion size Biju Das (1): pinctrl: renesas: rzg2l: Make reverse order of enable() for disable() Binbin Wu (1): selftests/x86/lam: Zero out buffer for readlink() Björn Töpel (2): selftests/bpf: Define SYS_PREFIX for riscv selftests/bpf: Define SYS_NANOSLEEP_KPROBE_NAME for riscv Bo Jiao (1): wifi: mt76: fix potential memory leak of beacon commands Brad Griffis (2): arm64: tegra: Fix P3767 card detect polarity arm64: tegra: Fix P3767 QSPI speed Brett Creeley (1): iavf: Fix promiscuous mode configuration flow messages Cezary Rojewski (1): ASoC: Intel: Skylake: Fix mem leak when parsing UUIDs fails Chancel Liu (1): ASoC: soc-pcm.c: Make sure DAI parameters cleared if the DAI becomes inactive Chao Yu (5): f2fs: compress: fix deadloop in f2fs_write_cache_pages() f2fs: compress: fix to avoid use-after-free on dic f2fs: compress: fix to avoid redundant compress extension f2fs: fix to drop meta_inode's page cache in f2fs_put_super() f2fs: fix to initialize map.m_pblk in f2fs_precache_extents() Charles Keepax (1): ASoC: intel: sof_sdw: Stop processing CODECs when enough are found Chen Ni (1): libnvdimm/of_pmem: Use devm_kstrdup instead of kstrdup and check its return value Chen Yu (1): genirq/matrix: Exclude managed interrupts in irq_matrix_allocated() Chen-Yu Tsai (2): regulator: mt6358: Fail probe on unknown chip ID dt-bindings: mfd: mt6397: Split out compatible for MediaTek MT6366 PMIC Chengchang Tang (3): RDMA/hns: Fix printing level of asynchronous events RDMA/hns: Fix uninitialized ucmd in hns_roce_create_qp_common() RDMA/hns: Fix signed-unsigned mixed comparisons Chengming Zhou (1): sched/fair: Fix cfs_rq_is_decayed() on !SMP Chris Packham (1): ARM64: dts: marvell: cn9310: Use appropriate label for spi1 pins Christophe JAILLET (14): wifi: ath: dfs_pattern_detector: Fix a memory initialization issue ACPI: sysfs: Fix create_pnp_modalias() and create_of_modalias() regmap: debugfs: Fix a erroneous check after snprintf() clk: imx: imx8: Fix an error handling path in clk_imx_acm_attach_pm_domains() clk: imx: imx8: Fix an error handling path if devm_clk_hw_register_mux_parent_data_table() fails clk: imx: imx8: Fix an error handling path in imx8_acm_clk_probe() accel/habanalabs/gaudi2: Fix incorrect string length computation in gaudi2_psoc_razwi_get_engines() drm/rockchip: cdn-dp: Fix some error handling paths in cdn_dp_probe() crypto: hisilicon/hpre - Fix a erroneous check after snprintf() fs: dlm: Fix the size of a buffer in dlm_create_debug_file() leds: trigger: ledtrig-cpu:: Fix 'output may be truncated' issue for 'cpu' dmaengine: pxa_dma: Remove an erroneous BUG_ON() in pxad_free_desc() media: i2c: max9286: Fix some redundant of_node_put() calls fs: dlm: Simplify buffer size computation in dlm_create_debug_file() Christophe Leroy (2): powerpc: Only define __parse_fpscr() when required powerpc/40x: Remove stale PTE_ATOMIC_UPDATES macro Claudiu Beznea (5): clk: renesas: rzg2l: Wait for status bit of SD mux before continuing clk: renesas: rzg2l: Lock around writes to mux register clk: renesas: rzg2l: Trust value returned by hardware clk: renesas: rzg2l: Use FIELD_GET() for PLL register fields clk: renesas: rzg2l: Fix computation formula Clément Léger (1): scripts/gdb: fix usage of MOD_TEXT not defined when CONFIG_MODULES=n Colin Ian King (1): rtla: Fix uninitialized variable found Cong Liu (1): drm/amd/display: Fix null pointer dereference in error message Conor Dooley (1): riscv: dts: allwinner: remove address-cells from intc node Cristian Ciocaltea (8): ASoC: cs35l41: Handle mdsync_down reg write errors ASoC: cs35l41: Handle mdsync_up reg write errors ASoC: cs35l41: Initialize completion object before requesting IRQ ASoC: cs35l41: Fix broken shared boost activation ASoC: cs35l41: Verify PM runtime resume errors in IRQ handler ASoC: cs35l41: Undo runtime PM changes at driver exit time ALSA: hda: cs35l41: Fix unbalanced pm_runtime_get() ALSA: hda: cs35l41: Undo runtime PM changes at driver exit time D. Wythe (3): net/smc: fix dangling sock under state SMC_APPFINCLOSEWAIT net/smc: allow cdc msg send rather than drop it with NULL sndbuf_desc net/smc: put sk reference if close work was canceled Dan Carpenter (13): thermal: core: prevent potential string overflow clk: keystone: pll: fix a couple NULL vs IS_ERR() checks clk: ti: fix double free in of_ti_divider_clk_setup() clk: mediatek: fix double free in mtk_clk_register_pllfh() drm/rockchip: Fix type promotion bug in rockchip_gem_iommu_map() PCI: endpoint: Fix double free in __pci_epc_create() dmaengine: ti: edma: handle irq_of_parse_and_map() errors media: ov13b10: Fix some error checking in probe Input: synaptics-rmi4 - fix use after free in rmi_unregister_function() watchdog: marvell_gti_wdt: Fix error code in probe() hsr: Prevent use after free in prp_create_tagged_frame() fbdev: imsttfb: fix double free in probe() fbdev: imsttfb: fix a resource leak in probe Dan Williams (10): cxl/pci: Remove unnecessary device reference management in sanitize work cxl/pci: Cleanup 'sanitize' to always poll cxl/pci: Remove inconsistent usage of dev_err_probe() cxl/pci: Clarify devm host for memdev relative setup cxl/pci: Fix sanitize notifier setup cxl/memdev: Fix sanitize vs decoder setup locking cxl/mem: Fix shutdown order virt: sevguest: Fix passing a stack buffer as a scatterlist target cxl/port: Fix @host confusion in cxl_dport_setup_regs() cxl/hdm: Remove broken error path Daniel Mentz (1): scsi: ufs: core: Leave space for '\0' in utf8 desc string Danila Tikhonov (1): clk: qcom: gcc-sm8150: Fix gcc_sdcc2_apps_clk_src Danny Kaehn (2): hid: cp2112: Fix duplicate workqueue initialization hid: cp2112: Fix IRQ shutdown stopping polling for all IRQs on chip Dario Binacchi (1): ARM: dts: stm32: stm32f7-pinctrl: don't use multiple blank lines Dave Ertman (1): ice: Fix SRIOV LAG disable on non-compliant aggregate David Heidelberg (2): arm64: dts: qcom: sdm845: Fix PSCI power domain names arm64: dts: qcom: sdm845: cheza doesn't support LMh node David Howells (2): iov_iter, x86: Be consistent about the __user tag on copy_mc_to_user() rxrpc: Fix two connection reaping bugs Devi Priya (1): clk: qcom: clk-rcg2: Fix clock rate overflow for high parent frequencies Dhruva Gole (1): firmware: ti_sci: Mark driver as non removable Dinghao Liu (2): mfd: dln2: Fix double put in dln2_probe i3c: Fix potential refcount leak in i3c_master_register_new_i3c_devs Diogo Ivo (1): net: ti: icss-iep: fix setting counter value Dirk Behme (1): clk: renesas: rcar-gen3: Extend SDnH divider table Dmitry Antipov (1): wifi: rtlwifi: fix EDCA limit set by BT coexistence Dmitry Baryshkov (8): drm/bridge: lt9611uxc: fix the race in the error path drm/msm/dsi: use msm_gem_kernel_put to free TX buffer drm/msm/dsi: free TX buffer in unbind arm64: dts: qcom: sc7280: link usb3_phy_wrapper_gcc_usb30_pipe_clk arm64: dts: qcom: sm8150: add ref clock to PCIe PHYs arm64: dts: qcom: sm8350: fix pinctrl for UART18 arm64: dts: qcom: sdm845-mtp: fix WiFi configuration soc: qcom: pmic_glink: fix connector type to be DisplayPort Dominique Martinet (1): Revert "mmc: core: Capture correct oemid-bits for eMMC cards" Doug Berger (1): rtc: brcmstb-waketimer: support level alarm_irq Douglas Anderson (1): drm: Call drm_atomic_helper_shutdown() at shutdown/remove time for misc drivers Dragos Bogdan (1): hwmon: (axi-fan-control) Fix possible NULL pointer dereference Emmanuel Grumbach (1): wifi: iwlwifi: honor the enable_ini value Eric Dumazet (19): udp: introduce udp->udp_flags udp: move udp->no_check6_tx to udp->udp_flags udp: move udp->no_check6_rx to udp->udp_flags udp: move udp->gro_enabled to udp->udp_flags udp: add missing WRITE_ONCE() around up->encap_rcv udp: move udp->accept_udp_{l4|fraglist} to udp->udp_flags udp: lockless UDP_ENCAP_L2TPINUDP / UDP_GRO udp: annotate data-races around udp->encap_type udplite: remove UDPLITE_BIT udplite: fix various data-races tcp_metrics: add missing barriers on delete tcp_metrics: properly set tp->snd_ssthresh in tcp_init_metrics() tcp_metrics: do not create an entry from tcp_init_metrics() chtls: fix tp->rcv_tstamp initialization tcp: fix cookie_init_timestamp() overflows virtio_net: use u64_stats_t infra to avoid data-races net: add DEV_STATS_READ() helper ipvlan: properly track tx_errors inet: shrink struct flowi_common Erik Kurzinger (1): drm/syncobj: fix DRM_SYNCOBJ_WAIT_FLAGS_WAIT_AVAILABLE Eugen Hristev (1): ASoC: mediatek: mt8186_mt6366_rt1019_rt5682s: trivial: fix error messages Fabio Estevam (2): arm64: dts: imx8qm-ss-img: Fix jpegenc compatible entry arm64: dts: imx8mp-debix-model-a: Remove USB hub reset-gpios Fei Shao (1): media: mtk-jpegenc: Fix bug in JPEG encode quality selection Felipe Negrelli Wolter (1): wifi: wfx: fix case where rates are out of order Felix Fietkau (4): wifi: mt76: mt7603: rework/fix rx pse hang check wifi: mt76: mt7603: improve watchdog reset reliablity wifi: mt76: mt7603: improve stuck beacon handling wifi: mt76: remove unused error path in mt76_connac_tx_complete_skb Fenghua Yu (1): dmaengine: idxd: Register dsa_bus_type before registering idxd sub-drivers Filipe Manana (1): btrfs: use u64 for buffer sizes in the tree search ioctls Filippo Storniolo (1): vsock/virtio: remove socket from connected/bound list on shutdown Florent Revest (1): kselftest: vm: fix mdwe's mmap_FIXED test case Florian Fainelli (1): pwm: brcmstb: Utilize appropriate clock APIs in suspend/resume Florian Westphal (1): netfilter: nat: fix ipv6 nat redirect with mapped and scoped addresses Francesco Dolcini (1): arm64: dts: ti: verdin-am62: disable MIPI DSI bridge Frederic Weisbecker (1): srcu: Fix callbacks acceleration mishandling Furong Xu (1): net: stmmac: xgmac: Enable support for multiple Flexible PPS outputs Gabriel Krisman Bertazi (2): io_uring/kbuf: Fix check of BID wrapping in provided buffers io_uring/kbuf: Allow the full buffer id space for provided buffers Gao Xiang (1): erofs: fix erofs_insert_workgroup() lockref usage Gaurav Jain (2): crypto: caam/qi2 - fix Chacha20 + Poly1305 self test failure crypto: caam/jr - fix Chacha20 + Poly1305 self test failure Gaurav Kohli (2): arm64: dts: qcom: msm8916: Fix iommu local address range arm64: dts: qcom: msm8939: Fix iommu local address range Geert Uytterhoeven (4): drm/ssd130x: Fix screen clearing ARM: dts: renesas: blanche: Fix typo in GP_11_2 pin name sh: bios: Revive earlyprintk support riscv: boot: Fix creation of loader.bin Geetha sowjanya (1): octeontx2-pf: Free pending and dropped SQEs Geliang Tang (2): selftests: mptcp: run userspace pm tests slower selftests: mptcp: fix wait_rm_addr/sf parameters George Kennedy (1): IB/mlx5: Fix init stage error handling to avoid double free of same QP and UAF George Shuklin (1): tg3: power down device only on SYSTEM_POWER_OFF Georgia Garcia (1): apparmor: fix invalid reference on profile->disconnected Giovanni Cabiddu (10): crypto: qat - fix state machines cleanup paths crypto: qat - ignore subsequent state up commands crypto: qat - fix unregistration of crypto algorithms crypto: qat - fix unregistration of compression algorithms crypto: qat - increase size of buffers crypto: qat - consolidate services structure crypto: qat - refactor fw config related functions crypto: qat - use masks for AE groups crypto: qat - fix ring to service map for QAT GEN4 crypto: qat - fix deadlock in backlog processing Gou Hao (1): ext4: move 'ix' sanity check to corrent position Guenter Roeck (2): Revert "hwmon: (sch56xx-common) Add DMI override table" Revert "hwmon: (sch56xx-common) Add automatic module loading on supported devices" Guoniu.zhou (1): media: ov5640: fix vblank unchange issue when work at dvp mode Gustavo A. R. Silva (7): gve: Use size_add() in call to struct_size() mlxsw: Use size_mul() in call to struct_size() tls: Use size_add() in call to struct_size() tipc: Use size_add() in calls to struct_size() net: spider_net: Use size_add() in call to struct_size() RDMA/core: Use size_{add,sub,mul}() in calls to struct_size() ASoC: SOF: ipc4-topology: Use size_add() in call to struct_size() Han Xu (1): spi: nxp-fspi: use the correct ioremap function Hangbin Liu (1): selftests: pmtu.sh: fix result checking Hangyu Hua (1): 9p/net: fix possible memory leak in p9_check_errors() Hans Verkuil (1): media: dvb-usb-v2: af9035: fix missing unlock Hans de Goede (4): HID: logitech-hidpp: Don't restart IO, instead defer hid_connect() only HID: logitech-hidpp: Revert "Don't restart communication if not necessary" HID: logitech-hidpp: Move get_wireless_feature_index() check to hidpp_connect_event() mfd: arizona-spi: Set pdata.hpdet_channel for ACPI enumerated devs Hao Chen (1): drivers/perf: hisi: use cpuhp_state_remove_instance_nocalls() for hisi_hns3_pmu uninit process Harald Freudenberger (1): s390/ap: re-init AP queues on config on Haren Myneni (1): powerpc/vas: Limit open window failure messages in log bufffer Harshit Mogalapalli (2): hte: tegra: Fix missing error code in tegra_hte_test_probe() drm/loongson: Fix error handling in lsdc_pixel_pll_setup() Harshitha Prem (1): wifi: ath12k: fix undefined behavior with __fls in dp Hayes Wang (1): r8152: break the loop when the budget is exhausted Heiner Kallweit (3): r8169: fix rare issue with broken rx after link-down on RTL8125 r8169: respect userspace disabling IFF_MULTICAST Revert "PCI/ASPM: Disable only ASPM_STATE_L1 when driver, disables L1" Heng Qi (4): virtio-net: fix mismatch of getting tx-frames virtio-net: consistently save parameters for per-queue virtio-net: fix per queue coalescing parameter setting virtio-net: fix the vq coalescing setting for vq resize Herbert Xu (2): KEYS: Include linux/errno.h in linux/verification.h certs: Break circular dependency when selftest is modular Herve Codina (1): mfd: core: Ensure disabled devices are skipped without aborting Hou Tao (1): bpf: Check map->usercnt after timer->timer is assigned Howard Hsu (2): wifi: mt76: mt7996: fix beamform mcu cmd configuration wifi: mt76: mt7996: fix beamformee ss subfield in EHT PHY cap Hui Wang (1): ASoC: fsl-asoc-card: Add comment for mclk in the codec_priv Ian Rogers (9): perf stat: Fix aggr mode initialization perf parse-events: Fix tracepoint name memory leak perf parse-events: Fix for term values that are raw events perf mem-events: Avoid uninitialized read perf machine: Avoid out of bounds LBR memory read libperf rc_check: Make implicit enabling work for GCC perf hist: Add missing puts to hist__account_cycles perf vendor events intel: Fix broadwellde tma_info_system_dram_bw_use metric perf vendor events intel: Add broadwellde two metrics Ilan Peer (4): wifi: mac80211: Fix setting vif links wifi: iwlwifi: mvm: Correctly set link configuration wifi: iwlwifi: mvm: Fix key flags for IGTK on AP interface wifi: iwlwifi: mvm: Don't always bind/link the P2P Device interface Ilkka Koskinen (2): perf vendor events arm64: Fix for AmpereOne metrics arm64/arm: arm_pmuv3: perf: Don't truncate 64-bit registers Ilpo Järvinen (2): selftests/resctrl: Ensure the benchmark commands fits to its array PCI: vmd: Correct PCI Header Type Register's multi-function check Irui Wang (1): media: mediatek: vcodec: Handle invalid encoder vsi Iulia Tanasescu (1): Bluetooth: ISO: Pass BIG encryption info through QoS Ivaylo Dimitrov (1): drivers/clocksource/timer-ti-dm: Don't call clk_get_rate() in stop function Jacob Keller (1): ice: fix pin assignment for E810-T without SMA control Jai Luthra (2): drm: bridge: it66121: Fix invalid connector dereference arm64: dts: ti: k3-am62a7-sk: Drop i2c-1 to 100Khz Jason Gunthorpe (1): iommufd: Add iopt_area_alloc() Jason-JH.Lin (4): drm/mediatek: Fix coverity issue with unintentional integer overflow drm/mediatek: Add mmsys_dev_num to mt8188 vdosys0 driver data drm/mediatek: Fix iommu fault by swapping FBs after updating plane state drm/mediatek: Fix iommu fault during crtc enabling Javier Carrasco (1): rtc: pcf85363: fix wrong mask/val parameters in regmap_update_bits call Jens Axboe (1): io_uring/net: ensure socket is marked connected on connect retry Jernej Skrabec (1): media: cedrus: Fix clock/reset sequence Jerome Brunet (2): ASoC: hdmi-codec: register hpd callback on component probe ASoC: dapm: fix clock get name Jia-Ju Bai (1): usb: dwc2: fix possible NULL pointer dereference caused by driver concurrency Jian Shen (1): net: page_pool: add missing free_percpu when page_pool_init fail Jiasheng Jiang (9): pstore/platform: Add check for kstrdup clk: mediatek: clk-mt6765: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt6779: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt6797: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt7629-eth: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt7629: Add check for mtk_alloc_clk_data clk: mediatek: clk-mt2701: Add check for mtk_alloc_clk_data media: vidtv: psi: Add check for kstrdup media: vidtv: mux: Add check and kfree for kstrdup Jingbo Xu (1): writeback, cgroup: switch inodes with dirty timestamps to release dying cgwbs Jinjie Ruan (10): wifi: rtw88: debug: Fix the NULL vs IS_ERR() bug for debugfs_create_file() wifi: rtw88: Remove duplicate NULL check before calling usb_kill/free_urb() kunit: Fix missed memory release in kunit_free_suite_set() kunit: Fix the wrong kfree of copy for kunit_filter_suites() kunit: Fix possible memory leak in kunit_filter_suites() kunit: test: Fix the possible memory leak in executor_test HID: uclogic: Fix user-memory-access bug in uclogic_params_ugee_v2_init_event_hooks() HID: uclogic: Fix a work->entry not empty bug in __queue_work() iio: frequency: adf4350: Use device managed functions and fix power down issue. misc: st_core: Do not call kfree_skb() under spin_lock_irqsave() Johannes Berg (18): wifi: cfg80211: add flush functions for wiphy work wifi: mac80211: move radar detect work to wiphy work wifi: mac80211: move scan work to wiphy work wifi: mac80211: move offchannel works to wiphy work wifi: mac80211: move sched-scan stop work to wiphy work wifi: mac80211: fix RCU usage warning in mesh fast-xmit wifi: cfg80211: fix off-by-one in element defrag wifi: mac80211: fix # of MSDU in A-MSDU calculation wifi: cfg80211: fix kernel-doc for wiphy_delayed_work_flush() wifi: mac80211: fix check for unusable RX result wifi: iwlwifi: mvm: use correct sta ID for IGTK/BIGTK wifi: mac80211: don't recreate driver link debugfs in reconfig wifi: iwlwifi: mvm: change iwl_mvm_flush_sta() API wifi: iwlwifi: mvm: fix iwl_mvm_mac_flush_sta() wifi: iwlwifi: mvm: remove TDLS stations from FW wifi: iwlwifi: increase number of RX buffers for EHT devices wifi: iwlwifi: mvm: fix netif csum flags wifi: iwlwifi: pcie: synchronize IRQs before NAPI Johnny Liu (1): gpu: host1x: Correct allocated size for contexts Jonas Blixt (1): USB: usbip: fix stub_dev hub disconnect Jonas Gorski (1): hwrng: geode - fix accessing registers Jonas Karlman (4): drm/rockchip: vop: Fix reset of state in duplicate state crtc funcs drm/rockchip: vop: Fix call to crtc reset helper drm/rockchip: vop2: Don't crash for invalid duplicate_state drm/rockchip: vop2: Add missing call to crtc reset helper Jonathan Neuschäfer (1): clk: npcm7xx: Fix incorrect kfree Josh Poimboeuf (4): x86/srso: Fix SBPB enablement for (possible) future fixed HW x86/srso: Print mitigation for retbleed IBPB case x86/srso: Fix vulnerability reporting for missing microcode x86/srso: Fix unret validation dependencies Juergen Gross (1): xenbus: fix error exit in xenbus_init() Junhao He (1): perf: hisi: Fix use-after-free when register pmu fails Junxian Huang (2): RDMA/hns: Fix unnecessary port_num transition in HW stats allocation RDMA/hns: Fix init failure of RoCE VF and HIP08 Kai Huang (1): x86/tdx: Zero out the missing RSI in TDX_HYPERCALL macro Kailang Yang (1): ALSA: hda/realtek: Add support dual speaker for Dell Kajol Jain (2): tools/perf: Update call stack check in builtin-lock.c perf vendor events: Update PMC used in PM_RUN_INST_CMPL event for power10 platform Kathiravan Thirumoorthy (3): clk: qcom: ipq5018: drop the CLK_SET_RATE_PARENT flag from GPLL clocks clk: qcom: ipq9574: drop the CLK_SET_RATE_PARENT flag from GPLL clocks clk: qcom: ipq5332: drop the CLK_SET_RATE_PARENT flag from GPLL clocks Katya Orlova (1): media: s3c-camif: Avoid inappropriate kfree() Kees Cook (1): string: Adjust strtomem() logic to allow for smaller sources Konrad Dybcio (20): clk: qcom: gcc-msm8996: Remove RPM bus clocks clk: qcom: mmcc-msm8998: Don't check halt bit on some branch clks clk: qcom: mmcc-msm8998: Fix the SMMU GDSC drm/msm/adreno: Fix SM6375 GPU ID drm/msm/a6xx: Fix unknown speedbin case arm64: dts: qcom: sc7280: Add missing LMH interrupts arm64: dts: qcom: qrb2210-rb1: Swap UART index arm64: dts: qcom: qrb2210-rb1: Fix regulators arm64: dts: qcom: sdm670: Fix pdc mapping interconnect: qcom: qdu1000: Set ACV enable_mask interconnect: qcom: sc7180: Set ACV enable_mask interconnect: qcom: sc7280: Set ACV enable_mask interconnect: qcom: sc8180x: Set ACV enable_mask interconnect: qcom: sc8280xp: Set ACV enable_mask interconnect: qcom: sdm670: Set ACV enable_mask interconnect: qcom: sdm845: Set ACV enable_mask interconnect: qcom: sm6350: Set ACV enable_mask interconnect: qcom: sm8150: Set ACV enable_mask interconnect: qcom: sm8250: Set ACV enable_mask interconnect: qcom: sm8350: Set ACV enable_mask Konstantin Meskhidze (1): drm/radeon: possible buffer overflow Krzysztof Kozlowski (4): arm64: dts: qcom: msm8992-libra: drop duplicated reserved memory arm64: dts: qcom: sc7280: drop incorrect EUD port on SoC side arm64: dts: qcom: sdx75-idp: align RPMh regulator nodes with bindings ARM: dts: qcom: mdm9615: populate vsdcc fixed regulator Kumar Kartikeya Dwivedi (2): bpf: Fix kfunc callback register type handling selftests/bpf: Make linked_list failure test more robust Kuninori Morimoto (2): ASoC: fsl: mpc5200_dma.c: Fix warning of Function parameter or member not described ASoC: ams-delta.c: use component after check Kuniyuki Iwashima (2): dccp: Call security_inet_conn_request() after setting IPv4 addresses. dccp/tcp: Call security_inet_conn_request() after setting IPv6 addresses. Kunwu.Chan (1): drm/amd/pm: Fix a memory leak on an error path Kursad Oney (1): ARM: 9321/1: memset: cast the constant byte to unsigned char Lalith Rajendran (1): platform/chrome: cros_ec_lpc: Separate host command and irq disable Laurent Pinchart (3): media: i2c: imx219: Convert to CCI register access helpers media: i2c: imx219: Replace register addresses with macros media: i2c: imx219: Drop IMX219_REG_CSI_LANE_MODE from common regs array Leon Hwang (2): selftests/bpf: Correct map_fd to data_fd in tailcalls bpf, x64: Fix tailcall infinite loop Leon Romanovsky (1): RDMA/hfi1: Workaround truncation compilation error Li Lingfeng (1): nbd: fix uaf in nbd_open Li Zhijian (1): cxl/region: Fix cxl_region_rwsem lock held when returning to user space Linus Walleij (1): watchdog: ixp4xx: Make sure restart always works Longfang Liu (1): crypto: hisilicon/qm - fix PF queue parameter issue Lorenzo Bianconi (1): net: ethernet: mtk_wed: fix EXT_INT_STATUS_RX_FBUF definitions for MT7986 SoC Luoyouming (2): RDMA/hns: Add check for SL RDMA/hns: The UD mode can only be configured with DCQCN Maarten Lankhorst (1): ASoC: SOF: core: Ensure sof_ops_free() is still called when probe never ran. Maciej Wieczor-Retman (1): selftests/pidfd: Fix ksft print formats Maciej Żenczykowski (1): netfilter: xt_recent: fix (increase) ipv6 literal buffer length Marc Kleine-Budde (3): can: dev: can_restart(): don't crash kernel if carrier is OK can: dev: can_restart(): fix race condition between controller restart and netif_carrier_on() can: dev: can_put_echo_skb(): don't crash kernel if can_priv::echo_skb is accessed out of bounds Marcel Ziswiler (1): Bluetooth: hci_sync: Fix Opcode prints in bt_dev_dbg/err Marcin Szycik (1): ice: Fix VF-VF direction matching in drop rule in switchdev Marek Behún (1): leds: turris-omnia: Do not use SMBUS calls Marek Marczykowski-Górecki (1): xen-pciback: Consider INTx disabled when MSI/MSI-X is enabled Marek Szyprowski (2): drm: bridge: samsung-dsim: Fix waiting for empty cmd transfer FIFO on older Exynos media: cec: meson: always include meson sub-directory in Makefile Marek Vasut (3): drm: bridge: samsung-dsim: Initialize ULPS EXIT for i.MX8M DSIM media: hantro: Check whether reset op is defined before use media: verisilicon: Do not enable G2 postproc downscale if source is narrower than destination Marijn Suijten (1): arm64: dts: qcom: sm6125: Pad APPS IOMMU address to 8 characters Mario Limonciello (5): crypto: ccp - Get a free page to use while fetching initial nonce crypto: ccp - Fix ioctl unit tests crypto: ccp - Fix DBC sample application error handling crypto: ccp - Fix sample application signature passing crypto: ccp - Fix some unfused tests Mark Rutland (1): arm64/arm: xen: enlighten: Fix KPTI checks Markus Schneider-Pargmann (1): thermal/drivers/mediatek: Fix probe for THERMAL_V2 Masahiro Yamada (2): modpost: fix tee MODULE_DEVICE_TABLE built on big-endian host modpost: fix ishtp MODULE_DEVICE_TABLE built on big-endian host Matti Lehtimäki (1): ARM: dts: qcom: apq8026-samsung-matisse-wifi: Fix inverted hall sensor Matti Vaittinen (1): tools: iio: iio_generic_buffer ensure alignment Maxime Ripard (1): drm/vc4: tests: Fix UAF in the mock helpers MeiChia Chiu (2): wifi: mt76: update beacon size limitation wifi: mt76: mt7915: fix beamforming availability check Michael Ellerman (1): powerpc: Hide empty pt_regs at base of the stack Michal Schmidt (1): ice: lag: in RCU, use atomic allocation Michał Mirosław (3): mfd: core: Un-constify mfd_cell.of_reg usb: chipidea: Fix DMA overwrite for Tegra usb: chipidea: Simplify Tegra DMA alignment code Michel Dänzer (3): drm/amd/display: Check all enabled planes in dm_check_crtc_cursor drm/amd/display: Refactor dm_get_plane_scale helper drm/amd/display: Bail from dm_check_crtc_cursor if no relevant change Mike Tipton (1): debugfs: Fix __rcu type comparison warning Ming Qian (3): media: imx-jpeg: initiate a drain of the capture queue in dynamic resolution change media: amphion: handle firmware debug message media: imx-jpeg: notify source chagne event when the first picture parsed Miri Korenblit (2): wifi: iwlwifi: don't use an uninitialized variable wifi: iwlwifi: empty overflow queue during flush Moudy Ho (1): media: platform: mtk-mdp3: fix uninitialized variable in mdp_path_config() Namhyung Kim (2): perf record: Fix BTF type checks in the off-cpu profiling perf tools: Do not ignore the default vmlinux.h Naresh Solanki (1): hwmon: (pmbus/mp2975) Move PGOOD fix NeilBrown (1): Fix termination state for idr_for_each_entry_ul() Nícolas F. R. A. Prado (1): thermal: core: Don't update trip points inside the hysteresis range Ondrej Zary (1): ACPI: video: Add acpi_backlight=vendor quirk for Toshiba Portégé R100 Paolo Abeni (1): mptcp: properly account fastopen data Patrick Thompson (1): net: r8169: Disable multicast filter for RTL8168H and RTL8107E Patrisious Haddad (1): IB/mlx5: Fix rdma counter binding for RAW QP Paul E. McKenney (1): x86/nmi: Fix out-of-order NMI nesting checks & false positive warning Peng Fan (1): clk: imx: imx8mq: correct error handling path Peter Chiu (4): wifi: mt76: mt7996: set correct wcid in txp wifi: mt76: mt7996: fix wmm queue mapping wifi: mt76: mt7996: fix rx rate report for CBW320-2 wifi: mt76: mt7996: fix TWT command format Peter Zijlstra (2): sched: Fix stop_one_cpu_nowait() vs hotplug perf: Optimize perf_cgroup_switch() Phil Sutter (2): netfilter: nf_tables: Drop pointless memset when dumping rules net: skb_find_text: Ignore patterns extending past 'to' Philip Yang (3): drm/amdgpu: Increase IH soft ring size for GFX v9.4.3 dGPU drm/amdkfd: Remove svm range validated_once flag drm/amdkfd: Handle errors from svm validate and map Pratyush Yadav (1): media: cadence: csi2rx: Unregister v4l2 async notifier Qais Yousef (2): sched/uclamp: Set max_spare_cap_cpu even if max_spare_cap is 0 sched/uclamp: Ignore (util == 0) optimization in feec() when p_util_max = 0 Qu Wenruo (1): btrfs: make found_logical_ret parameter mandatory for function queue_scrub_stripe() Raag Jadav (2): PM: sleep: Fix symbol export for _SIMPLE_ variants of _PM_OPS() pinctrl: baytrail: fix debounce disable case Rafał Miłecki (1): ARM: dts: BCM5301X: Explicitly disable unused switch CPU ports Randy Dunlap (2): clk: linux/clk-provider.h: fix kernel-doc warnings and typos drm: bridge: for GENERIC_PHY_MIPI_DPHY also select GENERIC_PHY Ratheesh Kannoth (2): octeontx2-pf: Fix error codes octeontx2-pf: Fix holes in error code Reinette Chatre (1): PCI/MSI: Provide stubs for IMS functions Reuben Hawkins (1): vfs: fix readahead(2) on block devices Robert Chiras (1): clk: imx: imx8qxp: Fix elcdif_pll clock Robert Richter (1): cxl/core/regs: Rename @dev to @host in struct cxl_register_map Robin Murphy (1): perf/arm-cmn: Fix DTC domain detection Roman Bacik (1): i2c: iproc: handle invalid slave state Rotem Saado (1): wifi: iwlwifi: yoyo: swap cdb and jacket bits values Sascha Hauer (1): PM / devfreq: rockchip-dfi: Make pmu regmap mandatory Sean Wang (3): wifi: mt76: move struct ieee80211_chanctx_conf up to struct mt76_vif wifi: mt76: mt7921: fix the wrong rate pickup for the chanctx driver wifi: mt76: mt7921: fix the wrong rate selected in fw for the chanctx driver Sebastian Andrzej Siewior (1): powerpc/imc-pmu: Use the correct spinlock initializer. Sergey Shtylyov (1): usb: host: xhci-plat: fix possible kernel oops while resuming Sergio Paracuellos (1): clk: ralink: mtmips: quiet unused variable warning Shayne Chen (1): wifi: mt76: fix per-band IEEE80211_CONF_MONITOR flag comparison Shigeru Yoshida (2): tipc: Change nla_policy for bearer-related names to NLA_NUL_STRING virtio/vsock: Fix uninit-value in virtio_transport_recv_pkt() Shuming Fan (1): ASoC: rt712-sdca: fix speaker route missing issue Siddharth Vadapalli (1): arm64: dts: ti: k3-j721s2-evm-gesi: Specify base dtb for overlay file Song Liu (1): bpf: Fix unnecessary -EBUSY from htab_lock_bucket Srinivasan Shanmugam (1): drm/radeon: Remove the references of radeon_gem_ pread & pwrite ioctls StanleyYP Wang (1): wifi: mt76: get rid of false alamrs of tx emission issues Stefan Wahren (1): hwrng: bcm2835 - Fix hwrng throughput regression Stephan Gerhold (1): arm64: dts: qcom: apq8016-sbc: Add missing ADV7533 regulators Steven Rostedt (Google) (1): eventfs: Check for NULL ef in eventfs_set_attr() Sudeep Holla (3): firmware: arm_ffa: Assign the missing IDR allocation ID to the FFA device firmware: arm_ffa: Allow the FF-A drivers to use 32bit mode of messaging clk: scmi: Free scmi_clk allocated when the clocks with invalid info are skipped Sumit Gupta (2): cpufreq: tegra194: fix warning due to missing opp_put firmware: tegra: Add suspend hook and reset BPMP IPC early on resume Theodore Ts'o (1): ext4: add missing initialization of call_notify_error in update_super_work() Thierry Reding (2): memory: tegra: Set BPMP msg flags to reset IPC channels arm64: tegra: Use correct interrupts for Tegra234 TKE Thomas Gleixner (2): cpu/SMT: Make SMT control more robust against enumeration failures x86/apic: Fake primary thread mask for XEN/PV Thomas Richter (1): perf trace: Use the right bpf_probe_read(_str) variant for reading user data Tomas Glozar (1): nd_btt: Make BTT lanes preemptible Tomi Valkeinen (12): drm/bridge: lt8912b: Fix bridge_detach drm/bridge: lt8912b: Fix crash on bridge detach drm/bridge: lt8912b: Manually disable HPD only if it was enabled drm/bridge: lt8912b: Add missing drm_bridge_attach call drm/bridge: tc358768: Fix use of uninitialized variable drm/bridge: tc358768: Fix bit updates drm/bridge: tc358768: Use struct videomode drm/bridge: tc358768: Print logical values, not raw register values drm/bridge: tc358768: Use dev for dbg prints, not priv->dev drm/bridge: tc358768: Rename dsibclk to hsbyteclk drm/bridge: tc358768: Clean up clock period code drm/bridge: tc358768: Fix tc358768_ns_to_cnt() Trond Myklebust (1): nfsd: Handle EOPENSTALE correctly in the filecache Tyrel Datwyler (1): scsi: ibmvfc: Fix erroneous use of rtas_busy_delay with hcall return code Uwe Kleine-König (4): soc: qcom: llcc: Handle a second device without data corruption backlight: pwm_bl: Disable PWM on shutdown, suspend and remove leds: pwm: Don't disable the PWM when the LED should be off pwm: sti: Reduce number of allocations and drop usage of chip_data Vaishnav Achath (1): spi: omap2-mcspi: Fix hardcoded reference clock Varadarajan Narayanan (5): clk: qcom: ipq5332: Drop set rate parent from gpll0 dependent clocks clk: qcom: config IPQ_APSS_6018 should depend on QCOM_SMEM clk: qcom: clk-alpha-pll: introduce stromer plus ops clk: qcom: apss-ipq-pll: Use stromer plus ops for stromer plus pll clk: qcom: apss-ipq-pll: Fix 'l' value for ipq5332_pll_config Vegard Nossum (1): cpupower: fix reference to nonexistent document Vincent Mailhol (2): can: etas_es58x: rework the version check logic to silence -Wformat-truncation can: etas_es58x: add missing a blank line after declaration Viresh Kumar (2): xen: Make struct privcmd_irqfd's layout architecture independent xen: irqfd: Use _IOW instead of the internal _IOC() macro Vlad Buslov (1): net/sched: act_ct: Always fill offloading tuple iifidx Vladimir Oltean (1): net: enetc: shorten enetc_setup_xdp_prog() error message to fit NETLINK_MAX_FMTMSG_LEN Wadim Egorov (1): arm64: dts: ti: k3-am625-beagleplay: Fix typo in ramoops reg Waiman Long (1): cgroup/cpuset: Fix load balance state in update_partition_sd_lb() Wang Yufen (1): powerpc/pseries: fix potential memory leak in init_cpu_associativity() WangJinchao (1): padata: Fix refcnt handling in padata_free_shell() Willem de Bruijn (1): llc: verify mac len before reading mac header Xiaogang Chen (1): drm/amdkfd: fix some race conditions in vram buffer alloc/free of svm code Xiaolei Wang (1): media: ov5640: Fix a memory leak when ov5640_probe fails Yafang Shao (1): bpf: Fix missed rcu read lock in bpf_task_under_cgroup() Yan Zhai (1): ipv6: avoid atomic fragment on GSO packets Yang Jihong (3): perf kwork: Fix incorrect and missing free atom in work_push_atom() perf kwork: Add the supported subcommands to the document perf kwork: Set ordered_events to true in 'struct perf_tool' Yang Yingliang (5): spi: omap2-mcspi: switch to use modern name interconnect: fix error handling in qnoc_probe() pcmcia: cs: fix possible hung task and memory leak pccardd() pcmcia: ds: fix refcount leak in pcmcia_device_add() pcmcia: ds: fix possible name leak in error path in pcmcia_device_add() Yazen Ghannam (1): x86/amd_nb: Use Family 19h Models 60h-7Fh Function 4 IDs Yedidya Benshimol (1): wifi: iwlwifi: mvm: update IGTK in mvmvif upon D3 resume Yi Yang (1): tty: tty_jobctrl: fix pid memleak in disassociate_ctty() Yicong Yang (1): drivers/perf: hisi_pcie: Check the type first in pmu::event_init() Yu Kuai (1): blk-core: use pr_warn_ratelimited() in bio_check_ro() Yujie Liu (1): tracing/kprobes: Fix the order of argument descriptions Yunfei Dong (1): media: mediatek: vcodec: using encoder device to alloc/free encoder memory Yuntao Wang (1): x86/boot: Fix incorrect startup_gdt_descr.size Yury Norov (3): numa: Generalize numa_map_to_online_node() sched/topology: Fix sched_numa_find_nth_cpu() in CPU-less case sched/topology: Fix sched_numa_find_nth_cpu() in non-NUMA case Zev Weiss (1): hwmon: (nct6775) Fix incorrect variable reuse in fan_div calculation Zhang Rui (1): hwmon: (coretemp) Fix potentially truncated sysfs attribute name Zhang Shurong (2): spi: tegra: Fix missing IRQ check in tegra_slink_probe() ASoC: fsl: Fix PM disable depth imbalance in fsl_easrc_probe Zheng Wang (1): media: bttv: fix use after free error due to btv->timeout timer Zheng Yejian (1): livepatch: Fix missing newline character in klp_resolve_symbols() Ziyang Xuan (1): Bluetooth: Make handle of hci_conn be unique wahrenst (1): ARM: 9323/1: mm: Fix ARCH_LOW_ADDRESS_LIMIT when CONFIG_ZONE_DMA Documentation/ABI/testing/sysfs-driver-qat | 2 + Documentation/admin-guide/hw-vuln/srso.rst | 24 +- .../devicetree/bindings/mfd/mt6397.txt | 4 +- .../bcm4708-buffalo-wzr-1166dhp-common.dtsi | 8 + .../dts/broadcom/bcm4708-luxul-xap-1510.dts | 8 + .../dts/broadcom/bcm4708-luxul-xwc-1000.dts | 8 + .../dts/broadcom/bcm4708-netgear-r6250.dts | 8 + .../dts/broadcom/bcm4708-smartrg-sr400ac.dts | 8 + .../broadcom/bcm47081-buffalo-wzr-600dhp2.dts | 8 + .../dts/broadcom/bcm47081-luxul-xap-1410.dts | 8 + .../dts/broadcom/bcm47081-luxul-xwr-1200.dts | 8 + .../dts/broadcom/bcm4709-netgear-r8000.dts | 8 + .../dts/broadcom/bcm47094-dlink-dir-885l.dts | 8 + .../dts/broadcom/bcm47094-dlink-dir-890l.dts | 8 + .../dts/broadcom/bcm47094-luxul-abr-4500.dts | 8 + .../dts/broadcom/bcm47094-luxul-xap-1610.dts | 8 + .../dts/broadcom/bcm47094-luxul-xbr-4500.dts | 8 + .../dts/broadcom/bcm47094-luxul-xwc-2000.dts | 8 + .../dts/broadcom/bcm47094-luxul-xwr-3100.dts | 8 + .../broadcom/bcm47094-luxul-xwr-3150-v1.dts | 8 + .../dts/broadcom/bcm53015-meraki-mr26.dts | 8 + .../dts/broadcom/bcm53016-meraki-mr32.dts | 8 + arch/arm/boot/dts/broadcom/bcm953012er.dts | 8 + .../qcom-apq8026-samsung-matisse-wifi.dts | 4 +- arch/arm/boot/dts/qcom/qcom-mdm9615.dtsi | 14 +- arch/arm/boot/dts/renesas/r8a7792-blanche.dts | 2 +- arch/arm/boot/dts/st/stm32f7-pinctrl.dtsi | 1 - arch/arm/boot/dts/ti/omap/am3517-evm.dts | 16 +- arch/arm/include/asm/arm_pmuv3.h | 48 +- arch/arm/include/asm/dma.h | 3 + arch/arm/lib/memset.S | 1 + arch/arm/xen/enlighten.c | 25 +- arch/arm64/boot/dts/freescale/imx8mm.dtsi | 1 + arch/arm64/boot/dts/freescale/imx8mn.dtsi | 1 + .../dts/freescale/imx8mp-debix-model-a.dts | 3 - .../boot/dts/freescale/imx8qm-ss-img.dtsi | 2 +- arch/arm64/boot/dts/marvell/cn9130-crb.dtsi | 4 +- arch/arm64/boot/dts/marvell/cn9130-db.dtsi | 4 +- .../arm64/boot/dts/nvidia/tegra234-p3767.dtsi | 4 +- arch/arm64/boot/dts/nvidia/tegra234.dtsi | 12 +- arch/arm64/boot/dts/qcom/apq8016-sbc.dts | 3 + arch/arm64/boot/dts/qcom/msm8916.dtsi | 2 +- arch/arm64/boot/dts/qcom/msm8939.dtsi | 2 +- arch/arm64/boot/dts/qcom/msm8976.dtsi | 8 +- .../boot/dts/qcom/msm8992-xiaomi-libra.dts | 5 - arch/arm64/boot/dts/qcom/qrb2210-rb1.dts | 90 ++-- arch/arm64/boot/dts/qcom/sc7280.dtsi | 32 +- arch/arm64/boot/dts/qcom/sdm670.dtsi | 3 +- arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi | 32 +- arch/arm64/boot/dts/qcom/sdm845-mtp.dts | 2 + arch/arm64/boot/dts/qcom/sdx75-idp.dts | 2 +- arch/arm64/boot/dts/qcom/sm6125.dtsi | 2 +- arch/arm64/boot/dts/qcom/sm8150.dtsi | 12 +- arch/arm64/boot/dts/qcom/sm8350.dtsi | 2 +- arch/arm64/boot/dts/ti/Makefile | 7 +- arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi | 1 + .../arm64/boot/dts/ti/k3-am625-beagleplay.dts | 2 +- arch/arm64/boot/dts/ti/k3-am62a7-sk.dts | 2 +- arch/arm64/include/asm/arm_pmuv3.h | 25 +- arch/arm64/include/asm/cputype.h | 3 +- arch/arm64/kvm/guest.c | 2 +- arch/powerpc/include/asm/nohash/32/pte-40x.h | 3 - arch/powerpc/kernel/process.c | 26 +- arch/powerpc/kernel/traps.c | 2 + arch/powerpc/kexec/core.c | 3 + arch/powerpc/perf/imc-pmu.c | 2 +- arch/powerpc/platforms/book3s/vas-api.c | 34 +- arch/powerpc/platforms/pseries/lpar.c | 4 +- arch/powerpc/platforms/pseries/vas.c | 4 +- arch/powerpc/sysdev/xive/native.c | 2 +- arch/riscv/boot/Makefile | 1 + arch/riscv/boot/dts/allwinner/sun20i-d1s.dtsi | 1 - arch/riscv/kernel/cpu.c | 11 +- arch/sh/Kconfig.debug | 11 + arch/x86/coco/tdx/tdcall.S | 1 + arch/x86/include/asm/nospec-branch.h | 4 +- arch/x86/include/asm/sparsemem.h | 2 + arch/x86/include/asm/uaccess.h | 2 +- arch/x86/kernel/amd_nb.c | 3 + arch/x86/kernel/apic/apic.c | 11 + arch/x86/kernel/cpu/bugs.c | 42 +- arch/x86/kernel/head64.c | 2 +- arch/x86/kernel/nmi.c | 13 +- arch/x86/lib/copy_mc.c | 8 +- arch/x86/mm/maccess.c | 19 +- arch/x86/mm/numa.c | 80 +++ arch/x86/net/bpf_jit_comp.c | 28 +- block/blk-core.c | 4 +- crypto/asymmetric_keys/Kconfig | 3 +- crypto/asymmetric_keys/Makefile | 3 +- crypto/asymmetric_keys/selftest.c | 13 +- crypto/asymmetric_keys/x509_parser.h | 9 - crypto/asymmetric_keys/x509_public_key.c | 8 +- drivers/accel/habanalabs/gaudi2/gaudi2.c | 4 +- drivers/acpi/device_sysfs.c | 10 +- drivers/acpi/numa/srat.c | 11 +- drivers/acpi/property.c | 19 +- drivers/acpi/video_detect.c | 26 + drivers/base/regmap/regmap-debugfs.c | 2 +- drivers/base/regmap/regmap.c | 16 +- drivers/block/nbd.c | 11 +- drivers/char/hw_random/bcm2835-rng.c | 2 +- drivers/char/hw_random/core.c | 6 + drivers/char/hw_random/geode-rng.c | 6 +- drivers/clk/clk-npcm7xx.c | 2 +- drivers/clk/clk-scmi.c | 1 + drivers/clk/imx/Kconfig | 1 + drivers/clk/imx/clk-imx8-acm.c | 16 +- drivers/clk/imx/clk-imx8mq.c | 17 +- drivers/clk/imx/clk-imx8qxp.c | 2 +- drivers/clk/keystone/pll.c | 15 +- drivers/clk/mediatek/clk-mt2701.c | 8 + drivers/clk/mediatek/clk-mt6765.c | 6 + drivers/clk/mediatek/clk-mt6779.c | 4 + drivers/clk/mediatek/clk-mt6797.c | 6 + drivers/clk/mediatek/clk-mt7629-eth.c | 4 + drivers/clk/mediatek/clk-mt7629.c | 6 + drivers/clk/mediatek/clk-pll.c | 6 +- drivers/clk/qcom/Kconfig | 1 + drivers/clk/qcom/apss-ipq-pll.c | 4 +- drivers/clk/qcom/clk-alpha-pll.c | 63 +++ drivers/clk/qcom/clk-alpha-pll.h | 1 + drivers/clk/qcom/clk-rcg2.c | 14 +- drivers/clk/qcom/gcc-ipq5018.c | 3 - drivers/clk/qcom/gcc-ipq5332.c | 4 - drivers/clk/qcom/gcc-ipq9574.c | 4 - drivers/clk/qcom/gcc-msm8996.c | 237 +-------- drivers/clk/qcom/gcc-sm8150.c | 2 +- drivers/clk/qcom/mmcc-msm8998.c | 7 +- drivers/clk/ralink/clk-mtmips.c | 20 +- drivers/clk/renesas/rcar-cpg-lib.c | 15 +- drivers/clk/renesas/rzg2l-cpg.c | 62 +-- drivers/clk/renesas/rzg2l-cpg.h | 2 +- drivers/clk/ti/divider.c | 8 +- drivers/clocksource/arm_arch_timer.c | 5 +- drivers/clocksource/timer-ti-dm.c | 36 +- drivers/cpufreq/tegra194-cpufreq.c | 2 + drivers/crypto/caam/caamalg.c | 3 +- drivers/crypto/caam/caamalg_qi2.c | 3 +- drivers/crypto/ccp/dbc.c | 2 +- drivers/crypto/hisilicon/hpre/hpre_main.c | 7 +- drivers/crypto/hisilicon/qm.c | 18 +- drivers/crypto/hisilicon/qm_common.h | 1 - drivers/crypto/hisilicon/sec2/sec_main.c | 5 + drivers/crypto/hisilicon/zip/zip_main.c | 5 + .../intel/qat/qat_4xxx/adf_4xxx_hw_data.c | 211 +++++--- drivers/crypto/intel/qat/qat_4xxx/adf_drv.c | 34 +- .../intel/qat/qat_common/adf_accel_devices.h | 3 +- .../crypto/intel/qat/qat_common/adf_admin.c | 39 +- .../intel/qat/qat_common/adf_cfg_services.h | 34 ++ .../intel/qat/qat_common/adf_cfg_strings.h | 1 + .../intel/qat/qat_common/adf_common_drv.h | 2 + .../crypto/intel/qat/qat_common/adf_init.c | 20 +- .../crypto/intel/qat/qat_common/adf_sysfs.c | 28 +- .../qat/qat_common/adf_transport_debug.c | 4 +- .../qat/qat_common/icp_qat_fw_init_admin.h | 1 + .../intel/qat/qat_common/qat_algs_send.c | 46 +- drivers/cxl/core/core.h | 1 + drivers/cxl/core/hdm.c | 40 +- drivers/cxl/core/mbox.c | 55 +- drivers/cxl/core/memdev.c | 157 +++--- drivers/cxl/core/port.c | 53 +- drivers/cxl/core/region.c | 241 +++++---- drivers/cxl/core/regs.c | 28 +- drivers/cxl/cxl.h | 4 +- drivers/cxl/cxlmem.h | 13 +- drivers/cxl/pci.c | 80 ++- drivers/devfreq/event/rockchip-dfi.c | 15 +- drivers/dma/idxd/Makefile | 6 +- drivers/dma/pxa_dma.c | 1 - drivers/dma/ti/edma.c | 4 +- drivers/firmware/arm_ffa/bus.c | 1 + drivers/firmware/arm_ffa/driver.c | 12 +- drivers/firmware/tegra/bpmp.c | 30 ++ drivers/firmware/ti_sci.c | 46 +- drivers/gpio/gpio-sim.c | 4 +- drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 2 + drivers/gpu/drm/amd/amdgpu/amdgpu_ih.h | 2 +- drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c | 12 +- drivers/gpu/drm/amd/amdgpu/gfx_v11_0.c | 12 +- drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 92 ++-- drivers/gpu/drm/amd/amdkfd/kfd_svm.h | 2 - .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 73 ++- .../drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c | 4 +- drivers/gpu/drm/aspeed/aspeed_gfx_drv.c | 7 + drivers/gpu/drm/bridge/Kconfig | 2 + drivers/gpu/drm/bridge/cadence/Kconfig | 1 + drivers/gpu/drm/bridge/ite-it66121.c | 12 +- drivers/gpu/drm/bridge/lontium-lt8912b.c | 22 +- drivers/gpu/drm/bridge/lontium-lt9611uxc.c | 10 +- drivers/gpu/drm/bridge/samsung-dsim.c | 20 +- drivers/gpu/drm/bridge/tc358768.c | 166 +++--- drivers/gpu/drm/drm_syncobj.c | 3 +- drivers/gpu/drm/loongson/lsdc_pixpll.c | 6 +- drivers/gpu/drm/mediatek/mtk_drm_crtc.c | 3 + drivers/gpu/drm/mediatek/mtk_drm_drv.c | 1 + drivers/gpu/drm/mediatek/mtk_drm_gem.c | 9 +- drivers/gpu/drm/mediatek/mtk_drm_plane.c | 41 +- drivers/gpu/drm/mediatek/mtk_dsi.c | 4 +- drivers/gpu/drm/mgag200/mgag200_drv.c | 8 + drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 2 +- drivers/gpu/drm/msm/adreno/adreno_device.c | 2 +- drivers/gpu/drm/msm/dsi/dsi.c | 1 + drivers/gpu/drm/msm/dsi/dsi.h | 1 + drivers/gpu/drm/msm/dsi/dsi_host.c | 16 +- drivers/gpu/drm/pl111/pl111_drv.c | 7 + drivers/gpu/drm/radeon/evergreen.c | 7 +- drivers/gpu/drm/radeon/radeon.h | 4 - drivers/gpu/drm/radeon/radeon_drv.c | 2 - drivers/gpu/drm/radeon/radeon_gem.c | 16 - drivers/gpu/drm/rockchip/cdn-dp-core.c | 15 +- drivers/gpu/drm/rockchip/rockchip_drm_gem.c | 2 +- drivers/gpu/drm/rockchip/rockchip_drm_vop.c | 8 +- drivers/gpu/drm/rockchip/rockchip_drm_vop2.c | 39 +- drivers/gpu/drm/solomon/ssd130x.c | 47 +- drivers/gpu/drm/stm/drv.c | 7 + drivers/gpu/drm/tilcdc/tilcdc_drv.c | 11 +- drivers/gpu/drm/tve200/tve200_drv.c | 7 + drivers/gpu/drm/vboxvideo/vbox_drv.c | 10 + drivers/gpu/drm/vc4/tests/vc4_mock_crtc.c | 2 +- drivers/gpu/drm/vc4/tests/vc4_mock_output.c | 2 +- drivers/gpu/host1x/context.c | 4 +- drivers/hid/hid-cp2112.c | 10 +- drivers/hid/hid-logitech-hidpp.c | 60 +-- drivers/hid/hid-uclogic-core-test.c | 7 + drivers/hid/hid-uclogic-params-test.c | 16 +- drivers/hte/hte-tegra194-test.c | 4 +- drivers/hwmon/axi-fan-control.c | 29 +- drivers/hwmon/coretemp.c | 2 +- drivers/hwmon/nct6775-core.c | 12 +- drivers/hwmon/pmbus/mp2975.c | 10 +- drivers/hwmon/sch5627.c | 21 +- drivers/hwmon/sch56xx-common.c | 64 +-- drivers/i2c/busses/i2c-bcm-iproc.c | 133 +++-- drivers/i3c/master.c | 4 +- drivers/iio/frequency/adf4350.c | 75 +-- drivers/infiniband/core/device.c | 2 +- drivers/infiniband/core/sa_query.c | 4 +- drivers/infiniband/core/sysfs.c | 10 +- drivers/infiniband/core/user_mad.c | 4 +- drivers/infiniband/hw/hfi1/efivar.c | 2 +- drivers/infiniband/hw/hns/hns_roce_ah.c | 13 +- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 34 +- drivers/infiniband/hw/hns/hns_roce_main.c | 22 +- drivers/infiniband/hw/hns/hns_roce_qp.c | 2 +- drivers/infiniband/hw/mlx5/main.c | 4 +- drivers/infiniband/hw/mlx5/qp.c | 27 + drivers/input/rmi4/rmi_bus.c | 2 +- drivers/interconnect/qcom/icc-rpm.c | 14 +- drivers/interconnect/qcom/osm-l3.c | 3 +- drivers/interconnect/qcom/qdu1000.c | 1 + drivers/interconnect/qcom/sc7180.c | 1 + drivers/interconnect/qcom/sc7280.c | 1 + drivers/interconnect/qcom/sc8180x.c | 1 + drivers/interconnect/qcom/sc8280xp.c | 1 + drivers/interconnect/qcom/sdm670.c | 1 + drivers/interconnect/qcom/sdm845.c | 1 + drivers/interconnect/qcom/sm6350.c | 1 + drivers/interconnect/qcom/sm8150.c | 1 + drivers/interconnect/qcom/sm8250.c | 1 + drivers/interconnect/qcom/sm8350.c | 1 + drivers/iommu/iommufd/io_pagetable.c | 18 +- drivers/iommu/iommufd/pages.c | 2 + drivers/irqchip/irq-sifive-plic.c | 7 +- drivers/leds/leds-pwm.c | 2 +- drivers/leds/leds-turris-omnia.c | 54 +- drivers/leds/trigger/ledtrig-cpu.c | 4 +- drivers/media/cec/platform/Makefile | 2 +- drivers/media/i2c/Kconfig | 1 + drivers/media/i2c/imx219.c | 503 ++++++++---------- drivers/media/i2c/max9286.c | 2 - drivers/media/i2c/ov13b10.c | 2 +- drivers/media/i2c/ov5640.c | 24 +- drivers/media/pci/bt8xx/bttv-driver.c | 1 + drivers/media/platform/amphion/vpu_defs.h | 1 + drivers/media/platform/amphion/vpu_helpers.c | 1 + drivers/media/platform/amphion/vpu_malone.c | 1 + drivers/media/platform/amphion/vpu_msgs.c | 31 +- drivers/media/platform/cadence/cdns-csi2rx.c | 7 +- .../platform/mediatek/jpeg/mtk_jpeg_enc_hw.c | 5 +- .../platform/mediatek/mdp3/mtk-mdp3-cmdq.c | 2 +- .../mediatek/vcodec/common/mtk_vcodec_util.c | 56 +- .../mediatek/vcodec/encoder/venc_vpu_if.c | 5 + .../media/platform/nxp/imx-jpeg/mxc-jpeg.c | 34 +- .../media/platform/nxp/imx-jpeg/mxc-jpeg.h | 1 + .../samsung/s3c-camif/camif-capture.c | 6 +- .../media/platform/verisilicon/hantro_drv.c | 3 +- .../platform/verisilicon/hantro_postproc.c | 2 +- .../platform/verisilicon/rockchip_vpu_hw.c | 2 +- drivers/media/test-drivers/vidtv/vidtv_mux.c | 7 +- drivers/media/test-drivers/vidtv/vidtv_psi.c | 45 +- drivers/media/usb/dvb-usb-v2/af9035.c | 13 +- drivers/memory/tegra/tegra234.c | 4 + drivers/mfd/arizona-spi.c | 3 + drivers/mfd/dln2.c | 1 - drivers/mfd/mfd-core.c | 17 +- drivers/misc/ti-st/st_core.c | 7 +- drivers/mmc/core/mmc.c | 2 +- drivers/net/can/dev/dev.c | 10 +- drivers/net/can/dev/skb.c | 6 +- drivers/net/can/usb/etas_es58x/es58x_core.c | 1 + drivers/net/can/usb/etas_es58x/es58x_core.h | 6 +- .../net/can/usb/etas_es58x/es58x_devlink.c | 57 +- drivers/net/ethernet/broadcom/tg3.c | 3 +- .../chelsio/inline_crypto/chtls/chtls_cm.c | 2 +- drivers/net/ethernet/freescale/enetc/enetc.c | 2 +- drivers/net/ethernet/google/gve/gve_main.c | 2 +- drivers/net/ethernet/intel/i40e/i40e_main.c | 10 +- drivers/net/ethernet/intel/iavf/iavf.h | 16 +- drivers/net/ethernet/intel/iavf/iavf_main.c | 43 +- .../net/ethernet/intel/iavf/iavf_virtchnl.c | 75 ++- drivers/net/ethernet/intel/ice/ice_lag.c | 18 +- drivers/net/ethernet/intel/ice/ice_ptp.c | 12 +- drivers/net/ethernet/intel/ice/ice_tc_lib.c | 114 +++- .../marvell/octeontx2/nic/otx2_common.c | 15 +- .../marvell/octeontx2/nic/otx2_common.h | 1 + .../ethernet/marvell/octeontx2/nic/otx2_pf.c | 81 +-- .../marvell/octeontx2/nic/otx2_struct.h | 34 +- .../marvell/octeontx2/nic/otx2_txrx.c | 42 ++ drivers/net/ethernet/mediatek/mtk_wed_regs.h | 4 +- .../mlxsw/spectrum_acl_bloom_filter.c | 2 +- drivers/net/ethernet/realtek/r8169_main.c | 10 +- .../net/ethernet/stmicro/stmmac/dwxgmac2.h | 2 +- .../ethernet/stmicro/stmmac/dwxgmac2_core.c | 14 +- drivers/net/ethernet/ti/icssg/icss_iep.c | 2 +- drivers/net/ethernet/toshiba/spider_net.c | 2 +- drivers/net/gtp.c | 4 +- drivers/net/ipvlan/ipvlan_core.c | 8 +- drivers/net/ipvlan/ipvlan_main.c | 1 + drivers/net/macsec.c | 6 +- drivers/net/usb/r8152.c | 18 +- drivers/net/virtio_net.c | 198 ++++--- drivers/net/wireless/ath/ath11k/mac.c | 8 + drivers/net/wireless/ath/ath11k/pci.c | 24 +- drivers/net/wireless/ath/ath12k/dp_rx.c | 2 +- drivers/net/wireless/ath/ath12k/dp_tx.c | 7 +- .../net/wireless/ath/dfs_pattern_detector.c | 2 +- drivers/net/wireless/intel/iwlwifi/cfg/bz.c | 14 +- drivers/net/wireless/intel/iwlwifi/cfg/sc.c | 10 +- drivers/net/wireless/intel/iwlwifi/dvm/tx.c | 5 +- .../wireless/intel/iwlwifi/fw/api/dbg-tlv.h | 1 + .../net/wireless/intel/iwlwifi/iwl-config.h | 5 +- .../net/wireless/intel/iwlwifi/iwl-dbg-tlv.h | 5 +- drivers/net/wireless/intel/iwlwifi/iwl-drv.c | 51 +- drivers/net/wireless/intel/iwlwifi/iwl-prph.h | 4 +- .../net/wireless/intel/iwlwifi/iwl-trans.h | 11 +- drivers/net/wireless/intel/iwlwifi/mvm/d3.c | 10 + .../intel/iwlwifi/mvm/ftm-responder.c | 9 +- drivers/net/wireless/intel/iwlwifi/mvm/link.c | 22 +- .../net/wireless/intel/iwlwifi/mvm/mac-ctxt.c | 8 +- .../net/wireless/intel/iwlwifi/mvm/mac80211.c | 182 +++---- .../net/wireless/intel/iwlwifi/mvm/mld-key.c | 16 +- .../wireless/intel/iwlwifi/mvm/mld-mac80211.c | 105 ++-- .../net/wireless/intel/iwlwifi/mvm/mld-sta.c | 2 +- drivers/net/wireless/intel/iwlwifi/mvm/mvm.h | 9 +- drivers/net/wireless/intel/iwlwifi/mvm/sta.c | 12 +- .../wireless/intel/iwlwifi/mvm/time-event.c | 31 +- drivers/net/wireless/intel/iwlwifi/mvm/tx.c | 22 +- .../wireless/intel/iwlwifi/pcie/trans-gen2.c | 1 + .../net/wireless/intel/iwlwifi/pcie/trans.c | 1 + drivers/net/wireless/intel/iwlwifi/queue/tx.c | 9 +- drivers/net/wireless/intel/iwlwifi/queue/tx.h | 2 +- drivers/net/wireless/mediatek/mt76/dma.c | 3 - drivers/net/wireless/mediatek/mt76/mac80211.c | 9 +- drivers/net/wireless/mediatek/mt76/mt76.h | 4 +- .../wireless/mediatek/mt76/mt7603/beacon.c | 76 ++- .../net/wireless/mediatek/mt76/mt7603/core.c | 2 + .../net/wireless/mediatek/mt76/mt7603/mac.c | 52 +- .../net/wireless/mediatek/mt76/mt7603/regs.h | 5 + .../net/wireless/mediatek/mt76/mt7615/mcu.c | 2 +- .../wireless/mediatek/mt76/mt7615/pci_mac.c | 2 +- .../wireless/mediatek/mt76/mt76_connac3_mac.h | 2 + .../wireless/mediatek/mt76/mt76_connac_mac.c | 24 +- .../wireless/mediatek/mt76/mt76_connac_mcu.c | 9 +- .../net/wireless/mediatek/mt76/mt7915/mac.c | 2 +- .../net/wireless/mediatek/mt76/mt7915/main.c | 8 +- .../net/wireless/mediatek/mt76/mt7915/mcu.c | 69 ++- .../net/wireless/mediatek/mt76/mt7915/mcu.h | 18 +- .../wireless/mediatek/mt76/mt7915/mt7915.h | 2 + .../net/wireless/mediatek/mt76/mt7921/main.c | 12 +- .../wireless/mediatek/mt76/mt7921/pci_mac.c | 2 +- drivers/net/wireless/mediatek/mt76/mt792x.h | 1 - .../net/wireless/mediatek/mt76/mt792x_core.c | 4 +- .../net/wireless/mediatek/mt76/mt7996/init.c | 9 +- .../net/wireless/mediatek/mt76/mt7996/mac.c | 12 +- .../net/wireless/mediatek/mt76/mt7996/main.c | 16 +- .../net/wireless/mediatek/mt76/mt7996/mcu.c | 67 ++- .../net/wireless/mediatek/mt76/mt7996/mcu.h | 11 +- .../wireless/realtek/rtlwifi/rtl8188ee/dm.c | 2 +- .../realtek/rtlwifi/rtl8192c/dm_common.c | 2 +- .../wireless/realtek/rtlwifi/rtl8723ae/dm.c | 2 +- drivers/net/wireless/realtek/rtw88/debug.c | 4 +- drivers/net/wireless/realtek/rtw88/usb.c | 9 +- drivers/net/wireless/silabs/wfx/data_tx.c | 71 +-- drivers/nvdimm/of_pmem.c | 8 +- drivers/nvdimm/region_devs.c | 8 +- drivers/nvme/host/ioctl.c | 7 +- drivers/pci/controller/vmd.c | 3 +- drivers/pci/endpoint/pci-epc-core.c | 1 - drivers/pci/pcie/aspm.c | 3 +- drivers/pcmcia/cs.c | 1 + drivers/pcmcia/ds.c | 14 +- drivers/perf/arm-cmn.c | 16 +- drivers/perf/arm_pmuv3.c | 6 +- drivers/perf/hisilicon/hisi_pcie_pmu.c | 7 +- drivers/perf/hisilicon/hisi_uncore_pa_pmu.c | 4 +- drivers/perf/hisilicon/hisi_uncore_sllc_pmu.c | 4 +- drivers/perf/hisilicon/hns3_pmu.c | 8 +- drivers/perf/riscv_pmu_sbi.c | 6 +- drivers/pinctrl/intel/pinctrl-baytrail.c | 11 +- drivers/pinctrl/renesas/pinctrl-rzg2l.c | 3 +- drivers/platform/chrome/cros_ec.c | 116 +++- drivers/platform/chrome/cros_ec.h | 4 + drivers/platform/chrome/cros_ec_lpc.c | 22 +- drivers/platform/x86/wmi.c | 36 +- drivers/pwm/pwm-brcmstb.c | 4 +- drivers/pwm/pwm-sti.c | 29 +- drivers/regulator/mt6358-regulator.c | 14 +- drivers/regulator/qcom-rpmh-regulator.c | 2 +- drivers/rtc/rtc-brcmstb-waketimer.c | 47 +- drivers/rtc/rtc-pcf85363.c | 2 +- drivers/s390/crypto/ap_bus.c | 21 +- drivers/s390/crypto/ap_bus.h | 1 + drivers/s390/crypto/ap_queue.c | 9 +- drivers/scsi/ibmvscsi/ibmvfc.c | 3 +- drivers/soc/qcom/llcc-qcom.c | 3 + drivers/soc/qcom/pmic_glink_altmode.c | 2 +- drivers/spi/Kconfig | 1 + drivers/spi/spi-nxp-fspi.c | 2 +- drivers/spi/spi-omap2-mcspi.c | 263 ++++----- drivers/spi/spi-tegra20-slink.c | 2 + .../staging/media/sunxi/cedrus/cedrus_hw.c | 24 +- drivers/thermal/mediatek/auxadc_thermal.c | 2 +- drivers/thermal/thermal_core.c | 6 +- drivers/thermal/thermal_trip.c | 19 +- drivers/tty/tty_jobctrl.c | 17 +- drivers/ufs/core/ufshcd.c | 2 +- drivers/usb/chipidea/host.c | 48 +- drivers/usb/dwc2/hcd.c | 2 +- drivers/usb/host/xhci-pci.c | 2 + drivers/usb/host/xhci-plat.c | 23 +- drivers/usb/usbip/stub_dev.c | 9 +- drivers/video/backlight/pwm_bl.c | 22 + drivers/video/fbdev/fsl-diu-fb.c | 2 +- drivers/video/fbdev/imsttfb.c | 35 +- drivers/virt/coco/sev-guest/sev-guest.c | 45 +- drivers/watchdog/ixp4xx_wdt.c | 28 +- drivers/watchdog/marvell_gti_wdt.c | 2 +- drivers/xen/privcmd.c | 2 +- drivers/xen/xen-pciback/conf_space.c | 19 +- .../xen/xen-pciback/conf_space_capability.c | 8 +- drivers/xen/xen-pciback/conf_space_header.c | 21 +- drivers/xen/xenbus/xenbus_probe.c | 2 +- fs/btrfs/ioctl.c | 10 +- fs/btrfs/scrub.c | 10 +- fs/debugfs/file.c | 2 +- fs/dlm/debug_fs.c | 13 +- fs/dlm/midcomms.c | 39 +- fs/erofs/utils.c | 8 +- fs/erofs/zdata.c | 1 + fs/ext4/extents.c | 10 +- fs/ext4/super.c | 3 +- fs/f2fs/data.c | 24 +- fs/f2fs/file.c | 1 + fs/f2fs/super.c | 35 +- fs/fs-writeback.c | 41 +- fs/nfsd/filecache.c | 27 +- fs/nfsd/vfs.c | 28 +- fs/nfsd/vfs.h | 4 +- fs/pstore/platform.c | 9 +- fs/tracefs/event_inode.c | 4 +- include/drm/bridge/samsung-dsim.h | 1 + include/linux/bpf.h | 5 + include/linux/clk-provider.h | 15 +- include/linux/cpuhotplug.h | 1 + include/linux/hisi_acc_qm.h | 7 + include/linux/hw_random.h | 1 + include/linux/idr.h | 6 +- include/linux/mfd/core.h | 2 +- include/linux/netdevice.h | 1 + include/linux/numa.h | 14 +- include/linux/objtool.h | 3 +- include/linux/pci.h | 34 +- include/linux/perf_event.h | 1 + include/linux/pm.h | 43 +- include/linux/string.h | 7 +- include/linux/topology.h | 2 +- include/linux/udp.h | 66 ++- include/linux/verification.h | 1 + include/net/bluetooth/hci.h | 3 + include/net/bluetooth/hci_core.h | 31 +- include/net/cfg80211.h | 21 + include/net/flow.h | 2 +- include/net/netfilter/nf_conntrack_act_ct.h | 30 +- include/net/tcp.h | 2 +- include/net/udp_tunnel.h | 9 +- include/net/udplite.h | 14 +- include/soc/tegra/bpmp.h | 6 + include/sound/cs35l41.h | 4 +- include/uapi/xen/privcmd.h | 4 +- io_uring/kbuf.c | 11 +- io_uring/net.c | 24 +- kernel/bpf/hashtab.c | 7 +- kernel/bpf/helpers.c | 32 +- kernel/bpf/trampoline.c | 4 +- kernel/bpf/verifier.c | 7 + kernel/cgroup/cpuset.c | 12 +- kernel/cpu.c | 18 +- kernel/events/core.c | 115 ++-- kernel/futex/core.c | 12 +- kernel/irq/matrix.c | 6 +- kernel/livepatch/core.c | 2 +- kernel/module/decompress.c | 8 +- kernel/padata.c | 6 +- kernel/rcu/srcutree.c | 31 +- kernel/sched/core.c | 10 +- kernel/sched/deadline.c | 2 + kernel/sched/fair.c | 35 +- kernel/sched/rt.c | 4 + kernel/sched/topology.c | 6 +- kernel/trace/trace_kprobe.c | 2 +- lib/kunit/executor.c | 23 +- lib/kunit/executor_test.c | 36 +- mm/mempolicy.c | 18 +- mm/readahead.c | 3 +- net/9p/client.c | 6 +- net/bluetooth/amp.c | 3 +- net/bluetooth/hci_conn.c | 58 +- net/bluetooth/hci_core.c | 3 + net/bluetooth/hci_event.c | 84 +-- net/bluetooth/hci_sync.c | 4 +- net/bluetooth/iso.c | 19 +- net/core/page_pool.c | 6 +- net/core/skbuff.c | 3 +- net/dccp/ipv4.c | 6 +- net/dccp/ipv6.c | 6 +- net/hsr/hsr_forward.c | 4 +- net/ipv4/syncookies.c | 20 +- net/ipv4/tcp_input.c | 9 +- net/ipv4/tcp_metrics.c | 15 +- net/ipv4/udp.c | 74 +-- net/ipv4/udp_offload.c | 4 +- net/ipv4/udp_tunnel_core.c | 2 +- net/ipv4/udplite.c | 1 - net/ipv4/xfrm4_input.c | 4 +- net/ipv6/ip6_output.c | 8 +- net/ipv6/syncookies.c | 7 +- net/ipv6/udp.c | 34 +- net/ipv6/udplite.c | 1 - net/ipv6/xfrm6_input.c | 4 +- net/l2tp/l2tp_core.c | 6 +- net/llc/llc_input.c | 10 +- net/llc/llc_s_ac.c | 3 + net/llc/llc_station.c | 3 + net/mac80211/driver-ops.c | 9 +- net/mac80211/drop.h | 3 + net/mac80211/ieee80211_i.h | 18 +- net/mac80211/iface.c | 2 +- net/mac80211/link.c | 2 +- net/mac80211/main.c | 25 +- net/mac80211/mesh_pathtbl.c | 2 +- net/mac80211/offchannel.c | 36 +- net/mac80211/rx.c | 2 +- net/mac80211/scan.c | 36 +- net/mac80211/sta_info.c | 2 +- net/mac80211/util.c | 11 +- net/mptcp/fastopen.c | 1 + net/netfilter/nf_nat_redirect.c | 27 +- net/netfilter/nf_tables_api.c | 4 - net/netfilter/xt_recent.c | 2 +- net/openvswitch/conntrack.c | 2 +- net/rxrpc/conn_object.c | 2 +- net/rxrpc/local_object.c | 2 +- net/sched/act_ct.c | 15 +- net/smc/af_smc.c | 4 +- net/smc/smc.h | 5 + net/smc/smc_cdc.c | 11 +- net/smc/smc_close.c | 5 +- net/tipc/link.c | 4 +- net/tipc/netlink.c | 4 +- net/tls/tls_sw.c | 2 +- net/vmw_vsock/virtio_transport_common.c | 18 +- net/wireless/core.c | 34 +- net/wireless/core.h | 3 +- net/wireless/scan.c | 4 +- net/wireless/sysfs.c | 4 +- scripts/Makefile.vmlinux_o | 3 +- scripts/gdb/linux/constants.py.in | 9 +- scripts/mod/file2alias.c | 14 +- security/apparmor/policy.c | 1 + security/apparmor/policy_unpack.c | 5 +- sound/pci/hda/cs35l41_hda.c | 9 +- sound/pci/hda/patch_realtek.c | 40 +- sound/soc/codecs/cs35l41-lib.c | 60 ++- sound/soc/codecs/cs35l41.c | 38 +- sound/soc/codecs/cs35l41.h | 1 - sound/soc/codecs/hdmi-codec.c | 27 +- sound/soc/codecs/rt712-sdca.c | 14 +- sound/soc/fsl/fsl-asoc-card.c | 1 + sound/soc/fsl/fsl_easrc.c | 8 +- sound/soc/fsl/mpc5200_dma.c | 3 + sound/soc/intel/boards/sof_sdw.c | 2 +- .../boards/sof_sdw_rt_sdca_jack_common.c | 8 + sound/soc/intel/skylake/skl-sst-utils.c | 1 + .../mt8186/mt8186-mt6366-rt1019-rt5682s.c | 4 +- sound/soc/soc-dapm.c | 2 +- sound/soc/soc-pcm.c | 21 +- sound/soc/sof/core.c | 6 +- sound/soc/sof/ipc4-topology.c | 3 +- sound/soc/ti/ams-delta.c | 4 +- tools/crypto/ccp/dbc.c | 17 +- tools/crypto/ccp/dbc.py | 8 +- tools/crypto/ccp/test_dbc.py | 45 +- tools/iio/iio_generic_buffer.c | 13 +- tools/lib/bpf/bpf_tracing.h | 2 - tools/lib/perf/include/internal/rc_check.h | 6 +- tools/objtool/objtool.c | 4 +- tools/perf/Documentation/perf-kwork.txt | 2 +- tools/perf/Makefile.perf | 4 + tools/perf/builtin-kwork.c | 13 +- tools/perf/builtin-lock.c | 17 +- tools/perf/builtin-stat.c | 2 +- .../arch/arm64/ampere/ampereone/metrics.json | 418 ++++++++------- .../pmu-events/arch/powerpc/power10/pmc.json | 2 +- .../arch/x86/broadwellde/bdwde-metrics.json | 14 +- tools/perf/util/bpf_off_cpu.c | 5 +- .../bpf_skel/augmented_raw_syscalls.bpf.c | 16 +- tools/perf/util/bpf_skel/vmlinux/.gitignore | 1 + tools/perf/util/hist.c | 10 +- tools/perf/util/machine.c | 22 +- tools/perf/util/mem-events.c | 3 +- tools/perf/util/parse-events.y | 9 +- .../cpupower/man/cpupower-powercap-info.1 | 2 +- tools/testing/cxl/test/mem.c | 4 +- .../selftests/bpf/prog_tests/linked_list.c | 10 +- .../bpf/prog_tests/module_fentry_shadow.c | 5 + .../selftests/bpf/prog_tests/tailcalls.c | 32 +- tools/testing/selftests/bpf/progs/bpf_misc.h | 3 + .../selftests/bpf/progs/linked_list_fail.c | 4 +- tools/testing/selftests/bpf/test_progs.h | 2 + tools/testing/selftests/mm/mdwe_test.c | 9 +- .../testing/selftests/net/mptcp/mptcp_join.sh | 18 +- tools/testing/selftests/net/pmtu.sh | 2 +- tools/testing/selftests/netfilter/Makefile | 2 +- .../testing/selftests/netfilter/xt_string.sh | 128 +++++ .../selftests/pidfd/pidfd_fdinfo_test.c | 2 +- tools/testing/selftests/pidfd/pidfd_test.c | 12 +- .../testing/selftests/resctrl/resctrl_tests.c | 5 + tools/testing/selftests/x86/lam.c | 6 +- tools/tracing/rtla/src/utils.c | 2 +- 650 files changed, 6536 insertions(+), 4426 deletions(-) create mode 100644 drivers/crypto/intel/qat/qat_common/adf_cfg_services.h create mode 100644 tools/perf/util/bpf_skel/vmlinux/.gitignore create mode 100755 tools/testing/selftests/netfilter/xt_string.sh -- 2.20.1

1 602

[PATCH OLK-5.10] ksmbd: delete asynchronous work from list
by Zizhi Wo 23 Nov '23

23 Nov '23

From: Namjae Jeon <linkinjeon(a)kernel.org> mainline inclusion from mainline-v6.3-rc6 commit 3a9b557f44ea8f216aab515a7db20e23f0eb51b9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I6KEWO?from=project-issue CVE: CVE-2023-1193 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When smb2_lock request is canceled by smb2_cancel or smb2_close(), ksmbd is missing deleting async_request_entry async_requests list. Because calling init_smb2_rsp_hdr() in smb2_lock() mark ->synchronous as true and then it will not be deleted in ksmbd_conn_try_dequeue_request(). This patch add release_async_work() to release the ones allocated for async work. Cc: stable(a)vger.kernel.org Signed-off-by: Namjae Jeon <linkinjeon(a)kernel.org> Signed-off-by: Steve French <stfrench(a)microsoft.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/ksmbd/connection.c | 12 +++++------- fs/ksmbd/ksmbd_work.h | 2 +- fs/ksmbd/smb2pdu.c | 28 +++++++++++++++++++++------- fs/ksmbd/smb2pdu.h | 1 + 4 files changed, 28 insertions(+), 15 deletions(-) diff --git a/fs/ksmbd/connection.c b/fs/ksmbd/connection.c index 96d6653dabc1..2a0e472f0826 100644 --- a/fs/ksmbd/connection.c +++ b/fs/ksmbd/connection.c @@ -106,10 +106,8 @@ void ksmbd_conn_enqueue_request(struct ksmbd_work *work) struct ksmbd_conn *conn = work->conn; struct list_head *requests_queue = NULL; - if (conn->ops->get_cmd_val(work) != SMB2_CANCEL_HE) { + if (conn->ops->get_cmd_val(work) != SMB2_CANCEL_HE) requests_queue = &conn->requests; - work->syncronous = true; - } if (requests_queue) { atomic_inc(&conn->req_running); @@ -130,14 +128,14 @@ int ksmbd_conn_try_dequeue_request(struct ksmbd_work *work) if (!work->multiRsp) atomic_dec(&conn->req_running); - spin_lock(&conn->request_lock); if (!work->multiRsp) { + spin_lock(&conn->request_lock); list_del_init(&work->request_entry); - if (work->syncronous == false) - list_del_init(&work->async_request_entry); + spin_unlock(&conn->request_lock); + if (work->asyncronous) + release_async_work(work); ret = 0; } - spin_unlock(&conn->request_lock); wake_up_all(&conn->req_running_q); return ret; diff --git a/fs/ksmbd/ksmbd_work.h b/fs/ksmbd/ksmbd_work.h index 5ece58e40c97..3b1fc8fbf7e1 100644 --- a/fs/ksmbd/ksmbd_work.h +++ b/fs/ksmbd/ksmbd_work.h @@ -68,7 +68,7 @@ struct ksmbd_work { /* Request is encrypted */ bool encrypted:1; /* Is this SYNC or ASYNC ksmbd_work */ - bool syncronous:1; + bool asyncronous:1; bool need_invalidate_rkey:1; unsigned int remote_key; diff --git a/fs/ksmbd/smb2pdu.c b/fs/ksmbd/smb2pdu.c index 29e9c94b2f7b..3d6302853b69 100644 --- a/fs/ksmbd/smb2pdu.c +++ b/fs/ksmbd/smb2pdu.c @@ -519,12 +519,6 @@ int init_smb2_rsp_hdr(struct ksmbd_work *work) rsp_hdr->SessionId = rcv_hdr->SessionId; memcpy(rsp_hdr->Signature, rcv_hdr->Signature, 16); - work->syncronous = true; - if (work->async_id) { - ksmbd_release_id(&conn->async_ida, work->async_id); - work->async_id = 0; - } - return 0; } @@ -682,7 +676,7 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) pr_err("Failed to alloc async message id\n"); return id; } - work->syncronous = false; + work->asyncronous = true; work->async_id = id; rsp_hdr->Id.AsyncId = cpu_to_le64(id); @@ -702,6 +696,24 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) return 0; } +void release_async_work(struct ksmbd_work *work) +{ + struct ksmbd_conn *conn = work->conn; + + spin_lock(&conn->request_lock); + list_del_init(&work->async_request_entry); + spin_unlock(&conn->request_lock); + + work->asyncronous = 0; + work->cancel_fn = NULL; + kfree(work->cancel_argv); + work->cancel_argv = NULL; + if (work->async_id) { + ksmbd_release_id(&conn->async_ida, work->async_id); + work->async_id = 0; + } +} + void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status) { struct smb2_hdr *rsp_hdr; @@ -7091,6 +7103,7 @@ int smb2_lock(struct ksmbd_work *work) work->send_no_response = 1; goto out; } + init_smb2_rsp_hdr(work); smb2_set_err_rsp(work); rsp->hdr.Status = @@ -7107,6 +7120,7 @@ int smb2_lock(struct ksmbd_work *work) spin_lock(&fp->f_lock); list_del(&work->fp_entry); spin_unlock(&fp->f_lock); + release_async_work(work); goto retry; } else if (!rc) { spin_lock(&work->conn->llist_lock); diff --git a/fs/ksmbd/smb2pdu.h b/fs/ksmbd/smb2pdu.h index ba255fd2f77c..089eb93a166f 100644 --- a/fs/ksmbd/smb2pdu.h +++ b/fs/ksmbd/smb2pdu.h @@ -1664,6 +1664,7 @@ int find_matching_smb2_dialect(int start_index, __le16 *cli_dialects, struct file_lock *smb_flock_init(struct file *f); int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg); +void release_async_work(struct ksmbd_work *work); void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status); struct channel *lookup_chann_list(struct ksmbd_session *sess, struct ksmbd_conn *conn); -- 2.39.2

2 1

[PATCH OLK-5.10] ksmbd: delete asynchronous work from list
by Zizhi Wo 23 Nov '23

23 Nov '23

From: Namjae Jeon <linkinjeon(a)kernel.org> mainline inclusion from mainline-v6.3-rc6 commit 3a9b557f44ea8f216aab515a7db20e23f0eb51b9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I6KEWO?from=project-issue CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When smb2_lock request is canceled by smb2_cancel or smb2_close(), ksmbd is missing deleting async_request_entry async_requests list. Because calling init_smb2_rsp_hdr() in smb2_lock() mark ->synchronous as true and then it will not be deleted in ksmbd_conn_try_dequeue_request(). This patch add release_async_work() to release the ones allocated for async work. Cc: stable(a)vger.kernel.org Signed-off-by: Namjae Jeon <linkinjeon(a)kernel.org> Signed-off-by: Steve French <stfrench(a)microsoft.com> Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/ksmbd/connection.c | 12 +++++------- fs/ksmbd/ksmbd_work.h | 2 +- fs/ksmbd/smb2pdu.c | 28 +++++++++++++++++++++------- fs/ksmbd/smb2pdu.h | 1 + 4 files changed, 28 insertions(+), 15 deletions(-) diff --git a/fs/ksmbd/connection.c b/fs/ksmbd/connection.c index 96d6653dabc1..2a0e472f0826 100644 --- a/fs/ksmbd/connection.c +++ b/fs/ksmbd/connection.c @@ -106,10 +106,8 @@ void ksmbd_conn_enqueue_request(struct ksmbd_work *work) struct ksmbd_conn *conn = work->conn; struct list_head *requests_queue = NULL; - if (conn->ops->get_cmd_val(work) != SMB2_CANCEL_HE) { + if (conn->ops->get_cmd_val(work) != SMB2_CANCEL_HE) requests_queue = &conn->requests; - work->syncronous = true; - } if (requests_queue) { atomic_inc(&conn->req_running); @@ -130,14 +128,14 @@ int ksmbd_conn_try_dequeue_request(struct ksmbd_work *work) if (!work->multiRsp) atomic_dec(&conn->req_running); - spin_lock(&conn->request_lock); if (!work->multiRsp) { + spin_lock(&conn->request_lock); list_del_init(&work->request_entry); - if (work->syncronous == false) - list_del_init(&work->async_request_entry); + spin_unlock(&conn->request_lock); + if (work->asyncronous) + release_async_work(work); ret = 0; } - spin_unlock(&conn->request_lock); wake_up_all(&conn->req_running_q); return ret; diff --git a/fs/ksmbd/ksmbd_work.h b/fs/ksmbd/ksmbd_work.h index 5ece58e40c97..3b1fc8fbf7e1 100644 --- a/fs/ksmbd/ksmbd_work.h +++ b/fs/ksmbd/ksmbd_work.h @@ -68,7 +68,7 @@ struct ksmbd_work { /* Request is encrypted */ bool encrypted:1; /* Is this SYNC or ASYNC ksmbd_work */ - bool syncronous:1; + bool asyncronous:1; bool need_invalidate_rkey:1; unsigned int remote_key; diff --git a/fs/ksmbd/smb2pdu.c b/fs/ksmbd/smb2pdu.c index 29e9c94b2f7b..3d6302853b69 100644 --- a/fs/ksmbd/smb2pdu.c +++ b/fs/ksmbd/smb2pdu.c @@ -519,12 +519,6 @@ int init_smb2_rsp_hdr(struct ksmbd_work *work) rsp_hdr->SessionId = rcv_hdr->SessionId; memcpy(rsp_hdr->Signature, rcv_hdr->Signature, 16); - work->syncronous = true; - if (work->async_id) { - ksmbd_release_id(&conn->async_ida, work->async_id); - work->async_id = 0; - } - return 0; } @@ -682,7 +676,7 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) pr_err("Failed to alloc async message id\n"); return id; } - work->syncronous = false; + work->asyncronous = true; work->async_id = id; rsp_hdr->Id.AsyncId = cpu_to_le64(id); @@ -702,6 +696,24 @@ int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg) return 0; } +void release_async_work(struct ksmbd_work *work) +{ + struct ksmbd_conn *conn = work->conn; + + spin_lock(&conn->request_lock); + list_del_init(&work->async_request_entry); + spin_unlock(&conn->request_lock); + + work->asyncronous = 0; + work->cancel_fn = NULL; + kfree(work->cancel_argv); + work->cancel_argv = NULL; + if (work->async_id) { + ksmbd_release_id(&conn->async_ida, work->async_id); + work->async_id = 0; + } +} + void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status) { struct smb2_hdr *rsp_hdr; @@ -7091,6 +7103,7 @@ int smb2_lock(struct ksmbd_work *work) work->send_no_response = 1; goto out; } + init_smb2_rsp_hdr(work); smb2_set_err_rsp(work); rsp->hdr.Status = @@ -7107,6 +7120,7 @@ int smb2_lock(struct ksmbd_work *work) spin_lock(&fp->f_lock); list_del(&work->fp_entry); spin_unlock(&fp->f_lock); + release_async_work(work); goto retry; } else if (!rc) { spin_lock(&work->conn->llist_lock); diff --git a/fs/ksmbd/smb2pdu.h b/fs/ksmbd/smb2pdu.h index ba255fd2f77c..089eb93a166f 100644 --- a/fs/ksmbd/smb2pdu.h +++ b/fs/ksmbd/smb2pdu.h @@ -1664,6 +1664,7 @@ int find_matching_smb2_dialect(int start_index, __le16 *cli_dialects, struct file_lock *smb_flock_init(struct file *f); int setup_async_work(struct ksmbd_work *work, void (*fn)(void **), void **arg); +void release_async_work(struct ksmbd_work *work); void smb2_send_interim_resp(struct ksmbd_work *work, __le32 status); struct channel *lookup_chann_list(struct ksmbd_session *sess, struct ksmbd_conn *conn); -- 2.39.2

2 1

[PATCH OLK-5.10 0/2] ext4: mitigatin cacheline false sharing in struct ext4_inode_info
by Zeng Heng 23 Nov '23

23 Nov '23

Yang Yingliang (2): ext4: mitigatin cacheline false sharing in struct ext4_inode_info enable MITIGATION_FALSE_SHARING by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/ext4/Kconfig | 9 +++++++++ fs/ext4/ext4.h | 5 +++++ fs/ext4/super.c | 4 ++++ 5 files changed, 20 insertions(+) -- 2.25.1

2 7

[PATCH OLK-5.10] LoongArch: disable 40bit user space by default
by Hongchen Zhang 23 Nov '23

23 Nov '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8IRUM ------------------------------------------ Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/configs/loongson3_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/loongarch/configs/loongson3_defconfig b/arch/loongarch/configs/loongson3_defconfig index d2beed31ab74..2f917ae20f76 100644 --- a/arch/loongarch/configs/loongson3_defconfig +++ b/arch/loongarch/configs/loongson3_defconfig @@ -44,6 +44,7 @@ CONFIG_CPU_HAS_LSX=y CONFIG_CPU_HAS_LASX=y CONFIG_NR_CPUS=256 CONFIG_NUMA=y +# CONFIG_VA_BITS_40 is not set CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y CONFIG_PARAVIRT_TIME_ACCOUNTING=y -- 2.33.0

2 5

[PATCH OLK-5.10] LoongArch: disable 40bit user space by default
by Hongchen Zhang 23 Nov '23

23 Nov '23

LoongArch inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8IRUM ------------------------------------------ Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/configs/loongson3_defconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/loongarch/configs/loongson3_defconfig b/arch/loongarch/configs/loongson3_defconfig index d2beed31ab74..2f917ae20f76 100644 --- a/arch/loongarch/configs/loongson3_defconfig +++ b/arch/loongarch/configs/loongson3_defconfig @@ -44,6 +44,7 @@ CONFIG_CPU_HAS_LSX=y CONFIG_CPU_HAS_LASX=y CONFIG_NR_CPUS=256 CONFIG_NUMA=y +# CONFIG_VA_BITS_40 is not set CONFIG_KEXEC=y CONFIG_CRASH_DUMP=y CONFIG_PARAVIRT_TIME_ACCOUNTING=y -- 2.33.0

2 5

[PATCH OLK-6.6 00/30] Backport linux 6.6.1 LTS patches
by Zheng Zengkai 22 Nov '23

22 Nov '23

Backport linux-6.6.1 LTS patches from upstream. git cherry-pick v6.6..v6.6.1~1 -s No conflicts. Andrey Konovalov (1): usb: raw-gadget: properly handle interrupted requests Badhri Jagan Sridharan (1): usb: typec: tcpm: Add additional checks for contaminant Cameron Williams (9): tty: 8250: Remove UC-257 and UC-431 tty: 8250: Add support for additional Brainboxes UC cards tty: 8250: Add support for Brainboxes UP cards tty: 8250: Add support for Intashield IS-100 tty: 8250: Fix port count of PX-257 tty: 8250: Fix up PX-803/PX-857 tty: 8250: Add support for additional Brainboxes PX cards tty: 8250: Add support for Intashield IX cards tty: 8250: Add Brainboxes Oxford Semiconductor-based quirks Daniel Starke (1): tty: n_gsm: fix race condition in status line change on dead connections Francesco Dolcini (1): dt-bindings: serial: rs485: Add rs485-rts-active-high Ian Rogers (1): perf evlist: Avoid frequency mode for the dummy event Janne Grunau (1): Bluetooth: hci_bcm4377: Mark bcm4378/bcm4387 as BROKEN_LE_CODED Jimmy Hu (1): usb: typec: tcpm: Fix NULL pointer dereference in tcpm_pd_svdm() Kai-Heng Feng (1): power: supply: core: Use blocking_notifier_call_chain to avoid RCU complaint LihaSika (1): usb: storage: set 1.50 as the lower bcdDevice for older "Super Top" compatibility Mark Hasemeyer (2): ALSA: hda: intel-dsp-config: Fix JSL Chromebook quirk detection ASoC: SOF: sof-pci-dev: Fix community key quirk detection Max McCarthy (1): ALSA: usb-audio: add quirk flag to enable native DSD for McIntosh devices Nicholas Kazlauskas (1): drm/amd/display: Don't use fsleep for PSR exit waits Siddharth Vadapalli (1): misc: pci_endpoint_test: Add deviceID for J721S2 PCIe EP device support Steven Rostedt (Google) (5): tracing: Have trace_event_file have ref counters eventfs: Remove "is_freed" union with rcu head eventfs: Save ownership and mode eventfs: Delete eventfs_inode when the last dentry is freed eventfs: Use simple_recursive_removal() to clean up dentries Tony Lindgren (1): serial: core: Fix runtime PM handling for pending tx Vicki Pfau (1): PCI: Prevent xHCI driver from claiming AMD VanGogh USB3 DRD device .../devicetree/bindings/serial/rs485.yaml | 4 + drivers/bluetooth/hci_bcm4377.c | 5 + drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c | 3 +- drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c | 3 +- drivers/misc/pci_endpoint_test.c | 4 + drivers/pci/quirks.c | 8 +- drivers/power/supply/power_supply_core.c | 8 +- drivers/tty/n_gsm.c | 2 + drivers/tty/serial/8250/8250_pci.c | 327 +++++++++++++++++- drivers/tty/serial/serial_core.c | 2 +- drivers/usb/gadget/legacy/raw_gadget.c | 26 +- drivers/usb/storage/unusual_cypress.h | 2 +- drivers/usb/typec/tcpm/tcpm.c | 5 + fs/tracefs/event_inode.c | 288 +++++++++------ include/linux/pci_ids.h | 1 + include/linux/power_supply.h | 2 +- include/linux/trace_events.h | 4 + kernel/trace/trace.c | 15 + kernel/trace/trace.h | 3 + kernel/trace/trace_events.c | 31 +- kernel/trace/trace_events_filter.c | 3 + sound/hda/intel-dsp-config.c | 6 + sound/soc/sof/sof-pci-dev.c | 7 + sound/usb/quirks.c | 2 + tools/perf/util/evlist.c | 5 +- 25 files changed, 622 insertions(+), 144 deletions(-) -- 2.20.1

1 30

Kernel SIG 双周例会时间提醒：11.24日下午14:00 // 答复: openEuler Kernel SIG 双周例会
by liaotao (C) 22 Nov '23

22 Nov '23

诚邀您参加openEuler Kernel SIG 双周例会，欢迎您自主申报例会议题~ -----原始约会----- 发件人: openEuler conference <public(a)openeuler.org> 发送时间: 2023年11月10日 14:57 收件人: dev@openeuler.org,kernel-discuss@openeuler.org,kernel@openeuler.org 主题: openEuler Kernel SIG 双周例会时间: 2023年11月24日星期五 14:00-15:45(UTC+08:00) 北京，重庆，香港特别行政区，乌鲁木齐。地点: 您好！ Kernel SIG 邀请您参加 2023-11-24 14:00 召开的WeLink会议(自动录制) 会议主题：openEuler Kernel SIG 双周例会会议内容： 1. 进展update 2. 议题征集中（新增议题可回复邮件申请，也可直接填至会议看板）会议链接：https://bmeeting.huaweicloud.com:36443/#/j/963265866 会议纪要：https://etherpad.openeuler.org/p/Kernel-meetings 温馨提醒：建议接入会议后修改参会人的姓名，也可以使用您在gitee.com的ID 更多资讯尽在：https://openeuler.org/zh/ Hello! openEuler Kernel SIG invites you to attend the WeLink conference(auto recording) will be held at 2023-11-24 14:00, The subject of the conference is openEuler Kernel SIG 双周例会, Summary: 1. 进展update 2. 议题征集中（新增议题可回复邮件申请，也可直接填至会议看板） You can join the meeting at https://bmeeting.huaweicloud.com:36443/#/j/963265866. Add topics at https://etherpad.openeuler.org/p/Kernel-meetings. Note: You are advised to change the participant name after joining the conference or use your ID at gitee.com. More information: https://openeuler.org/en/

1 0

[PATCH OLK-5.10 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 22 Nov '23

22 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[src-openEuler/rdma-core v2] libhns: Get dmac from kernel driver
by Chengchang Tang 22 Nov '23

22 Nov '23

From: Junxian Huang <huangjunxian6(a)hisilicon.com> driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HZ7W -------------------------------------------------------------------------- As dmac is already resolved in kernel while creating AH, there is no need to repeat the resolving in userspace. Prioritizes getting dmac from kernel driver, unless kernel driver didn't response one. Signed-off-by: Junxian Huang <huangjunxian6(a)hisilicon.com> --- kernel-headers/rdma/hns-abi.h | 2 +- providers/hns/hns_roce_u_verbs.c | 10 +++++++--- 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/kernel-headers/rdma/hns-abi.h b/kernel-headers/rdma/hns-abi.h index 785c4e1..8581df9 100644 --- a/kernel-headers/rdma/hns-abi.h +++ b/kernel-headers/rdma/hns-abi.h @@ -135,7 +135,7 @@ struct hns_roce_ib_create_qp_resp { struct hns_roce_ib_create_ah_resp { __u8 priority; __u8 tc_mode; - __u8 reserved[6]; + __u8 dmac[6]; }; struct hns_roce_ib_modify_qp_resp { diff --git a/providers/hns/hns_roce_u_verbs.c b/providers/hns/hns_roce_u_verbs.c index 5e46f89..c906632 100644 --- a/providers/hns/hns_roce_u_verbs.c +++ b/providers/hns/hns_roce_u_verbs.c @@ -2210,9 +2210,13 @@ struct ibv_ah *hns_roce_u_create_ah(struct ibv_pd *pd, struct ibv_ah_attr *attr) if (ibv_cmd_create_ah(pd, &ah->ibv_ah, attr, &resp.ibv_resp, sizeof(resp))) goto err; - if (hr_dev->link_type != HNS_DEV_LINK_TYPE_UB && - ibv_resolve_eth_l2_from_gid(pd->context, attr, ah->av.mac, NULL)) - goto err; + if (hr_dev->link_type != HNS_DEV_LINK_TYPE_UB) { + if (memcmp(ah->av.mac, resp.dmac, ETH_ALEN)) + memcpy(ah->av.mac, resp.dmac, ETH_ALEN); + else if (ibv_resolve_eth_l2_from_gid(pd->context, attr, + ah->av.mac, NULL)) + goto err; + } if (resp.tc_mode == HNS_ROCE_TC_MAP_MODE_DSCP) ah->av.sl = resp.priority; -- 2.30.0

1 0

[OLK-5.10 v3 0/8] A set of bugfixes for openEuler.
by Chengchang Tang 22 Nov '23

22 Nov '23

From: Juan Zhou <zhoujuan51(a)h-partners.com> Junxian Huang (5): RDMA/hns: Fix incorrect congest type configuration RDMA/hns: Fix a missing default value for invalid congest type RDMA/hns: Cleanup of RoCE Bonding driver RDMA/hns: Add a max length of gid table RDMA/hns: Response dmac to userspace Luoyouming (2): RDMA/hns: Fix congestions control algorithm type for UD RDMA/hns: Fix a missing validation check for sl wenglianfa (1): RDMA/hns: Fix simultaneous reset and resource deregistration drivers/infiniband/hw/hns/hns_roce_ah.c | 10 +++ drivers/infiniband/hw/hns/hns_roce_bond.c | 10 +-- drivers/infiniband/hw/hns/hns_roce_cq.c | 26 +++++- drivers/infiniband/hw/hns/hns_roce_db.c | 30 +++++-- drivers/infiniband/hw/hns/hns_roce_device.h | 34 +++++++- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 27 ++++-- drivers/infiniband/hw/hns/hns_roce_main.c | 8 ++ drivers/infiniband/hw/hns/hns_roce_mr.c | 92 ++++++++++++++++++++- drivers/infiniband/hw/hns/hns_roce_qp.c | 61 ++++++++++---- drivers/infiniband/hw/hns/hns_roce_srq.c | 39 +++++++-- include/uapi/rdma/hns-abi.h | 2 +- 11 files changed, 292 insertions(+), 47 deletions(-) -- 2.30.0

1 8

[PATCH OLK-5.10 00/20] psi irq in cgroupv1 and psi fine grained
by Lu Jialin 22 Nov '23

22 Nov '23

Chen Wandun (3): mm: disable psi cgroup v1 by default mm: add config isolation for psi under cgroup v1 psi: dont alloc memory for psi by default Chengming Zhou (3): sched/psi: Fix periodic aggregation shut off sched/psi: Optimize task switch inside shared cgroups again sched/psi: Add PSI_IRQ to track IRQ/SOFTIRQ pressure Hao Jia (1): sched/psi: Zero the memory of struct psi_group Johannes Weiner (1): sched/psi: Remove NR_ONCPU task accounting Lu Jialin (11): psi: support irq.pressure under cgroup v1 psi: enable CONFIG_PSI_CGROUP_V1 in openeuler_defconfig psi: update psi irqtime when the irq delta is nozero memcg: split async memcg reclaim from reclaim_high psi: add struct psi_group_ext PSI: Introduce fine grained stall time collect for cgroup reclaim PSI: Introduce avgs and total calculation for cgroup reclaim PSI: Introduce pressure.stat in psi PSI: add more memory fine grained stall tracking in pressure.stat add cpu fine grained stall tracking in pressure.stat PSI: enable CONFIG_PSI_FINE_GRAINED in openeuler_defconfig Suren Baghdasaryan (1): psi: Fix "defined but not used" warnings when CONFIG_PROC_FS=n Documentation/admin-guide/cgroup-v2.rst | 6 + arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + block/blk-cgroup.c | 2 +- block/blk-core.c | 2 +- include/linux/cgroup-defs.h | 6 +- include/linux/cgroup.h | 2 +- include/linux/psi.h | 9 +- include/linux/psi_types.h | 99 +++- include/linux/sched.h | 4 + init/Kconfig | 20 + kernel/cgroup/cgroup.c | 64 ++- kernel/sched/core.c | 2 + kernel/sched/cpuacct.c | 6 +- kernel/sched/fair.c | 6 - kernel/sched/psi.c | 690 ++++++++++++++++++++---- kernel/sched/stats.h | 9 + mm/compaction.c | 2 +- mm/filemap.c | 4 +- mm/memcontrol.c | 92 ++-- mm/page_alloc.c | 6 + mm/page_io.c | 3 + mm/vmscan.c | 5 +- 23 files changed, 872 insertions(+), 171 deletions(-) -- 2.34.1

1 20

[PATCH openEuler-22.03-LTS 00/17] Add error handle for add_disk
by Li Nan 22 Nov '23

22 Nov '23

Christoph Hellwig (5): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference Luis Chamberlain (4): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling Tetsuo Handa (1): block: check minor range in device_add_disk() Yu Kuai (1): block: fix memory leak for elevator on add_disk failure Zhong Jinghua (6): block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path block/blk.h | 5 +- include/linux/genhd.h | 13 +++ block/blk-integrity.c | 12 ++- block/genhd.c | 246 +++++++++++++++++++++++++----------------- 4 files changed, 173 insertions(+), 103 deletions(-) -- 2.39.2

1 17

[PATCH OLK-6.6 v2] dm ioctl: add DMINFO() to track dm device create/remove
by Li Lingfeng 21 Nov '23

21 Nov '23

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVC5 CVE: NA -------------------------------- Add DMINFO() to help tracking device creation/removal success. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/md/dm-ioctl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 21ebb6c39394..5efe0193b2e8 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -327,6 +327,9 @@ static struct dm_table *__hash_remove(struct hash_cell *hc) table = NULL; if (hc->new_map) table = hc->new_map; + + DMINFO("%s[%i]: %s (%s) is removed successfully", + current->comm, current->pid, hc->md->disk->disk_name, hc->name); dm_put(hc->md); free_cell(hc); @@ -880,6 +883,7 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si { int r, m = DM_ANY_MINOR; struct mapped_device *md; + struct hash_cell *hc; r = check_name(param->name); if (r) @@ -903,6 +907,13 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si __dev_status(md, param); + mutex_lock(&dm_hash_cells_mutex); + hc = dm_get_mdptr(md); + if (hc) + DMINFO("%s[%i]: %s (%s) is created successfully", + current->comm, current->pid, md->disk->disk_name, hc->name); + + mutex_unlock(&dm_hash_cells_mutex); dm_put(md); return 0; -- 2.31.1

2 5

[PATCH V2 OLK-5.10] vhost-vdpa: allow set feature VHOST_F_LOG_ALL when been negotiated.
by Jiang Dongxu 21 Nov '23

21 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- It's not allowed to change the features after vhost-vdpa devices have been negotiated. But log start/end is allowed. Add exception to feature VHOST_F_LOG_ALL. Signed-off-by: jiangdongxu <jiangdongxu1(a)huawei.com> --- drivers/vhost/vdpa.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 8e3bf64123ae..eec8027dfc4f 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -425,16 +425,19 @@ static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep) u64 features; int i; + if (copy_from_user(&features, featurep, sizeof(features))) + return -EFAULT; + + actual_features = ops->get_driver_features(vdpa); + /* * It's not allowed to change the features after they have - * been negotiated. + * been negotiated. But log start/end is allowed. */ - if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) + if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) && + (features & ~(BIT_ULL(VHOST_F_LOG_ALL))) != actual_features) return -EBUSY; - if (copy_from_user(&features, featurep, sizeof(features))) - return -EFAULT; - if (vdpa_set_features(vdpa, features)) return -EINVAL; -- 2.27.0

2 5

[OLK-6.6 v2] dm ioctl: add DMINFO() to track dm device create/remove
by Li Lingfeng 21 Nov '23

21 Nov '23

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVC5 CVE: NA -------------------------------- Add DMINFO() to help tracking device creation/removal success. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/md/dm-ioctl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 21ebb6c39394..5efe0193b2e8 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -327,6 +327,9 @@ static struct dm_table *__hash_remove(struct hash_cell *hc) table = NULL; if (hc->new_map) table = hc->new_map; + + DMINFO("%s[%i]: %s (%s) is removed successfully", + current->comm, current->pid, hc->md->disk->disk_name, hc->name); dm_put(hc->md); free_cell(hc); @@ -880,6 +883,7 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si { int r, m = DM_ANY_MINOR; struct mapped_device *md; + struct hash_cell *hc; r = check_name(param->name); if (r) @@ -903,6 +907,13 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si __dev_status(md, param); + mutex_lock(&dm_hash_cells_mutex); + hc = dm_get_mdptr(md); + if (hc) + DMINFO("%s[%i]: %s (%s) is created successfully", + current->comm, current->pid, md->disk->disk_name, hc->name); + + mutex_unlock(&dm_hash_cells_mutex); dm_put(md); return 0; -- 2.31.1

1 0

[PATCH OLK-5.10] vhost-vdpa: allow set feature VHOST_F_LOG_ALL when been negotiated.
by Jiang Dongxu 21 Nov '23

21 Nov '23

From: jiangdongxu <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- It's not allowed to change the features after vhost-vdpa devices have been negotiated. But log start/end is allowed. Add exception to feature VHOST_F_LOG_ALL. Signed-off-by: jiangdongxu <jiangdongxu1(a)huawei.com> --- drivers/vhost/vdpa.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 8e3bf64123ae..eec8027dfc4f 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -425,16 +425,19 @@ static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep) u64 features; int i; + if (copy_from_user(&features, featurep, sizeof(features))) + return -EFAULT; + + actual_features = ops->get_driver_features(vdpa); + /* * It's not allowed to change the features after they have - * been negotiated. + * been negotiated. But log start/end is allowed. */ - if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) + if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) && + (features & ~(BIT_ULL(VHOST_F_LOG_ALL))) != actual_features) return -EBUSY; - if (copy_from_user(&features, featurep, sizeof(features))) - return -EFAULT; - if (vdpa_set_features(vdpa, features)) return -EINVAL; -- 2.27.0

2 5

[PATCH OLK-6.6] dm ioctl: add DMINFO() to track dm device create/remove
by Li Lingfeng 21 Nov '23

21 Nov '23

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVC5 CVE: NA -------------------------------- Add DMINFO() to help tracking device creation/removal success. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/md/dm-ioctl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 21ebb6c39394..5efe0193b2e8 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -327,6 +327,9 @@ static struct dm_table *__hash_remove(struct hash_cell *hc) table = NULL; if (hc->new_map) table = hc->new_map; + + DMINFO("%s[%i]: %s (%s) is removed successfully", + current->comm, current->pid, hc->md->disk->disk_name, hc->name); dm_put(hc->md); free_cell(hc); @@ -880,6 +883,7 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si { int r, m = DM_ANY_MINOR; struct mapped_device *md; + struct hash_cell *hc; r = check_name(param->name); if (r) @@ -903,6 +907,13 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si __dev_status(md, param); + mutex_lock(&dm_hash_cells_mutex); + hc = dm_get_mdptr(md); + if (hc) + DMINFO("%s[%i]: %s (%s) is created successfully", + current->comm, current->pid, md->disk->disk_name, hc->name); + + mutex_unlock(&dm_hash_cells_mutex); dm_put(md); return 0; -- 2.31.1

2 1

[PATCH OLK-5.10] vhost_vdpa: add reset state params to indicate reset level
by Jiang Dongxu 21 Nov '23

21 Nov '23

From: jiangdongxu1 <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- Add the interface parameter state to the vdpa reset interface, which respectively identifies the reset when the device is turned on/off and the virtio reset issued by the virtual machine. Signed-off-by: jiangdongxu1 <jiangdongxu1(a)huawei.com> --- drivers/vhost/vdpa.c | 10 +++++----- include/linux/vdpa.h | 16 +++++++++++++--- 2 files changed, 18 insertions(+), 8 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index eec8027dfc4f..eed51d004531 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -212,13 +212,13 @@ static void vhost_vdpa_unsetup_vq_irq(struct vhost_vdpa *v, u16 qid) irq_bypass_unregister_producer(&vq->call_ctx.producer); } -static int vhost_vdpa_reset(struct vhost_vdpa *v) +static int vhost_vdpa_reset(struct vhost_vdpa *v, int state) { struct vdpa_device *vdpa = v->vdpa; v->in_batch = 0; - return vdpa_reset(vdpa); + return vdpa_reset(vdpa, state); } static long vhost_vdpa_bind_mm(struct vhost_vdpa *v) @@ -297,7 +297,7 @@ static long vhost_vdpa_set_status(struct vhost_vdpa *v, u8 __user *statusp) vhost_vdpa_unsetup_vq_irq(v, i); if (status == 0) { - ret = vdpa_reset(vdpa); + ret = vdpa_reset(vdpa, VDPA_DEV_RESET_VIRTIO); if (ret) return ret; } else @@ -1496,7 +1496,7 @@ static int vhost_vdpa_open(struct inode *inode, struct file *filep) return r; nvqs = v->nvqs; - r = vhost_vdpa_reset(v); + r = vhost_vdpa_reset(v, VDPA_DEV_RESET_OPEN); if (r) goto err; @@ -1542,7 +1542,7 @@ static int vhost_vdpa_release(struct inode *inode, struct file *filep) mutex_lock(&d->mutex); filep->private_data = NULL; vhost_vdpa_clean_irq(v); - vhost_vdpa_reset(v); + vhost_vdpa_reset(v, VDPA_DEV_RESET_CLOSE); vhost_dev_stop(&v->vdev); vhost_vdpa_unbind_mm(v); vhost_vdpa_config_put(v); diff --git a/include/linux/vdpa.h b/include/linux/vdpa.h index 83df2380eeb9..ef53829c7a7a 100644 --- a/include/linux/vdpa.h +++ b/include/linux/vdpa.h @@ -120,6 +120,12 @@ struct vdpa_map_file { u64 offset; }; +enum vdpa_reset_state { + VDPA_DEV_RESET_VIRTIO = 0, + VDPA_DEV_RESET_OPEN = 1, + VDPA_DEV_RESET_CLOSE = 2, +}; + /** * struct vdpa_config_ops - operations for configuring a vDPA device. * Note: vDPA device drivers are required to implement all of the @@ -218,6 +224,10 @@ struct vdpa_map_file { * @status: virtio device status * @reset: Reset device * @vdev: vdpa device + * @state: state for reset + * VDPA_DEV_RESET_VIRTIO for virtio reset + * VDPA_DEV_RESET_OPEN for vhost-vdpa device open + * VDPA_DEV_RESET_CLOSE for vhost-vdpa device close * Returns integer: success (0) or error (< 0) * @suspend: Suspend the device (optional) * @vdev: vdpa device @@ -359,7 +369,7 @@ struct vdpa_config_ops { u32 (*get_vendor_id)(struct vdpa_device *vdev); u8 (*get_status)(struct vdpa_device *vdev); void (*set_status)(struct vdpa_device *vdev, u8 status); - int (*reset)(struct vdpa_device *vdev); + int (*reset)(struct vdpa_device *vdev, int state); int (*suspend)(struct vdpa_device *vdev); int (*resume)(struct vdpa_device *vdev); size_t (*get_config_size)(struct vdpa_device *vdev); @@ -482,14 +492,14 @@ static inline struct device *vdpa_get_dma_dev(struct vdpa_device *vdev) return vdev->dma_dev; } -static inline int vdpa_reset(struct vdpa_device *vdev) +static inline int vdpa_reset(struct vdpa_device *vdev, int state) { const struct vdpa_config_ops *ops = vdev->config; int ret; down_write(&vdev->cf_lock); vdev->features_valid = false; - ret = ops->reset(vdev); + ret = ops->reset(vdev, state); up_write(&vdev->cf_lock); return ret; } -- 2.27.0

2 4

[PATCH OLK-5.10] vhost-vdpa: allow set feature VHOST_F_LOG_ALL when been negotiated.
by Jiang Dongxu 21 Nov '23

21 Nov '23

From: jiangdongxu1 <jiangdongxu1(a)huawei.com> virt inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I86ITO ---------------------------------------------------------------------- It's not allowed to change the features after vhost-vdpa devices have been negotiated. But log start/end is allowed. Add exception to feature VHOST_F_LOG_ALL. --- drivers/vhost/vdpa.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c index 8e3bf64123ae..eec8027dfc4f 100644 --- a/drivers/vhost/vdpa.c +++ b/drivers/vhost/vdpa.c @@ -425,16 +425,19 @@ static long vhost_vdpa_set_features(struct vhost_vdpa *v, u64 __user *featurep) u64 features; int i; + if (copy_from_user(&features, featurep, sizeof(features))) + return -EFAULT; + + actual_features = ops->get_driver_features(vdpa); + /* * It's not allowed to change the features after they have - * been negotiated. + * been negotiated. But log start/end is allowed. */ - if (ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) + if ((ops->get_status(vdpa) & VIRTIO_CONFIG_S_FEATURES_OK) && + (features & ~(BIT_ULL(VHOST_F_LOG_ALL))) != actual_features) return -EBUSY; - if (copy_from_user(&features, featurep, sizeof(features))) - return -EFAULT; - if (vdpa_set_features(vdpa, features)) return -EINVAL; -- 2.27.0

2 5

[PATCH OLK-6.6 v2 0/3] Add initial openeuler_defconfig for arm64 and
by Zheng Zengkai 21 Nov '23

21 Nov '23

If new options are introduced, but openeuler_defconfig is not explicitly configured, the actual compiled version may be configured according to the default settings, which may be different from the author's expectation. Therefore, some commands/scripts are added to help developers to check and update the defconfig. It is also convenient for continuous integration tools to check the consistency of defconfig. Based on OLK-5.10 openeuler_defconfig, use the commands mentioned above to generate the initial openeuler_defconfig for OLK-6.6 kernel build. Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zheng Zengkai (2): config: add initial openeuler_defconfig for arm64 config: add initial openeuler_defconfig for x86 arch/arm64/configs/openeuler_defconfig | 7942 ++++++++++++++++++++ arch/x86/configs/openeuler_defconfig | 9158 ++++++++++++++++++++++++ scripts/kconfig/Makefile | 22 + scripts/kconfig/makeconfig.sh | 24 + 4 files changed, 17146 insertions(+) create mode 100644 arch/arm64/configs/openeuler_defconfig create mode 100644 arch/x86/configs/openeuler_defconfig create mode 100644 scripts/kconfig/makeconfig.sh -- 2.20.1

2 5

[PATCH OLK-5.10 v3 RESEND 0/3] Introduce CPU inspect feature
by Yu Liao 21 Nov '23

21 Nov '23

v2 -> v3: - fix compilation failure on arm32 - enable CPU inspect for arm64 by default This patches series introduce CPU-inspect feature. CPU-inspect is designed to provide a framework for early detection of SDC by proactively executing CPU inspection test cases. Yu Liao (3): cpuinspect: add CPU-inspect infrastructure cpuinspect: add ATF inspector openeuler_defconfig: enable CPU inspect for arm64 by default arch/arm64/configs/openeuler_defconfig | 7 + drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cpuinspect/Kconfig | 24 +++ drivers/cpuinspect/Makefile | 8 + drivers/cpuinspect/cpuinspect.c | 165 ++++++++++++++++ drivers/cpuinspect/cpuinspect.h | 46 +++++ drivers/cpuinspect/inspector-atf.c | 70 +++++++ drivers/cpuinspect/inspector.c | 124 ++++++++++++ drivers/cpuinspect/sysfs.c | 257 +++++++++++++++++++++++++ include/linux/cpuinspect.h | 40 ++++ 11 files changed, 744 insertions(+) create mode 100644 drivers/cpuinspect/Kconfig create mode 100644 drivers/cpuinspect/Makefile create mode 100644 drivers/cpuinspect/cpuinspect.c create mode 100644 drivers/cpuinspect/cpuinspect.h create mode 100644 drivers/cpuinspect/inspector-atf.c create mode 100644 drivers/cpuinspect/inspector.c create mode 100644 drivers/cpuinspect/sysfs.c create mode 100644 include/linux/cpuinspect.h -- 2.33.0

1 3

[PATCH OLK-5.10] iommu/arm-smmu-v3: Add config isolations for BTM
by Zhang Zekun 20 Nov '23

20 Nov '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8I0NL ------------------------------- Add config isolation for SMMU BTM (Broadcast TLB Maintenance), because VMIDS are all shared and will be invalid all together, which might cause performance issues. Besides, If the hardware fail to invalid TLB, it can cause unexpected errors, you need to do some tests to make sure BTM work as expect. Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- drivers/iommu/Kconfig | 10 ++++++++++ drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 2 ++ 2 files changed, 12 insertions(+) diff --git a/drivers/iommu/Kconfig b/drivers/iommu/Kconfig index b630e58c49b6..f768fbc60dd3 100644 --- a/drivers/iommu/Kconfig +++ b/drivers/iommu/Kconfig @@ -438,6 +438,16 @@ config SMMU_BYPASS_DEV This feature will be replaced by ACPI IORT RMR node, which will be upstreamed in mainline. +config SMMU_BTM_SUPPORT + bool "SMMU BTM (Broadcast TLB Maintenance) support" + depends on ARM_SMMU_V3 + help + ARM SMMU BTM feature can support for receiving broadcast TLBI + operations issued by Arm PEs in the system, which can help to + speed up the tlb invalidation speed in SVA scenarios. + + This feature need support for BTM features, if not sure, say NO. + endif # IOMMU_SUPPORT config IOVA_MAX_GLOBAL_MAGS diff --git a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c index d93ce123df49..09777b05b089 100644 --- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c @@ -4954,8 +4954,10 @@ static int arm_smmu_device_hw_probe(struct arm_smmu_device *smmu) * broadcasted TLB invalidations that target EL2-E2H world. Don't enable * BTM in that case. */ +#ifdef CONFIG_SMMU_BTM_SUPPORT if (reg & IDR0_BTM && (!vhe || reg & IDR0_HYP)) smmu->features |= ARM_SMMU_FEAT_BTM; +#endif /* * The coherency feature as set by FW is used in preference to the ID -- 2.17.1

2 1

[PATCH OLK-5.10 v3 0/3] Introduce CPU inspect feature
by Yu Liao 20 Nov '23

20 Nov '23

v2 -> v3: - fix compilation failure on arm32 - enable CPU inspect for arm64 by default This patches series introduce CPU-inspect feature. CPU-inspect is designed to provide a framework for early detection of SDC by proactively executing CPU inspection test cases. Yu Liao (3): cpuinspect: add CPU-inspect infrastructure cpuinspect: add ATF inspector openeuler_defconfig: enable CPU inspect for arm64 by default arch/arm64/configs/openeuler_defconfig | 7 + drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cpuinspect/Kconfig | 24 +++ drivers/cpuinspect/Makefile | 8 + drivers/cpuinspect/cpuinspect.c | 165 ++++++++++++++++ drivers/cpuinspect/cpuinspect.h | 46 +++++ drivers/cpuinspect/inspector-atf.c | 70 +++++++ drivers/cpuinspect/inspector.c | 124 ++++++++++++ drivers/cpuinspect/sysfs.c | 257 +++++++++++++++++++++++++ include/linux/cpuinspect.h | 40 ++++ 11 files changed, 744 insertions(+) create mode 100644 drivers/cpuinspect/Kconfig create mode 100644 drivers/cpuinspect/Makefile create mode 100644 drivers/cpuinspect/cpuinspect.c create mode 100644 drivers/cpuinspect/cpuinspect.h create mode 100644 drivers/cpuinspect/inspector-atf.c create mode 100644 drivers/cpuinspect/inspector.c create mode 100644 drivers/cpuinspect/sysfs.c create mode 100644 include/linux/cpuinspect.h -- 2.33.0

1 3

[PATCH OLK-5.10 v2 0/2] Introduce CPU inspect feature
by Yu Liao 20 Nov '23

20 Nov '23

This patches series introduce CPU-inspect feature. CPU-inspect is designed to provide a framework for early detection of SDC by proactively executing CPU inspection test cases. v1 -> v2: fix some issues Yu Liao (2): cpuinspect: add CPU-inspect infrastructure cpuinspect: add ATF inspector drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cpuinspect/Kconfig | 24 +++ drivers/cpuinspect/Makefile | 8 + drivers/cpuinspect/cpuinspect.c | 165 ++++++++++++++++++ drivers/cpuinspect/cpuinspect.h | 46 ++++++ drivers/cpuinspect/inspector-atf.c | 70 ++++++++ drivers/cpuinspect/inspector.c | 124 ++++++++++++++ drivers/cpuinspect/sysfs.c | 257 +++++++++++++++++++++++++++++ include/linux/cpuinspect.h | 40 +++++ 10 files changed, 737 insertions(+) create mode 100644 drivers/cpuinspect/Kconfig create mode 100644 drivers/cpuinspect/Makefile create mode 100644 drivers/cpuinspect/cpuinspect.c create mode 100644 drivers/cpuinspect/cpuinspect.h create mode 100644 drivers/cpuinspect/inspector-atf.c create mode 100644 drivers/cpuinspect/inspector.c create mode 100644 drivers/cpuinspect/sysfs.c create mode 100644 include/linux/cpuinspect.h -- 2.33.0

2 3

[PATCH OLK-6.6] dm ioctl: add DMINFO() to track dm device create/remove
by Li Lingfeng 20 Nov '23

20 Nov '23

From: Luo Meng <luomeng12(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVC5 CVE: NA -------------------------------- Add DMINFO() to help tracking device creation/removal success. Signed-off-by: Luo Meng <luomeng12(a)huawei.com> Reviewed-by: Zhang Yi <yi.zhang(a)huawei.com> Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/md/dm-ioctl.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/md/dm-ioctl.c b/drivers/md/dm-ioctl.c index 21ebb6c39394..5efe0193b2e8 100644 --- a/drivers/md/dm-ioctl.c +++ b/drivers/md/dm-ioctl.c @@ -327,6 +327,9 @@ static struct dm_table *__hash_remove(struct hash_cell *hc) table = NULL; if (hc->new_map) table = hc->new_map; + + DMINFO("%s[%i]: %s (%s) is removed successfully", + current->comm, current->pid, hc->md->disk->disk_name, hc->name); dm_put(hc->md); free_cell(hc); @@ -880,6 +883,7 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si { int r, m = DM_ANY_MINOR; struct mapped_device *md; + struct hash_cell *hc; r = check_name(param->name); if (r) @@ -903,6 +907,13 @@ static int dev_create(struct file *filp, struct dm_ioctl *param, size_t param_si __dev_status(md, param); + mutex_lock(&dm_hash_cells_mutex); + hc = dm_get_mdptr(md); + if (hc) + DMINFO("%s[%i]: %s (%s) is created successfully", + current->comm, current->pid, md->disk->disk_name, hc->name); + + mutex_unlock(&dm_hash_cells_mutex); dm_put(md); return 0; -- 2.31.1

1 0

[PATCH openEuler-22.03-LTS-SP2 0/3] add pid enable_swap
by liubo 20 Nov '23

20 Nov '23

Liu Yuntao (1): add "enable_swap" ability Ni Cunshu (1): mm: add preferred_swap ability liubo (1): preferred_swap: share memory can specify swap device fs/proc/base.c | 188 +++++++++++++++++++++++++++++++++ include/linux/mm_types.h | 4 + include/linux/sched/coredump.h | 1 + include/linux/swap.h | 13 +++ kernel/fork.c | 4 + mm/Kconfig | 7 ++ mm/swap_slots.c | 64 ++++++++++- mm/swap_state.c | 1 + mm/swapfile.c | 86 ++++++++++++++- mm/vmscan.c | 38 +++++++ 10 files changed, 403 insertions(+), 3 deletions(-) -- 2.33.0

2 4

[PATCH OLK-5.10 v2 0/2] Fix syntax issues in comments and print
by Li Lingfeng 20 Nov '23

20 Nov '23

Fix syntax issues in comments and print statements v1->v2: Fix syntax issues in comments of the code section that opening an exclusive opened block device for write. Li Lingfeng (2): fs: Fix syntax issues in comments and print statements. fs: Fix syntax issues in comments fs/block_dev.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) -- 2.31.1

2 3

[PATCH OLK-5.10 RESEND 0/2] block: warn once for each partition in bio_check_ro()
by Li Nan 20 Nov '23

20 Nov '23

From: Yu Kuai <yukuai3(a)huawei.com> Yu Kuai (2): block: warn once for each partition in bio_check_ro() block: fix kabi broken in struct hd_part block/blk-core.c | 5 +++++ include/linux/genhd.h | 1 + 2 files changed, 6 insertions(+) -- 2.39.2

1 0

[PATCH OLK-5.10] fs: Fix syntax issues in comments and print statements.
by Li Lingfeng 20 Nov '23

20 Nov '23

hulk inclusion category: cleanup bugzilla: https://gitee.com/openeuler/kernel/issues/I8HWD1 -------------------------------- There are syntax errors in the comments and print statements of the code section that detects opening write opened block devices exclusively, which need to be fixed. Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/block_dev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/block_dev.c b/fs/block_dev.c index a0e4d3ec300e..6389551aa29e 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1669,11 +1669,11 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, void *holder, #ifdef CONFIG_BLK_DEV_DUMPINFO /* * Open an write opened block device exclusively, the - * writing process may probability corrupt the device, + * writing process may probably corrupt the device, * such as a mounted file system, give a hint here. */ if (is_conflict_excl_open(bdev, claiming, mode)) - blkdev_dump_conflict_opener(bdev, "VFS: Open an write opened " + blkdev_dump_conflict_opener(bdev, "VFS: Open a write opened " "block device exclusively"); #endif bd_finish_claiming(bdev, claiming, holder); -- 2.31.1

2 1

[PATCH openEuler-1.0-LTS] mm/migrate.c: fix potential indeterminate pte entry in migrate_vma_insert_page()
by Tong Tiangen 20 Nov '23

20 Nov '23

From: Miaohe Lin <linmiaohe(a)huawei.com> mainline inclusion from mainline-v5.13-rc1 commit 34f5e9b9d1990d286199084efa752530ee3d8297 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8HVSL Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- If the zone device page does not belong to un-addressable device memory, the variable entry will be uninitialized and lead to indeterminate pte entry ultimately. Fix this unexpected case and warn about it. Link: https://lkml.kernel.org/r/20210325131524.48181-4-linmiaohe@huawei.com Fixes: df6ad69838fc ("mm/device-public-memory: device memory cache coherent with CPU") Signed-off-by: Miaohe Lin <linmiaohe(a)huawei.com> Reviewed-by: David Hildenbrand <david(a)redhat.com> Cc: Alistair Popple <apopple(a)nvidia.com> Cc: Jerome Glisse <jglisse(a)redhat.com> Cc: Rafael Aquini <aquini(a)redhat.com> Cc: Yang Shi <shy828301(a)gmail.com> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> conflict: mm/migrate.c Signed-off-by: Tong Tiangen <tongtiangen(a)huawei.com> --- mm/migrate.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/mm/migrate.c b/mm/migrate.c index dc6416ccef44..56a2033d443c 100644 --- a/mm/migrate.c +++ b/mm/migrate.c @@ -2688,6 +2688,13 @@ static void migrate_vma_insert_page(struct migrate_vma *migrate, if (vma->vm_flags & VM_WRITE) entry = pte_mkwrite(pte_mkdirty(entry)); entry = pte_mkdevmap(entry); + } else { + /* + * For now we not support migrating to MEMORY_DEVICE_FS_DAX + * and MEMORY_DEVICE_PCI_P2PDMA device memory. + */ + pr_warn_once("Unsupported ZONE_DEVICE page type.\n"); + goto abort; } } else { entry = mk_pte(page, vma->vm_page_prot); -- 2.25.1

2 1

[PATCH OLK-5.10 0/2] block: warn once for each partition in bio_check_ro()
by Li Nan 20 Nov '23

20 Nov '23

From: Yu Kuai <yukuai3(a)huawei.com> Yu Kuai (2): block: warn once for each partition in bio_check_ro() block: fix kabi broken in struct hd_part block/blk-core.c | 5 +++++ include/linux/genhd.h | 1 + 2 files changed, 6 insertions(+) -- 2.39.2

2 3

[PATCH OLK-5.10] net/tls: do not free tls_rec on async operation in bpf_exec_tx_verdict()
by Liu Jian 18 Nov '23

18 Nov '23

stable inclusion from stable-v5.10.195 commit a5096cc6e7836711541b7cd2d6da48d36fe420e9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8H5DK CVE: CVE-2023-6176 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit cfaa80c91f6f99b9342b6557f0f0e1143e434066 ] I got the below warning when do fuzzing test: BUG: KASAN: null-ptr-deref in scatterwalk_copychunks+0x320/0x470 Read of size 4 at addr 0000000000000008 by task kworker/u8:1/9 CPU: 0 PID: 9 Comm: kworker/u8:1 Tainted: G OE Hardware name: linux,dummy-virt (DT) Workqueue: pencrypt_parallel padata_parallel_worker Call trace: dump_backtrace+0x0/0x420 show_stack+0x34/0x44 dump_stack+0x1d0/0x248 __kasan_report+0x138/0x140 kasan_report+0x44/0x6c __asan_load4+0x94/0xd0 scatterwalk_copychunks+0x320/0x470 skcipher_next_slow+0x14c/0x290 skcipher_walk_next+0x2fc/0x480 skcipher_walk_first+0x9c/0x110 skcipher_walk_aead_common+0x380/0x440 skcipher_walk_aead_encrypt+0x54/0x70 ccm_encrypt+0x13c/0x4d0 crypto_aead_encrypt+0x7c/0xfc pcrypt_aead_enc+0x28/0x84 padata_parallel_worker+0xd0/0x2dc process_one_work+0x49c/0xbdc worker_thread+0x124/0x880 kthread+0x210/0x260 ret_from_fork+0x10/0x18 This is because the value of rec_seq of tls_crypto_info configured by the user program is too large, for example, 0xffffffffffffff. In addition, TLS is asynchronously accelerated. When tls_do_encryption() returns -EINPROGRESS and sk->sk_err is set to EBADMSG due to rec_seq overflow, skmsg is released before the asynchronous encryption process ends. As a result, the UAF problem occurs during the asynchronous processing of the encryption module. If the operation is asynchronous and the encryption module returns EINPROGRESS, do not free the record information. Fixes: 635d93981786 ("net/tls: free record only on encryption error") Signed-off-by: Liu Jian <liujian56(a)huawei.com> Reviewed-by: Sabrina Dubroca <sd(a)queasysnail.net> Link: https://lore.kernel.org/r/20230909081434.2324940-1-liujian56@huawei.com Signed-off-by: Paolo Abeni <pabeni(a)redhat.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/tls/tls_sw.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c index 96c95ea728ac..84e1f2af1a83 100644 --- a/net/tls/tls_sw.c +++ b/net/tls/tls_sw.c @@ -815,7 +815,7 @@ static int bpf_exec_tx_verdict(struct sk_msg *msg, struct sock *sk, psock = sk_psock_get(sk); if (!psock || !policy) { err = tls_push_record(sk, flags, record_type); - if (err && sk->sk_err == EBADMSG) { + if (err && err != -EINPROGRESS && sk->sk_err == EBADMSG) { *copied -= sk_msg_free(sk, msg); tls_free_open_rec(sk); err = -sk->sk_err; @@ -844,7 +844,7 @@ static int bpf_exec_tx_verdict(struct sk_msg *msg, struct sock *sk, switch (psock->eval) { case __SK_PASS: err = tls_push_record(sk, flags, record_type); - if (err && sk->sk_err == EBADMSG) { + if (err && err != -EINPROGRESS && sk->sk_err == EBADMSG) { *copied -= sk_msg_free(sk, msg); tls_free_open_rec(sk); err = -sk->sk_err; -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS-SP2] mm/hugetlb: fix parameter passed to allocate bootmem memory
by Wupeng Ma 18 Nov '23

18 Nov '23

From: Jason Zeng <jason.zeng(a)intel.com> Intel inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7HUWC CVE: NA ----------------------------------------------- Intel-SIG: hugetlb: fix parameter passed to allocate bootmem memory __alloc_bootmem_huge_page_inner() should use 'min_addr" as the 3rd parameter to invoke memblock_alloc_try_nid_raw() and memblock_alloc_try_nid_raw_flags. Fixes: 74bfdf157f1f ("mm/hugetlb: Hugetlb use non-mirrored memory if memory reliable is enabled") Signed-off-by: Jason Zeng <jason.zeng(a)intel.com> [Wupeng: cherry-pick from a78d9e8141261369833de3cd2aa4e7e7880bddaf] Signed-off-by: Ma Wupeng <mawupeng1(a)huawei.com> --- mm/hugetlb.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 7aca3351b9b9..8cf3d5bb4881 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -2713,10 +2713,10 @@ static void *__init __alloc_bootmem_huge_page_inner(phys_addr_t size, int nid) { if (!mem_reliable_is_enabled()) - return memblock_alloc_try_nid_raw(size, align, max_addr, + return memblock_alloc_try_nid_raw(size, align, min_addr, max_addr, nid); - return memblock_alloc_try_nid_raw_flags(size, align, max_addr, max_addr, + return memblock_alloc_try_nid_raw_flags(size, align, min_addr, max_addr, nid, MEMBLOCK_NOMIRROR); } -- 2.25.1

2 1

[PATCH OLK-6.6 0/3] Add initial openeuler_defconfig for arm64 and x86
by Zheng Zengkai 17 Nov '23

17 Nov '23

If new options are introduced, but openeuler_defconfig is not explicitly configured, the actual compiled version may be configured according to the default settings, which may be different from the author's expectation. Therefore, some commands/scripts are added to help developers to check and update the defconfig. It is also convenient for continuous integration tools to check the consistency of defconfig. Based on OLK-5.10 openeuler_defconfig, use the commands mentioned above to generate the initial openeuler_defconfig for OLK-6.6 kernel build. Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zheng Zengkai (2): config: add initial openeuler_defconfig for arm64 config: add initial openeuler_defconfig for x86 arch/arm64/configs/openeuler_defconfig | 7942 ++++++++++++++++++++ arch/x86/configs/openeuler_defconfig | 9158 ++++++++++++++++++++++++ scripts/kconfig/Makefile | 22 + scripts/kconfig/makeconfig.sh | 24 + 4 files changed, 17146 insertions(+) create mode 100644 arch/arm64/configs/openeuler_defconfig create mode 100644 arch/x86/configs/openeuler_defconfig create mode 100644 scripts/kconfig/makeconfig.sh -- 2.20.1

1 3

[PATCH OLK-6.6 0/3] Add initial openeuler_defconfig for arm64 and x86
by Zheng Zengkai 17 Nov '23

17 Nov '23

If new options are introduced, but openeuler_defconfig is not explicitly configured, the actual compiled version may be configured according to the default settings, which may be different from the author's expectation. Therefore, some commands/scripts are added to help developers to check and update the defconfig. It is also convenient for continuous integration tools to check the consistency of defconfig. Based on OLK-5.10 openeuler_defconfig, use the commands mentioned above to generate the initial openeuler_defconfig for OLK-6.6 kernel build. Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zheng Zengkai (2): config: add initial openeuler_defconfig for arm64 config: add initial openeuler_defconfig for x86 arch/arm64/configs/openeuler_defconfig | 7942 ++++++++++++++++++++ arch/x86/configs/openeuler_defconfig | 9158 ++++++++++++++++++++++++ scripts/kconfig/Makefile | 22 + scripts/kconfig/makeconfig.sh | 24 + 4 files changed, 17146 insertions(+) create mode 100644 arch/arm64/configs/openeuler_defconfig create mode 100644 arch/x86/configs/openeuler_defconfig create mode 100644 scripts/kconfig/makeconfig.sh -- 2.20.1

1 3

[PATCH OLK-6.6 0/3] Add initial openeuler_defconfig for arm64 and x86
by Zheng Zengkai 17 Nov '23

17 Nov '23

If new options are introduced, but openeuler_defconfig is not explicitly configured, the actual compiled version may be configured according to the default settings, which may be different from the author's expectation. Therefore, some commands/scripts are added to help developers to check and update the defconfig. It is also convenient for continuous integration tools to check the consistency of defconfig. Based on OLK-5.10 openeuler_defconfig, use the commands mentioned above to generate the initial openeuler_defconfig for OLK-6.6 kernel build. Xie XiuQi (1): kconfig: Add script to check & update openeuler_defconfig Zheng Zengkai (2): config: add initial openeuler_defconfig for arm64 config: add initial openeuler_defconfig for x86 arch/arm64/configs/openeuler_defconfig | 7942 ++++++++++++++++++++ arch/x86/configs/openeuler_defconfig | 9158 ++++++++++++++++++++++++ scripts/kconfig/Makefile | 22 + scripts/kconfig/makeconfig.sh | 24 + 4 files changed, 17146 insertions(+) create mode 100644 arch/arm64/configs/openeuler_defconfig create mode 100644 arch/x86/configs/openeuler_defconfig create mode 100644 scripts/kconfig/makeconfig.sh -- 2.20.1

1 3

[PATCH openEuler-1.0-LTS v3] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 17 Nov '23

17 Nov '23

From: Florian Westphal <fw(a)strlen.de> mainline inclusion from mainline-v6.5-rc1 commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Conflicts: net/netfilter/nf_conntrack_proto_dccp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 52 +++++++++++++++++++++++-- 1 file changed, 49 insertions(+), 3 deletions(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 3ba1f4d9934f..fe17c2ddb6c8 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -433,17 +433,47 @@ static u64 dccp_ack_seq(const struct dccp_hdr *dh) ntohl(dhack->dccph_ack_nr_low); } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + static int dccp_packet(struct nf_conn *ct, const struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; - dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); + dh = skb_header_pointer(skb, dataoff, sizeof(*dh), &_dh.dh); BUG_ON(dh == NULL); + + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (type == DCCP_PKT_RESET && @@ -526,10 +556,20 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, struct sk_buff *skb, unsigned int dataoff, u_int8_t pf, unsigned int hooknum) { + static const unsigned long require_seq48 = 1 << DCCP_PKT_REQUEST | + 1 << DCCP_PKT_RESPONSE | + 1 << DCCP_PKT_CLOSEREQ | + 1 << DCCP_PKT_CLOSE | + 1 << DCCP_PKT_RESET | + 1 << DCCP_PKT_SYNC | + 1 << DCCP_PKT_SYNCACK; struct dccp_hdr _dh, *dh; unsigned int dccp_len = skb->len - dataoff; unsigned int cscov; const char *msg; + u8 type; + + BUILD_BUG_ON(DCCP_PKT_INVALID >= BITS_PER_LONG); dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); if (dh == NULL) { @@ -559,11 +599,17 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, goto out_invalid; } - if (dh->dccph_type >= DCCP_PKT_INVALID) { + type = dh->dccph_type; + if (type >= DCCP_PKT_INVALID) { msg = "nf_ct_dccp: reserved packet type "; goto out_invalid; } + if (test_bit(type, &require_seq48) && !dh->dccph_x) { + msg = "nf_ct_dccp: type lacks 48bit sequence numbers"; + goto out_invalid; + } + return NF_ACCEPT; out_invalid: -- 2.34.1

2 1

[PATCH OLK-5.10] LoongArch: fix two cpu hotplug problem
by Hongchen Zhang 17 Nov '23

17 Nov '23

From: Bibo Mao <maobibo(a)loongson.cn> LoongArch inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8H1QC ------------------------------------------ 1. When cpu is hotplug out, cpu is in idle state and function arch_cpu_idle_dead is called. Timer interrupt for this processor should be disabled, else there will be timer interrupt for the dead cpu. Also this prevents vcpu to schedule out during halt-polling flow when system is running in vm mode, since there is pending timer interrupt. This patch adds detailed implementation for timer shutdown interface, so that timer will be disabled when cpu is plug-out. 2. for woken-up cpu, entry address is 8 bytes, we should check first low 4 bytes and then high 4 bytes. Signed-off-by: Bibo Mao <maobibo(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/kernel/smp.c | 17 +++++++++++++++-- arch/loongarch/kernel/time.c | 25 +++++++++---------------- 2 files changed, 24 insertions(+), 18 deletions(-) diff --git a/arch/loongarch/kernel/smp.c b/arch/loongarch/kernel/smp.c index f5a94559d441..abf6484fac70 100644 --- a/arch/loongarch/kernel/smp.c +++ b/arch/loongarch/kernel/smp.c @@ -310,16 +310,29 @@ void play_dead(void) register void (*init_fn)(void); idle_task_exit(); - local_irq_enable(); + /* + * vcpu can be woken up from idle emulation in vm if irq is disabled + */ + if (!cpu_has_hypervisor) + local_irq_enable(); set_csr_ecfg(ECFGF_IPI); __this_cpu_write(cpu_state, CPU_DEAD); __smp_mb(); do { __asm__ __volatile__("idle 0\n\t"); - addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); + /* + * mailbox info is wroten from other CPU with IPI send method + * in function csr_mail_send, only 4 bytes can be wroten with + * IPI send method in one time. + * + * High 4 bytes is sent and then low 4 bytes for 8 bytes mail + * sending method. Here low 4 bytes is read by the first. + */ + addr = iocsr_read32(LOONGARCH_IOCSR_MBUF0); } while (addr == 0); + addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); init_fn = (void *)TO_CACHE(addr); iocsr_write32(0xffffffff, LOONGARCH_IOCSR_IPI_CLEAR); diff --git a/arch/loongarch/kernel/time.c b/arch/loongarch/kernel/time.c index 18fa38705da7..b2e8108bee10 100644 --- a/arch/loongarch/kernel/time.c +++ b/arch/loongarch/kernel/time.c @@ -59,21 +59,6 @@ static int constant_set_state_oneshot(struct clock_event_device *evt) return 0; } -static int constant_set_state_oneshot_stopped(struct clock_event_device *evt) -{ - unsigned long timer_config; - - raw_spin_lock(&state_lock); - - timer_config = csr_read64(LOONGARCH_CSR_TCFG); - timer_config &= ~CSR_TCFG_EN; - csr_write64(timer_config, LOONGARCH_CSR_TCFG); - - raw_spin_unlock(&state_lock); - - return 0; -} - static int constant_set_state_periodic(struct clock_event_device *evt) { unsigned long period; @@ -93,6 +78,14 @@ static int constant_set_state_periodic(struct clock_event_device *evt) static int constant_set_state_shutdown(struct clock_event_device *evt) { + unsigned long timer_config; + + raw_spin_lock(&state_lock); + timer_config = csr_read64(LOONGARCH_CSR_TCFG); + timer_config &= ~CSR_TCFG_EN; + csr_write64(timer_config, LOONGARCH_CSR_TCFG); + raw_spin_unlock(&state_lock); + return 0; } @@ -161,7 +154,7 @@ int constant_clockevent_init(void) cd->rating = 320; cd->cpumask = cpumask_of(cpu); cd->set_state_oneshot = constant_set_state_oneshot; - cd->set_state_oneshot_stopped = constant_set_state_oneshot_stopped; + cd->set_state_oneshot_stopped = constant_set_state_shutdown; cd->set_state_periodic = constant_set_state_periodic; cd->set_state_shutdown = constant_set_state_shutdown; cd->set_next_event = constant_timer_next_event; -- 2.33.0

2 1

[PATCH openEuler-22.03-LTS 00/22] Add error handle for add_disk
by Li Nan 16 Nov '23

16 Nov '23

To make applying the mainline patch easier, reorder the nbd commit about the first_minor check here. Christoph Hellwig (5): block: fold register_disk into device_add_disk block: call blk_integrity_add earlier in device_add_disk block: add the events* attributes to disk_attrs block: fix error unwinding in device_add_disk block: clear ->slave_dir when dropping the main slave_dir reference Luis Chamberlain (4): block: return errors from blk_integrity_add block: return errors from disk_alloc_events block: add error handling for device_add_disk / add_disk block: fix device_add_disk() kobject_create_and_add() error handling Tetsuo Handa (1): block: check minor range in device_add_disk() Wen Yang (1): Revert "Revert "block: nbd: add sanity check for first_minor"" Yu Kuai (3): nbd: fix max value for 'first_minor' nbd: fix possible overflow for 'first_minor' in nbd_dev_add() block: fix memory leak for elevator on add_disk failure Zhang Wensheng (1): nbd: fix possible overflow on 'first_minor' in nbd_dev_add() Zhong Jinghua (7): nbd: Reorganize the messy commit log about the first_minor check block: return errors from blk_register_region block: Fix the kabi change in device_add_disk block: Fix the kabi change on blk_register_region block: call blk_get_queue earlier in __device_add_disk block: Fix minor range check in device_add_disk() block: Set memalloc_noio to false in the error path block/blk.h | 5 +- include/linux/genhd.h | 13 +++ block/blk-integrity.c | 12 ++- block/genhd.c | 246 +++++++++++++++++++++++++----------------- drivers/block/nbd.c | 5 +- 5 files changed, 175 insertions(+), 106 deletions(-) -- 2.39.2

2 23

[PATCH OLK-5.10 0/2] Introduce CPU inspect feature
by Yu Liao 16 Nov '23

16 Nov '23

This patches series introduce CPU-inspect feature. CPU-inspect is designed to provide a framework for early detection of SDC by proactively executing CPU inspection test cases. Yu Liao (2): cpuinspect: add CPU-inspect infrastructure cpuinspect: add ATF inspector drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/cpuinspect/Kconfig | 24 +++ drivers/cpuinspect/Makefile | 8 + drivers/cpuinspect/cpuinspect.c | 166 ++++++++++++++++++++ drivers/cpuinspect/cpuinspect.h | 46 ++++++ drivers/cpuinspect/inspector-atf.c | 70 +++++++++ drivers/cpuinspect/inspector.c | 124 +++++++++++++++ drivers/cpuinspect/sysfs.c | 236 +++++++++++++++++++++++++++++ include/linux/cpuinspect.h | 40 +++++ 10 files changed, 717 insertions(+) create mode 100644 drivers/cpuinspect/Kconfig create mode 100644 drivers/cpuinspect/Makefile create mode 100644 drivers/cpuinspect/cpuinspect.c create mode 100644 drivers/cpuinspect/cpuinspect.h create mode 100644 drivers/cpuinspect/inspector-atf.c create mode 100644 drivers/cpuinspect/inspector.c create mode 100644 drivers/cpuinspect/sysfs.c create mode 100644 include/linux/cpuinspect.h -- 2.33.0

2 3

[PATCH openEuler-1.0-LTS] sched/rt: Fix double enqueue caused by rt_effective_prio
by Xia Fukun 16 Nov '23

16 Nov '23

From: Peter Zijlstra <peterz(a)infradead.org> mainline inclusion from mainline-v6.5-rc7 commit f558c2b834ec27e75d37b1c860c139e7b7c3a8e4 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8H4SN CVE: NA -------------------------------- Double enqueues in rt runqueues (list) have been reported while running a simple test that spawns a number of threads doing a short sleep/run pattern while being concurrently setscheduled between rt and fair class. WARNING: CPU: 3 PID: 2825 at kernel/sched/rt.c:1294 enqueue_task_rt+0x355/0x360 CPU: 3 PID: 2825 Comm: setsched__13 RIP: 0010:enqueue_task_rt+0x355/0x360 Call Trace: __sched_setscheduler+0x581/0x9d0 _sched_setscheduler+0x63/0xa0 do_sched_setscheduler+0xa0/0x150 __x64_sys_sched_setscheduler+0x1a/0x30 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xae list_add double add: new=ffff9867cb629b40, prev=ffff9867cb629b40, next=ffff98679fc67ca0. kernel BUG at lib/list_debug.c:31! invalid opcode: 0000 [#1] PREEMPT_RT SMP PTI CPU: 3 PID: 2825 Comm: setsched__13 RIP: 0010:__list_add_valid+0x41/0x50 Call Trace: enqueue_task_rt+0x291/0x360 __sched_setscheduler+0x581/0x9d0 _sched_setscheduler+0x63/0xa0 do_sched_setscheduler+0xa0/0x150 __x64_sys_sched_setscheduler+0x1a/0x30 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xae __sched_setscheduler() uses rt_effective_prio() to handle proper queuing of priority boosted tasks that are setscheduled while being boosted. rt_effective_prio() is however called twice per each __sched_setscheduler() call: first directly by __sched_setscheduler() before dequeuing the task and then by __setscheduler() to actually do the priority change. If the priority of the pi_top_task is concurrently being changed however, it might happen that the two calls return different results. If, for example, the first call returned the same rt priority the task was running at and the second one a fair priority, the task won't be removed by the rt list (on_list still set) and then enqueued in the fair runqueue. When eventually setscheduled back to rt it will be seen as enqueued already and the WARNING/BUG be issued. Fix this by calling rt_effective_prio() only once and then reusing the return value. While at it refactor code as well for clarity. Concurrent priority inheritance handling is still safe and will eventually converge to a new state by following the inheritance chain(s). Fixes: 0782e63bc6fe ("sched: Handle priority boosted tasks proper in setscheduler()") [squashed Peterz changes; added changelog] Reported-by: Mark Simmons <msimmons(a)redhat.com> Signed-off-by: Juri Lelli <juri.lelli(a)redhat.com> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Link: https://lkml.kernel.org/r/20210803104501.38333-1-juri.lelli@redhat.com Signed-off-by: Xia Fukun <xiafukun(a)huawei.com> --- kernel/sched/core.c | 82 +++++++++++++++++++-------------------------- 1 file changed, 35 insertions(+), 47 deletions(-) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 7825ceaae0c4..ce0a9026450d 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -784,12 +784,18 @@ void deactivate_task(struct rq *rq, struct task_struct *p, int flags) dequeue_task(rq, p, flags); } -/* - * __normal_prio - return the priority that is based on the static prio - */ -static inline int __normal_prio(struct task_struct *p) +static inline int __normal_prio(int policy, int rt_prio, int nice) { - return p->static_prio; + int prio; + + if (dl_policy(policy)) + prio = MAX_DL_PRIO - 1; + else if (rt_policy(policy)) + prio = MAX_RT_PRIO - 1 - rt_prio; + else + prio = NICE_TO_PRIO(nice); + + return prio; } /* @@ -801,15 +807,7 @@ static inline int __normal_prio(struct task_struct *p) */ static inline int normal_prio(struct task_struct *p) { - int prio; - - if (task_has_dl_policy(p)) - prio = MAX_DL_PRIO-1; - else if (task_has_rt_policy(p)) - prio = MAX_RT_PRIO-1 - p->rt_priority; - else - prio = __normal_prio(p); - return prio; + return __normal_prio(p->policy, p->rt_priority, PRIO_TO_NICE(p->static_prio)); } /* @@ -2327,7 +2325,7 @@ int sched_fork(unsigned long clone_flags, struct task_struct *p) } else if (PRIO_TO_NICE(p->static_prio) < 0) p->static_prio = NICE_TO_PRIO(0); - p->prio = p->normal_prio = __normal_prio(p); + p->prio = p->normal_prio = p->static_prio; set_load_weight(p, false); /* @@ -3795,6 +3793,18 @@ int default_wake_function(wait_queue_entry_t *curr, unsigned mode, int wake_flag } EXPORT_SYMBOL(default_wake_function); +static void __setscheduler_prio(struct task_struct *p, int prio) +{ + if (dl_prio(prio)) + p->sched_class = &dl_sched_class; + else if (rt_prio(prio)) + p->sched_class = &rt_sched_class; + else + p->sched_class = &fair_sched_class; + + p->prio = prio; +} + #ifdef CONFIG_RT_MUTEXES static inline int __rt_effective_prio(struct task_struct *pi_task, int prio) @@ -3909,22 +3919,19 @@ void rt_mutex_setprio(struct task_struct *p, struct task_struct *pi_task) queue_flag |= ENQUEUE_REPLENISH; } else p->dl.dl_boosted = 0; - p->sched_class = &dl_sched_class; } else if (rt_prio(prio)) { if (dl_prio(oldprio)) p->dl.dl_boosted = 0; if (oldprio < prio) queue_flag |= ENQUEUE_HEAD; - p->sched_class = &rt_sched_class; } else { if (dl_prio(oldprio)) p->dl.dl_boosted = 0; if (rt_prio(oldprio)) p->rt.timeout = 0; - p->sched_class = &fair_sched_class; } - p->prio = prio; + __setscheduler_prio(p, prio); if (queued) enqueue_task(rq, p, queue_flag); @@ -4158,27 +4165,6 @@ static void __setscheduler_params(struct task_struct *p, set_load_weight(p, true); } -/* Actually do priority change: must hold pi & rq lock. */ -static void __setscheduler(struct rq *rq, struct task_struct *p, - const struct sched_attr *attr, bool keep_boost) -{ - __setscheduler_params(p, attr); - - /* - * Keep a potential priority boosting if called from - * sched_setscheduler(). - */ - p->prio = normal_prio(p); - if (keep_boost) - p->prio = rt_effective_prio(p, p->prio); - - if (dl_prio(p->prio)) - p->sched_class = &dl_sched_class; - else if (rt_prio(p->prio)) - p->sched_class = &rt_sched_class; - else - p->sched_class = &fair_sched_class; -} /* * Check the target process has a UID that matches the current process's: @@ -4200,10 +4186,8 @@ static int __sched_setscheduler(struct task_struct *p, const struct sched_attr *attr, bool user, bool pi) { - int newprio = dl_policy(attr->sched_policy) ? MAX_DL_PRIO - 1 : - MAX_RT_PRIO - 1 - attr->sched_priority; - int retval, oldprio, oldpolicy = -1, queued, running; - int new_effective_prio, policy = attr->sched_policy; + int oldpolicy = -1, policy = attr->sched_policy; + int retval, oldprio, newprio, queued, running; const struct sched_class *prev_class; struct rq_flags rf; int reset_on_fork; @@ -4399,6 +4383,7 @@ static int __sched_setscheduler(struct task_struct *p, p->sched_reset_on_fork = reset_on_fork; oldprio = p->prio; + newprio = __normal_prio(policy, attr->sched_priority, attr->sched_nice); if (pi) { /* * Take priority boosted tasks into account. If the new @@ -4407,8 +4392,8 @@ static int __sched_setscheduler(struct task_struct *p, * the runqueue. This will be done when the task deboost * itself. */ - new_effective_prio = rt_effective_prio(p, newprio); - if (new_effective_prio == oldprio) + newprio = rt_effective_prio(p, newprio); + if (newprio == oldprio) queue_flags &= ~DEQUEUE_MOVE; } @@ -4420,7 +4405,10 @@ static int __sched_setscheduler(struct task_struct *p, put_prev_task(rq, p); prev_class = p->sched_class; - __setscheduler(rq, p, attr, pi); + if (!(attr->sched_flags & SCHED_FLAG_KEEP_PARAMS)) { + __setscheduler_params(p, attr); + __setscheduler_prio(p, newprio); + } if (queued) { /* -- 2.34.1

2 1

[PATCH OLK-5.10] LoongArch: fix two cpu hotplug problem
by Hongchen Zhang 16 Nov '23

16 Nov '23

From: Bibo Mao <maobibo(a)loongson.cn> LoongArch inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8H1QC ------------------------------------------ 1. When cpu is hotplug out, cpu is in idle state and function arch_cpu_idle_dead is called. Timer interrupt for this processor should be disabled, else there will be timer interrupt for the dead cpu. Also this prevents vcpu to schedule out during halt-polling flow when system is running in vm mode, since there is pending timer interrupt. This patch adds detailed implementation for timer shutdown interface, so that timer will be disabled when cpu is plug-out. 2. for woken-up cpu, entry address is 8 bytes, we should check first low 4 bytes and then high 4 bytes. Signed-off-by: Bibo Mao <maobibo(a)loongson.cn> Signed-off-by: Hongchen Zhang <zhanghongchen(a)loongson.cn> --- arch/loongarch/kernel/smp.c | 17 +++++++++++++++-- arch/loongarch/kernel/time.c | 25 +++++++++---------------- 2 files changed, 24 insertions(+), 18 deletions(-) diff --git a/arch/loongarch/kernel/smp.c b/arch/loongarch/kernel/smp.c index f5a94559d441..abf6484fac70 100644 --- a/arch/loongarch/kernel/smp.c +++ b/arch/loongarch/kernel/smp.c @@ -310,16 +310,29 @@ void play_dead(void) register void (*init_fn)(void); idle_task_exit(); - local_irq_enable(); + /* + * vcpu can be woken up from idle emulation in vm if irq is disabled + */ + if (!cpu_has_hypervisor) + local_irq_enable(); set_csr_ecfg(ECFGF_IPI); __this_cpu_write(cpu_state, CPU_DEAD); __smp_mb(); do { __asm__ __volatile__("idle 0\n\t"); - addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); + /* + * mailbox info is wroten from other CPU with IPI send method + * in function csr_mail_send, only 4 bytes can be wroten with + * IPI send method in one time. + * + * High 4 bytes is sent and then low 4 bytes for 8 bytes mail + * sending method. Here low 4 bytes is read by the first. + */ + addr = iocsr_read32(LOONGARCH_IOCSR_MBUF0); } while (addr == 0); + addr = iocsr_read64(LOONGARCH_IOCSR_MBUF0); init_fn = (void *)TO_CACHE(addr); iocsr_write32(0xffffffff, LOONGARCH_IOCSR_IPI_CLEAR); diff --git a/arch/loongarch/kernel/time.c b/arch/loongarch/kernel/time.c index 18fa38705da7..b2e8108bee10 100644 --- a/arch/loongarch/kernel/time.c +++ b/arch/loongarch/kernel/time.c @@ -59,21 +59,6 @@ static int constant_set_state_oneshot(struct clock_event_device *evt) return 0; } -static int constant_set_state_oneshot_stopped(struct clock_event_device *evt) -{ - unsigned long timer_config; - - raw_spin_lock(&state_lock); - - timer_config = csr_read64(LOONGARCH_CSR_TCFG); - timer_config &= ~CSR_TCFG_EN; - csr_write64(timer_config, LOONGARCH_CSR_TCFG); - - raw_spin_unlock(&state_lock); - - return 0; -} - static int constant_set_state_periodic(struct clock_event_device *evt) { unsigned long period; @@ -93,6 +78,14 @@ static int constant_set_state_periodic(struct clock_event_device *evt) static int constant_set_state_shutdown(struct clock_event_device *evt) { + unsigned long timer_config; + + raw_spin_lock(&state_lock); + timer_config = csr_read64(LOONGARCH_CSR_TCFG); + timer_config &= ~CSR_TCFG_EN; + csr_write64(timer_config, LOONGARCH_CSR_TCFG); + raw_spin_unlock(&state_lock); + return 0; } @@ -161,7 +154,7 @@ int constant_clockevent_init(void) cd->rating = 320; cd->cpumask = cpumask_of(cpu); cd->set_state_oneshot = constant_set_state_oneshot; - cd->set_state_oneshot_stopped = constant_set_state_oneshot_stopped; + cd->set_state_oneshot_stopped = constant_set_state_shutdown; cd->set_state_periodic = constant_set_state_periodic; cd->set_state_shutdown = constant_set_state_shutdown; cd->set_next_event = constant_timer_next_event; -- 2.33.0

2 1

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

2 20

[PATCH OLK-5.10] SCSI: hisi_raid: support SPxxx series RAID/HBA controllers
by zhanglei 16 Nov '23

16 Nov '23

driver inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I89D3P CVE: NA ------------------------------------------ This commit is to support SPxxx RAID/HBA controllers. RAID controllers support RAID 0/1/5/6/10/50/60 modes. HBA controlllers support RAID 0/1/10 modes. RAID/HBA support SAS/SATA HDD/SSD. Signed-off-by: zhanglei <zhanglei48(a)huawei.com> --- Documentation/scsi/hisi_raid.rst | 84 + MAINTAINERS | 7 + arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/hisi_raid/Kconfig | 14 + drivers/scsi/hisi_raid/Makefile | 7 + drivers/scsi/hisi_raid/hiraid.h | 760 +++++ drivers/scsi/hisi_raid/hiraid_main.c | 3982 ++++++++++++++++++++++++ 10 files changed, 4858 insertions(+) create mode 100644 Documentation/scsi/hisi_raid.rst create mode 100644 drivers/scsi/hisi_raid/Kconfig create mode 100644 drivers/scsi/hisi_raid/Makefile create mode 100644 drivers/scsi/hisi_raid/hiraid.h create mode 100644 drivers/scsi/hisi_raid/hiraid_main.c diff --git a/Documentation/scsi/hisi_raid.rst b/Documentation/scsi/hisi_raid.rst new file mode 100644 index 000000000000..523a6763a7fd --- /dev/null +++ b/Documentation/scsi/hisi_raid.rst @@ -0,0 +1,84 @@ +.. SPDX-License-Identifier: GPL-2.0 + +============================================== +hisi_raid - HUAWEI SCSI RAID Controller driver +============================================== + +This file describes the hisi_raid SCSI driver for HUAWEI +RAID controllers. The hisi_raid driver is the first +generation RAID driver. + +For hisi_raid controller support, enable the hisi_raid driver +when configuring the kernel. + +hisi_raid specific entries in /sys +================================= + +hisi_raid host attributes +------------------------ + - /sys/class/scsi_host/host*/csts_pp + - /sys/class/scsi_host/host*/csts_shst + - /sys/class/scsi_host/host*/csts_cfs + - /sys/class/scsi_host/host*/csts_rdy + - /sys/class/scsi_host/host*/fw_version + + The host csts_pp attribute is a read only attribute. This attribute + indicates whether the controller is processing commands. If this attribute + is set to ‘1’, then the controller is processing commands normally. If + this attribute is cleared to ‘0’, then the controller has temporarily stopped + processing commands in order to handle an event (e.g., firmware activation). + + The host csts_shst attribute is a read only attribute. This attribute + indicates status of shutdown processing.The shutdown status values are defined + as: + ====== ============================== + Value Definition + ====== ============================== + 00b Normal operation + 01b Shutdown processing occurring + 10b Shutdown processing complete + 11b Reserved + ====== ============================== + The host csts_cfs attribute is a read only attribute. This attribute is set to + ’1’ when a fatal controller error occurred that could not be communicated in the + appropriate Completion Queue. This bit is cleared to ‘0’ when a fatal controller + error has not occurred. + + The host csts_rdy attribute is a read only attribute. This attribute is set to + ‘1’ when the controller is ready to process submission queue entries. + + The fw_version attribute is read-only and will return the driver version and the + controller firmware version. + +hisi_raid scsi device attributes +------------------------------ + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_level + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_state + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_resync + + The device raid_level attribute is a read only attribute. This attribute indicates + RAID level of scsi device(will dispaly "NA" if scsi device is not virtual disk type). + + The device raid_state attribute is read-only and indicates RAID status of scsi + device(will dispaly "NA" if scsi device is not virtual disk type). + + The device raid_resync attribute is read-only and indicates RAID rebuild processing + of scsi device(will dispaly "NA" if scsi device is not virtual disk type). + +Supported devices +================= + + =================== ======= ======================================= + PCI ID (pci.ids) OEM Product + =================== ======= ======================================= + 19E5:3858 HUAWEI SP186-M-8i(HBA:8Ports) + 19E5:3858 HUAWEI SP186-M-16i(HBA:16Ports) + 19E5:3858 HUAWEI SP186-M-32i(HBA:32Ports) + 19E5:3858 HUAWEI SP186-M-40i(HBA:40Ports) + 19E5:3758 HUAWEI SP686C-M-16i(RAID:16Ports,2G cache) + 19E5:3758 HUAWEI SP686C-M-16i(RAID:16Ports,4G cache) + 19E5:3758 HUAWEI SP686C-MH-32i(RAID:32Ports,4G cache) + 19E5:3758 HUAWEI SP686C-M-40i(RAID:40Ports,2G cache) + 19E5:3758 HUAWEI SP686C-M-40i(RAID:40Ports,4G cache) + =================== ======= ======================================= + diff --git a/MAINTAINERS b/MAINTAINERS index a7815fd1072f..8324f56a2096 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -8070,6 +8070,13 @@ M: Yonglong Liu <liuyonglong(a)huawei.com> S: Supported F: drivers/ptp/ptp_hisi.c +HISI_RAID SCSI RAID DRIVERS +M: Zhang Lei <zhanglei48(a)huawei.com> +L: linux-scsi(a)vger.kernel.org +S: Maintained +F: Documentation/scsi/hisi_raid.rst +F: drivers/scsi/hisi_raid/ + HMM - Heterogeneous Memory Management M: Jérôme Glisse <jglisse(a)redhat.com> L: linux-mm(a)kvack.org diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index ec758f0530c1..b9a50ef6d768 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -2413,6 +2413,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_SCSI_HISI_RAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_MYRB is not set diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 5171aa50736b..43b5294326e6 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -2369,6 +2369,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_SCSI_HISI_RAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_BUSLOGIC is not set diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index a9da1b2dec4a..41ef664cf0ed 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -473,6 +473,7 @@ source "drivers/scsi/megaraid/Kconfig.megaraid" source "drivers/scsi/sssraid/Kconfig" source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/smartpqi/Kconfig" +source "drivers/scsi/hisi_raid/Kconfig" source "drivers/scsi/ufs/Kconfig" config SCSI_HPTIOP diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index c2a1efa16912..8f26dbb5ee37 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -101,6 +101,7 @@ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ obj-$(CONFIG_SCSI_MPT3SAS) += mpt3sas/ +obj-$(CONFIG_SCSI_HISI_RAID) += hisi_raid/ obj-$(CONFIG_SCSI_UFSHCD) += ufs/ obj-$(CONFIG_SCSI_ACARD) += atp870u.o obj-$(CONFIG_SCSI_SUNESP) += esp_scsi.o sun_esp.o diff --git a/drivers/scsi/hisi_raid/Kconfig b/drivers/scsi/hisi_raid/Kconfig new file mode 100644 index 000000000000..d402dc45a7c1 --- /dev/null +++ b/drivers/scsi/hisi_raid/Kconfig @@ -0,0 +1,14 @@ +# +# Kernel configuration file for the hisi_raid +# + +config SCSI_HISI_RAID + tristate "Huawei Hisi_Raid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports hisi_raid SPxxx serial RAID controller, which has + PCI Express Gen4 interface with host and supports SAS/SATA HDD/SSD. + To compile this driver as a module, choose M here: the module will + be called hisi_raid. diff --git a/drivers/scsi/hisi_raid/Makefile b/drivers/scsi/hisi_raid/Makefile new file mode 100644 index 000000000000..b71a675f4190 --- /dev/null +++ b/drivers/scsi/hisi_raid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the hisi_raid drivers. +# + +obj-$(CONFIG_SCSI_HISI_RAID) += hiraid.o + +hiraid-objs := hiraid_main.o diff --git a/drivers/scsi/hisi_raid/hiraid.h b/drivers/scsi/hisi_raid/hiraid.h new file mode 100644 index 000000000000..1ebc3dd3f2ec --- /dev/null +++ b/drivers/scsi/hisi_raid/hiraid.h @@ -0,0 +1,760 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2022 Huawei Technologies Co., Ltd */ + +#ifndef __HIRAID_H_ +#define __HIRAID_H_ + +#define HIRAID_HDD_PD_QD 64 +#define HIRAID_HDD_VD_QD 256 +#define HIRAID_SSD_PD_QD 64 +#define HIRAID_SSD_VD_QD 256 + +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +#define HIRAID_DEFAULT_MAX_CHANNEL 4 +#define HIRAID_DEFAULT_MAX_ID 240 +#define HIRAID_DEFAULT_MAX_LUN_PER_HOST 8 + +#define FUA_MASK 0x08 + +#define HIRAID_IO_SQES 7 +#define HIRAID_IO_CQES 4 +#define PRP_ENTRY_SIZE 8 + +#define EXTRA_POOL_SIZE 256 +#define MAX_EXTRA_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 16 + +#define HIRAID_AQ_DEPTH 128 +#define HIRAID_ASYN_COMMANDS 16 +#define HIRAID_AQ_BLK_MQ_DEPTH (HIRAID_AQ_DEPTH - HIRAID_ASYN_COMMANDS) +#define HIRAID_AQ_MQ_TAG_DEPTH (HIRAID_AQ_BLK_MQ_DEPTH - 1) + +#define HIRAID_ADMIN_QUEUE_NUM 1 +#define HIRAID_PTHRU_CMDS_PERQ 1 +#define HIRAID_TOTAL_PTCMDS(qn) (HIRAID_PTHRU_CMDS_PERQ * (qn)) + +#define HIRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define HIRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define HIRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define HIRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) +#define HIRAID_DEV_DISK_TYPE(attr) ((attr) & 0x1e) + +#define HIRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define HIRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +#define HIRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define HIRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define HIRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define HIRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define HIRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define HIRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define IO_SQE_SIZE sizeof(struct hiraid_scsi_io_cmd) +#define ADMIN_SQE_SIZE sizeof(struct hiraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct hiraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define IO_6_DEFAULT_TX_LEN 256 + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 + +#define MAX_REALTIME_BGTASK_NUM 32 + +#define PCI_VENDOR_ID_HUAWEI_LOGIC 0x19E5 +#define HIRAID_SERVER_DEVICE_HBA_DID 0x3858 +#define HIRAID_SERVER_DEVICE_RAID_DID 0x3758 + +enum { + HIRAID_SC_SUCCESS = 0x0, + HIRAID_SC_INVALID_OPCODE = 0x1, + HIRAID_SC_INVALID_FIELD = 0x2, + + HIRAID_SC_ABORT_LIMIT = 0x103, + HIRAID_SC_ABORT_MISSING = 0x104, + HIRAID_SC_ASYNC_LIMIT = 0x105, + + HIRAID_SC_DNR = 0x4000, +}; + +enum { + HIRAID_REG_CAP = 0x0000, + HIRAID_REG_CC = 0x0014, + HIRAID_REG_CSTS = 0x001c, + HIRAID_REG_AQA = 0x0024, + HIRAID_REG_ASQ = 0x0028, + HIRAID_REG_ACQ = 0x0030, + HIRAID_REG_DBS = 0x1000, +}; + +enum { + HIRAID_CC_ENABLE = 1 << 0, + HIRAID_CC_CSS_NVM = 0 << 4, + HIRAID_CC_MPS_SHIFT = 7, + HIRAID_CC_AMS_SHIFT = 11, + HIRAID_CC_SHN_SHIFT = 14, + HIRAID_CC_IOSQES_SHIFT = 16, + HIRAID_CC_IOCQES_SHIFT = 20, + HIRAID_CC_AMS_RR = 0 << HIRAID_CC_AMS_SHIFT, + HIRAID_CC_SHN_NONE = 0 << HIRAID_CC_SHN_SHIFT, + HIRAID_CC_IOSQES = HIRAID_IO_SQES << HIRAID_CC_IOSQES_SHIFT, + HIRAID_CC_IOCQES = HIRAID_IO_CQES << HIRAID_CC_IOCQES_SHIFT, + HIRAID_CC_SHN_NORMAL = 1 << HIRAID_CC_SHN_SHIFT, + HIRAID_CC_SHN_MASK = 3 << HIRAID_CC_SHN_SHIFT, + HIRAID_CSTS_CFS_SHIFT = 1, + HIRAID_CSTS_SHST_SHIFT = 2, + HIRAID_CSTS_PP_SHIFT = 5, + HIRAID_CSTS_RDY = 1 << 0, + HIRAID_CSTS_SHST_CMPLT = 2 << 2, + HIRAID_CSTS_SHST_MASK = 3 << 2, + HIRAID_CSTS_CFS_MASK = 1 << HIRAID_CSTS_CFS_SHIFT, + HIRAID_CSTS_PP_MASK = 1 << HIRAID_CSTS_PP_SHIFT, +}; + +enum { + HIRAID_ADMIN_DELETE_SQ = 0x00, + HIRAID_ADMIN_CREATE_SQ = 0x01, + HIRAID_ADMIN_DELETE_CQ = 0x04, + HIRAID_ADMIN_CREATE_CQ = 0x05, + HIRAID_ADMIN_ABORT_CMD = 0x08, + HIRAID_ADMIN_SET_FEATURES = 0x09, + HIRAID_ADMIN_ASYNC_EVENT = 0x0c, + HIRAID_ADMIN_GET_INFO = 0xc6, + HIRAID_ADMIN_RESET = 0xc8, +}; + +enum { + HIRAID_GET_CTRL_INFO = 0, + HIRAID_GET_DEVLIST_INFO = 1, +}; + +enum hiraid_rst_type { + HIRAID_RESET_TARGET = 0, + HIRAID_RESET_BUS = 1, +}; + +enum { + HIRAID_ASYN_EVENT_ERROR = 0, + HIRAID_ASYN_EVENT_NOTICE = 2, + HIRAID_ASYN_EVENT_VS = 7, +}; + +enum { + HIRAID_ASYN_DEV_CHANGED = 0x00, + HIRAID_ASYN_FW_ACT_START = 0x01, + HIRAID_ASYN_HOST_PROBING = 0x10, +}; + +enum { + HIRAID_ASYN_TIMESYN = 0x00, + HIRAID_ASYN_FW_ACT_FINISH = 0x02, + HIRAID_ASYN_EVENT_MIN = 0x80, + HIRAID_ASYN_EVENT_MAX = 0xff, +}; + +enum { + HIRAID_CMD_WRITE = 0x01, + HIRAID_CMD_READ = 0x02, + + HIRAID_CMD_NONRW_NONE = 0x80, + HIRAID_CMD_NONRW_TODEV = 0x81, + HIRAID_CMD_NONRW_FROMDEV = 0x82, +}; + +enum { + HIRAID_QUEUE_PHYS_CONTIG = (1 << 0), + HIRAID_CQ_IRQ_ENABLED = (1 << 1), + + HIRAID_FEATURE_NUM_QUEUES = 0x07, + HIRAID_FEATURE_ASYNC_EVENT = 0x0b, + HIRAID_FEATURE_TIMESTAMP = 0x0e, +}; + +enum hiraid_dev_state { + DEV_NEW, + DEV_LIVE, + DEV_RESETTING, + DEV_DELETING, + DEV_DEAD, +}; + +enum { + HIRAID_CARD_HBA, + HIRAID_CARD_RAID, +}; + +enum hiraid_cmd_type { + HIRAID_CMD_ADMIN, + HIRAID_CMD_PTHRU, +}; + +enum { + SQE_FLAG_SGL_METABUF = (1 << 6), + SQE_FLAG_SGL_METASEG = (1 << 7), + SQE_FLAG_SGL_ALL = SQE_FLAG_SGL_METABUF | SQE_FLAG_SGL_METASEG, +}; + +enum hiraid_cmd_state { + CMD_IDLE = 0, + CMD_FLIGHT = 1, + CMD_COMPLETE = 2, + CMD_TIMEOUT = 3, + CMD_TMO_COMPLETE = 4, +}; + +enum { + HIRAID_BSG_ADMIN, + HIRAID_BSG_IOPTHRU, +}; + +enum { + HIRAID_SAS_HDD_VD = 0x04, + HIRAID_SATA_HDD_VD = 0x08, + HIRAID_SAS_SSD_VD = 0x0c, + HIRAID_SATA_SSD_VD = 0x10, + HIRAID_NVME_SSD_VD = 0x14, + HIRAID_SAS_HDD_PD = 0x06, + HIRAID_SATA_HDD_PD = 0x0a, + HIRAID_SAS_SSD_PD = 0x0e, + HIRAID_SATA_SSD_PD = 0x12, + HIRAID_NVME_SSD_PD = 0x16, +}; + +enum { + DISPATCH_BY_CPU, + DISPATCH_BY_DISK, +}; + +struct hiraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __le16 cmd_id; + __le16 status; +}; + +struct hiraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_boot; + __u8 mdts; + __u8 acl; + __u8 asynevent; + __u8 card_type; + __u8 pt_use_sgl; + __u8 rsvd; + __le32 rtd3e; + __u8 sn[32]; + __u8 fw_version[16]; + __u8 rsvd1[4020]; +}; + +struct hiraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct hiraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_extra_pool[MAX_EXTRA_POOL_NUM]; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + u32 db_stride; + u32 __iomem *dbs; + struct rw_semaphore dev_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + u32 scsi_qd; + u32 instance; + struct hiraid_ctrl_info *ctrl_info; + struct hiraid_dev_info *dev_info; + + struct hiraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct hiraid_cmd *io_ptcmds; + struct list_head io_pt_list; + spinlock_t io_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + struct work_struct fwact_work; + + enum hiraid_dev_state state; + spinlock_t state_lock; + + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + u32 last_qcnt; + u8 hdd_dispatch; + + struct request_queue *bsg_queue; +}; + +struct hiraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union hiraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct hiraid_sgl_desc sgl; +}; + +struct hiraid_admin_com_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le32 cdw2[4]; + union hiraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct hiraid_features { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[2]; + union hiraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct hiraid_create_cq { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct hiraid_create_sq { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct hiraid_delete_queue { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct hiraid_get_info { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u32 rsvd2[4]; + union hiraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct hiraid_usr_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union hiraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct hiraid_abort_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct hiraid_reset_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct hiraid_admin_command { + union { + struct hiraid_admin_com_cmd common; + struct hiraid_features features; + struct hiraid_create_cq create_cq; + struct hiraid_create_sq create_sq; + struct hiraid_delete_queue delete_queue; + struct hiraid_get_info get_info; + struct hiraid_abort_cmd abort; + struct hiraid_reset_cmd reset; + struct hiraid_usr_cmd usr_cmd; + }; +}; + +struct hiraid_scsi_io_com_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union hiraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct hiraid_scsi_rw_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union hiraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct hiraid_scsi_nonrw_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union hiraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buf_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct hiraid_scsi_io_cmd { + union { + struct hiraid_scsi_io_com_cmd common; + struct hiraid_scsi_rw_cmd rw; + struct hiraid_scsi_nonrw_cmd nonrw; + }; +}; + +struct hiraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct hiraid_passthru_io_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct hiraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct hiraid_passthru_common_cmd admcmd; + struct hiraid_passthru_io_cmd pthrucmd; + }; +}; + +struct hiraid_cmd { + u16 qid; + u16 cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum hiraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct hiraid_queue { + struct hiraid_dev *hdev; + spinlock_t sq_lock; + + spinlock_t cq_lock ____cacheline_aligned_in_smp; + + void *sq_cmds; + + struct hiraid_completion *cqes; + + dma_addr_t sq_buffer_phy; + dma_addr_t cq_buffer_phy; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + atomic_t inflight; + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + struct dma_pool *prp_small_pool; +}; + +struct hiraid_mapmange { + struct hiraid_queue *hiraidq; + enum hiraid_cmd_state state; + u16 cid; + int page_cnt; + u32 sge_cnt; + u32 len; + bool use_sgl; + dma_addr_t first_dma; + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + struct scatterlist *sgl; + void *list[0]; +}; + +struct hiraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct hiraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + +struct hiraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +struct hiraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct hiraid_dev_info devinfo[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct hiraid_sdev_hostdata { + u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 hwq; + u16 pend_count; +}; + +#endif + diff --git a/drivers/scsi/hisi_raid/hiraid_main.c b/drivers/scsi/hisi_raid/hiraid_main.c new file mode 100644 index 000000000000..b9ffa642479c --- /dev/null +++ b/drivers/scsi/hisi_raid/hiraid_main.c @@ -0,0 +1,3982 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2022 Huawei Technologies Co., Ltd */ + +/* Huawei Raid Series Linux Driver */ + +#define pr_fmt(fmt) "hiraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/version.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> +#include <target/target_core_backend.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> +#include <scsi/sg.h> + +#include "hiraid.h" + +static u32 admin_tmout = 60; +module_param(admin_tmout, uint, 0644); +MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); + +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); + +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); + +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); + +static bool work_mode; +module_param(work_mode, bool, 0444); +MODULE_PARM_DESC(work_mode, "work mode switch, default false for multi hw queues"); + +#define MAX_IO_QUEUES 128 +#define MIN_IO_QUEUES 1 + +static int ioq_num_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < MIN_IO_QUEUES || n > MAX_IO_QUEUES) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops max_hwq_num_ops = { + .set = ioq_num_set, + .get = param_get_uint, +}; + +static u32 max_hwq_num = 128; +module_param_cb(max_hwq_num, &max_hwq_num_ops, &max_hwq_num, 0444); +MODULE_PARM_DESC(max_hwq_num, "max num of hw io queues, should >= 1, default 128"); + +static int io_queue_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops io_queue_depth_ops = { + .set = io_queue_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &io_queue_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static u32 log_debug_switch; +module_param(log_debug_switch, uint, 0644); +MODULE_PARM_DESC(log_debug_switch, "set log state, default zero for switch off"); + +static int extra_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_EXTRA_POOL_NUM) + n = MAX_EXTRA_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = extra_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char extra_pool_num = 4; +module_param_cb(extra_pool_num, &small_pool_num_ops, &extra_pool_num, 0644); +MODULE_PARM_DESC(extra_pool_num, "set prp extra pool num, default 4, MAX 16"); + +static void hiraid_handle_async_notice(struct hiraid_dev *hdev, u32 result); +static void hiraid_handle_async_vs(struct hiraid_dev *hdev, u32 result, u32 result1); + +static struct class *hiraid_class; + +#define HIRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *work_queue; + +#define dev_log_dbg(dev, fmt, ...) do { \ + if (unlikely(log_debug_switch)) \ + dev_info(dev, "[%s] " fmt, \ + __func__, ##__VA_ARGS__); \ +} while (0) + +#define HIRAID_DRV_VERSION "1.1.0.0" + +#define ADMIN_TIMEOUT (admin_tmout * HZ) +#define USRCMD_TIMEOUT (180 * HZ) +#define CTL_RST_TIME (600 * HZ) + +#define HIRAID_WAIT_ABNL_CMD_TIMEOUT 6 +#define HIRAID_WAIT_RST_IO_TIMEOUT 10 + +#define HIRAID_DMA_MSK_BIT_MAX 64 + +#define IOQ_PT_DATA_LEN 4096 +#define IOQ_PT_SGL_DATA_LEN (1024 * 1024) + +#define MAX_CAN_QUEUE (4096 - 1) +#define MIN_CAN_QUEUE (1024 - 1) + +enum SENSE_STATE_CODE { + SENSE_STATE_OK = 0, + SENSE_STATE_NEED_CHECK, + SENSE_STATE_ERROR, + SENSE_STATE_EP_PCIE_ERROR, + SENSE_STATE_NAC_DMA_ERROR, + SENSE_STATE_ABORTED, + SENSE_STATE_NEED_RETRY +}; + +enum { + FW_EH_OK = 0, + FW_EH_DEV_NONE = 0x701 +}; + +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", "FORMATTING", "SANITIZING", + "INITIALIZING", "INITIALIZE_FAIL", "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + +static int hiraid_remap_bar(struct hiraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + HIRAID_REG_DBS; + + return 0; +} + +static int hiraid_dev_map(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "hiraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = hiraid_remap_bar(hdev, HIRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void hiraid_dev_unmap(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int hiraid_pci_enable(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = HIRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, "enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + HIRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "read CSTS register failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + HIRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, HIRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); + hdev->db_stride = 1 << HIRAID_CAP_STRIDE(hdev->cap); + + maskbit = HIRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > HIRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, "err, dma mask invalid[%llu], set to default\n", maskbit); + maskbit = HIRAID_DMA_MSK_BIT_MAX; + } + + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) { + dev_err(hdev->dev, "set dma mask[32] and coherent failed\n"); + goto disable; + } + dev_info(hdev->dev, "set dma mask[32] success\n"); + } else { + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, "allocate one IRQ for setup admin queue failed\n"); + goto disable; + } + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + + +/* + * It is fact that first prp and last prp may be not full page. + * The size to count total nprps for the io equal to size + page_size, + * it may be a slightly overestimate. + * + * 8B per prp address. It may be there is one prp_list address per page, + * prp_list address does not count in io data prps. So divisor equal to + * PAGE_SIZE - 8, it may be a slightly overestimate. + */ +static int hiraid_prp_pagenum(struct hiraid_dev *hdev) +{ + u32 size = 1U << ((hdev->ctrl_info->mdts) * 1U) << 12; + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, hdev->page_size - PRP_ENTRY_SIZE); +} + +/* + * Calculates the number of pages needed for the SGL segments. For example a 4k + * page can accommodate 256 SGL descriptors. + */ +static int hiraid_sgl_pagenum(struct hiraid_dev *hdev) +{ + u32 nsge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + + return DIV_ROUND_UP(nsge * sizeof(struct hiraid_sgl_desc), hdev->page_size); +} + +static inline void **hiraid_mapbuf_list(struct hiraid_mapmange *mapbuf) +{ + return mapbuf->list; +} + +static u32 hiraid_get_max_cmd_size(struct hiraid_dev *hdev) +{ + u32 alloc_size = sizeof(__le64 *) * max(hiraid_prp_pagenum(hdev), hiraid_sgl_pagenum(hdev)); + + dev_info(hdev->dev, "mapbuf size[%lu], alloc_size[%u]\n", + sizeof(struct hiraid_mapmange), alloc_size); + + return sizeof(struct hiraid_mapmange) + alloc_size; +} + +static int hiraid_build_passthru_prp(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + struct scatterlist *sg = mapbuf->sgl; + __le64 *phy_regpage, *prior_list; + u64 buf_addr = sg_dma_address(sg); + int buf_length = sg_dma_len(sg); + u32 page_size = hdev->page_size; + int offset = buf_addr & (page_size - 1); + void **list = hiraid_mapbuf_list(mapbuf); + int maplen = mapbuf->len; + struct dma_pool *pool; + dma_addr_t buffer_phy; + int i; + + maplen -= (page_size - offset); + if (maplen <= 0) { + mapbuf->first_dma = 0; + return 0; + } + + buf_length -= (page_size - offset); + if (buf_length) { + buf_addr += (page_size - offset); + } else { + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + if (maplen <= page_size) { + mapbuf->first_dma = buf_addr; + return 0; + } + + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate first admin prp_list memory failed\n"); + mapbuf->first_dma = buf_addr; + mapbuf->page_cnt = -1; + return -ENOMEM; + } + list[0] = phy_regpage; + mapbuf->first_dma = buffer_phy; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + prior_list = phy_regpage; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th admin prp list memory failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = phy_regpage; + phy_regpage[0] = prior_list[i - 1]; + prior_list[i - 1] = cpu_to_le64(buffer_phy); + i = 1; + } + phy_regpage[i++] = cpu_to_le64(buf_addr); + buf_addr += page_size; + buf_length -= page_size; + maplen -= page_size; + if (maplen <= 0) + break; + if (buf_length > 0) + continue; + if (unlikely(buf_length < 0)) + goto bad_admin_sgl; + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + return 0; + +bad_admin_sgl: + dev_err(hdev->dev, "setup prps, invalid admin SGL for payload[%d] nents[%d]\n", + mapbuf->len, mapbuf->sge_cnt); + return -EIO; +} + +static int hiraid_build_prp(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + struct scatterlist *sg = mapbuf->sgl; + __le64 *phy_regpage, *prior_list; + u64 buf_addr = sg_dma_address(sg); + int buf_length = sg_dma_len(sg); + u32 page_size = hdev->page_size; + int offset = buf_addr & (page_size - 1); + void **list = hiraid_mapbuf_list(mapbuf); + int maplen = mapbuf->len; + struct dma_pool *pool; + dma_addr_t buffer_phy; + int nprps, i; + + maplen -= (page_size - offset); + if (maplen <= 0) { + mapbuf->first_dma = 0; + return 0; + } + + buf_length -= (page_size - offset); + if (buf_length) { + buf_addr += (page_size - offset); + } else { + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + if (maplen <= page_size) { + mapbuf->first_dma = buf_addr; + return 0; + } + + nprps = DIV_ROUND_UP(maplen, page_size); + if (nprps <= (EXTRA_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = mapbuf->hiraidq->prp_small_pool; + mapbuf->page_cnt = 0; + } else { + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + } + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate first prp_list memory failed\n"); + mapbuf->first_dma = buf_addr; + mapbuf->page_cnt = -1; + return -ENOMEM; + } + list[0] = phy_regpage; + mapbuf->first_dma = buffer_phy; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + prior_list = phy_regpage; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th prp list memory failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = phy_regpage; + phy_regpage[0] = prior_list[i - 1]; + prior_list[i - 1] = cpu_to_le64(buffer_phy); + i = 1; + } + phy_regpage[i++] = cpu_to_le64(buf_addr); + buf_addr += page_size; + buf_length -= page_size; + maplen -= page_size; + if (maplen <= 0) + break; + if (buf_length > 0) + continue; + if (unlikely(buf_length < 0)) + goto bad_sgl; + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, "setup prps, invalid SGL for payload[%d] nents[%d]\n", + mapbuf->len, mapbuf->sge_cnt); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct hiraid_sgl_desc)) + +static void hiraid_submit_cmd(struct hiraid_queue *hiraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(hiraidq->qid); + unsigned long flags; + struct hiraid_admin_com_cmd *acd = (struct hiraid_admin_com_cmd *)cmd; + + spin_lock_irqsave(&hiraidq->sq_lock, flags); + memcpy((hiraidq->sq_cmds + sqes * hiraidq->sq_tail), cmd, sqes); + if (++hiraidq->sq_tail == hiraidq->q_depth) + hiraidq->sq_tail = 0; + + writel(hiraidq->sq_tail, hiraidq->q_db); + spin_unlock_irqrestore(&hiraidq->sq_lock, flags); + + dev_log_dbg(hiraidq->hdev->dev, "cid[%d] qid[%d] opcode[0x%x] flags[0x%x] hdid[%u]\n", + le16_to_cpu(acd->cmd_id), hiraidq->qid, acd->opcode, acd->flags, + le32_to_cpu(acd->hdid)); +} + +static inline bool hiraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + return true; + default: + return false; + } +} + +/* + * checks if prps can be built for the IO cmd + */ +static bool hiraid_is_prp(struct hiraid_dev *hdev, struct scatterlist *sgl, u32 nsge) +{ + struct scatterlist *sg = sgl; + u32 page_mask = hdev->page_size - 1; + bool is_prp = true; + u32 i = 0; + + for_each_sg(sgl, sg, nsge, i) { + /* + * Data length of the middle sge multiple of page_size, + * address page_size aligned. + */ + if (i != 0 && i != nsge - 1) { + if ((sg_dma_len(sg) & page_mask) || + (sg_dma_address(sg) & page_mask)) { + is_prp = false; + break; + } + } + + /* + * The first sge addr plus the data length meets + * the page_size alignment. + */ + if (nsge > 1 && i == 0) { + if ((sg_dma_address(sg) + sg_dma_len(sg)) & page_mask) { + is_prp = false; + break; + } + } + + /* The last sge addr meets the page_size alignment. */ + if (nsge > 1 && i == (nsge - 1)) { + if (sg_dma_address(sg) & page_mask) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + HIRAID_SGL_FMT_DATA_DESC = 0x00, + HIRAID_SGL_FMT_SEG_DESC = 0x02, + HIRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + HIRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + HIRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void hiraid_sgl_set_data(struct hiraid_sgl_desc *sge, struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = HIRAID_SGL_FMT_DATA_DESC << 4; +} + +static void hiraid_sgl_set_seg(struct hiraid_sgl_desc *sge, dma_addr_t buffer_phy, int entries) +{ + sge->addr = cpu_to_le64(buffer_phy); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = HIRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = HIRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int hiraid_build_passthru_sgl(struct hiraid_dev *hdev, + struct hiraid_admin_command *admin_cmd, + struct hiraid_mapmange *mapbuf) +{ + struct hiraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = mapbuf->sgl; + void **list = hiraid_mapbuf_list(mapbuf); + struct dma_pool *pool; + int nsge = mapbuf->sge_cnt; + dma_addr_t buffer_phy; + int i = 0; + + admin_cmd->common.flags |= SQE_FLAG_SGL_METABUF; + + if (nsge == 1) { + hiraid_sgl_set_data(&admin_cmd->common.dptr.sgl, sg); + return 0; + } + + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate first admin sgl_list failed\n"); + mapbuf->page_cnt = -1; + return -ENOMEM; + } + + list[0] = sg_list; + mapbuf->first_dma = buffer_phy; + hiraid_sgl_set_seg(&admin_cmd->common.dptr.sgl, buffer_phy, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th admin sgl_list failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + hiraid_sgl_set_seg(link, buffer_phy, nsge); + } + + hiraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + + +static int hiraid_build_sgl(struct hiraid_dev *hdev, struct hiraid_scsi_io_cmd *io_cmd, + struct hiraid_mapmange *mapbuf) +{ + struct hiraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = mapbuf->sgl; + void **list = hiraid_mapbuf_list(mapbuf); + struct dma_pool *pool; + int nsge = mapbuf->sge_cnt; + dma_addr_t buffer_phy; + int i = 0; + + io_cmd->common.flags |= SQE_FLAG_SGL_METABUF; + + if (nsge == 1) { + hiraid_sgl_set_data(&io_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (EXTRA_POOL_SIZE / sizeof(struct hiraid_sgl_desc))) { + pool = mapbuf->hiraidq->prp_small_pool; + mapbuf->page_cnt = 0; + } else { + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate first sgl_list failed\n"); + mapbuf->page_cnt = -1; + return -ENOMEM; + } + + list[0] = sg_list; + mapbuf->first_dma = buffer_phy; + hiraid_sgl_set_seg(&io_cmd->common.dptr.sgl, buffer_phy, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th sgl_list failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + hiraid_sgl_set_seg(link, buffer_phy, nsge); + } + + hiraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define HIRAID_RW_FUA BIT(14) + +static int hiraid_setup_rw_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_rw_cmd *io_cmd, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + io_cmd->opcode = HIRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + io_cmd->opcode = HIRAID_CMD_READ; + } else { + dev_err(hdev->dev, "invalid RW_IO for unsupported data direction[%d]\n", + scmd->sc_data_direction); + WARN_ON(1); + return -EINVAL; + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, "invalid IO for illegal transfer data length[%u]\n", datalength); + WARN_ON(1); + return -EINVAL; + } + + io_cmd->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + io_cmd->nlb = cpu_to_le16((u16)(datalength - 1)); + io_cmd->control = cpu_to_le16(control); + + return 0; +} + +static int hiraid_setup_nonrw_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_nonrw_cmd *io_cmd, struct scsi_cmnd *scmd) +{ + io_cmd->buf_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + io_cmd->opcode = HIRAID_CMD_NONRW_NONE; + break; + case DMA_TO_DEVICE: + io_cmd->opcode = HIRAID_CMD_NONRW_TODEV; + break; + case DMA_FROM_DEVICE: + io_cmd->opcode = HIRAID_CMD_NONRW_FROMDEV; + break; + default: + dev_err(hdev->dev, "invalid NON_IO for unsupported data direction[%d]\n", + scmd->sc_data_direction); + WARN_ON(1); + return -EINVAL; + } + + return 0; +} + +static int hiraid_setup_io_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_io_cmd *io_cmd, struct scsi_cmnd *scmd) +{ + memcpy(io_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + io_cmd->common.cdb_len = scmd->cmd_len; + + if (hiraid_is_rw_scmd(scmd)) + return hiraid_setup_rw_cmd(hdev, &io_cmd->rw, scmd); + else + return hiraid_setup_nonrw_cmd(hdev, &io_cmd->nonrw, scmd); +} + +static inline void hiraid_init_mapbuff(struct hiraid_mapmange *mapbuf) +{ + mapbuf->sge_cnt = 0; + mapbuf->page_cnt = -1; + mapbuf->use_sgl = false; + WRITE_ONCE(mapbuf->state, CMD_IDLE); +} + +static void hiraid_free_mapbuf(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t buffer_phy, next_buffer_phy; + struct hiraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + buffer_phy = mapbuf->first_dma; + if (mapbuf->page_cnt == 0) + dma_pool_free(mapbuf->hiraidq->prp_small_pool, + hiraid_mapbuf_list(mapbuf)[0], buffer_phy); + + for (i = 0; i < mapbuf->page_cnt; i++) { + addr = hiraid_mapbuf_list(mapbuf)[i]; + + if (mapbuf->use_sgl) { + sg_list = addr; + next_buffer_phy = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_buffer_phy = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, buffer_phy); + buffer_phy = next_buffer_phy; + } + + mapbuf->sense_buffer_virt = NULL; + mapbuf->page_cnt = -1; +} + +static int hiraid_io_map_data(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf, + struct scsi_cmnd *scmd, struct hiraid_scsi_io_cmd *io_cmd) +{ + int ret; + + ret = scsi_dma_map(scmd); + if (unlikely(ret < 0)) + return ret; + mapbuf->sge_cnt = ret; + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(mapbuf->sge_cnt == 0)) + return 0; + + mapbuf->len = scsi_bufflen(scmd); + mapbuf->sgl = scsi_sglist(scmd); + mapbuf->use_sgl = !hiraid_is_prp(hdev, mapbuf->sgl, mapbuf->sge_cnt); + + if (mapbuf->use_sgl) { + ret = hiraid_build_sgl(hdev, io_cmd, mapbuf); + } else { + ret = hiraid_build_prp(hdev, mapbuf); + io_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(mapbuf->sgl)); + io_cmd->common.dptr.prp2 = cpu_to_le64(mapbuf->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void hiraid_check_status(struct hiraid_mapmange *mapbuf, struct scsi_cmnd *scmd, + struct hiraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case SENSE_STATE_OK: + set_host_byte(scmd, DID_OK); + break; + case SENSE_STATE_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, + mapbuf->sense_buffer_virt, SCSI_SENSE_BUFFERSIZE); + scmd->result = (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case SENSE_STATE_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case SENSE_STATE_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + dev_warn_ratelimited(mapbuf->hiraidq->hdev->dev, "cid[%d] qid[%d] sdev[%d:%d] opcode[%.2x] bad status[0x%x]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->sq_id), scmd->device->channel, + scmd->device->id, scmd->cmnd[0], le16_to_cpu(cqe->status)); + break; + } +} + +static inline void hiraid_query_scmd_tag(struct scsi_cmnd *scmd, u16 *qid, u16 *cid, + struct hiraid_dev *hdev, struct hiraid_sdev_hostdata *hostdata) +{ + u32 tag = blk_mq_unique_tag(blk_mq_rq_from_pdu((void *)scmd)); + + if (work_mode) { + if ((hdev->hdd_dispatch == DISPATCH_BY_DISK) && (hostdata->hwq != 0)) + *qid = hostdata->hwq; + else + *qid = raw_smp_processor_id() % (hdev->online_queues - 1) + 1; + } else { + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + } + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int hiraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_scsi_io_cmd io_cmd; + struct hiraid_queue *ioq; + u16 hwq, cid; + int ret; + + if (unlikely(hdev->state == DEV_RESETTING)) + return SCSI_MLQUEUE_HOST_BUSY; + + if (unlikely(hdev->state != DEV_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scmd->scsi_done(scmd); + return 0; + } + + if (log_debug_switch) + scsi_print_command(scmd); + + hostdata = sdev->hostdata; + hiraid_query_scmd_tag(scmd, &hwq, &cid, hdev, hostdata); + ioq = &hdev->queues[hwq]; + + if (unlikely(atomic_inc_return(&ioq->inflight) > + (hdev->ioq_depth - HIRAID_PTHRU_CMDS_PERQ))) { + atomic_dec(&ioq->inflight); + return SCSI_MLQUEUE_HOST_BUSY; + } + + memset(&io_cmd, 0, sizeof(io_cmd)); + io_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + io_cmd.rw.cmd_id = cpu_to_le16(cid); + + ret = hiraid_setup_io_cmd(hdev, &io_cmd, scmd); + if (unlikely(ret)) { + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + atomic_dec(&ioq->inflight); + return 0; + } + + ret = cid * SCSI_SENSE_BUFFERSIZE; + if (work_mode) { + mapbuf->sense_buffer_virt = hdev->sense_buffer_virt + ret; + mapbuf->sense_buffer_phy = hdev->sense_buffer_phy + ret; + } else { + mapbuf->sense_buffer_virt = ioq->sense_buffer_virt + ret; + mapbuf->sense_buffer_phy = ioq->sense_buffer_phy + ret; + } + io_cmd.common.sense_addr = cpu_to_le64(mapbuf->sense_buffer_phy); + io_cmd.common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + hiraid_init_mapbuff(mapbuf); + + mapbuf->hiraidq = ioq; + mapbuf->cid = cid; + ret = hiraid_io_map_data(hdev, mapbuf, scmd, &io_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "io map data err\n"); + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + ret = 0; + goto deinit_iobuf; + } + + WRITE_ONCE(mapbuf->state, CMD_FLIGHT); + hiraid_submit_cmd(ioq, &io_cmd); + + return 0; + +deinit_iobuf: + atomic_dec(&ioq->inflight); + hiraid_free_mapbuf(hdev, mapbuf); + return ret; +} + +static int hiraid_match_dev(struct hiraid_dev *hdev, u16 idx, struct scsi_device *sdev) +{ + if (HIRAID_DEV_INFO_FLAG_VALID(hdev->dev_info[idx].flag)) { + if (sdev->channel == hdev->dev_info[idx].channel && + sdev->id == le16_to_cpu(hdev->dev_info[idx].target) && + sdev->lun < hdev->dev_info[idx].lun) { + dev_info(hdev->dev, "match device success, channel:target:lun[%d:%d:%d]\n", + hdev->dev_info[idx].channel, + hdev->dev_info[idx].target, + hdev->dev_info[idx].lun); + return 1; + } + } + + return 0; +} + +static int hiraid_disk_qd(u8 attr) +{ + switch (HIRAID_DEV_DISK_TYPE(attr)) { + case HIRAID_SAS_HDD_VD: + case HIRAID_SATA_HDD_VD: + return HIRAID_HDD_VD_QD; + case HIRAID_SAS_SSD_VD: + case HIRAID_SATA_SSD_VD: + case HIRAID_NVME_SSD_VD: + return HIRAID_SSD_VD_QD; + case HIRAID_SAS_HDD_PD: + case HIRAID_SATA_HDD_PD: + return HIRAID_HDD_PD_QD; + case HIRAID_SAS_SSD_PD: + case HIRAID_SATA_SSD_PD: + case HIRAID_NVME_SSD_PD: + return HIRAID_SSD_PD_QD; + default: + return MAX_CMD_PER_DEV; + } +} + +static bool hiraid_disk_is_hdd(u8 attr) +{ + switch (HIRAID_DEV_DISK_TYPE(attr)) { + case HIRAID_SAS_HDD_VD: + case HIRAID_SATA_HDD_VD: + case HIRAID_SAS_HDD_PD: + case HIRAID_SATA_HDD_PD: + return true; + default: + return false; + } +} + +static int hiraid_slave_alloc(struct scsi_device *sdev) +{ + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->dev_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (hiraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->dev_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->dev_info[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->dev_info[idx].max_io_kb); + hostdata->attr = hdev->dev_info[idx].attr; + hostdata->flag = hdev->dev_info[idx].flag; + hostdata->rg_id = 0xff; + sdev->hostdata = hostdata; + up_read(&hdev->dev_rwsem); + return 0; +} + +static void hiraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int hiraid_slave_configure(struct scsi_device *sdev) +{ + unsigned int timeout = scmd_tmout_rawdisk * HZ; + struct hiraid_dev *hdev = shost_priv(sdev->host); + struct hiraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + int qd = MAX_CMD_PER_DEV; + + if (hostdata) { + if (HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (HIRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; + qd = hiraid_disk_qd(hostdata->attr); + + if (hiraid_disk_is_hdd(hostdata->attr)) + hostdata->hwq = hostdata->hdid % (hdev->online_queues - 1) + 1; + else + hostdata->hwq = 0; + } else { + dev_err(hdev->dev, "err, sdev->hostdata is null\n"); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + scsi_change_queue_depth(sdev, qd); + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + dev_info(hdev->dev, "sdev->channel:id:lun[%d:%d:%lld] scmd_timeout[%d]s maxsec[%d]\n", + sdev->channel, sdev->id, sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void hiraid_shost_init(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = work_mode ? 1 : hdev->online_queues - 1; + hdev->shost->can_queue = hdev->scsi_qd; + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = hiraid_get_max_cmd_size(hdev); +} + +static int hiraid_alloc_queue(struct hiraid_dev *hdev, u16 qid, u16 depth) +{ + struct hiraid_queue *hiraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "warn: queue[%d] is exist\n", qid); + return 0; + } + + hiraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &hiraidq->cq_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hiraidq->cqes) + return -ENOMEM; + + hiraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &hiraidq->sq_buffer_phy, GFP_KERNEL); + if (!hiraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + /* + * if single hw queue, we do not need to alloc sense buffer for every queue, + * we have alloced all on hiraid_alloc_resources. + */ + if (work_mode) + goto initq; + + /* alloc sense buffer */ + hiraidq->sense_buffer_virt = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &hiraidq->sense_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hiraidq->sense_buffer_virt) { + ret = -ENOMEM; + goto free_sq_cmds; + } + +initq: + spin_lock_init(&hiraidq->sq_lock); + spin_lock_init(&hiraidq->cq_lock); + hiraidq->hdev = hdev; + hiraidq->q_depth = depth; + hiraidq->qid = qid; + hiraidq->cq_vector = -1; + hdev->queue_count++; + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), (void *)hiraidq->sq_cmds, + hiraidq->sq_buffer_phy); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)hiraidq->cqes, + hiraidq->cq_buffer_phy); + return ret; +} + +static int hiraid_wait_control_ready(struct hiraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((HIRAID_CAP_TIMEOUT(cap) + 1) * HIRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? HIRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int hiraid_shutdown_control(struct hiraid_dev *hdev) +{ + unsigned long timeout = le32_to_cpu(hdev->ctrl_info->rtd3e) / 1000000 * HZ + jiffies; + + hdev->ctrl_config &= ~HIRAID_CC_SHN_MASK; + hdev->ctrl_config |= HIRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + while ((readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_SHST_MASK) != + HIRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "device shutdown incomplete, abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int hiraid_disable_control(struct hiraid_dev *hdev) +{ + hdev->ctrl_config &= ~HIRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~HIRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + return hiraid_wait_control_ready(hdev, hdev->cap, false); +} + +static int hiraid_enable_control(struct hiraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = HIRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, "minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, HIRAID_CAP_MPSMAX(cap) + 12, PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = HIRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << HIRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= HIRAID_CC_AMS_RR | HIRAID_CC_SHN_NONE; + hdev->ctrl_config |= HIRAID_CC_IOSQES | HIRAID_CC_IOCQES; + hdev->ctrl_config |= HIRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + return hiraid_wait_control_ready(hdev, cap, true); +} + +static void hiraid_init_queue(struct hiraid_queue *hiraidq, u16 qid) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + + memset((void *)hiraidq->cqes, 0, CQ_SIZE(hiraidq->q_depth)); + + hiraidq->sq_tail = 0; + hiraidq->cq_head = 0; + hiraidq->cq_phase = 1; + hiraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + hiraidq->prp_small_pool = hdev->prp_extra_pool[qid % extra_pool_num]; + hdev->online_queues++; + atomic_set(&hiraidq->inflight, 0); +} + +static inline bool hiraid_cqe_pending(struct hiraid_queue *hiraidq) +{ + return (le16_to_cpu(hiraidq->cqes[hiraidq->cq_head].status) & 1) == + hiraidq->cq_phase; +} + +static void hiraid_complete_io_cmnd(struct hiraid_queue *ioq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct hiraid_mapmange *mapbuf; + struct request *req; + unsigned long elapsed; + + atomic_dec(&ioq->inflight); + + if (work_mode) + tags = hdev->shost->tag_set.tags[0]; + else + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, le16_to_cpu(cqe->cmd_id)); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "invalid id[%d] completed on queue[%d]\n", + le16_to_cpu(cqe->cmd_id), ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + mapbuf = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", + le16_to_cpu(cqe->cmd_id), ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_COMPLETE) != CMD_FLIGHT) { + dev_warn(hdev->dev, "cid[%d] qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", + le16_to_cpu(cqe->cmd_id), ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(mapbuf->state, CMD_TMO_COMPLETE); + + if (mapbuf->sge_cnt) { + mapbuf->sge_cnt = 0; + scsi_dma_unmap(scmd); + } + hiraid_free_mapbuf(hdev, mapbuf); + + return; + } + + hiraid_check_status(mapbuf, scmd, cqe); + if (mapbuf->sge_cnt) { + mapbuf->sge_cnt = 0; + scsi_dma_unmap(scmd); + } + hiraid_free_mapbuf(hdev, mapbuf); + scmd->scsi_done(scmd); +} + +static void hiraid_complete_admin_cmnd(struct hiraid_queue *adminq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = adminq->hdev; + struct hiraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + le16_to_cpu(cqe->cmd_id); + if (unlikely(adm_cmd->state == CMD_IDLE)) { + dev_warn(adminq->hdev->dev, "invalid id[%d] completed on queue[%d]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void hiraid_send_async_event(struct hiraid_dev *hdev, u16 cid); + +static void hiraid_complete_async_event(struct hiraid_queue *hiraidq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "recv async event, cid[%d] status[0x%x] result[0x%x]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->status) >> 1, result); + + hiraid_send_async_event(hdev, le16_to_cpu(cqe->cmd_id)); + + if ((le16_to_cpu(cqe->status) >> 1) != HIRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case HIRAID_ASYN_EVENT_NOTICE: + hiraid_handle_async_notice(hdev, result); + break; + case HIRAID_ASYN_EVENT_VS: + hiraid_handle_async_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "unsupported async event type[%u]\n", result & 0x7); + break; + } +} + +static void hiraid_complete_pthru_cmnd(struct hiraid_queue *ioq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = ioq->hdev; + struct hiraid_cmd *ptcmd; + + ptcmd = hdev->io_ptcmds + (ioq->qid - 1) * HIRAID_PTHRU_CMDS_PERQ + + le16_to_cpu(cqe->cmd_id) - hdev->scsi_qd; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void hiraid_handle_cqe(struct hiraid_queue *hiraidq, u16 idx) +{ + struct hiraid_completion *cqe = &hiraidq->cqes[idx]; + struct hiraid_dev *hdev = hiraidq->hdev; + u16 cid = le16_to_cpu(cqe->cmd_id); + + if (unlikely(!work_mode && (cid >= hiraidq->q_depth))) { + dev_err(hdev->dev, "invalid command id[%d] completed on queue[%d]\n", + cid, cqe->sq_id); + return; + } + + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] result[0x%x] sqid[%d] status[0x%x]\n", + cid, hiraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(hiraidq->qid == 0 && cid >= HIRAID_AQ_BLK_MQ_DEPTH)) { + hiraid_complete_async_event(hiraidq, cqe); + return; + } + + if (unlikely(hiraidq->qid && cid >= hdev->scsi_qd)) { + hiraid_complete_pthru_cmnd(hiraidq, cqe); + return; + } + + if (hiraidq->qid) + hiraid_complete_io_cmnd(hiraidq, cqe); + else + hiraid_complete_admin_cmnd(hiraidq, cqe); +} + +static void hiraid_complete_cqes(struct hiraid_queue *hiraidq, u16 start, u16 end) +{ + while (start != end) { + hiraid_handle_cqe(hiraidq, start); + if (++start == hiraidq->q_depth) + start = 0; + } +} + +static inline void hiraid_update_cq_head(struct hiraid_queue *hiraidq) +{ + if (++hiraidq->cq_head == hiraidq->q_depth) { + hiraidq->cq_head = 0; + hiraidq->cq_phase = !hiraidq->cq_phase; + } +} + +static inline bool hiraid_process_cq(struct hiraid_queue *hiraidq, u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = hiraidq->cq_head; + while (!found && hiraid_cqe_pending(hiraidq)) { + if (le16_to_cpu(hiraidq->cqes[hiraidq->cq_head].cmd_id) == tag) + found = true; + hiraid_update_cq_head(hiraidq); + } + *end = hiraidq->cq_head; + + if (*start != *end) + writel(hiraidq->cq_head, hiraidq->q_db + hiraidq->hdev->db_stride); + + return found; +} + +static bool hiraid_poll_cq(struct hiraid_queue *hiraidq, int cid) +{ + u16 start, end; + bool found; + + if (!hiraid_cqe_pending(hiraidq)) + return 0; + + spin_lock_irq(&hiraidq->cq_lock); + found = hiraid_process_cq(hiraidq, &start, &end, cid); + spin_unlock_irq(&hiraidq->cq_lock); + + hiraid_complete_cqes(hiraidq, start, end); + return found; +} + +static irqreturn_t hiraid_handle_irq(int irq, void *data) +{ + struct hiraid_queue *hiraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&hiraidq->cq_lock); + if (hiraidq->cq_head != hiraidq->last_cq_head) + ret = IRQ_HANDLED; + + hiraid_process_cq(hiraidq, &start, &end, -1); + hiraidq->last_cq_head = hiraidq->cq_head; + spin_unlock(&hiraidq->cq_lock); + + if (start != end) { + hiraid_complete_cqes(hiraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int hiraid_setup_admin_queue(struct hiraid_dev *hdev) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "start disable controller\n"); + + ret = hiraid_disable_control(hdev); + if (ret) + return ret; + + ret = hiraid_alloc_queue(hdev, 0, HIRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + HIRAID_REG_AQA); + lo_hi_writeq(adminq->sq_buffer_phy, hdev->bar + HIRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_buffer_phy, hdev->bar + HIRAID_REG_ACQ); + + dev_info(hdev->dev, "start enable controller\n"); + + ret = hiraid_enable_control(hdev); + if (ret) { + ret = -ENODEV; + return ret; + } + + adminq->cq_vector = 0; + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, hiraid_handle_irq, NULL, + adminq, "hiraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + adminq->cq_vector = -1; + return ret; + } + + hiraid_init_queue(adminq, 0); + + dev_info(hdev->dev, "setup admin queue success, queuecount[%d] online[%d] pagesize[%d]\n", + hdev->queue_count, hdev->online_queues, hdev->page_size); + + return 0; +} + +static u32 hiraid_get_bar_size(struct hiraid_dev *hdev, u32 nr_ioqs) +{ + return (HIRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int hiraid_create_admin_cmds(struct hiraid_dev *hdev) +{ + u16 i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(HIRAID_AQ_BLK_MQ_DEPTH, sizeof(struct hiraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < HIRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "alloc admin cmds success, num[%d]\n", HIRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void hiraid_free_admin_cmds(struct hiraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct hiraid_cmd *hiraid_get_cmd(struct hiraid_dev *hdev, enum hiraid_cmd_type type) +{ + struct hiraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == HIRAID_CMD_PTHRU) { + head = &hdev->io_pt_list; + slock = &hdev->io_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct hiraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, CMD_FLIGHT); + + return cmd; +} + +static void hiraid_put_cmd(struct hiraid_dev *hdev, struct hiraid_cmd *cmd, + enum hiraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == HIRAID_CMD_PTHRU) { + head = &hdev->io_pt_list; + slock = &hdev->io_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + +static bool hiraid_admin_need_reset(struct hiraid_admin_command *cmd) +{ + switch (cmd->common.opcode) { + case HIRAID_ADMIN_DELETE_SQ: + case HIRAID_ADMIN_CREATE_SQ: + case HIRAID_ADMIN_DELETE_CQ: + case HIRAID_ADMIN_CREATE_CQ: + case HIRAID_ADMIN_SET_FEATURES: + return false; + default: + return true; + } +} + +static int hiraid_reset_work_sync(struct hiraid_dev *hdev); +static inline void hiraid_admin_timeout(struct hiraid_dev *hdev, struct hiraid_cmd *cmd) +{ + /* command may be returned because controller reset */ + if (READ_ONCE(cmd->state) == CMD_COMPLETE) + return; + if (hiraid_reset_work_sync(hdev) == -EBUSY) + flush_work(&hdev->reset_work); +} + +static int hiraid_put_admin_sync_request(struct hiraid_dev *hdev, struct hiraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct hiraid_cmd *adm_cmd = hiraid_get_cmd(hdev, HIRAID_CMD_ADMIN); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.cmd_id = cpu_to_le16(adm_cmd->cid); + hiraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + adm_cmd->cid, adm_cmd->qid, cmd->usr_cmd.opcode, + cmd->usr_cmd.info_0.subopcode); + + /* reset controller if admin timeout */ + if (hiraid_admin_need_reset(cmd)) + hiraid_admin_timeout(hdev, adm_cmd); + + hiraid_put_cmd(hdev, adm_cmd, HIRAID_CMD_ADMIN); + return -ETIME; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + hiraid_put_cmd(hdev, adm_cmd, HIRAID_CMD_ADMIN); + + return adm_cmd->status; +} + +/** + * hiraid_create_cq - send cmd to controller for create controller cq + */ +static int hiraid_create_complete_queue(struct hiraid_dev *hdev, u16 qid, + struct hiraid_queue *hiraidq, u16 cq_vector) +{ + struct hiraid_admin_command admin_cmd; + int flags = HIRAID_QUEUE_PHYS_CONTIG | HIRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = HIRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(hiraidq->cq_buffer_phy); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(hiraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +/** + * hiraid_create_sq - send cmd to controller for create controller sq + */ +static int hiraid_create_send_queue(struct hiraid_dev *hdev, u16 qid, + struct hiraid_queue *hiraidq) +{ + struct hiraid_admin_command admin_cmd; + int flags = HIRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = HIRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(hiraidq->sq_buffer_phy); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(hiraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void hiraid_free_all_queues(struct hiraid_dev *hdev) +{ + int i; + struct hiraid_queue *hq; + + for (i = 0; i < hdev->queue_count; i++) { + hq = &hdev->queues[i]; + dma_free_coherent(hdev->dev, CQ_SIZE(hq->q_depth), + (void *)hq->cqes, hq->cq_buffer_phy); + dma_free_coherent(hdev->dev, SQ_SIZE(hq->qid, hq->q_depth), + hq->sq_cmds, hq->sq_buffer_phy); + if (!work_mode) + dma_free_coherent(hdev->dev, SENSE_SIZE(hq->q_depth), + hq->sense_buffer_virt, hq->sense_buffer_phy); + } + + hdev->queue_count = 0; +} + +static void hiraid_free_sense_buffer(struct hiraid_dev *hdev) +{ + if (hdev->sense_buffer_virt) { + dma_free_coherent(hdev->dev, + SENSE_SIZE(hdev->scsi_qd + max_hwq_num * HIRAID_PTHRU_CMDS_PERQ), + hdev->sense_buffer_virt, hdev->sense_buffer_phy); + hdev->sense_buffer_virt = NULL; + } +} + +static int hiraid_delete_queue(struct hiraid_dev *hdev, u8 opcode, u16 qid) +{ + struct hiraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = opcode; + admin_cmd.delete_queue.qid = cpu_to_le16(qid); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "delete %s:[%d] failed\n", + (opcode == HIRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", qid); + + return ret; +} + +static int hiraid_delete_complete_queue(struct hiraid_dev *hdev, u16 cqid) +{ + return hiraid_delete_queue(hdev, HIRAID_ADMIN_DELETE_CQ, cqid); +} + +static int hiraid_delete_send_queue(struct hiraid_dev *hdev, u16 sqid) +{ + return hiraid_delete_queue(hdev, HIRAID_ADMIN_DELETE_SQ, sqid); +} + +static int hiraid_create_queue(struct hiraid_queue *hiraidq, u16 qid) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = hiraid_create_complete_queue(hdev, qid, hiraidq, cq_vector); + if (ret) + return ret; + + ret = hiraid_create_send_queue(hdev, qid, hiraidq); + if (ret) + goto delete_cq; + + hiraidq->cq_vector = cq_vector; + ret = pci_request_irq(hdev->pdev, cq_vector, hiraid_handle_irq, NULL, + hiraidq, "hiraid%d_q%d", hdev->instance, qid); + if (ret) { + hiraidq->cq_vector = -1; + dev_err(hdev->dev, "request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + hiraid_init_queue(hiraidq, qid); + + return 0; + +delete_sq: + hiraid_delete_send_queue(hdev, qid); +delete_cq: + hiraid_delete_complete_queue(hdev, qid); + + return ret; +} + +static int hiraid_create_io_queues(struct hiraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = hiraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "create queue[%d] failed\n", i); + break; + } + } + + if (!hdev->last_qcnt) + hdev->last_qcnt = hdev->online_queues; + + dev_info(hdev->dev, "queue_count[%d] online_queue[%d] last_online[%d]", + hdev->queue_count, hdev->online_queues, hdev->last_qcnt); + + return ret >= 0 ? 0 : ret; +} + +static int hiraid_set_features(struct hiraid_dev *hdev, u32 fid, u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct hiraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = HIRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_configure_timestamp(struct hiraid_dev *hdev) +{ + __le64 timestamp; + int ret; + + timestamp = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = hiraid_set_features(hdev, HIRAID_FEATURE_TIMESTAMP, 0, + &timestamp, sizeof(timestamp), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed[%d]\n", ret); + return ret; +} + +static int hiraid_get_queue_cnt(struct hiraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = hiraid_set_features(hdev, HIRAID_FEATURE_NUM_QUEUES, q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "set queue count failed, status[%d]\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "illegal qcount: zero, nr_ioqs[%d], cnt[%d]\n", nr_ioqs, *cnt); + return -EIO; + } + return 0; +} + +static int hiraid_setup_io_queues(struct hiraid_dev *hdev) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 i, size, nr_ioqs; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + /* alloc IO sense buffer for single hw queue mode */ + if (work_mode && !hdev->sense_buffer_virt) { + hdev->sense_buffer_virt = dma_alloc_coherent(hdev->dev, + SENSE_SIZE(hdev->scsi_qd + max_hwq_num * HIRAID_PTHRU_CMDS_PERQ), + &hdev->sense_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hdev->sense_buffer_virt) + return -ENOMEM; + } + + nr_ioqs = min(num_online_cpus(), max_hwq_num); + ret = hiraid_get_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = hiraid_get_bar_size(hdev, nr_ioqs); + ret = hiraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + hdev->online_queues--; + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, hiraid_handle_irq, NULL, + adminq, "hiraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + hdev->online_queues++; + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = hiraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "max_qid[%d] queuecount[%d] onlinequeue[%d] ioqdepth[%d]\n", + hdev->max_qid, hdev->queue_count, hdev->online_queues, hdev->ioq_depth); + + return hiraid_create_io_queues(hdev); +} + +static void hiraid_delete_io_queues(struct hiraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = HIRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "err, io queue has been delete\n"); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (hiraid_delete_queue(hdev, opcode, i)) + break; + + opcode = HIRAID_ADMIN_DELETE_CQ; + } +} + +static void hiraid_pci_disable(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void hiraid_disable_admin_queue(struct hiraid_dev *hdev, bool shutdown) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + hiraid_shutdown_control(hdev); + else + hiraid_disable_control(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "err, admin queue has been delete\n"); + return; + } + + spin_lock_irq(&adminq->cq_lock); + hiraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + hiraid_complete_cqes(adminq, start, end); +} + +static int hiraid_create_prp_pools(struct hiraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < extra_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_extra_pool[i] = dma_pool_create(poolname, hdev->dev, EXTRA_POOL_SIZE, + EXTRA_POOL_SIZE, 0); + + if (!hdev->prp_extra_pool[i]) { + dev_err(hdev->dev, "create prp extra pool[%d] failed\n", i); + goto destroy_prp_extra_pool; + } + } + + return 0; + +destroy_prp_extra_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_extra_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void hiraid_free_prp_pools(struct hiraid_dev *hdev) +{ + int i; + + for (i = 0; i < extra_pool_num; i++) + dma_pool_destroy(hdev->prp_extra_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int hiraid_request_devices(struct hiraid_dev *hdev, struct hiraid_dev_info *dev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct hiraid_admin_command admin_cmd; + struct hiraid_dev_list *list_buf; + dma_addr_t buffer_phy = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = HIRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = HIRAID_GET_DEVLIST_INFO; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "get device list failed, nd[%u] idx[%u] ret[%d]\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "get dev list ndev num[%u]\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devinfo[i].hdid); + dev_info(hdev->dev, "devices[%d], hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + i, hdid, le16_to_cpu(list_buf->devinfo[i].target), + list_buf->devinfo[i].channel, + list_buf->devinfo[i].lun, + list_buf->devinfo[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", hdid); + continue; + } + memcpy(&dev[hdid - 1], &list_buf->devinfo[i], + sizeof(struct hiraid_dev_info)); + } + idx += ndev; + + if (ndev < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, buffer_phy); + return ret; +} + +static void hiraid_send_async_event(struct hiraid_dev *hdev, u16 cid) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = HIRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.cmd_id = cpu_to_le16(cid); + + hiraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send async event to controller, cid[%d]\n", cid); +} + +static inline void hiraid_init_async_event(struct hiraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->asynevent; i++) + hiraid_send_async_event(hdev, i + HIRAID_AQ_BLK_MQ_DEPTH); +} + +static int hiraid_add_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "add device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (sdev) { + dev_warn(hdev->dev, "device is already exist, channel[%d] targetid[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + return 0; +} + +static int hiraid_rescan_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "rescan device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit rescan it, channel[%d] target_id[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int hiraid_delete_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "remove device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit remove it, channel[%d] target_id[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int hiraid_dev_list_init(struct hiraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + + hdev->dev_info = kzalloc_node(nd * sizeof(struct hiraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->dev_info) + return -ENOMEM; + + return 0; +} + +static int hiraid_luntarget_sort(const void *l, const void *r) +{ + const struct hiraid_dev_info *ln = l; + const struct hiraid_dev_info *rn = r; + int l_attr = HIRAID_DEV_INFO_ATTR_BOOT(ln->attr); + int r_attr = HIRAID_DEV_INFO_ATTR_BOOT(rn->attr); + + /* boot first */ + if (l_attr != r_attr) + return (r_attr - l_attr); + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + +static void hiraid_scan_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = + container_of(work, struct hiraid_dev, scan_work); + struct hiraid_dev_info *dev, *old_dev, *new_dev; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + int count = 0; + + dev = kcalloc(nd, sizeof(struct hiraid_dev_info), GFP_KERNEL); + if (!dev) + return; + + new_dev = kcalloc(nd, sizeof(struct hiraid_dev_info), GFP_KERNEL); + if (!new_dev) + goto free_list; + + ret = hiraid_request_devices(hdev, dev); + if (ret) + goto free_all; + old_dev = hdev->dev_info; + for (i = 0; i < nd; i++) { + org_flag = old_dev[i].flag; + flag = dev[i].flag; + + dev_log_dbg(hdev->dev, "i[%d] org_flag[0x%x] flag[0x%x]\n", i, org_flag, flag); + + if (HIRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!HIRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->dev_rwsem); + memcpy(&old_dev[i], &dev[i], + sizeof(struct hiraid_dev_info)); + memcpy(&new_dev[count++], &dev[i], + sizeof(struct hiraid_dev_info)); + up_write(&hdev->dev_rwsem); + } else if (HIRAID_DEV_INFO_FLAG_CHANGE(flag)) { + hiraid_rescan_device(hdev, &dev[i]); + } + } else { + if (HIRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->dev_rwsem); + old_dev[i].flag &= 0xfe; + up_write(&hdev->dev_rwsem); + hiraid_delete_device(hdev, &old_dev[i]); + } + } + } + + dev_info(hdev->dev, "scan work add device num[%d]\n", count); + + sort(new_dev, count, sizeof(new_dev[0]), hiraid_luntarget_sort, NULL); + + for (i = 0; i < count; i++) + hiraid_add_device(hdev, &new_dev[i]); + +free_all: + kfree(new_dev); +free_list: + kfree(dev); +} + +static void hiraid_timesyn_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = + container_of(work, struct hiraid_dev, timesyn_work); + + hiraid_configure_timestamp(hdev); +} + +static int hiraid_init_control_info(struct hiraid_dev *hdev); +static void hiraid_fwactive_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = container_of(work, struct hiraid_dev, fwact_work); + + if (hiraid_init_control_info(hdev)) + dev_err(hdev->dev, "get controller info failed after fw activation\n"); +} + +static void hiraid_queue_scan(struct hiraid_dev *hdev) +{ + queue_work(work_queue, &hdev->scan_work); +} + +static void hiraid_handle_async_notice(struct hiraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case HIRAID_ASYN_DEV_CHANGED: + hiraid_queue_scan(hdev); + break; + case HIRAID_ASYN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; + case HIRAID_ASYN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result[%08x]\n", result); + } +} + +static void hiraid_handle_async_vs(struct hiraid_dev *hdev, u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case HIRAID_ASYN_TIMESYN: + queue_work(work_queue, &hdev->timesyn_work); + break; + case HIRAID_ASYN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(work_queue, &hdev->fwact_work); + break; + case HIRAID_ASYN_EVENT_MIN ... HIRAID_ASYN_EVENT_MAX: + dev_info(hdev->dev, "recv card event[%d] param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; + default: + dev_warn(hdev->dev, "async event result[0x%x]\n", result); + } +} + +static int hiraid_alloc_resources(struct hiraid_dev *hdev) +{ + int ret, nqueue; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) + return -ENOMEM; + + ret = hiraid_create_prp_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = min(num_possible_cpus(), max_hwq_num) + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct hiraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = hiraid_create_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "total queues num[%d]\n", nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + hiraid_free_prp_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); + + return ret; +} + +static void hiraid_free_resources(struct hiraid_dev *hdev) +{ + hiraid_free_admin_cmds(hdev); + kfree(hdev->queues); + hiraid_free_prp_pools(hdev); + kfree(hdev->ctrl_info); +} + +static void hiraid_bsg_buf_unmap(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_mapmange *mapbuf = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (mapbuf->sge_cnt) + dma_unmap_sg(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir); + + hiraid_free_mapbuf(hdev, mapbuf); +} + +static int hiraid_bsg_buf_map(struct hiraid_dev *hdev, struct bsg_job *job, + struct hiraid_admin_command *cmd) +{ + struct hiraid_bsg_request *bsg_req = job->request; + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_mapmange *mapbuf = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + /* No data to DMA, it may be scsi no-rw command */ + mapbuf->sge_cnt = job->request_payload.sg_cnt; + mapbuf->sgl = job->request_payload.sg_list; + mapbuf->len = job->request_payload.payload_len; + mapbuf->page_cnt = -1; + if (unlikely(mapbuf->sge_cnt == 0)) + goto out; + + mapbuf->use_sgl = !hiraid_is_prp(hdev, mapbuf->sgl, mapbuf->sge_cnt); + + ret = dma_map_sg_attrs(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + if ((mapbuf->use_sgl == (bool)true) && (bsg_req->msgcode == HIRAID_BSG_IOPTHRU) && + (hdev->ctrl_info->pt_use_sgl != (bool)false)) { + ret = hiraid_build_passthru_sgl(hdev, cmd, mapbuf); + } else { + mapbuf->use_sgl = false; + + ret = hiraid_build_passthru_prp(hdev, mapbuf); + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(mapbuf->sgl)); + cmd->common.dptr.prp2 = cpu_to_le64(mapbuf->first_dma); + } + + if (ret) + goto unmap; + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir); +out: + return ret; +} + +static int hiraid_get_control_info(struct hiraid_dev *hdev, struct hiraid_ctrl_info *ctrl_info) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = HIRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = HIRAID_GET_CTRL_INFO; + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct hiraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_init_control_info(struct hiraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = hiraid_get_control_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed[%d]\n", ret); + + dev_info(hdev->dev, "device_num = %d\n", hdev->ctrl_info->nd); + dev_info(hdev->dev, "max_cmd = %d\n", hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "max_channel = %d\n", hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "max_tgt_id = %d\n", hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "max_lun = %d\n", hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "max_num_sge = %d\n", hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "lun_num_boot = %d\n", hdev->ctrl_info->lun_num_boot); + dev_info(hdev->dev, "max_data_transfer_size = %d\n", hdev->ctrl_info->mdts); + dev_info(hdev->dev, "abort_cmd_limit = %d\n", hdev->ctrl_info->acl); + dev_info(hdev->dev, "asyn_event_num = %d\n", hdev->ctrl_info->asynevent); + dev_info(hdev->dev, "card_type = %d\n", hdev->ctrl_info->card_type); + dev_info(hdev->dev, "pt_use_sgl = %d\n", hdev->ctrl_info->pt_use_sgl); + dev_info(hdev->dev, "rtd3e = %d\n", hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "serial_num = %s\n", hdev->ctrl_info->sn); + dev_info(hdev->dev, "fw_verion = %s\n", hdev->ctrl_info->fw_version); + + if (!hdev->ctrl_info->asynevent) + hdev->ctrl_info->asynevent = 1; + if (hdev->ctrl_info->asynevent > HIRAID_ASYN_COMMANDS) + hdev->ctrl_info->asynevent = HIRAID_ASYN_COMMANDS; + + hdev->scsi_qd = work_mode ? + le16_to_cpu(hdev->ctrl_info->max_cmds) : (hdev->ioq_depth - HIRAID_PTHRU_CMDS_PERQ); + + return 0; +} + +static int hiraid_user_send_admcmd(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct hiraid_bsg_request *bsg_req = job->request; + struct hiraid_passthru_common_cmd *ptcmd = &(bsg_req->admcmd); + struct hiraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(ptcmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", + hdev->state); + return -EBUSY; + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = ptcmd->opcode; + admin_cmd.common.flags = ptcmd->flags; + admin_cmd.common.hdid = cpu_to_le32(ptcmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(ptcmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(ptcmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(ptcmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(ptcmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(ptcmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(ptcmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(ptcmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(ptcmd->cdw15); + + status = hiraid_bsg_buf_map(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "err, map data failed\n"); + return status; + } + + status = hiraid_put_admin_sync_request(hdev, &admin_cmd, &result[0], &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + if (status) + dev_info(hdev->dev, "opcode[0x%x] subopcode[0x%x] status[0x%x] result0[0x%x];" + "result1[0x%x]\n", ptcmd->opcode, ptcmd->info_0.subopcode, status, + result[0], result[1]); + + hiraid_bsg_buf_unmap(hdev, job); + + return status; +} + +static int hiraid_alloc_io_ptcmds(struct hiraid_dev *hdev) +{ + u32 i; + u32 ptnum = HIRAID_TOTAL_PTCMDS(hdev->online_queues - 1); + + INIT_LIST_HEAD(&hdev->io_pt_list); + spin_lock_init(&hdev->io_pt_lock); + + hdev->io_ptcmds = kcalloc_node(ptnum, sizeof(struct hiraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->io_ptcmds) { + dev_err(hdev->dev, "alloc io pthrunum failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->io_ptcmds[i].qid = i / HIRAID_PTHRU_CMDS_PERQ + 1; + hdev->io_ptcmds[i].cid = i % HIRAID_PTHRU_CMDS_PERQ + hdev->scsi_qd; + list_add_tail(&(hdev->io_ptcmds[i].list), &hdev->io_pt_list); + } + + dev_info(hdev->dev, "alloc io pthru cmd success, pthrunum[%d]\n", ptnum); + + return 0; +} + +static void hiraid_free_io_ptcmds(struct hiraid_dev *hdev) +{ + kfree(hdev->io_ptcmds); + hdev->io_ptcmds = NULL; + + INIT_LIST_HEAD(&hdev->io_pt_list); +} + +static int hiraid_put_io_sync_request(struct hiraid_dev *hdev, struct hiraid_scsi_io_cmd *io_cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t buffer_phy; + struct hiraid_queue *ioq; + void *sense_addr = NULL; + struct hiraid_cmd *pt_cmd = hiraid_get_cmd(hdev, HIRAID_CMD_PTHRU); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + if (work_mode) { + ret = ((pt_cmd->qid - 1) * HIRAID_PTHRU_CMDS_PERQ + pt_cmd->cid) * + SCSI_SENSE_BUFFERSIZE; + sense_addr = hdev->sense_buffer_virt + ret; + buffer_phy = hdev->sense_buffer_phy + ret; + } else { + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense_buffer_virt + ret; + buffer_phy = ioq->sense_buffer_phy + ret; + } + + io_cmd->common.sense_addr = cpu_to_le64(buffer_phy); + io_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + io_cmd->common.cmd_id = cpu_to_le16(pt_cmd->cid); + + hiraid_submit_cmd(ioq, io_cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + pt_cmd->cid, pt_cmd->qid, io_cmd->common.opcode, + (le32_to_cpu(io_cmd->common.cdw3[0]) & 0xffff)); + + hiraid_admin_timeout(hdev, pt_cmd); + + hiraid_put_cmd(hdev, pt_cmd, HIRAID_CMD_PTHRU); + return -ETIME; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + hiraid_put_cmd(hdev, pt_cmd, HIRAID_CMD_PTHRU); + + return pt_cmd->status; +} + +static int hiraid_user_send_ptcmd(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct hiraid_bsg_request *bsg_req = (struct hiraid_bsg_request *)(job->request); + struct hiraid_passthru_io_cmd *cmd = &(bsg_req->pthrucmd); + struct hiraid_scsi_io_cmd pthru_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + // data len is 4k before use sgl, now len is 1M + u32 io_pt_data_len = (hdev->ctrl_info->pt_use_sgl == (bool)true) ? + IOQ_PT_SGL_DATA_LEN : IOQ_PT_DATA_LEN; + + if (cmd->data_len > io_pt_data_len) { + dev_err(hdev->dev, "data len bigger than %d\n", io_pt_data_len); + return -EFAULT; + } + + if (hdev->state != DEV_LIVE) { + dev_err(hdev->dev, "err, host state[%d] is not live\n", hdev->state); + return -EBUSY; + } + + memset(&pthru_cmd, 0, sizeof(pthru_cmd)); + pthru_cmd.common.opcode = cmd->opcode; + pthru_cmd.common.flags = cmd->flags; + pthru_cmd.common.hdid = cpu_to_le32(cmd->nsid); + pthru_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + pthru_cmd.common.cdb_len = cmd->info_0.cdb_len; + pthru_cmd.common.rsvd2 = cmd->info_0.rsvd0; + pthru_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + pthru_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + pthru_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + pthru_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + pthru_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + pthru_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + pthru_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + pthru_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + pthru_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(pthru_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + pthru_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + pthru_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + pthru_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + pthru_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = hiraid_bsg_buf_map(hdev, job, (struct hiraid_admin_command *)&pthru_cmd); + if (status) { + dev_err(hdev->dev, "err, map data failed\n"); + return status; + } + + status = hiraid_put_io_sync_request(hdev, &pthru_cmd, job->reply, &job->reply_len, timeout); + + if (status) + dev_info(hdev->dev, "opcode[0x%x] subopcode[0x%x] status[0x%x] replylen[%d]\n", + cmd->opcode, cmd->info_1.subopcode, status, job->reply_len); + + hiraid_bsg_buf_unmap(hdev, job); + + return status; +} + +static bool hiraid_check_scmd_finished(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_queue *hiraidq; + + hiraidq = mapbuf->hiraidq; + if (!hiraidq) + return false; + if (READ_ONCE(mapbuf->state) == CMD_COMPLETE || hiraid_poll_cq(hiraidq, mapbuf->cid)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", + mapbuf->cid, hiraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return hiraid_timed_out(struct scsi_cmnd *scmd) +{ + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (hiraid_check_scmd_finished(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_TIMEOUT) == CMD_FLIGHT) + return BLK_EH_DONE; + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int hiraid_send_abort_cmd(struct hiraid_dev *hdev, u32 hdid, u16 qid, u16 cid) +{ + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = HIRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int hiraid_send_reset_cmd(struct hiraid_dev *hdev, u8 type, u32 hdid) +{ + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = HIRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool hiraid_dev_state_trans(struct hiraid_dev *hdev, enum hiraid_dev_state new_state) +{ + unsigned long flags; + enum hiraid_dev_state old_state; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + old_state = hdev->state; + switch (new_state) { + case DEV_LIVE: + switch (old_state) { + case DEV_NEW: + case DEV_RESETTING: + change = true; + break; + default: + break; + } + break; + case DEV_RESETTING: + switch (old_state) { + case DEV_LIVE: + change = true; + break; + default: + break; + } + break; + case DEV_DELETING: + if (old_state != DEV_DELETING) + change = true; + break; + case DEV_DEAD: + switch (old_state) { + case DEV_NEW: + case DEV_LIVE: + case DEV_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = new_state; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "oldstate[%d]->newstate[%d], change[%d]\n", + old_state, new_state, change); + + return change; +} + +static void hiraid_drain_pending_ios(struct hiraid_dev *hdev); + +static void hiraid_flush_running_cmds(struct hiraid_dev *hdev) +{ + int i, j; + + scsi_block_requests(hdev->shost); + hiraid_drain_pending_ios(hdev); + scsi_unblock_requests(hdev->shost); + + j = HIRAID_AQ_BLK_MQ_DEPTH; + for (i = 0; i < j; i++) { + if (READ_ONCE(hdev->adm_cmds[i].state) == CMD_FLIGHT) { + dev_info(hdev->dev, "flush admin, cid[%d]\n", i); + hdev->adm_cmds[i].status = 0xFFFF; + WRITE_ONCE(hdev->adm_cmds[i].state, CMD_COMPLETE); + complete(&(hdev->adm_cmds[i].cmd_done)); + } + } + + j = HIRAID_TOTAL_PTCMDS(hdev->online_queues - 1); + for (i = 0; i < j; i++) { + if (READ_ONCE(hdev->io_ptcmds[i].state) == CMD_FLIGHT) { + hdev->io_ptcmds[i].status = 0xFFFF; + WRITE_ONCE(hdev->io_ptcmds[i].state, CMD_COMPLETE); + complete(&(hdev->io_ptcmds[i].cmd_done)); + } + } +} + +static int hiraid_dev_disable(struct hiraid_dev *hdev, bool shutdown) +{ + int ret = -ENODEV; + struct hiraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + hiraid_shutdown_control(hdev); + else + ret = hiraid_disable_control(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "warn: queue has been delete\n"); + return ret; + } + + spin_lock_irq(&adminq->cq_lock); + hiraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + hiraid_complete_cqes(adminq, start, end); + + hiraid_pci_disable(hdev); + + hiraid_flush_running_cmds(hdev); + + return ret; +} + +static void hiraid_reset_work(struct work_struct *work) +{ + int ret = 0; + struct hiraid_dev *hdev = container_of(work, struct hiraid_dev, reset_work); + + if (hdev->state != DEV_RESETTING) { + dev_err(hdev->dev, "err, host is not reset state\n"); + return; + } + + dev_info(hdev->dev, "enter host reset\n"); + + if (hdev->ctrl_config & HIRAID_CC_ENABLE) { + dev_info(hdev->dev, "start dev_disable\n"); + ret = hiraid_dev_disable(hdev, false); + } + + if (ret) + goto out; + + ret = hiraid_pci_enable(hdev); + if (ret) + goto out; + + ret = hiraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = hiraid_setup_io_queues(hdev); + if (ret || hdev->online_queues != hdev->last_qcnt) + goto pci_disable; + + hiraid_dev_state_trans(hdev, DEV_LIVE); + + hiraid_init_async_event(hdev); + + hiraid_queue_scan(hdev); + + return; + +pci_disable: + hiraid_pci_disable(hdev); +out: + hiraid_dev_state_trans(hdev, DEV_DEAD); + dev_err(hdev->dev, "err, host reset failed\n"); +} + +static int hiraid_reset_work_sync(struct hiraid_dev *hdev) +{ + if (!hiraid_dev_state_trans(hdev, DEV_RESETTING)) { + dev_info(hdev->dev, "can't change to reset state\n"); + return -EBUSY; + } + + if (!queue_work(work_queue, &hdev->reset_work)) { + dev_err(hdev->dev, "err, host is already in reset state\n"); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != DEV_LIVE) + return -ENODEV; + + return 0; +} + +static int hiraid_wait_io_completion(struct hiraid_mapmange *mapbuf) +{ + u16 times = 0; + + do { + if (READ_ONCE(mapbuf->state) == CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= HIRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= HIRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static bool hiraid_tgt_rst_pending_io_count(struct request *rq, void *data, bool reserved) +{ + unsigned int id = *(unsigned int *)data; + struct scsi_cmnd *scmd = blk_mq_rq_to_pdu(rq); + struct hiraid_mapmange *mapbuf; + struct hiraid_sdev_hostdata *hostdata; + + if (scmd) { + mapbuf = scsi_cmd_priv(scmd); + if ((mapbuf->state == CMD_FLIGHT) || (mapbuf->state == CMD_TIMEOUT)) { + if ((scmd->device) && (scmd->device->id == id)) { + hostdata = scmd->device->hostdata; + hostdata->pend_count++; + } + } + } + return true; +} +static bool hiraid_clean_pending_io(struct request *rq, void *data, bool reserved) +{ + struct hiraid_dev *hdev = data; + struct scsi_cmnd *scmd; + struct hiraid_mapmange *mapbuf; + + if (unlikely(!rq || !blk_mq_request_started(rq))) + return true; + + scmd = blk_mq_rq_to_pdu(rq); + mapbuf = scsi_cmd_priv(scmd); + + if ((cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_COMPLETE) != CMD_FLIGHT) && + (cmpxchg(&mapbuf->state, CMD_TIMEOUT, CMD_COMPLETE) != CMD_TIMEOUT)) + return true; + + set_host_byte(scmd, DID_NO_CONNECT); + if (mapbuf->sge_cnt) + scsi_dma_unmap(scmd); + hiraid_free_mapbuf(hdev, mapbuf); + dev_warn_ratelimited(hdev->dev, "back unfinished CQE, cid[%d] qid[%d]\n", + mapbuf->cid, mapbuf->hiraidq->qid); + scmd->scsi_done(scmd); + + return true; +} + +static void hiraid_drain_pending_ios(struct hiraid_dev *hdev) +{ + blk_mq_tagset_busy_iter(&hdev->shost->tag_set, hiraid_clean_pending_io, (void *)(hdev)); +} + +static int wait_tgt_reset_io_done(struct scsi_cmnd *scmd) +{ + u16 timeout = 0; + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + + hostdata = scmd->device->hostdata; + + do { + hostdata->pend_count = 0; + blk_mq_tagset_busy_iter(&hdev->shost->tag_set, hiraid_tgt_rst_pending_io_count, + (void *)(&scmd->device->id)); + + if (!hostdata->pend_count) + return 0; + + msleep(500); + timeout++; + } while (timeout <= HIRAID_WAIT_RST_IO_TIMEOUT); + + return -ETIMEDOUT; +} + +static int hiraid_abort(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != DEV_LIVE || !hiraid_wait_io_completion(mapbuf) || + hiraid_check_scmd_finished(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + cid = mapbuf->cid; + hwq = mapbuf->hiraidq->qid; + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, send abort\n", cid, hwq); + ret = hiraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != -ETIME) { + ret = hiraid_wait_io_completion(mapbuf); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", cid, hwq); + return FAILED; +} + +static int hiraid_scsi_reset(struct scsi_cmnd *scmd, enum hiraid_rst_type rst) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_sdev_hostdata *hostdata; + int ret; + + if (hdev->state != DEV_LIVE) + return SUCCESS; + + hostdata = scmd->device->hostdata; + + dev_warn(hdev->dev, "sdev[%d:%d] send %s reset\n", scmd->device->channel, scmd->device->id, + rst ? "bus" : "target"); + ret = hiraid_send_reset_cmd(hdev, rst, hostdata->hdid); + if ((ret == 0) || (ret == FW_EH_DEV_NONE && rst == HIRAID_RESET_TARGET)) { + if (rst == HIRAID_RESET_TARGET) { + ret = wait_tgt_reset_io_done(scmd); + if (ret) { + dev_warn(hdev->dev, "sdev[%d:%d] target has %d peding cmd, target reset failed\n", + scmd->device->channel, scmd->device->id, + hostdata->pend_count); + return FAILED; + } + } + dev_warn(hdev->dev, "sdev[%d:%d] %s reset success\n", + scmd->device->channel, scmd->device->id, rst ? "bus" : "target"); + return SUCCESS; + } + + dev_warn(hdev->dev, "sdev[%d:%d] %s reset failed\n", + scmd->device->channel, scmd->device->id, rst ? "bus" : "target"); + return FAILED; +} + +static int hiraid_target_reset(struct scsi_cmnd *scmd) +{ + return hiraid_scsi_reset(scmd, HIRAID_RESET_TARGET); +} + +static int hiraid_bus_reset(struct scsi_cmnd *scmd) +{ + return hiraid_scsi_reset(scmd, HIRAID_RESET_BUS); +} + +static int hiraid_host_reset(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + + if (hdev->state != DEV_LIVE) + return SUCCESS; + + dev_warn(hdev->dev, "sdev[%d:%d] send host reset\n", + scmd->device->channel, scmd->device->id); + if (hiraid_reset_work_sync(hdev) == -EBUSY) + flush_work(&hdev->reset_work); + + if (hdev->state != DEV_LIVE) { + dev_warn(hdev->dev, "sdev[%d:%d] host reset failed\n", + scmd->device->channel, scmd->device->id); + return FAILED; + } + + dev_warn(hdev->dev, "sdev[%d:%d] host reset success\n", + scmd->device->channel, scmd->device->id); + + return SUCCESS; +} + +static pci_ers_result_t hiraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "pci error detected, state[%d]\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + hiraid_dev_state_trans(hdev, DEV_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t hiraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(work_queue, &hdev->reset_work)) { + dev_err(hdev->dev, "err, the device is resetting state\n"); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void hiraid_reset_pci_finish(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter hiraid reset finish\n"); +} + +static ssize_t csts_pp_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_PP_MASK); + ret >>= HIRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_SHST_MASK); + ret >>= HIRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_CFS_MASK); + ret >>= HIRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fw_version); +} + +static ssize_t hdd_dispatch_store(struct device *cdev, struct device_attribute *attr, + const char *buf, size_t count) +{ + int val = 0; + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + if (kstrtoint(buf, 0, &val) != 0) + return -EINVAL; + if (val < DISPATCH_BY_CPU || val > DISPATCH_BY_DISK) + return -EINVAL; + hdev->hdd_dispatch = val; + + return strlen(buf); +} +static ssize_t hdd_dispatch_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%d\n", hdev->hdd_dispatch); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); +static DEVICE_ATTR_RW(hdd_dispatch); + +static struct device_attribute *hiraid_host_attrs[] = { + &dev_attr_csts_rdy, + &dev_attr_csts_pp, + &dev_attr_csts_cfs, + &dev_attr_fw_version, + &dev_attr_csts_shst, + &dev_attr_hdd_dispatch, + NULL, +}; + +static int hiraid_get_vd_info(struct hiraid_dev *hdev, struct hiraid_vd_info *vd_info, u16 vid) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", hdev->state); + return -EBUSY; + } + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, USRCMD_TIMEOUT); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct hiraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_get_bgtask(struct hiraid_dev *hdev, struct hiraid_bgtask *bgtask) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", hdev->state); + return -EBUSY; + } + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, USRCMD_TIMEOUT); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct hiraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_bgtask *bgtask; + struct hiraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct hiraid_bgtask *)vd_info; + ret = hiraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static ssize_t dispatch_hwq_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct hiraid_sdev_hostdata *hostdata; + + hostdata = to_scsi_device(dev)->hostdata; + return snprintf(buf, PAGE_SIZE, "%d\n", hostdata->hwq); +} + +static ssize_t dispatch_hwq_store(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + int val; + struct hiraid_dev *hdev; + struct scsi_device *sdev; + struct hiraid_sdev_hostdata *hostdata; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + if (kstrtoint(buf, 0, &val) != 0) + return -EINVAL; + if (val <= 0 || val >= hdev->online_queues) + return -EINVAL; + if (!hiraid_disk_is_hdd(hostdata->attr)) + return -EINVAL; + + hostdata->hwq = val; + return strlen(buf); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); +static DEVICE_ATTR_RW(dispatch_hwq); + +static struct device_attribute *hiraid_dev_attrs[] = { + &dev_attr_raid_state, + &dev_attr_raid_level, + &dev_attr_raid_resync, + &dev_attr_dispatch_hwq, + NULL, +}; + +static struct pci_error_handlers hiraid_err_handler = { + .error_detected = hiraid_pci_error_detected, + .slot_reset = hiraid_pci_slot_reset, + .reset_done = hiraid_reset_pci_finish, +}; + +static int hiraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct hiraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "start sysfs host reset cmd\n"); + ret = hiraid_reset_work_sync(hdev); + dev_info(hdev->dev, "stop sysfs host reset cmd[%d]\n", ret); + + return ret; +} + +static int hiraid_scan_finished(struct Scsi_Host *shost, unsigned long time) +{ + struct hiraid_dev *hdev = shost_priv(shost); + + hiraid_scan_work(&hdev->scan_work); + + return 1; +} + +static struct scsi_host_template hiraid_driver_template = { + .module = THIS_MODULE, + .name = "hiraid", + .proc_name = "hiraid", + .queuecommand = hiraid_queue_command, + .slave_alloc = hiraid_slave_alloc, + .slave_destroy = hiraid_slave_destroy, + .slave_configure = hiraid_slave_configure, + .scan_finished = hiraid_scan_finished, + .eh_timed_out = hiraid_timed_out, + .eh_abort_handler = hiraid_abort, + .eh_target_reset_handler = hiraid_target_reset, + .eh_bus_reset_handler = hiraid_bus_reset, + .eh_host_reset_handler = hiraid_host_reset, + .change_queue_depth = scsi_change_queue_depth, + .this_id = -1, + .unchecked_isa_dma = 0, + .shost_attrs = hiraid_host_attrs, + .sdev_attrs = hiraid_dev_attrs, + .host_reset = hiraid_sysfs_host_reset, +}; + +static void hiraid_shutdown(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + hiraid_delete_io_queues(hdev); + hiraid_disable_admin_queue(hdev, true); +} + +static bool hiraid_bsg_is_valid(struct bsg_job *job) +{ + u64 timeout = 0; + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_bsg_request *bsg_req = job->request; + struct hiraid_dev *hdev = shost_priv(dev_to_shost(job->dev)); + + if (bsg_req == NULL || job->request_len != sizeof(struct hiraid_bsg_request)) + return false; + + switch (bsg_req->msgcode) { + case HIRAID_BSG_ADMIN: + timeout = msecs_to_jiffies(bsg_req->admcmd.timeout_ms); + break; + case HIRAID_BSG_IOPTHRU: + timeout = msecs_to_jiffies(bsg_req->pthrucmd.timeout_ms); + break; + default: + dev_info(hdev->dev, "bsg unsupport msgcode[%d]\n", bsg_req->msgcode); + return false; + } + + if ((timeout + CTL_RST_TIME) > rq->timeout) { + dev_err(hdev->dev, "bsg invalid time\n"); + return false; + } + + return true; +} + +/* bsg dispatch user command */ +static int hiraid_bsg_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct hiraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_bsg_request *bsg_req = job->request; + int ret = -ENOMSG; + + job->reply_len = 0; + + if (!hiraid_bsg_is_valid(job)) { + bsg_job_done(job, ret, 0); + return 0; + } + + dev_log_dbg(hdev->dev, "bsg msgcode[%d] msglen[%d] timeout[%d];" + "reqnsge[%d], reqlen[%d]\n", + bsg_req->msgcode, job->request_len, rq->timeout, + job->request_payload.sg_cnt, job->request_payload.payload_len); + + switch (bsg_req->msgcode) { + case HIRAID_BSG_ADMIN: + ret = hiraid_user_send_admcmd(hdev, job); + break; + case HIRAID_BSG_IOPTHRU: + ret = hiraid_user_send_ptcmd(hdev, job); + break; + default: + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void hiraid_unregist_bsg(struct hiraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} +static int hiraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct hiraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&hiraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + hdev->instance = shost->host_no; + pci_set_drvdata(pdev, hdev); + + ret = hiraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->dev_rwsem); + INIT_WORK(&hdev->scan_work, hiraid_scan_work); + INIT_WORK(&hdev->timesyn_work, hiraid_timesyn_work); + INIT_WORK(&hdev->reset_work, hiraid_reset_work); + INIT_WORK(&hdev->fwact_work, hiraid_fwactive_work); + spin_lock_init(&hdev->state_lock); + + ret = hiraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = hiraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = hiraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = hiraid_init_control_info(hdev); + if (ret) + goto disable_admin_q; + + ret = hiraid_setup_io_queues(hdev); + if (ret) + goto disable_admin_q; + + hiraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "add shost to system failed, ret[%d]\n", ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "hiraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, hiraid_bsg_dispatch, + NULL, hiraid_get_max_cmd_size(hdev)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == HIRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn: only admin queue can be used\n"); + return 0; + } + + hdev->state = DEV_LIVE; + + hiraid_init_async_event(hdev); + + ret = hiraid_dev_list_init(hdev); + if (ret) + goto unregist_bsg; + + ret = hiraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "time synchronization failed\n"); + + ret = hiraid_alloc_io_ptcmds(hdev); + if (ret) + goto unregist_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +unregist_bsg: + hiraid_unregist_bsg(hdev); +remove_io_queues: + hiraid_delete_io_queues(hdev); +disable_admin_q: + hiraid_free_sense_buffer(hdev); + hiraid_disable_admin_queue(hdev, false); +pci_disable: + hiraid_free_all_queues(hdev); + hiraid_pci_disable(hdev); +resources_free: + hiraid_free_resources(hdev); +dev_unmap: + hiraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void hiraid_remove(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter hiraid remove\n"); + + hiraid_dev_state_trans(hdev, DEV_DELETING); + flush_work(&hdev->reset_work); + + if (!pci_device_is_present(pdev)) + hiraid_flush_running_cmds(hdev); + + hiraid_unregist_bsg(hdev); + scsi_remove_host(shost); + hiraid_free_io_ptcmds(hdev); + kfree(hdev->dev_info); + hiraid_delete_io_queues(hdev); + hiraid_free_sense_buffer(hdev); + hiraid_disable_admin_queue(hdev, false); + hiraid_free_all_queues(hdev); + hiraid_pci_disable(hdev); + hiraid_free_resources(hdev); + hiraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit hiraid remove\n"); +} + +static const struct pci_device_id hiraid_hw_card_ids[] = { + { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI_LOGIC, HIRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI_LOGIC, HIRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, hiraid_hw_card_ids); + +static struct pci_driver hiraid_driver = { + .name = "hiraid", + .id_table = hiraid_hw_card_ids, + .probe = hiraid_probe, + .remove = hiraid_remove, + .shutdown = hiraid_shutdown, + .err_handler = &hiraid_err_handler, +}; + +static int __init hiraid_init(void) +{ + int ret; + + work_queue = alloc_workqueue("hiraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS, 0); + if (!work_queue) + return -ENOMEM; + + hiraid_class = class_create(THIS_MODULE, "hiraid"); + if (IS_ERR(hiraid_class)) { + ret = PTR_ERR(hiraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&hiraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(hiraid_class); +destroy_wq: + destroy_workqueue(work_queue); + + return ret; +} + +static void __exit hiraid_exit(void) +{ + pci_unregister_driver(&hiraid_driver); + class_destroy(hiraid_class); + destroy_workqueue(work_queue); +} + +MODULE_AUTHOR("Huawei Technologies CO., Ltd"); +MODULE_DESCRIPTION("Huawei RAID driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(HIRAID_DRV_VERSION); +module_init(hiraid_init); +module_exit(hiraid_exit); -- 2.22.0.windows.1

2 1

[PATCH openEuler-1.0-LTS] Revert "tcp: fix delayed ACKs for MSS boundary condition"
by Dong Chenchen 16 Nov '23

16 Nov '23

From: dongchenchen <dongchenchen2(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8GYWB CVE: NA -------------------------------- This reverts commit 389055ab28760dd7b25c6996c6647b0a37e0a34e. Signed-off-by: dongchenchen <dongchenchen2(a)huawei.com> --- net/ipv4/tcp_input.c | 13 ------------- 1 file changed, 13 deletions(-) diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index f8b1ace50f7a..a12598dabb80 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -172,19 +172,6 @@ static void tcp_measure_rcv_mss(struct sock *sk, const struct sk_buff *skb) if (unlikely(len > icsk->icsk_ack.rcv_mss + MAX_TCP_OPTION_SPACE)) tcp_gro_dev_warn(sk, skb, len); - /* If the skb has a len of exactly 1*MSS and has the PSH bit - * set then it is likely the end of an application write. So - * more data may not be arriving soon, and yet the data sender - * may be waiting for an ACK if cwnd-bound or using TX zero - * copy. So we set ICSK_ACK_PUSHED here so that - * tcp_cleanup_rbuf() will send an ACK immediately if the app - * reads all of the data and is not ping-pong. If len > MSS - * then this logic does not matter (and does not hurt) because - * tcp_cleanup_rbuf() will always ACK immediately if the app - * reads data and there is more than an MSS of unACKed data. - */ - if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_PSH) - icsk->icsk_ack.pending |= ICSK_ACK_PUSHED; } else { /* Otherwise, we make more careful check taking into account, * that SACKs block is variable. -- 2.25.1

2 1

[PATCH OLK-5.10] SCSI: hisi_raid: support SPxxx series RAID/HBA controllers
by z00848923 16 Nov '23

16 Nov '23

driver inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I89D3P CVE: NA ------------------------------------------ This commit is to support SPxxx RAID/HBA controllers. RAID controllers support RAID 0/1/5/6/10/50/60 modes. HBA controlllers support RAID 0/1/10 modes. RAID/HBA support SAS/SATA HDD/SSD. Signed-off-by: z00848923 <zhanglei48(a)huawei.com> --- Documentation/scsi/hisi_raid.rst | 84 + MAINTAINERS | 7 + arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + drivers/scsi/Kconfig | 1 + drivers/scsi/Makefile | 1 + drivers/scsi/hisi_raid/Kconfig | 14 + drivers/scsi/hisi_raid/Makefile | 7 + drivers/scsi/hisi_raid/hiraid.h | 760 +++++ drivers/scsi/hisi_raid/hiraid_main.c | 3982 ++++++++++++++++++++++++ 10 files changed, 4858 insertions(+) create mode 100644 Documentation/scsi/hisi_raid.rst create mode 100644 drivers/scsi/hisi_raid/Kconfig create mode 100644 drivers/scsi/hisi_raid/Makefile create mode 100644 drivers/scsi/hisi_raid/hiraid.h create mode 100644 drivers/scsi/hisi_raid/hiraid_main.c diff --git a/Documentation/scsi/hisi_raid.rst b/Documentation/scsi/hisi_raid.rst new file mode 100644 index 000000000000..523a6763a7fd --- /dev/null +++ b/Documentation/scsi/hisi_raid.rst @@ -0,0 +1,84 @@ +.. SPDX-License-Identifier: GPL-2.0 + +============================================== +hisi_raid - HUAWEI SCSI RAID Controller driver +============================================== + +This file describes the hisi_raid SCSI driver for HUAWEI +RAID controllers. The hisi_raid driver is the first +generation RAID driver. + +For hisi_raid controller support, enable the hisi_raid driver +when configuring the kernel. + +hisi_raid specific entries in /sys +================================= + +hisi_raid host attributes +------------------------ + - /sys/class/scsi_host/host*/csts_pp + - /sys/class/scsi_host/host*/csts_shst + - /sys/class/scsi_host/host*/csts_cfs + - /sys/class/scsi_host/host*/csts_rdy + - /sys/class/scsi_host/host*/fw_version + + The host csts_pp attribute is a read only attribute. This attribute + indicates whether the controller is processing commands. If this attribute + is set to ‘1’, then the controller is processing commands normally. If + this attribute is cleared to ‘0’, then the controller has temporarily stopped + processing commands in order to handle an event (e.g., firmware activation). + + The host csts_shst attribute is a read only attribute. This attribute + indicates status of shutdown processing.The shutdown status values are defined + as: + ====== ============================== + Value Definition + ====== ============================== + 00b Normal operation + 01b Shutdown processing occurring + 10b Shutdown processing complete + 11b Reserved + ====== ============================== + The host csts_cfs attribute is a read only attribute. This attribute is set to + ’1’ when a fatal controller error occurred that could not be communicated in the + appropriate Completion Queue. This bit is cleared to ‘0’ when a fatal controller + error has not occurred. + + The host csts_rdy attribute is a read only attribute. This attribute is set to + ‘1’ when the controller is ready to process submission queue entries. + + The fw_version attribute is read-only and will return the driver version and the + controller firmware version. + +hisi_raid scsi device attributes +------------------------------ + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_level + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_state + - /sys/class/scsi_device/X\:X\:X\:X/device/raid_resync + + The device raid_level attribute is a read only attribute. This attribute indicates + RAID level of scsi device(will dispaly "NA" if scsi device is not virtual disk type). + + The device raid_state attribute is read-only and indicates RAID status of scsi + device(will dispaly "NA" if scsi device is not virtual disk type). + + The device raid_resync attribute is read-only and indicates RAID rebuild processing + of scsi device(will dispaly "NA" if scsi device is not virtual disk type). + +Supported devices +================= + + =================== ======= ======================================= + PCI ID (pci.ids) OEM Product + =================== ======= ======================================= + 19E5:3858 HUAWEI SP186-M-8i(HBA:8Ports) + 19E5:3858 HUAWEI SP186-M-16i(HBA:16Ports) + 19E5:3858 HUAWEI SP186-M-32i(HBA:32Ports) + 19E5:3858 HUAWEI SP186-M-40i(HBA:40Ports) + 19E5:3758 HUAWEI SP686C-M-16i(RAID:16Ports,2G cache) + 19E5:3758 HUAWEI SP686C-M-16i(RAID:16Ports,4G cache) + 19E5:3758 HUAWEI SP686C-MH-32i(RAID:32Ports,4G cache) + 19E5:3758 HUAWEI SP686C-M-40i(RAID:40Ports,2G cache) + 19E5:3758 HUAWEI SP686C-M-40i(RAID:40Ports,4G cache) + =================== ======= ======================================= + diff --git a/MAINTAINERS b/MAINTAINERS index a7815fd1072f..8324f56a2096 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -8070,6 +8070,13 @@ M: Yonglong Liu <liuyonglong(a)huawei.com> S: Supported F: drivers/ptp/ptp_hisi.c +HISI_RAID SCSI RAID DRIVERS +M: Zhang Lei <zhanglei48(a)huawei.com> +L: linux-scsi(a)vger.kernel.org +S: Maintained +F: Documentation/scsi/hisi_raid.rst +F: drivers/scsi/hisi_raid/ + HMM - Heterogeneous Memory Management M: Jérôme Glisse <jglisse(a)redhat.com> L: linux-mm(a)kvack.org diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index ec758f0530c1..b9a50ef6d768 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -2413,6 +2413,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_SCSI_HISI_RAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_MYRB is not set diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index 5171aa50736b..43b5294326e6 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -2369,6 +2369,7 @@ CONFIG_SCSI_MPT2SAS_MAX_SGE=128 CONFIG_SCSI_MPT3SAS_MAX_SGE=128 CONFIG_SCSI_MPT2SAS=m CONFIG_SCSI_SMARTPQI=m +CONFIG_SCSI_HISI_RAID=m # CONFIG_SCSI_UFSHCD is not set # CONFIG_SCSI_HPTIOP is not set # CONFIG_SCSI_BUSLOGIC is not set diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index a9da1b2dec4a..41ef664cf0ed 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -473,6 +473,7 @@ source "drivers/scsi/megaraid/Kconfig.megaraid" source "drivers/scsi/sssraid/Kconfig" source "drivers/scsi/mpt3sas/Kconfig" source "drivers/scsi/smartpqi/Kconfig" +source "drivers/scsi/hisi_raid/Kconfig" source "drivers/scsi/ufs/Kconfig" config SCSI_HPTIOP diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index c2a1efa16912..8f26dbb5ee37 100644 --- a/drivers/scsi/Makefile +++ b/drivers/scsi/Makefile @@ -101,6 +101,7 @@ obj-$(CONFIG_MEGARAID_LEGACY) += megaraid.o obj-$(CONFIG_MEGARAID_NEWGEN) += megaraid/ obj-$(CONFIG_MEGARAID_SAS) += megaraid/ obj-$(CONFIG_SCSI_MPT3SAS) += mpt3sas/ +obj-$(CONFIG_SCSI_HISI_RAID) += hisi_raid/ obj-$(CONFIG_SCSI_UFSHCD) += ufs/ obj-$(CONFIG_SCSI_ACARD) += atp870u.o obj-$(CONFIG_SCSI_SUNESP) += esp_scsi.o sun_esp.o diff --git a/drivers/scsi/hisi_raid/Kconfig b/drivers/scsi/hisi_raid/Kconfig new file mode 100644 index 000000000000..d402dc45a7c1 --- /dev/null +++ b/drivers/scsi/hisi_raid/Kconfig @@ -0,0 +1,14 @@ +# +# Kernel configuration file for the hisi_raid +# + +config SCSI_HISI_RAID + tristate "Huawei Hisi_Raid Adapter" + depends on PCI && SCSI + select BLK_DEV_BSGLIB + depends on ARM64 || X86_64 + help + This driver supports hisi_raid SPxxx serial RAID controller, which has + PCI Express Gen4 interface with host and supports SAS/SATA HDD/SSD. + To compile this driver as a module, choose M here: the module will + be called hisi_raid. diff --git a/drivers/scsi/hisi_raid/Makefile b/drivers/scsi/hisi_raid/Makefile new file mode 100644 index 000000000000..b71a675f4190 --- /dev/null +++ b/drivers/scsi/hisi_raid/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the hisi_raid drivers. +# + +obj-$(CONFIG_SCSI_HISI_RAID) += hiraid.o + +hiraid-objs := hiraid_main.o diff --git a/drivers/scsi/hisi_raid/hiraid.h b/drivers/scsi/hisi_raid/hiraid.h new file mode 100644 index 000000000000..1ebc3dd3f2ec --- /dev/null +++ b/drivers/scsi/hisi_raid/hiraid.h @@ -0,0 +1,760 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright(c) 2022 Huawei Technologies Co., Ltd */ + +#ifndef __HIRAID_H_ +#define __HIRAID_H_ + +#define HIRAID_HDD_PD_QD 64 +#define HIRAID_HDD_VD_QD 256 +#define HIRAID_SSD_PD_QD 64 +#define HIRAID_SSD_VD_QD 256 + +#define BGTASK_TYPE_REBUILD 4 +#define USR_CMD_READ 0xc2 +#define USR_CMD_RDLEN 0x1000 +#define USR_CMD_VDINFO 0x704 +#define USR_CMD_BGTASK 0x504 +#define VDINFO_PARAM_LEN 0x04 + +#define HIRAID_DEFAULT_MAX_CHANNEL 4 +#define HIRAID_DEFAULT_MAX_ID 240 +#define HIRAID_DEFAULT_MAX_LUN_PER_HOST 8 + +#define FUA_MASK 0x08 + +#define HIRAID_IO_SQES 7 +#define HIRAID_IO_CQES 4 +#define PRP_ENTRY_SIZE 8 + +#define EXTRA_POOL_SIZE 256 +#define MAX_EXTRA_POOL_NUM 16 +#define MAX_CMD_PER_DEV 64 +#define MAX_CDB_LEN 16 + +#define HIRAID_AQ_DEPTH 128 +#define HIRAID_ASYN_COMMANDS 16 +#define HIRAID_AQ_BLK_MQ_DEPTH (HIRAID_AQ_DEPTH - HIRAID_ASYN_COMMANDS) +#define HIRAID_AQ_MQ_TAG_DEPTH (HIRAID_AQ_BLK_MQ_DEPTH - 1) + +#define HIRAID_ADMIN_QUEUE_NUM 1 +#define HIRAID_PTHRU_CMDS_PERQ 1 +#define HIRAID_TOTAL_PTCMDS(qn) (HIRAID_PTHRU_CMDS_PERQ * (qn)) + +#define HIRAID_DEV_INFO_ATTR_BOOT(attr) ((attr) & 0x01) +#define HIRAID_DEV_INFO_ATTR_VD(attr) (((attr) & 0x02) == 0x0) +#define HIRAID_DEV_INFO_ATTR_PT(attr) (((attr) & 0x22) == 0x02) +#define HIRAID_DEV_INFO_ATTR_RAWDISK(attr) ((attr) & 0x20) +#define HIRAID_DEV_DISK_TYPE(attr) ((attr) & 0x1e) + +#define HIRAID_DEV_INFO_FLAG_VALID(flag) ((flag) & 0x01) +#define HIRAID_DEV_INFO_FLAG_CHANGE(flag) ((flag) & 0x02) + +#define HIRAID_CAP_MQES(cap) ((cap) & 0xffff) +#define HIRAID_CAP_STRIDE(cap) (((cap) >> 32) & 0xf) +#define HIRAID_CAP_MPSMIN(cap) (((cap) >> 48) & 0xf) +#define HIRAID_CAP_MPSMAX(cap) (((cap) >> 52) & 0xf) +#define HIRAID_CAP_TIMEOUT(cap) (((cap) >> 24) & 0xff) +#define HIRAID_CAP_DMAMASK(cap) (((cap) >> 37) & 0xff) + +#define IO_SQE_SIZE sizeof(struct hiraid_scsi_io_cmd) +#define ADMIN_SQE_SIZE sizeof(struct hiraid_admin_command) +#define SQE_SIZE(qid) (((qid) > 0) ? IO_SQE_SIZE : ADMIN_SQE_SIZE) +#define CQ_SIZE(depth) ((depth) * sizeof(struct hiraid_completion)) +#define SQ_SIZE(qid, depth) ((depth) * SQE_SIZE(qid)) + +#define SENSE_SIZE(depth) ((depth) * SCSI_SENSE_BUFFERSIZE) + +#define IO_6_DEFAULT_TX_LEN 256 + +#define MAX_DEV_ENTRY_PER_PAGE_4K 340 + +#define MAX_REALTIME_BGTASK_NUM 32 + +#define PCI_VENDOR_ID_HUAWEI_LOGIC 0x19E5 +#define HIRAID_SERVER_DEVICE_HBA_DID 0x3858 +#define HIRAID_SERVER_DEVICE_RAID_DID 0x3758 + +enum { + HIRAID_SC_SUCCESS = 0x0, + HIRAID_SC_INVALID_OPCODE = 0x1, + HIRAID_SC_INVALID_FIELD = 0x2, + + HIRAID_SC_ABORT_LIMIT = 0x103, + HIRAID_SC_ABORT_MISSING = 0x104, + HIRAID_SC_ASYNC_LIMIT = 0x105, + + HIRAID_SC_DNR = 0x4000, +}; + +enum { + HIRAID_REG_CAP = 0x0000, + HIRAID_REG_CC = 0x0014, + HIRAID_REG_CSTS = 0x001c, + HIRAID_REG_AQA = 0x0024, + HIRAID_REG_ASQ = 0x0028, + HIRAID_REG_ACQ = 0x0030, + HIRAID_REG_DBS = 0x1000, +}; + +enum { + HIRAID_CC_ENABLE = 1 << 0, + HIRAID_CC_CSS_NVM = 0 << 4, + HIRAID_CC_MPS_SHIFT = 7, + HIRAID_CC_AMS_SHIFT = 11, + HIRAID_CC_SHN_SHIFT = 14, + HIRAID_CC_IOSQES_SHIFT = 16, + HIRAID_CC_IOCQES_SHIFT = 20, + HIRAID_CC_AMS_RR = 0 << HIRAID_CC_AMS_SHIFT, + HIRAID_CC_SHN_NONE = 0 << HIRAID_CC_SHN_SHIFT, + HIRAID_CC_IOSQES = HIRAID_IO_SQES << HIRAID_CC_IOSQES_SHIFT, + HIRAID_CC_IOCQES = HIRAID_IO_CQES << HIRAID_CC_IOCQES_SHIFT, + HIRAID_CC_SHN_NORMAL = 1 << HIRAID_CC_SHN_SHIFT, + HIRAID_CC_SHN_MASK = 3 << HIRAID_CC_SHN_SHIFT, + HIRAID_CSTS_CFS_SHIFT = 1, + HIRAID_CSTS_SHST_SHIFT = 2, + HIRAID_CSTS_PP_SHIFT = 5, + HIRAID_CSTS_RDY = 1 << 0, + HIRAID_CSTS_SHST_CMPLT = 2 << 2, + HIRAID_CSTS_SHST_MASK = 3 << 2, + HIRAID_CSTS_CFS_MASK = 1 << HIRAID_CSTS_CFS_SHIFT, + HIRAID_CSTS_PP_MASK = 1 << HIRAID_CSTS_PP_SHIFT, +}; + +enum { + HIRAID_ADMIN_DELETE_SQ = 0x00, + HIRAID_ADMIN_CREATE_SQ = 0x01, + HIRAID_ADMIN_DELETE_CQ = 0x04, + HIRAID_ADMIN_CREATE_CQ = 0x05, + HIRAID_ADMIN_ABORT_CMD = 0x08, + HIRAID_ADMIN_SET_FEATURES = 0x09, + HIRAID_ADMIN_ASYNC_EVENT = 0x0c, + HIRAID_ADMIN_GET_INFO = 0xc6, + HIRAID_ADMIN_RESET = 0xc8, +}; + +enum { + HIRAID_GET_CTRL_INFO = 0, + HIRAID_GET_DEVLIST_INFO = 1, +}; + +enum hiraid_rst_type { + HIRAID_RESET_TARGET = 0, + HIRAID_RESET_BUS = 1, +}; + +enum { + HIRAID_ASYN_EVENT_ERROR = 0, + HIRAID_ASYN_EVENT_NOTICE = 2, + HIRAID_ASYN_EVENT_VS = 7, +}; + +enum { + HIRAID_ASYN_DEV_CHANGED = 0x00, + HIRAID_ASYN_FW_ACT_START = 0x01, + HIRAID_ASYN_HOST_PROBING = 0x10, +}; + +enum { + HIRAID_ASYN_TIMESYN = 0x00, + HIRAID_ASYN_FW_ACT_FINISH = 0x02, + HIRAID_ASYN_EVENT_MIN = 0x80, + HIRAID_ASYN_EVENT_MAX = 0xff, +}; + +enum { + HIRAID_CMD_WRITE = 0x01, + HIRAID_CMD_READ = 0x02, + + HIRAID_CMD_NONRW_NONE = 0x80, + HIRAID_CMD_NONRW_TODEV = 0x81, + HIRAID_CMD_NONRW_FROMDEV = 0x82, +}; + +enum { + HIRAID_QUEUE_PHYS_CONTIG = (1 << 0), + HIRAID_CQ_IRQ_ENABLED = (1 << 1), + + HIRAID_FEATURE_NUM_QUEUES = 0x07, + HIRAID_FEATURE_ASYNC_EVENT = 0x0b, + HIRAID_FEATURE_TIMESTAMP = 0x0e, +}; + +enum hiraid_dev_state { + DEV_NEW, + DEV_LIVE, + DEV_RESETTING, + DEV_DELETING, + DEV_DEAD, +}; + +enum { + HIRAID_CARD_HBA, + HIRAID_CARD_RAID, +}; + +enum hiraid_cmd_type { + HIRAID_CMD_ADMIN, + HIRAID_CMD_PTHRU, +}; + +enum { + SQE_FLAG_SGL_METABUF = (1 << 6), + SQE_FLAG_SGL_METASEG = (1 << 7), + SQE_FLAG_SGL_ALL = SQE_FLAG_SGL_METABUF | SQE_FLAG_SGL_METASEG, +}; + +enum hiraid_cmd_state { + CMD_IDLE = 0, + CMD_FLIGHT = 1, + CMD_COMPLETE = 2, + CMD_TIMEOUT = 3, + CMD_TMO_COMPLETE = 4, +}; + +enum { + HIRAID_BSG_ADMIN, + HIRAID_BSG_IOPTHRU, +}; + +enum { + HIRAID_SAS_HDD_VD = 0x04, + HIRAID_SATA_HDD_VD = 0x08, + HIRAID_SAS_SSD_VD = 0x0c, + HIRAID_SATA_SSD_VD = 0x10, + HIRAID_NVME_SSD_VD = 0x14, + HIRAID_SAS_HDD_PD = 0x06, + HIRAID_SATA_HDD_PD = 0x0a, + HIRAID_SAS_SSD_PD = 0x0e, + HIRAID_SATA_SSD_PD = 0x12, + HIRAID_NVME_SSD_PD = 0x16, +}; + +enum { + DISPATCH_BY_CPU, + DISPATCH_BY_DISK, +}; + +struct hiraid_completion { + __le32 result; + union { + struct { + __u8 sense_len; + __u8 resv[3]; + }; + __le32 result1; + }; + __le16 sq_head; + __le16 sq_id; + __le16 cmd_id; + __le16 status; +}; + +struct hiraid_ctrl_info { + __le32 nd; + __le16 max_cmds; + __le16 max_channel; + __le32 max_tgt_id; + __le16 max_lun; + __le16 max_num_sge; + __le16 lun_num_boot; + __u8 mdts; + __u8 acl; + __u8 asynevent; + __u8 card_type; + __u8 pt_use_sgl; + __u8 rsvd; + __le32 rtd3e; + __u8 sn[32]; + __u8 fw_version[16]; + __u8 rsvd1[4020]; +}; + +struct hiraid_dev { + struct pci_dev *pdev; + struct device *dev; + struct Scsi_Host *shost; + struct hiraid_queue *queues; + struct dma_pool *prp_page_pool; + struct dma_pool *prp_extra_pool[MAX_EXTRA_POOL_NUM]; + void __iomem *bar; + u32 max_qid; + u32 num_vecs; + u32 queue_count; + u32 ioq_depth; + u32 db_stride; + u32 __iomem *dbs; + struct rw_semaphore dev_rwsem; + int numa_node; + u32 page_size; + u32 ctrl_config; + u32 online_queues; + u64 cap; + u32 scsi_qd; + u32 instance; + struct hiraid_ctrl_info *ctrl_info; + struct hiraid_dev_info *dev_info; + + struct hiraid_cmd *adm_cmds; + struct list_head adm_cmd_list; + spinlock_t adm_cmd_lock; + + struct hiraid_cmd *io_ptcmds; + struct list_head io_pt_list; + spinlock_t io_pt_lock; + + struct work_struct scan_work; + struct work_struct timesyn_work; + struct work_struct reset_work; + struct work_struct fwact_work; + + enum hiraid_dev_state state; + spinlock_t state_lock; + + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + u32 last_qcnt; + u8 hdd_dispatch; + + struct request_queue *bsg_queue; +}; + +struct hiraid_sgl_desc { + __le64 addr; + __le32 length; + __u8 rsvd[3]; + __u8 type; +}; + +union hiraid_data_ptr { + struct { + __le64 prp1; + __le64 prp2; + }; + struct hiraid_sgl_desc sgl; +}; + +struct hiraid_admin_com_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le32 cdw2[4]; + union hiraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct hiraid_features { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[2]; + union hiraid_data_ptr dptr; + __le32 fid; + __le32 dword11; + __le32 dword12; + __le32 dword13; + __le32 dword14; + __le32 dword15; +}; + +struct hiraid_create_cq { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 cqid; + __le16 qsize; + __le16 cq_flags; + __le16 irq_vector; + __u32 rsvd12[4]; +}; + +struct hiraid_create_sq { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[5]; + __le64 prp1; + __u64 rsvd8; + __le16 sqid; + __le16 qsize; + __le16 sq_flags; + __le16 cqid; + __u32 rsvd12[4]; +}; + +struct hiraid_delete_queue { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __u32 rsvd1[9]; + __le16 qid; + __u16 rsvd10; + __u32 rsvd11[5]; +}; + +struct hiraid_get_info { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u32 rsvd2[4]; + union hiraid_data_ptr dptr; + __u8 type; + __u8 rsvd10[3]; + __le32 cdw11; + __u32 rsvd12[4]; +}; + +struct hiraid_usr_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + union { + struct { + __le16 subopcode; + __le16 rsvd1; + } info_0; + __le32 cdw2; + }; + union { + struct { + __le16 data_len; + __le16 param_len; + } info_1; + __le32 cdw3; + }; + __u64 metadata; + union hiraid_data_ptr dptr; + __le32 cdw10; + __le32 cdw11; + __le32 cdw12; + __le32 cdw13; + __le32 cdw14; + __le32 cdw15; +}; + +struct hiraid_abort_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[4]; + __le16 sqid; + __le16 cid; + __u32 rsvd11[5]; +}; + +struct hiraid_reset_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __u64 rsvd2[4]; + __u8 type; + __u8 rsvd10[3]; + __u32 rsvd11[5]; +}; + +struct hiraid_admin_command { + union { + struct hiraid_admin_com_cmd common; + struct hiraid_features features; + struct hiraid_create_cq create_cq; + struct hiraid_create_sq create_sq; + struct hiraid_delete_queue delete_queue; + struct hiraid_get_info get_info; + struct hiraid_abort_cmd abort; + struct hiraid_reset_cmd reset; + struct hiraid_usr_cmd usr_cmd; + }; +}; + +struct hiraid_scsi_io_com_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __le32 cdw3[3]; + union hiraid_data_ptr dptr; + __le32 cdw10[6]; + __u8 cdb[32]; + __le64 sense_addr; + __le32 cdw26[6]; +}; + +struct hiraid_scsi_rw_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_len; + __u8 rsvd2; + __u32 rsvd3[3]; + union hiraid_data_ptr dptr; + __le64 slba; + __le16 nlb; + __le16 control; + __u32 rsvd13[3]; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct hiraid_scsi_nonrw_cmd { + __u8 opcode; + __u8 flags; + __le16 cmd_id; + __le32 hdid; + __le16 sense_len; + __u8 cdb_length; + __u8 rsvd2; + __u32 rsvd3[3]; + union hiraid_data_ptr dptr; + __u32 rsvd10[5]; + __le32 buf_len; + __u8 cdb[32]; + __le64 sense_addr; + __u32 rsvd26[6]; +}; + +struct hiraid_scsi_io_cmd { + union { + struct hiraid_scsi_io_com_cmd common; + struct hiraid_scsi_rw_cmd rw; + struct hiraid_scsi_nonrw_cmd nonrw; + }; +}; + +struct hiraid_passthru_common_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 data_len; + __u16 param_len; + } info_1; + __u32 cdw3; + }; + __u64 metadata; + + __u64 addr; + __u64 prp2; + + __u32 cdw10; + __u32 cdw11; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 cdw15; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct hiraid_passthru_io_cmd { + __u8 opcode; + __u8 flags; + __u16 rsvd0; + __u32 nsid; + union { + struct { + __u16 res_sense_len; + __u8 cdb_len; + __u8 rsvd0; + } info_0; + __u32 cdw2; + }; + union { + struct { + __u16 subopcode; + __u16 rsvd1; + } info_1; + __u32 cdw3; + }; + union { + struct { + __u16 rsvd; + __u16 param_len; + } info_2; + __u32 cdw4; + }; + __u32 cdw5; + __u64 addr; + __u64 prp2; + union { + struct { + __u16 eid; + __u16 sid; + } info_3; + __u32 cdw10; + }; + union { + struct { + __u16 did; + __u8 did_flag; + __u8 rsvd2; + } info_4; + __u32 cdw11; + }; + __u32 cdw12; + __u32 cdw13; + __u32 cdw14; + __u32 data_len; + __u32 cdw16; + __u32 cdw17; + __u32 cdw18; + __u32 cdw19; + __u32 cdw20; + __u32 cdw21; + __u32 cdw22; + __u32 cdw23; + __u64 sense_addr; + __u32 cdw26[4]; + __u32 timeout_ms; + __u32 result0; + __u32 result1; +}; + +struct hiraid_bsg_request { + u32 msgcode; + u32 control; + union { + struct hiraid_passthru_common_cmd admcmd; + struct hiraid_passthru_io_cmd pthrucmd; + }; +}; + +struct hiraid_cmd { + u16 qid; + u16 cid; + u32 result0; + u32 result1; + u16 status; + void *priv; + enum hiraid_cmd_state state; + struct completion cmd_done; + struct list_head list; +}; + +struct hiraid_queue { + struct hiraid_dev *hdev; + spinlock_t sq_lock; + + spinlock_t cq_lock ____cacheline_aligned_in_smp; + + void *sq_cmds; + + struct hiraid_completion *cqes; + + dma_addr_t sq_buffer_phy; + dma_addr_t cq_buffer_phy; + u32 __iomem *q_db; + u8 cq_phase; + u8 sqes; + u16 qid; + u16 sq_tail; + u16 cq_head; + u16 last_cq_head; + u16 q_depth; + s16 cq_vector; + atomic_t inflight; + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + struct dma_pool *prp_small_pool; +}; + +struct hiraid_mapmange { + struct hiraid_queue *hiraidq; + enum hiraid_cmd_state state; + u16 cid; + int page_cnt; + u32 sge_cnt; + u32 len; + bool use_sgl; + dma_addr_t first_dma; + void *sense_buffer_virt; + dma_addr_t sense_buffer_phy; + struct scatterlist *sgl; + void *list[0]; +}; + +struct hiraid_vd_info { + __u8 name[32]; + __le16 id; + __u8 rg_id; + __u8 rg_level; + __u8 sg_num; + __u8 sg_disk_num; + __u8 vd_status; + __u8 vd_type; + __u8 rsvd1[4056]; +}; + +struct bgtask_info { + __u8 type; + __u8 progress; + __u8 rate; + __u8 rsvd0; + __le16 vd_id; + __le16 time_left; + __u8 rsvd1[4]; +}; + +struct hiraid_bgtask { + __u8 sw; + __u8 task_num; + __u8 rsvd[6]; + struct bgtask_info bgtask[MAX_REALTIME_BGTASK_NUM]; +}; + +struct hiraid_dev_info { + __le32 hdid; + __le16 target; + __u8 channel; + __u8 lun; + __u8 attr; + __u8 flag; + __le16 max_io_kb; +}; + +struct hiraid_dev_list { + __le32 dev_num; + __u32 rsvd0[3]; + struct hiraid_dev_info devinfo[MAX_DEV_ENTRY_PER_PAGE_4K]; +}; + +struct hiraid_sdev_hostdata { + u32 hdid; + u16 max_io_kb; + u8 attr; + u8 flag; + u8 rg_id; + u8 hwq; + u16 pend_count; +}; + +#endif + diff --git a/drivers/scsi/hisi_raid/hiraid_main.c b/drivers/scsi/hisi_raid/hiraid_main.c new file mode 100644 index 000000000000..b9ffa642479c --- /dev/null +++ b/drivers/scsi/hisi_raid/hiraid_main.c @@ -0,0 +1,3982 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright(c) 2022 Huawei Technologies Co., Ltd */ + +/* Huawei Raid Series Linux Driver */ + +#define pr_fmt(fmt) "hiraid: " fmt + +#include <linux/sched/signal.h> +#include <linux/version.h> +#include <linux/pci.h> +#include <linux/aer.h> +#include <linux/module.h> +#include <linux/ioport.h> +#include <linux/device.h> +#include <linux/delay.h> +#include <linux/interrupt.h> +#include <linux/cdev.h> +#include <linux/sysfs.h> +#include <linux/gfp.h> +#include <linux/types.h> +#include <linux/ratelimit.h> +#include <linux/once.h> +#include <linux/debugfs.h> +#include <linux/io-64-nonatomic-lo-hi.h> +#include <linux/blkdev.h> +#include <linux/bsg-lib.h> +#include <asm/unaligned.h> +#include <linux/sort.h> +#include <target/target_core_backend.h> + +#include <scsi/scsi.h> +#include <scsi/scsi_cmnd.h> +#include <scsi/scsi_device.h> +#include <scsi/scsi_host.h> +#include <scsi/scsi_transport.h> +#include <scsi/scsi_dbg.h> +#include <scsi/sg.h> + +#include "hiraid.h" + +static u32 admin_tmout = 60; +module_param(admin_tmout, uint, 0644); +MODULE_PARM_DESC(admin_tmout, "admin commands timeout (seconds)"); + +static u32 scmd_tmout_rawdisk = 180; +module_param(scmd_tmout_rawdisk, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_rawdisk, "scsi commands timeout for rawdisk(seconds)"); + +static u32 scmd_tmout_vd = 180; +module_param(scmd_tmout_vd, uint, 0644); +MODULE_PARM_DESC(scmd_tmout_vd, "scsi commands timeout for vd(seconds)"); + +static bool max_io_force; +module_param(max_io_force, bool, 0644); +MODULE_PARM_DESC(max_io_force, "force max_hw_sectors_kb = 1024, default false(performance first)"); + +static bool work_mode; +module_param(work_mode, bool, 0444); +MODULE_PARM_DESC(work_mode, "work mode switch, default false for multi hw queues"); + +#define MAX_IO_QUEUES 128 +#define MIN_IO_QUEUES 1 + +static int ioq_num_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < MIN_IO_QUEUES || n > MAX_IO_QUEUES) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops max_hwq_num_ops = { + .set = ioq_num_set, + .get = param_get_uint, +}; + +static u32 max_hwq_num = 128; +module_param_cb(max_hwq_num, &max_hwq_num_ops, &max_hwq_num, 0444); +MODULE_PARM_DESC(max_hwq_num, "max num of hw io queues, should >= 1, default 128"); + +static int io_queue_depth_set(const char *val, const struct kernel_param *kp) +{ + int n = 0; + int ret; + + ret = kstrtoint(val, 10, &n); + if (ret != 0 || n < 2) + return -EINVAL; + + return param_set_int(val, kp); +} + +static const struct kernel_param_ops io_queue_depth_ops = { + .set = io_queue_depth_set, + .get = param_get_uint, +}; + +static u32 io_queue_depth = 1024; +module_param_cb(io_queue_depth, &io_queue_depth_ops, &io_queue_depth, 0644); +MODULE_PARM_DESC(io_queue_depth, "set io queue depth, should >= 2"); + +static u32 log_debug_switch; +module_param(log_debug_switch, uint, 0644); +MODULE_PARM_DESC(log_debug_switch, "set log state, default zero for switch off"); + +static int extra_pool_num_set(const char *val, const struct kernel_param *kp) +{ + u8 n = 0; + int ret; + + ret = kstrtou8(val, 10, &n); + if (ret != 0) + return -EINVAL; + if (n > MAX_EXTRA_POOL_NUM) + n = MAX_EXTRA_POOL_NUM; + if (n < 1) + n = 1; + *((u8 *)kp->arg) = n; + + return 0; +} + +static const struct kernel_param_ops small_pool_num_ops = { + .set = extra_pool_num_set, + .get = param_get_byte, +}; + +/* It was found that the spindlock of a single pool conflicts + * a lot with multiple CPUs.So multiple pools are introduced + * to reduce the conflictions. + */ +static unsigned char extra_pool_num = 4; +module_param_cb(extra_pool_num, &small_pool_num_ops, &extra_pool_num, 0644); +MODULE_PARM_DESC(extra_pool_num, "set prp extra pool num, default 4, MAX 16"); + +static void hiraid_handle_async_notice(struct hiraid_dev *hdev, u32 result); +static void hiraid_handle_async_vs(struct hiraid_dev *hdev, u32 result, u32 result1); + +static struct class *hiraid_class; + +#define HIRAID_CAP_TIMEOUT_UNIT_MS (HZ / 2) + +static struct workqueue_struct *work_queue; + +#define dev_log_dbg(dev, fmt, ...) do { \ + if (unlikely(log_debug_switch)) \ + dev_info(dev, "[%s] " fmt, \ + __func__, ##__VA_ARGS__); \ +} while (0) + +#define HIRAID_DRV_VERSION "1.1.0.0" + +#define ADMIN_TIMEOUT (admin_tmout * HZ) +#define USRCMD_TIMEOUT (180 * HZ) +#define CTL_RST_TIME (600 * HZ) + +#define HIRAID_WAIT_ABNL_CMD_TIMEOUT 6 +#define HIRAID_WAIT_RST_IO_TIMEOUT 10 + +#define HIRAID_DMA_MSK_BIT_MAX 64 + +#define IOQ_PT_DATA_LEN 4096 +#define IOQ_PT_SGL_DATA_LEN (1024 * 1024) + +#define MAX_CAN_QUEUE (4096 - 1) +#define MIN_CAN_QUEUE (1024 - 1) + +enum SENSE_STATE_CODE { + SENSE_STATE_OK = 0, + SENSE_STATE_NEED_CHECK, + SENSE_STATE_ERROR, + SENSE_STATE_EP_PCIE_ERROR, + SENSE_STATE_NAC_DMA_ERROR, + SENSE_STATE_ABORTED, + SENSE_STATE_NEED_RETRY +}; + +enum { + FW_EH_OK = 0, + FW_EH_DEV_NONE = 0x701 +}; + +static const char * const raid_levels[] = {"0", "1", "5", "6", "10", "50", "60", "NA"}; + +static const char * const raid_states[] = { + "NA", "NORMAL", "FAULT", "DEGRADE", "NOT_FORMATTED", "FORMATTING", "SANITIZING", + "INITIALIZING", "INITIALIZE_FAIL", "DELETING", "DELETE_FAIL", "WRITE_PROTECT" +}; + +static int hiraid_remap_bar(struct hiraid_dev *hdev, u32 size) +{ + struct pci_dev *pdev = hdev->pdev; + + if (size > pci_resource_len(pdev, 0)) { + dev_err(hdev->dev, "input size[%u] exceed bar0 length[%llu]\n", + size, pci_resource_len(pdev, 0)); + return -ENOMEM; + } + + if (hdev->bar) + iounmap(hdev->bar); + + hdev->bar = ioremap(pci_resource_start(pdev, 0), size); + if (!hdev->bar) { + dev_err(hdev->dev, "ioremap for bar0 failed\n"); + return -ENOMEM; + } + hdev->dbs = hdev->bar + HIRAID_REG_DBS; + + return 0; +} + +static int hiraid_dev_map(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret; + + ret = pci_request_mem_regions(pdev, "hiraid"); + if (ret) { + dev_err(hdev->dev, "fail to request memory regions\n"); + return ret; + } + + ret = hiraid_remap_bar(hdev, HIRAID_REG_DBS + 4096); + if (ret) { + pci_release_mem_regions(pdev); + return ret; + } + + return 0; +} + +static void hiraid_dev_unmap(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + + if (hdev->bar) { + iounmap(hdev->bar); + hdev->bar = NULL; + } + pci_release_mem_regions(pdev); +} + +static int hiraid_pci_enable(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + int ret = -ENOMEM; + u64 maskbit = HIRAID_DMA_MSK_BIT_MAX; + + if (pci_enable_device_mem(pdev)) { + dev_err(hdev->dev, "enable pci device memory resources failed\n"); + return ret; + } + pci_set_master(pdev); + + if (readl(hdev->bar + HIRAID_REG_CSTS) == U32_MAX) { + ret = -ENODEV; + dev_err(hdev->dev, "read CSTS register failed\n"); + goto disable; + } + + hdev->cap = lo_hi_readq(hdev->bar + HIRAID_REG_CAP); + hdev->ioq_depth = min_t(u32, HIRAID_CAP_MQES(hdev->cap) + 1, io_queue_depth); + hdev->db_stride = 1 << HIRAID_CAP_STRIDE(hdev->cap); + + maskbit = HIRAID_CAP_DMAMASK(hdev->cap); + if (maskbit < 32 || maskbit > HIRAID_DMA_MSK_BIT_MAX) { + dev_err(hdev->dev, "err, dma mask invalid[%llu], set to default\n", maskbit); + maskbit = HIRAID_DMA_MSK_BIT_MAX; + } + + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(maskbit))) { + if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32))) { + dev_err(hdev->dev, "set dma mask[32] and coherent failed\n"); + goto disable; + } + dev_info(hdev->dev, "set dma mask[32] success\n"); + } else { + dev_info(hdev->dev, "set dma mask[%llu] success\n", maskbit); + } + + ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_ALL_TYPES); + if (ret < 0) { + dev_err(hdev->dev, "allocate one IRQ for setup admin queue failed\n"); + goto disable; + } + + pci_enable_pcie_error_reporting(pdev); + pci_save_state(pdev); + + return 0; + +disable: + pci_disable_device(pdev); + return ret; +} + + +/* + * It is fact that first prp and last prp may be not full page. + * The size to count total nprps for the io equal to size + page_size, + * it may be a slightly overestimate. + * + * 8B per prp address. It may be there is one prp_list address per page, + * prp_list address does not count in io data prps. So divisor equal to + * PAGE_SIZE - 8, it may be a slightly overestimate. + */ +static int hiraid_prp_pagenum(struct hiraid_dev *hdev) +{ + u32 size = 1U << ((hdev->ctrl_info->mdts) * 1U) << 12; + u32 nprps = DIV_ROUND_UP(size + hdev->page_size, hdev->page_size); + + return DIV_ROUND_UP(PRP_ENTRY_SIZE * nprps, hdev->page_size - PRP_ENTRY_SIZE); +} + +/* + * Calculates the number of pages needed for the SGL segments. For example a 4k + * page can accommodate 256 SGL descriptors. + */ +static int hiraid_sgl_pagenum(struct hiraid_dev *hdev) +{ + u32 nsge = le16_to_cpu(hdev->ctrl_info->max_num_sge); + + return DIV_ROUND_UP(nsge * sizeof(struct hiraid_sgl_desc), hdev->page_size); +} + +static inline void **hiraid_mapbuf_list(struct hiraid_mapmange *mapbuf) +{ + return mapbuf->list; +} + +static u32 hiraid_get_max_cmd_size(struct hiraid_dev *hdev) +{ + u32 alloc_size = sizeof(__le64 *) * max(hiraid_prp_pagenum(hdev), hiraid_sgl_pagenum(hdev)); + + dev_info(hdev->dev, "mapbuf size[%lu], alloc_size[%u]\n", + sizeof(struct hiraid_mapmange), alloc_size); + + return sizeof(struct hiraid_mapmange) + alloc_size; +} + +static int hiraid_build_passthru_prp(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + struct scatterlist *sg = mapbuf->sgl; + __le64 *phy_regpage, *prior_list; + u64 buf_addr = sg_dma_address(sg); + int buf_length = sg_dma_len(sg); + u32 page_size = hdev->page_size; + int offset = buf_addr & (page_size - 1); + void **list = hiraid_mapbuf_list(mapbuf); + int maplen = mapbuf->len; + struct dma_pool *pool; + dma_addr_t buffer_phy; + int i; + + maplen -= (page_size - offset); + if (maplen <= 0) { + mapbuf->first_dma = 0; + return 0; + } + + buf_length -= (page_size - offset); + if (buf_length) { + buf_addr += (page_size - offset); + } else { + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + if (maplen <= page_size) { + mapbuf->first_dma = buf_addr; + return 0; + } + + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate first admin prp_list memory failed\n"); + mapbuf->first_dma = buf_addr; + mapbuf->page_cnt = -1; + return -ENOMEM; + } + list[0] = phy_regpage; + mapbuf->first_dma = buffer_phy; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + prior_list = phy_regpage; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th admin prp list memory failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = phy_regpage; + phy_regpage[0] = prior_list[i - 1]; + prior_list[i - 1] = cpu_to_le64(buffer_phy); + i = 1; + } + phy_regpage[i++] = cpu_to_le64(buf_addr); + buf_addr += page_size; + buf_length -= page_size; + maplen -= page_size; + if (maplen <= 0) + break; + if (buf_length > 0) + continue; + if (unlikely(buf_length < 0)) + goto bad_admin_sgl; + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + return 0; + +bad_admin_sgl: + dev_err(hdev->dev, "setup prps, invalid admin SGL for payload[%d] nents[%d]\n", + mapbuf->len, mapbuf->sge_cnt); + return -EIO; +} + +static int hiraid_build_prp(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + struct scatterlist *sg = mapbuf->sgl; + __le64 *phy_regpage, *prior_list; + u64 buf_addr = sg_dma_address(sg); + int buf_length = sg_dma_len(sg); + u32 page_size = hdev->page_size; + int offset = buf_addr & (page_size - 1); + void **list = hiraid_mapbuf_list(mapbuf); + int maplen = mapbuf->len; + struct dma_pool *pool; + dma_addr_t buffer_phy; + int nprps, i; + + maplen -= (page_size - offset); + if (maplen <= 0) { + mapbuf->first_dma = 0; + return 0; + } + + buf_length -= (page_size - offset); + if (buf_length) { + buf_addr += (page_size - offset); + } else { + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + if (maplen <= page_size) { + mapbuf->first_dma = buf_addr; + return 0; + } + + nprps = DIV_ROUND_UP(maplen, page_size); + if (nprps <= (EXTRA_POOL_SIZE / PRP_ENTRY_SIZE)) { + pool = mapbuf->hiraidq->prp_small_pool; + mapbuf->page_cnt = 0; + } else { + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + } + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate first prp_list memory failed\n"); + mapbuf->first_dma = buf_addr; + mapbuf->page_cnt = -1; + return -ENOMEM; + } + list[0] = phy_regpage; + mapbuf->first_dma = buffer_phy; + i = 0; + for (;;) { + if (i == page_size / PRP_ENTRY_SIZE) { + prior_list = phy_regpage; + + phy_regpage = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!phy_regpage) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th prp list memory failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = phy_regpage; + phy_regpage[0] = prior_list[i - 1]; + prior_list[i - 1] = cpu_to_le64(buffer_phy); + i = 1; + } + phy_regpage[i++] = cpu_to_le64(buf_addr); + buf_addr += page_size; + buf_length -= page_size; + maplen -= page_size; + if (maplen <= 0) + break; + if (buf_length > 0) + continue; + if (unlikely(buf_length < 0)) + goto bad_sgl; + sg = sg_next(sg); + buf_addr = sg_dma_address(sg); + buf_length = sg_dma_len(sg); + } + + return 0; + +bad_sgl: + dev_err(hdev->dev, "setup prps, invalid SGL for payload[%d] nents[%d]\n", + mapbuf->len, mapbuf->sge_cnt); + return -EIO; +} + +#define SGES_PER_PAGE (PAGE_SIZE / sizeof(struct hiraid_sgl_desc)) + +static void hiraid_submit_cmd(struct hiraid_queue *hiraidq, const void *cmd) +{ + u32 sqes = SQE_SIZE(hiraidq->qid); + unsigned long flags; + struct hiraid_admin_com_cmd *acd = (struct hiraid_admin_com_cmd *)cmd; + + spin_lock_irqsave(&hiraidq->sq_lock, flags); + memcpy((hiraidq->sq_cmds + sqes * hiraidq->sq_tail), cmd, sqes); + if (++hiraidq->sq_tail == hiraidq->q_depth) + hiraidq->sq_tail = 0; + + writel(hiraidq->sq_tail, hiraidq->q_db); + spin_unlock_irqrestore(&hiraidq->sq_lock, flags); + + dev_log_dbg(hiraidq->hdev->dev, "cid[%d] qid[%d] opcode[0x%x] flags[0x%x] hdid[%u]\n", + le16_to_cpu(acd->cmd_id), hiraidq->qid, acd->opcode, acd->flags, + le32_to_cpu(acd->hdid)); +} + +static inline bool hiraid_is_rw_scmd(struct scsi_cmnd *scmd) +{ + switch (scmd->cmnd[0]) { + case READ_6: + case READ_10: + case READ_12: + case READ_16: + case WRITE_6: + case WRITE_10: + case WRITE_12: + case WRITE_16: + return true; + default: + return false; + } +} + +/* + * checks if prps can be built for the IO cmd + */ +static bool hiraid_is_prp(struct hiraid_dev *hdev, struct scatterlist *sgl, u32 nsge) +{ + struct scatterlist *sg = sgl; + u32 page_mask = hdev->page_size - 1; + bool is_prp = true; + u32 i = 0; + + for_each_sg(sgl, sg, nsge, i) { + /* + * Data length of the middle sge multiple of page_size, + * address page_size aligned. + */ + if (i != 0 && i != nsge - 1) { + if ((sg_dma_len(sg) & page_mask) || + (sg_dma_address(sg) & page_mask)) { + is_prp = false; + break; + } + } + + /* + * The first sge addr plus the data length meets + * the page_size alignment. + */ + if (nsge > 1 && i == 0) { + if ((sg_dma_address(sg) + sg_dma_len(sg)) & page_mask) { + is_prp = false; + break; + } + } + + /* The last sge addr meets the page_size alignment. */ + if (nsge > 1 && i == (nsge - 1)) { + if (sg_dma_address(sg) & page_mask) { + is_prp = false; + break; + } + } + } + + return is_prp; +} + +enum { + HIRAID_SGL_FMT_DATA_DESC = 0x00, + HIRAID_SGL_FMT_SEG_DESC = 0x02, + HIRAID_SGL_FMT_LAST_SEG_DESC = 0x03, + HIRAID_KEY_SGL_FMT_DATA_DESC = 0x04, + HIRAID_TRANSPORT_SGL_DATA_DESC = 0x05 +}; + +static void hiraid_sgl_set_data(struct hiraid_sgl_desc *sge, struct scatterlist *sg) +{ + sge->addr = cpu_to_le64(sg_dma_address(sg)); + sge->length = cpu_to_le32(sg_dma_len(sg)); + sge->type = HIRAID_SGL_FMT_DATA_DESC << 4; +} + +static void hiraid_sgl_set_seg(struct hiraid_sgl_desc *sge, dma_addr_t buffer_phy, int entries) +{ + sge->addr = cpu_to_le64(buffer_phy); + if (entries <= SGES_PER_PAGE) { + sge->length = cpu_to_le32(entries * sizeof(*sge)); + sge->type = HIRAID_SGL_FMT_LAST_SEG_DESC << 4; + } else { + sge->length = cpu_to_le32(PAGE_SIZE); + sge->type = HIRAID_SGL_FMT_SEG_DESC << 4; + } +} + +static int hiraid_build_passthru_sgl(struct hiraid_dev *hdev, + struct hiraid_admin_command *admin_cmd, + struct hiraid_mapmange *mapbuf) +{ + struct hiraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = mapbuf->sgl; + void **list = hiraid_mapbuf_list(mapbuf); + struct dma_pool *pool; + int nsge = mapbuf->sge_cnt; + dma_addr_t buffer_phy; + int i = 0; + + admin_cmd->common.flags |= SQE_FLAG_SGL_METABUF; + + if (nsge == 1) { + hiraid_sgl_set_data(&admin_cmd->common.dptr.sgl, sg); + return 0; + } + + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate first admin sgl_list failed\n"); + mapbuf->page_cnt = -1; + return -ENOMEM; + } + + list[0] = sg_list; + mapbuf->first_dma = buffer_phy; + hiraid_sgl_set_seg(&admin_cmd->common.dptr.sgl, buffer_phy, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th admin sgl_list failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + hiraid_sgl_set_seg(link, buffer_phy, nsge); + } + + hiraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + + +static int hiraid_build_sgl(struct hiraid_dev *hdev, struct hiraid_scsi_io_cmd *io_cmd, + struct hiraid_mapmange *mapbuf) +{ + struct hiraid_sgl_desc *sg_list, *link, *old_sg_list; + struct scatterlist *sg = mapbuf->sgl; + void **list = hiraid_mapbuf_list(mapbuf); + struct dma_pool *pool; + int nsge = mapbuf->sge_cnt; + dma_addr_t buffer_phy; + int i = 0; + + io_cmd->common.flags |= SQE_FLAG_SGL_METABUF; + + if (nsge == 1) { + hiraid_sgl_set_data(&io_cmd->common.dptr.sgl, sg); + return 0; + } + + if (nsge <= (EXTRA_POOL_SIZE / sizeof(struct hiraid_sgl_desc))) { + pool = mapbuf->hiraidq->prp_small_pool; + mapbuf->page_cnt = 0; + } else { + pool = hdev->prp_page_pool; + mapbuf->page_cnt = 1; + } + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate first sgl_list failed\n"); + mapbuf->page_cnt = -1; + return -ENOMEM; + } + + list[0] = sg_list; + mapbuf->first_dma = buffer_phy; + hiraid_sgl_set_seg(&io_cmd->common.dptr.sgl, buffer_phy, nsge); + do { + if (i == SGES_PER_PAGE) { + old_sg_list = sg_list; + link = &old_sg_list[SGES_PER_PAGE - 1]; + + sg_list = dma_pool_alloc(pool, GFP_ATOMIC, &buffer_phy); + if (!sg_list) { + dev_err_ratelimited(hdev->dev, "allocate [%d]th sgl_list failed\n", + mapbuf->page_cnt + 1); + return -ENOMEM; + } + list[mapbuf->page_cnt++] = sg_list; + + i = 0; + memcpy(&sg_list[i++], link, sizeof(*link)); + hiraid_sgl_set_seg(link, buffer_phy, nsge); + } + + hiraid_sgl_set_data(&sg_list[i++], sg); + sg = sg_next(sg); + } while (--nsge > 0); + + return 0; +} + +#define HIRAID_RW_FUA BIT(14) + +static int hiraid_setup_rw_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_rw_cmd *io_cmd, + struct scsi_cmnd *scmd) +{ + u32 start_lba_lo, start_lba_hi; + u32 datalength = 0; + u16 control = 0; + + start_lba_lo = 0; + start_lba_hi = 0; + + if (scmd->sc_data_direction == DMA_TO_DEVICE) { + io_cmd->opcode = HIRAID_CMD_WRITE; + } else if (scmd->sc_data_direction == DMA_FROM_DEVICE) { + io_cmd->opcode = HIRAID_CMD_READ; + } else { + dev_err(hdev->dev, "invalid RW_IO for unsupported data direction[%d]\n", + scmd->sc_data_direction); + WARN_ON(1); + return -EINVAL; + } + + /* 6-byte READ(0x08) or WRITE(0x0A) cdb */ + if (scmd->cmd_len == 6) { + datalength = (u32)(scmd->cmnd[4] == 0 ? + IO_6_DEFAULT_TX_LEN : scmd->cmnd[4]); + start_lba_lo = (u32)get_unaligned_be24(&scmd->cmnd[1]); + + start_lba_lo &= 0x1FFFFF; + } + + /* 10-byte READ(0x28) or WRITE(0x2A) cdb */ + else if (scmd->cmd_len == 10) { + datalength = (u32)get_unaligned_be16(&scmd->cmnd[7]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + + /* 12-byte READ(0xA8) or WRITE(0xAA) cdb */ + else if (scmd->cmd_len == 12) { + datalength = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + /* 16-byte READ(0x88) or WRITE(0x8A) cdb */ + else if (scmd->cmd_len == 16) { + datalength = get_unaligned_be32(&scmd->cmnd[10]); + start_lba_lo = get_unaligned_be32(&scmd->cmnd[6]); + start_lba_hi = get_unaligned_be32(&scmd->cmnd[2]); + + if (scmd->cmnd[1] & FUA_MASK) + control |= HIRAID_RW_FUA; + } + + if (unlikely(datalength > U16_MAX || datalength == 0)) { + dev_err(hdev->dev, "invalid IO for illegal transfer data length[%u]\n", datalength); + WARN_ON(1); + return -EINVAL; + } + + io_cmd->slba = cpu_to_le64(((u64)start_lba_hi << 32) | start_lba_lo); + /* 0base for nlb */ + io_cmd->nlb = cpu_to_le16((u16)(datalength - 1)); + io_cmd->control = cpu_to_le16(control); + + return 0; +} + +static int hiraid_setup_nonrw_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_nonrw_cmd *io_cmd, struct scsi_cmnd *scmd) +{ + io_cmd->buf_len = cpu_to_le32(scsi_bufflen(scmd)); + + switch (scmd->sc_data_direction) { + case DMA_NONE: + io_cmd->opcode = HIRAID_CMD_NONRW_NONE; + break; + case DMA_TO_DEVICE: + io_cmd->opcode = HIRAID_CMD_NONRW_TODEV; + break; + case DMA_FROM_DEVICE: + io_cmd->opcode = HIRAID_CMD_NONRW_FROMDEV; + break; + default: + dev_err(hdev->dev, "invalid NON_IO for unsupported data direction[%d]\n", + scmd->sc_data_direction); + WARN_ON(1); + return -EINVAL; + } + + return 0; +} + +static int hiraid_setup_io_cmd(struct hiraid_dev *hdev, + struct hiraid_scsi_io_cmd *io_cmd, struct scsi_cmnd *scmd) +{ + memcpy(io_cmd->common.cdb, scmd->cmnd, scmd->cmd_len); + io_cmd->common.cdb_len = scmd->cmd_len; + + if (hiraid_is_rw_scmd(scmd)) + return hiraid_setup_rw_cmd(hdev, &io_cmd->rw, scmd); + else + return hiraid_setup_nonrw_cmd(hdev, &io_cmd->nonrw, scmd); +} + +static inline void hiraid_init_mapbuff(struct hiraid_mapmange *mapbuf) +{ + mapbuf->sge_cnt = 0; + mapbuf->page_cnt = -1; + mapbuf->use_sgl = false; + WRITE_ONCE(mapbuf->state, CMD_IDLE); +} + +static void hiraid_free_mapbuf(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf) +{ + const int last_prp = hdev->page_size / sizeof(__le64) - 1; + dma_addr_t buffer_phy, next_buffer_phy; + struct hiraid_sgl_desc *sg_list; + __le64 *prp_list; + void *addr; + int i; + + buffer_phy = mapbuf->first_dma; + if (mapbuf->page_cnt == 0) + dma_pool_free(mapbuf->hiraidq->prp_small_pool, + hiraid_mapbuf_list(mapbuf)[0], buffer_phy); + + for (i = 0; i < mapbuf->page_cnt; i++) { + addr = hiraid_mapbuf_list(mapbuf)[i]; + + if (mapbuf->use_sgl) { + sg_list = addr; + next_buffer_phy = + le64_to_cpu((sg_list[SGES_PER_PAGE - 1]).addr); + } else { + prp_list = addr; + next_buffer_phy = le64_to_cpu(prp_list[last_prp]); + } + + dma_pool_free(hdev->prp_page_pool, addr, buffer_phy); + buffer_phy = next_buffer_phy; + } + + mapbuf->sense_buffer_virt = NULL; + mapbuf->page_cnt = -1; +} + +static int hiraid_io_map_data(struct hiraid_dev *hdev, struct hiraid_mapmange *mapbuf, + struct scsi_cmnd *scmd, struct hiraid_scsi_io_cmd *io_cmd) +{ + int ret; + + ret = scsi_dma_map(scmd); + if (unlikely(ret < 0)) + return ret; + mapbuf->sge_cnt = ret; + + /* No data to DMA, it may be scsi no-rw command */ + if (unlikely(mapbuf->sge_cnt == 0)) + return 0; + + mapbuf->len = scsi_bufflen(scmd); + mapbuf->sgl = scsi_sglist(scmd); + mapbuf->use_sgl = !hiraid_is_prp(hdev, mapbuf->sgl, mapbuf->sge_cnt); + + if (mapbuf->use_sgl) { + ret = hiraid_build_sgl(hdev, io_cmd, mapbuf); + } else { + ret = hiraid_build_prp(hdev, mapbuf); + io_cmd->common.dptr.prp1 = + cpu_to_le64(sg_dma_address(mapbuf->sgl)); + io_cmd->common.dptr.prp2 = cpu_to_le64(mapbuf->first_dma); + } + + if (ret) + scsi_dma_unmap(scmd); + + return ret; +} + +static void hiraid_check_status(struct hiraid_mapmange *mapbuf, struct scsi_cmnd *scmd, + struct hiraid_completion *cqe) +{ + scsi_set_resid(scmd, 0); + + switch ((le16_to_cpu(cqe->status) >> 1) & 0x7f) { + case SENSE_STATE_OK: + set_host_byte(scmd, DID_OK); + break; + case SENSE_STATE_NEED_CHECK: + set_host_byte(scmd, DID_OK); + scmd->result |= le16_to_cpu(cqe->status) >> 8; + if (scmd->result & SAM_STAT_CHECK_CONDITION) { + memset(scmd->sense_buffer, 0, SCSI_SENSE_BUFFERSIZE); + memcpy(scmd->sense_buffer, + mapbuf->sense_buffer_virt, SCSI_SENSE_BUFFERSIZE); + scmd->result = (scmd->result & 0x00ffffff) | (DRIVER_SENSE << 24); + } + break; + case SENSE_STATE_ABORTED: + set_host_byte(scmd, DID_ABORT); + break; + case SENSE_STATE_NEED_RETRY: + set_host_byte(scmd, DID_REQUEUE); + break; + default: + set_host_byte(scmd, DID_BAD_TARGET); + dev_warn_ratelimited(mapbuf->hiraidq->hdev->dev, "cid[%d] qid[%d] sdev[%d:%d] opcode[%.2x] bad status[0x%x]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->sq_id), scmd->device->channel, + scmd->device->id, scmd->cmnd[0], le16_to_cpu(cqe->status)); + break; + } +} + +static inline void hiraid_query_scmd_tag(struct scsi_cmnd *scmd, u16 *qid, u16 *cid, + struct hiraid_dev *hdev, struct hiraid_sdev_hostdata *hostdata) +{ + u32 tag = blk_mq_unique_tag(blk_mq_rq_from_pdu((void *)scmd)); + + if (work_mode) { + if ((hdev->hdd_dispatch == DISPATCH_BY_DISK) && (hostdata->hwq != 0)) + *qid = hostdata->hwq; + else + *qid = raw_smp_processor_id() % (hdev->online_queues - 1) + 1; + } else { + *qid = blk_mq_unique_tag_to_hwq(tag) + 1; + } + *cid = blk_mq_unique_tag_to_tag(tag); +} + +static int hiraid_queue_command(struct Scsi_Host *shost, struct scsi_cmnd *scmd) +{ + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_dev *hdev = shost_priv(shost); + struct scsi_device *sdev = scmd->device; + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_scsi_io_cmd io_cmd; + struct hiraid_queue *ioq; + u16 hwq, cid; + int ret; + + if (unlikely(hdev->state == DEV_RESETTING)) + return SCSI_MLQUEUE_HOST_BUSY; + + if (unlikely(hdev->state != DEV_LIVE)) { + set_host_byte(scmd, DID_NO_CONNECT); + scmd->scsi_done(scmd); + return 0; + } + + if (log_debug_switch) + scsi_print_command(scmd); + + hostdata = sdev->hostdata; + hiraid_query_scmd_tag(scmd, &hwq, &cid, hdev, hostdata); + ioq = &hdev->queues[hwq]; + + if (unlikely(atomic_inc_return(&ioq->inflight) > + (hdev->ioq_depth - HIRAID_PTHRU_CMDS_PERQ))) { + atomic_dec(&ioq->inflight); + return SCSI_MLQUEUE_HOST_BUSY; + } + + memset(&io_cmd, 0, sizeof(io_cmd)); + io_cmd.rw.hdid = cpu_to_le32(hostdata->hdid); + io_cmd.rw.cmd_id = cpu_to_le16(cid); + + ret = hiraid_setup_io_cmd(hdev, &io_cmd, scmd); + if (unlikely(ret)) { + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + atomic_dec(&ioq->inflight); + return 0; + } + + ret = cid * SCSI_SENSE_BUFFERSIZE; + if (work_mode) { + mapbuf->sense_buffer_virt = hdev->sense_buffer_virt + ret; + mapbuf->sense_buffer_phy = hdev->sense_buffer_phy + ret; + } else { + mapbuf->sense_buffer_virt = ioq->sense_buffer_virt + ret; + mapbuf->sense_buffer_phy = ioq->sense_buffer_phy + ret; + } + io_cmd.common.sense_addr = cpu_to_le64(mapbuf->sense_buffer_phy); + io_cmd.common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + + hiraid_init_mapbuff(mapbuf); + + mapbuf->hiraidq = ioq; + mapbuf->cid = cid; + ret = hiraid_io_map_data(hdev, mapbuf, scmd, &io_cmd); + if (unlikely(ret)) { + dev_err(hdev->dev, "io map data err\n"); + set_host_byte(scmd, DID_ERROR); + scmd->scsi_done(scmd); + ret = 0; + goto deinit_iobuf; + } + + WRITE_ONCE(mapbuf->state, CMD_FLIGHT); + hiraid_submit_cmd(ioq, &io_cmd); + + return 0; + +deinit_iobuf: + atomic_dec(&ioq->inflight); + hiraid_free_mapbuf(hdev, mapbuf); + return ret; +} + +static int hiraid_match_dev(struct hiraid_dev *hdev, u16 idx, struct scsi_device *sdev) +{ + if (HIRAID_DEV_INFO_FLAG_VALID(hdev->dev_info[idx].flag)) { + if (sdev->channel == hdev->dev_info[idx].channel && + sdev->id == le16_to_cpu(hdev->dev_info[idx].target) && + sdev->lun < hdev->dev_info[idx].lun) { + dev_info(hdev->dev, "match device success, channel:target:lun[%d:%d:%d]\n", + hdev->dev_info[idx].channel, + hdev->dev_info[idx].target, + hdev->dev_info[idx].lun); + return 1; + } + } + + return 0; +} + +static int hiraid_disk_qd(u8 attr) +{ + switch (HIRAID_DEV_DISK_TYPE(attr)) { + case HIRAID_SAS_HDD_VD: + case HIRAID_SATA_HDD_VD: + return HIRAID_HDD_VD_QD; + case HIRAID_SAS_SSD_VD: + case HIRAID_SATA_SSD_VD: + case HIRAID_NVME_SSD_VD: + return HIRAID_SSD_VD_QD; + case HIRAID_SAS_HDD_PD: + case HIRAID_SATA_HDD_PD: + return HIRAID_HDD_PD_QD; + case HIRAID_SAS_SSD_PD: + case HIRAID_SATA_SSD_PD: + case HIRAID_NVME_SSD_PD: + return HIRAID_SSD_PD_QD; + default: + return MAX_CMD_PER_DEV; + } +} + +static bool hiraid_disk_is_hdd(u8 attr) +{ + switch (HIRAID_DEV_DISK_TYPE(attr)) { + case HIRAID_SAS_HDD_VD: + case HIRAID_SATA_HDD_VD: + case HIRAID_SAS_HDD_PD: + case HIRAID_SATA_HDD_PD: + return true; + default: + return false; + } +} + +static int hiraid_slave_alloc(struct scsi_device *sdev) +{ + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_dev *hdev; + u16 idx; + + hdev = shost_priv(sdev->host); + hostdata = kzalloc(sizeof(*hostdata), GFP_KERNEL); + if (!hostdata) { + dev_err(hdev->dev, "alloc scsi host data memory failed\n"); + return -ENOMEM; + } + + down_read(&hdev->dev_rwsem); + for (idx = 0; idx < le32_to_cpu(hdev->ctrl_info->nd); idx++) { + if (hiraid_match_dev(hdev, idx, sdev)) + goto scan_host; + } + up_read(&hdev->dev_rwsem); + + kfree(hostdata); + return -ENXIO; + +scan_host: + hostdata->hdid = le32_to_cpu(hdev->dev_info[idx].hdid); + hostdata->max_io_kb = le16_to_cpu(hdev->dev_info[idx].max_io_kb); + hostdata->attr = hdev->dev_info[idx].attr; + hostdata->flag = hdev->dev_info[idx].flag; + hostdata->rg_id = 0xff; + sdev->hostdata = hostdata; + up_read(&hdev->dev_rwsem); + return 0; +} + +static void hiraid_slave_destroy(struct scsi_device *sdev) +{ + kfree(sdev->hostdata); + sdev->hostdata = NULL; +} + +static int hiraid_slave_configure(struct scsi_device *sdev) +{ + unsigned int timeout = scmd_tmout_rawdisk * HZ; + struct hiraid_dev *hdev = shost_priv(sdev->host); + struct hiraid_sdev_hostdata *hostdata = sdev->hostdata; + u32 max_sec = sdev->host->max_sectors; + int qd = MAX_CMD_PER_DEV; + + if (hostdata) { + if (HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + timeout = scmd_tmout_vd * HZ; + else if (HIRAID_DEV_INFO_ATTR_RAWDISK(hostdata->attr)) + timeout = scmd_tmout_rawdisk * HZ; + max_sec = hostdata->max_io_kb << 1; + qd = hiraid_disk_qd(hostdata->attr); + + if (hiraid_disk_is_hdd(hostdata->attr)) + hostdata->hwq = hostdata->hdid % (hdev->online_queues - 1) + 1; + else + hostdata->hwq = 0; + } else { + dev_err(hdev->dev, "err, sdev->hostdata is null\n"); + } + + blk_queue_rq_timeout(sdev->request_queue, timeout); + sdev->eh_timeout = timeout; + scsi_change_queue_depth(sdev, qd); + + if ((max_sec == 0) || (max_sec > sdev->host->max_sectors)) + max_sec = sdev->host->max_sectors; + + if (!max_io_force) + blk_queue_max_hw_sectors(sdev->request_queue, max_sec); + + dev_info(hdev->dev, "sdev->channel:id:lun[%d:%d:%lld] scmd_timeout[%d]s maxsec[%d]\n", + sdev->channel, sdev->id, sdev->lun, timeout / HZ, max_sec); + + return 0; +} + +static void hiraid_shost_init(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u8 domain, bus; + u32 dev_func; + + domain = pci_domain_nr(pdev->bus); + bus = pdev->bus->number; + dev_func = pdev->devfn; + + hdev->shost->nr_hw_queues = work_mode ? 1 : hdev->online_queues - 1; + hdev->shost->can_queue = hdev->scsi_qd; + + hdev->shost->sg_tablesize = le16_to_cpu(hdev->ctrl_info->max_num_sge); + /* 512B per sector */ + hdev->shost->max_sectors = (1U << ((hdev->ctrl_info->mdts) * 1U) << 12) / 512; + hdev->shost->cmd_per_lun = MAX_CMD_PER_DEV; + hdev->shost->max_channel = le16_to_cpu(hdev->ctrl_info->max_channel) - 1; + hdev->shost->max_id = le32_to_cpu(hdev->ctrl_info->max_tgt_id); + hdev->shost->max_lun = le16_to_cpu(hdev->ctrl_info->max_lun); + + hdev->shost->this_id = -1; + hdev->shost->unique_id = (domain << 16) | (bus << 8) | dev_func; + hdev->shost->max_cmd_len = MAX_CDB_LEN; + hdev->shost->hostt->cmd_size = hiraid_get_max_cmd_size(hdev); +} + +static int hiraid_alloc_queue(struct hiraid_dev *hdev, u16 qid, u16 depth) +{ + struct hiraid_queue *hiraidq = &hdev->queues[qid]; + int ret = 0; + + if (hdev->queue_count > qid) { + dev_info(hdev->dev, "warn: queue[%d] is exist\n", qid); + return 0; + } + + hiraidq->cqes = dma_alloc_coherent(hdev->dev, CQ_SIZE(depth), + &hiraidq->cq_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hiraidq->cqes) + return -ENOMEM; + + hiraidq->sq_cmds = dma_alloc_coherent(hdev->dev, SQ_SIZE(qid, depth), + &hiraidq->sq_buffer_phy, GFP_KERNEL); + if (!hiraidq->sq_cmds) { + ret = -ENOMEM; + goto free_cqes; + } + + /* + * if single hw queue, we do not need to alloc sense buffer for every queue, + * we have alloced all on hiraid_alloc_resources. + */ + if (work_mode) + goto initq; + + /* alloc sense buffer */ + hiraidq->sense_buffer_virt = dma_alloc_coherent(hdev->dev, SENSE_SIZE(depth), + &hiraidq->sense_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hiraidq->sense_buffer_virt) { + ret = -ENOMEM; + goto free_sq_cmds; + } + +initq: + spin_lock_init(&hiraidq->sq_lock); + spin_lock_init(&hiraidq->cq_lock); + hiraidq->hdev = hdev; + hiraidq->q_depth = depth; + hiraidq->qid = qid; + hiraidq->cq_vector = -1; + hdev->queue_count++; + + return 0; + +free_sq_cmds: + dma_free_coherent(hdev->dev, SQ_SIZE(qid, depth), (void *)hiraidq->sq_cmds, + hiraidq->sq_buffer_phy); +free_cqes: + dma_free_coherent(hdev->dev, CQ_SIZE(depth), (void *)hiraidq->cqes, + hiraidq->cq_buffer_phy); + return ret; +} + +static int hiraid_wait_control_ready(struct hiraid_dev *hdev, u64 cap, bool enabled) +{ + unsigned long timeout = + ((HIRAID_CAP_TIMEOUT(cap) + 1) * HIRAID_CAP_TIMEOUT_UNIT_MS) + jiffies; + u32 bit = enabled ? HIRAID_CSTS_RDY : 0; + + while ((readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_RDY) != bit) { + usleep_range(1000, 2000); + if (fatal_signal_pending(current)) + return -EINTR; + + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "device not ready; aborting %s\n", + enabled ? "initialisation" : "reset"); + return -ENODEV; + } + } + return 0; +} + +static int hiraid_shutdown_control(struct hiraid_dev *hdev) +{ + unsigned long timeout = le32_to_cpu(hdev->ctrl_info->rtd3e) / 1000000 * HZ + jiffies; + + hdev->ctrl_config &= ~HIRAID_CC_SHN_MASK; + hdev->ctrl_config |= HIRAID_CC_SHN_NORMAL; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + while ((readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_SHST_MASK) != + HIRAID_CSTS_SHST_CMPLT) { + msleep(100); + if (fatal_signal_pending(current)) + return -EINTR; + if (time_after(jiffies, timeout)) { + dev_err(hdev->dev, "device shutdown incomplete, abort shutdown\n"); + return -ENODEV; + } + } + return 0; +} + +static int hiraid_disable_control(struct hiraid_dev *hdev) +{ + hdev->ctrl_config &= ~HIRAID_CC_SHN_MASK; + hdev->ctrl_config &= ~HIRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + return hiraid_wait_control_ready(hdev, hdev->cap, false); +} + +static int hiraid_enable_control(struct hiraid_dev *hdev) +{ + u64 cap = hdev->cap; + u32 dev_page_min = HIRAID_CAP_MPSMIN(cap) + 12; + u32 page_shift = PAGE_SHIFT; + + if (page_shift < dev_page_min) { + dev_err(hdev->dev, "minimum device page size[%u], too large for host[%u]\n", + 1U << dev_page_min, 1U << page_shift); + return -ENODEV; + } + + page_shift = min_t(unsigned int, HIRAID_CAP_MPSMAX(cap) + 12, PAGE_SHIFT); + hdev->page_size = 1U << page_shift; + + hdev->ctrl_config = HIRAID_CC_CSS_NVM; + hdev->ctrl_config |= (page_shift - 12) << HIRAID_CC_MPS_SHIFT; + hdev->ctrl_config |= HIRAID_CC_AMS_RR | HIRAID_CC_SHN_NONE; + hdev->ctrl_config |= HIRAID_CC_IOSQES | HIRAID_CC_IOCQES; + hdev->ctrl_config |= HIRAID_CC_ENABLE; + writel(hdev->ctrl_config, hdev->bar + HIRAID_REG_CC); + + return hiraid_wait_control_ready(hdev, cap, true); +} + +static void hiraid_init_queue(struct hiraid_queue *hiraidq, u16 qid) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + + memset((void *)hiraidq->cqes, 0, CQ_SIZE(hiraidq->q_depth)); + + hiraidq->sq_tail = 0; + hiraidq->cq_head = 0; + hiraidq->cq_phase = 1; + hiraidq->q_db = &hdev->dbs[qid * 2 * hdev->db_stride]; + hiraidq->prp_small_pool = hdev->prp_extra_pool[qid % extra_pool_num]; + hdev->online_queues++; + atomic_set(&hiraidq->inflight, 0); +} + +static inline bool hiraid_cqe_pending(struct hiraid_queue *hiraidq) +{ + return (le16_to_cpu(hiraidq->cqes[hiraidq->cq_head].status) & 1) == + hiraidq->cq_phase; +} + +static void hiraid_complete_io_cmnd(struct hiraid_queue *ioq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = ioq->hdev; + struct blk_mq_tags *tags; + struct scsi_cmnd *scmd; + struct hiraid_mapmange *mapbuf; + struct request *req; + unsigned long elapsed; + + atomic_dec(&ioq->inflight); + + if (work_mode) + tags = hdev->shost->tag_set.tags[0]; + else + tags = hdev->shost->tag_set.tags[ioq->qid - 1]; + req = blk_mq_tag_to_rq(tags, le16_to_cpu(cqe->cmd_id)); + if (unlikely(!req || !blk_mq_request_started(req))) { + dev_warn(hdev->dev, "invalid id[%d] completed on queue[%d]\n", + le16_to_cpu(cqe->cmd_id), ioq->qid); + return; + } + + scmd = blk_mq_rq_to_pdu(req); + mapbuf = scsi_cmd_priv(scmd); + + elapsed = jiffies - scmd->jiffies_at_alloc; + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] finish IO cost %3ld.%3ld seconds\n", + le16_to_cpu(cqe->cmd_id), ioq->qid, elapsed / HZ, elapsed % HZ); + + if (cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_COMPLETE) != CMD_FLIGHT) { + dev_warn(hdev->dev, "cid[%d] qid[%d] enters abnormal handler, cost %3ld.%3ld seconds\n", + le16_to_cpu(cqe->cmd_id), ioq->qid, elapsed / HZ, elapsed % HZ); + WRITE_ONCE(mapbuf->state, CMD_TMO_COMPLETE); + + if (mapbuf->sge_cnt) { + mapbuf->sge_cnt = 0; + scsi_dma_unmap(scmd); + } + hiraid_free_mapbuf(hdev, mapbuf); + + return; + } + + hiraid_check_status(mapbuf, scmd, cqe); + if (mapbuf->sge_cnt) { + mapbuf->sge_cnt = 0; + scsi_dma_unmap(scmd); + } + hiraid_free_mapbuf(hdev, mapbuf); + scmd->scsi_done(scmd); +} + +static void hiraid_complete_admin_cmnd(struct hiraid_queue *adminq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = adminq->hdev; + struct hiraid_cmd *adm_cmd; + + adm_cmd = hdev->adm_cmds + le16_to_cpu(cqe->cmd_id); + if (unlikely(adm_cmd->state == CMD_IDLE)) { + dev_warn(adminq->hdev->dev, "invalid id[%d] completed on queue[%d]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->sq_id)); + return; + } + + adm_cmd->status = le16_to_cpu(cqe->status) >> 1; + adm_cmd->result0 = le32_to_cpu(cqe->result); + adm_cmd->result1 = le32_to_cpu(cqe->result1); + + complete(&adm_cmd->cmd_done); +} + +static void hiraid_send_async_event(struct hiraid_dev *hdev, u16 cid); + +static void hiraid_complete_async_event(struct hiraid_queue *hiraidq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + u32 result = le32_to_cpu(cqe->result); + + dev_info(hdev->dev, "recv async event, cid[%d] status[0x%x] result[0x%x]\n", + le16_to_cpu(cqe->cmd_id), le16_to_cpu(cqe->status) >> 1, result); + + hiraid_send_async_event(hdev, le16_to_cpu(cqe->cmd_id)); + + if ((le16_to_cpu(cqe->status) >> 1) != HIRAID_SC_SUCCESS) + return; + switch (result & 0x7) { + case HIRAID_ASYN_EVENT_NOTICE: + hiraid_handle_async_notice(hdev, result); + break; + case HIRAID_ASYN_EVENT_VS: + hiraid_handle_async_vs(hdev, result, le32_to_cpu(cqe->result1)); + break; + default: + dev_warn(hdev->dev, "unsupported async event type[%u]\n", result & 0x7); + break; + } +} + +static void hiraid_complete_pthru_cmnd(struct hiraid_queue *ioq, struct hiraid_completion *cqe) +{ + struct hiraid_dev *hdev = ioq->hdev; + struct hiraid_cmd *ptcmd; + + ptcmd = hdev->io_ptcmds + (ioq->qid - 1) * HIRAID_PTHRU_CMDS_PERQ + + le16_to_cpu(cqe->cmd_id) - hdev->scsi_qd; + + ptcmd->status = le16_to_cpu(cqe->status) >> 1; + ptcmd->result0 = le32_to_cpu(cqe->result); + ptcmd->result1 = le32_to_cpu(cqe->result1); + + complete(&ptcmd->cmd_done); +} + +static inline void hiraid_handle_cqe(struct hiraid_queue *hiraidq, u16 idx) +{ + struct hiraid_completion *cqe = &hiraidq->cqes[idx]; + struct hiraid_dev *hdev = hiraidq->hdev; + u16 cid = le16_to_cpu(cqe->cmd_id); + + if (unlikely(!work_mode && (cid >= hiraidq->q_depth))) { + dev_err(hdev->dev, "invalid command id[%d] completed on queue[%d]\n", + cid, cqe->sq_id); + return; + } + + dev_log_dbg(hdev->dev, "cid[%d] qid[%d] result[0x%x] sqid[%d] status[0x%x]\n", + cid, hiraidq->qid, le32_to_cpu(cqe->result), + le16_to_cpu(cqe->sq_id), le16_to_cpu(cqe->status)); + + if (unlikely(hiraidq->qid == 0 && cid >= HIRAID_AQ_BLK_MQ_DEPTH)) { + hiraid_complete_async_event(hiraidq, cqe); + return; + } + + if (unlikely(hiraidq->qid && cid >= hdev->scsi_qd)) { + hiraid_complete_pthru_cmnd(hiraidq, cqe); + return; + } + + if (hiraidq->qid) + hiraid_complete_io_cmnd(hiraidq, cqe); + else + hiraid_complete_admin_cmnd(hiraidq, cqe); +} + +static void hiraid_complete_cqes(struct hiraid_queue *hiraidq, u16 start, u16 end) +{ + while (start != end) { + hiraid_handle_cqe(hiraidq, start); + if (++start == hiraidq->q_depth) + start = 0; + } +} + +static inline void hiraid_update_cq_head(struct hiraid_queue *hiraidq) +{ + if (++hiraidq->cq_head == hiraidq->q_depth) { + hiraidq->cq_head = 0; + hiraidq->cq_phase = !hiraidq->cq_phase; + } +} + +static inline bool hiraid_process_cq(struct hiraid_queue *hiraidq, u16 *start, u16 *end, int tag) +{ + bool found = false; + + *start = hiraidq->cq_head; + while (!found && hiraid_cqe_pending(hiraidq)) { + if (le16_to_cpu(hiraidq->cqes[hiraidq->cq_head].cmd_id) == tag) + found = true; + hiraid_update_cq_head(hiraidq); + } + *end = hiraidq->cq_head; + + if (*start != *end) + writel(hiraidq->cq_head, hiraidq->q_db + hiraidq->hdev->db_stride); + + return found; +} + +static bool hiraid_poll_cq(struct hiraid_queue *hiraidq, int cid) +{ + u16 start, end; + bool found; + + if (!hiraid_cqe_pending(hiraidq)) + return 0; + + spin_lock_irq(&hiraidq->cq_lock); + found = hiraid_process_cq(hiraidq, &start, &end, cid); + spin_unlock_irq(&hiraidq->cq_lock); + + hiraid_complete_cqes(hiraidq, start, end); + return found; +} + +static irqreturn_t hiraid_handle_irq(int irq, void *data) +{ + struct hiraid_queue *hiraidq = data; + irqreturn_t ret = IRQ_NONE; + u16 start, end; + + spin_lock(&hiraidq->cq_lock); + if (hiraidq->cq_head != hiraidq->last_cq_head) + ret = IRQ_HANDLED; + + hiraid_process_cq(hiraidq, &start, &end, -1); + hiraidq->last_cq_head = hiraidq->cq_head; + spin_unlock(&hiraidq->cq_lock); + + if (start != end) { + hiraid_complete_cqes(hiraidq, start, end); + ret = IRQ_HANDLED; + } + return ret; +} + +static int hiraid_setup_admin_queue(struct hiraid_dev *hdev) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + u32 aqa; + int ret; + + dev_info(hdev->dev, "start disable controller\n"); + + ret = hiraid_disable_control(hdev); + if (ret) + return ret; + + ret = hiraid_alloc_queue(hdev, 0, HIRAID_AQ_DEPTH); + if (ret) + return ret; + + aqa = adminq->q_depth - 1; + aqa |= aqa << 16; + writel(aqa, hdev->bar + HIRAID_REG_AQA); + lo_hi_writeq(adminq->sq_buffer_phy, hdev->bar + HIRAID_REG_ASQ); + lo_hi_writeq(adminq->cq_buffer_phy, hdev->bar + HIRAID_REG_ACQ); + + dev_info(hdev->dev, "start enable controller\n"); + + ret = hiraid_enable_control(hdev); + if (ret) { + ret = -ENODEV; + return ret; + } + + adminq->cq_vector = 0; + ret = pci_request_irq(hdev->pdev, adminq->cq_vector, hiraid_handle_irq, NULL, + adminq, "hiraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + adminq->cq_vector = -1; + return ret; + } + + hiraid_init_queue(adminq, 0); + + dev_info(hdev->dev, "setup admin queue success, queuecount[%d] online[%d] pagesize[%d]\n", + hdev->queue_count, hdev->online_queues, hdev->page_size); + + return 0; +} + +static u32 hiraid_get_bar_size(struct hiraid_dev *hdev, u32 nr_ioqs) +{ + return (HIRAID_REG_DBS + ((nr_ioqs + 1) * 8 * hdev->db_stride)); +} + +static int hiraid_create_admin_cmds(struct hiraid_dev *hdev) +{ + u16 i; + + INIT_LIST_HEAD(&hdev->adm_cmd_list); + spin_lock_init(&hdev->adm_cmd_lock); + + hdev->adm_cmds = kcalloc_node(HIRAID_AQ_BLK_MQ_DEPTH, sizeof(struct hiraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->adm_cmds) { + dev_err(hdev->dev, "alloc admin cmds failed\n"); + return -ENOMEM; + } + + for (i = 0; i < HIRAID_AQ_BLK_MQ_DEPTH; i++) { + hdev->adm_cmds[i].qid = 0; + hdev->adm_cmds[i].cid = i; + list_add_tail(&(hdev->adm_cmds[i].list), &hdev->adm_cmd_list); + } + + dev_info(hdev->dev, "alloc admin cmds success, num[%d]\n", HIRAID_AQ_BLK_MQ_DEPTH); + + return 0; +} + +static void hiraid_free_admin_cmds(struct hiraid_dev *hdev) +{ + kfree(hdev->adm_cmds); + hdev->adm_cmds = NULL; + INIT_LIST_HEAD(&hdev->adm_cmd_list); +} + +static struct hiraid_cmd *hiraid_get_cmd(struct hiraid_dev *hdev, enum hiraid_cmd_type type) +{ + struct hiraid_cmd *cmd = NULL; + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == HIRAID_CMD_PTHRU) { + head = &hdev->io_pt_list; + slock = &hdev->io_pt_lock; + } + + spin_lock_irqsave(slock, flags); + if (list_empty(head)) { + spin_unlock_irqrestore(slock, flags); + dev_err(hdev->dev, "err, cmd[%d] list empty\n", type); + return NULL; + } + cmd = list_entry(head->next, struct hiraid_cmd, list); + list_del_init(&cmd->list); + spin_unlock_irqrestore(slock, flags); + + WRITE_ONCE(cmd->state, CMD_FLIGHT); + + return cmd; +} + +static void hiraid_put_cmd(struct hiraid_dev *hdev, struct hiraid_cmd *cmd, + enum hiraid_cmd_type type) +{ + unsigned long flags; + struct list_head *head = &hdev->adm_cmd_list; + spinlock_t *slock = &hdev->adm_cmd_lock; + + if (type == HIRAID_CMD_PTHRU) { + head = &hdev->io_pt_list; + slock = &hdev->io_pt_lock; + } + + spin_lock_irqsave(slock, flags); + WRITE_ONCE(cmd->state, CMD_IDLE); + list_add_tail(&cmd->list, head); + spin_unlock_irqrestore(slock, flags); +} + +static bool hiraid_admin_need_reset(struct hiraid_admin_command *cmd) +{ + switch (cmd->common.opcode) { + case HIRAID_ADMIN_DELETE_SQ: + case HIRAID_ADMIN_CREATE_SQ: + case HIRAID_ADMIN_DELETE_CQ: + case HIRAID_ADMIN_CREATE_CQ: + case HIRAID_ADMIN_SET_FEATURES: + return false; + default: + return true; + } +} + +static int hiraid_reset_work_sync(struct hiraid_dev *hdev); +static inline void hiraid_admin_timeout(struct hiraid_dev *hdev, struct hiraid_cmd *cmd) +{ + /* command may be returned because controller reset */ + if (READ_ONCE(cmd->state) == CMD_COMPLETE) + return; + if (hiraid_reset_work_sync(hdev) == -EBUSY) + flush_work(&hdev->reset_work); +} + +static int hiraid_put_admin_sync_request(struct hiraid_dev *hdev, struct hiraid_admin_command *cmd, + u32 *result0, u32 *result1, u32 timeout) +{ + struct hiraid_cmd *adm_cmd = hiraid_get_cmd(hdev, HIRAID_CMD_ADMIN); + + if (!adm_cmd) { + dev_err(hdev->dev, "err, get admin cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&adm_cmd->cmd_done); + + cmd->common.cmd_id = cpu_to_le16(adm_cmd->cid); + hiraid_submit_cmd(&hdev->queues[0], cmd); + + if (!wait_for_completion_timeout(&adm_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + adm_cmd->cid, adm_cmd->qid, cmd->usr_cmd.opcode, + cmd->usr_cmd.info_0.subopcode); + + /* reset controller if admin timeout */ + if (hiraid_admin_need_reset(cmd)) + hiraid_admin_timeout(hdev, adm_cmd); + + hiraid_put_cmd(hdev, adm_cmd, HIRAID_CMD_ADMIN); + return -ETIME; + } + + if (result0) + *result0 = adm_cmd->result0; + if (result1) + *result1 = adm_cmd->result1; + + hiraid_put_cmd(hdev, adm_cmd, HIRAID_CMD_ADMIN); + + return adm_cmd->status; +} + +/** + * hiraid_create_cq - send cmd to controller for create controller cq + */ +static int hiraid_create_complete_queue(struct hiraid_dev *hdev, u16 qid, + struct hiraid_queue *hiraidq, u16 cq_vector) +{ + struct hiraid_admin_command admin_cmd; + int flags = HIRAID_QUEUE_PHYS_CONTIG | HIRAID_CQ_IRQ_ENABLED; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_cq.opcode = HIRAID_ADMIN_CREATE_CQ; + admin_cmd.create_cq.prp1 = cpu_to_le64(hiraidq->cq_buffer_phy); + admin_cmd.create_cq.cqid = cpu_to_le16(qid); + admin_cmd.create_cq.qsize = cpu_to_le16(hiraidq->q_depth - 1); + admin_cmd.create_cq.cq_flags = cpu_to_le16(flags); + admin_cmd.create_cq.irq_vector = cpu_to_le16(cq_vector); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +/** + * hiraid_create_sq - send cmd to controller for create controller sq + */ +static int hiraid_create_send_queue(struct hiraid_dev *hdev, u16 qid, + struct hiraid_queue *hiraidq) +{ + struct hiraid_admin_command admin_cmd; + int flags = HIRAID_QUEUE_PHYS_CONTIG; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.create_sq.opcode = HIRAID_ADMIN_CREATE_SQ; + admin_cmd.create_sq.prp1 = cpu_to_le64(hiraidq->sq_buffer_phy); + admin_cmd.create_sq.sqid = cpu_to_le16(qid); + admin_cmd.create_sq.qsize = cpu_to_le16(hiraidq->q_depth - 1); + admin_cmd.create_sq.sq_flags = cpu_to_le16(flags); + admin_cmd.create_sq.cqid = cpu_to_le16(qid); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +static void hiraid_free_all_queues(struct hiraid_dev *hdev) +{ + int i; + struct hiraid_queue *hq; + + for (i = 0; i < hdev->queue_count; i++) { + hq = &hdev->queues[i]; + dma_free_coherent(hdev->dev, CQ_SIZE(hq->q_depth), + (void *)hq->cqes, hq->cq_buffer_phy); + dma_free_coherent(hdev->dev, SQ_SIZE(hq->qid, hq->q_depth), + hq->sq_cmds, hq->sq_buffer_phy); + if (!work_mode) + dma_free_coherent(hdev->dev, SENSE_SIZE(hq->q_depth), + hq->sense_buffer_virt, hq->sense_buffer_phy); + } + + hdev->queue_count = 0; +} + +static void hiraid_free_sense_buffer(struct hiraid_dev *hdev) +{ + if (hdev->sense_buffer_virt) { + dma_free_coherent(hdev->dev, + SENSE_SIZE(hdev->scsi_qd + max_hwq_num * HIRAID_PTHRU_CMDS_PERQ), + hdev->sense_buffer_virt, hdev->sense_buffer_phy); + hdev->sense_buffer_virt = NULL; + } +} + +static int hiraid_delete_queue(struct hiraid_dev *hdev, u8 opcode, u16 qid) +{ + struct hiraid_admin_command admin_cmd; + int ret; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.delete_queue.opcode = opcode; + admin_cmd.delete_queue.qid = cpu_to_le16(qid); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) + dev_err(hdev->dev, "delete %s:[%d] failed\n", + (opcode == HIRAID_ADMIN_DELETE_CQ) ? "cq" : "sq", qid); + + return ret; +} + +static int hiraid_delete_complete_queue(struct hiraid_dev *hdev, u16 cqid) +{ + return hiraid_delete_queue(hdev, HIRAID_ADMIN_DELETE_CQ, cqid); +} + +static int hiraid_delete_send_queue(struct hiraid_dev *hdev, u16 sqid) +{ + return hiraid_delete_queue(hdev, HIRAID_ADMIN_DELETE_SQ, sqid); +} + +static int hiraid_create_queue(struct hiraid_queue *hiraidq, u16 qid) +{ + struct hiraid_dev *hdev = hiraidq->hdev; + u16 cq_vector; + int ret; + + cq_vector = (hdev->num_vecs == 1) ? 0 : qid; + ret = hiraid_create_complete_queue(hdev, qid, hiraidq, cq_vector); + if (ret) + return ret; + + ret = hiraid_create_send_queue(hdev, qid, hiraidq); + if (ret) + goto delete_cq; + + hiraidq->cq_vector = cq_vector; + ret = pci_request_irq(hdev->pdev, cq_vector, hiraid_handle_irq, NULL, + hiraidq, "hiraid%d_q%d", hdev->instance, qid); + if (ret) { + hiraidq->cq_vector = -1; + dev_err(hdev->dev, "request queue[%d] irq failed\n", qid); + goto delete_sq; + } + + hiraid_init_queue(hiraidq, qid); + + return 0; + +delete_sq: + hiraid_delete_send_queue(hdev, qid); +delete_cq: + hiraid_delete_complete_queue(hdev, qid); + + return ret; +} + +static int hiraid_create_io_queues(struct hiraid_dev *hdev) +{ + u32 i, max; + int ret = 0; + + max = min(hdev->max_qid, hdev->queue_count - 1); + for (i = hdev->online_queues; i <= max; i++) { + ret = hiraid_create_queue(&hdev->queues[i], i); + if (ret) { + dev_err(hdev->dev, "create queue[%d] failed\n", i); + break; + } + } + + if (!hdev->last_qcnt) + hdev->last_qcnt = hdev->online_queues; + + dev_info(hdev->dev, "queue_count[%d] online_queue[%d] last_online[%d]", + hdev->queue_count, hdev->online_queues, hdev->last_qcnt); + + return ret >= 0 ? 0 : ret; +} + +static int hiraid_set_features(struct hiraid_dev *hdev, u32 fid, u32 dword11, void *buffer, + size_t buflen, u32 *result) +{ + struct hiraid_admin_command admin_cmd; + int ret; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + + if (buffer && buflen) { + data_ptr = dma_alloc_coherent(hdev->dev, buflen, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memcpy(data_ptr, buffer, buflen); + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.features.opcode = HIRAID_ADMIN_SET_FEATURES; + admin_cmd.features.fid = cpu_to_le32(fid); + admin_cmd.features.dword11 = cpu_to_le32(dword11); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, result, NULL, 0); + + if (data_ptr) + dma_free_coherent(hdev->dev, buflen, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_configure_timestamp(struct hiraid_dev *hdev) +{ + __le64 timestamp; + int ret; + + timestamp = cpu_to_le64(ktime_to_ms(ktime_get_real())); + ret = hiraid_set_features(hdev, HIRAID_FEATURE_TIMESTAMP, 0, + &timestamp, sizeof(timestamp), NULL); + + if (ret) + dev_err(hdev->dev, "set timestamp failed[%d]\n", ret); + return ret; +} + +static int hiraid_get_queue_cnt(struct hiraid_dev *hdev, u32 *cnt) +{ + u32 q_cnt = (*cnt - 1) | ((*cnt - 1) << 16); + u32 nr_ioqs, result; + int status; + + status = hiraid_set_features(hdev, HIRAID_FEATURE_NUM_QUEUES, q_cnt, NULL, 0, &result); + if (status) { + dev_err(hdev->dev, "set queue count failed, status[%d]\n", + status); + return -EIO; + } + + nr_ioqs = min(result & 0xffff, result >> 16) + 1; + *cnt = min(*cnt, nr_ioqs); + if (*cnt == 0) { + dev_err(hdev->dev, "illegal qcount: zero, nr_ioqs[%d], cnt[%d]\n", nr_ioqs, *cnt); + return -EIO; + } + return 0; +} + +static int hiraid_setup_io_queues(struct hiraid_dev *hdev) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + struct pci_dev *pdev = hdev->pdev; + u32 i, size, nr_ioqs; + int ret; + + struct irq_affinity affd = { + .pre_vectors = 1 + }; + + /* alloc IO sense buffer for single hw queue mode */ + if (work_mode && !hdev->sense_buffer_virt) { + hdev->sense_buffer_virt = dma_alloc_coherent(hdev->dev, + SENSE_SIZE(hdev->scsi_qd + max_hwq_num * HIRAID_PTHRU_CMDS_PERQ), + &hdev->sense_buffer_phy, GFP_KERNEL | __GFP_ZERO); + if (!hdev->sense_buffer_virt) + return -ENOMEM; + } + + nr_ioqs = min(num_online_cpus(), max_hwq_num); + ret = hiraid_get_queue_cnt(hdev, &nr_ioqs); + if (ret < 0) + return ret; + + size = hiraid_get_bar_size(hdev, nr_ioqs); + ret = hiraid_remap_bar(hdev, size); + if (ret) + return -ENOMEM; + + adminq->q_db = hdev->dbs; + + pci_free_irq(pdev, 0, adminq); + pci_free_irq_vectors(pdev); + hdev->online_queues--; + + ret = pci_alloc_irq_vectors_affinity(pdev, 1, (nr_ioqs + 1), + PCI_IRQ_ALL_TYPES | PCI_IRQ_AFFINITY, &affd); + if (ret <= 0) + return -EIO; + + hdev->num_vecs = ret; + hdev->max_qid = max(ret - 1, 1); + + ret = pci_request_irq(pdev, adminq->cq_vector, hiraid_handle_irq, NULL, + adminq, "hiraid%d_q%d", hdev->instance, adminq->qid); + if (ret) { + dev_err(hdev->dev, "request admin irq failed\n"); + adminq->cq_vector = -1; + return ret; + } + + hdev->online_queues++; + + for (i = hdev->queue_count; i <= hdev->max_qid; i++) { + ret = hiraid_alloc_queue(hdev, i, hdev->ioq_depth); + if (ret) + break; + } + dev_info(hdev->dev, "max_qid[%d] queuecount[%d] onlinequeue[%d] ioqdepth[%d]\n", + hdev->max_qid, hdev->queue_count, hdev->online_queues, hdev->ioq_depth); + + return hiraid_create_io_queues(hdev); +} + +static void hiraid_delete_io_queues(struct hiraid_dev *hdev) +{ + u16 queues = hdev->online_queues - 1; + u8 opcode = HIRAID_ADMIN_DELETE_SQ; + u16 i, pass; + + if (!pci_device_is_present(hdev->pdev)) { + dev_err(hdev->dev, "pci_device is not present, skip disable io queues\n"); + return; + } + + if (hdev->online_queues < 2) { + dev_err(hdev->dev, "err, io queue has been delete\n"); + return; + } + + for (pass = 0; pass < 2; pass++) { + for (i = queues; i > 0; i--) + if (hiraid_delete_queue(hdev, opcode, i)) + break; + + opcode = HIRAID_ADMIN_DELETE_CQ; + } +} + +static void hiraid_pci_disable(struct hiraid_dev *hdev) +{ + struct pci_dev *pdev = hdev->pdev; + u32 i; + + for (i = 0; i < hdev->online_queues; i++) + pci_free_irq(pdev, hdev->queues[i].cq_vector, &hdev->queues[i]); + pci_free_irq_vectors(pdev); + if (pci_is_enabled(pdev)) { + pci_disable_pcie_error_reporting(pdev); + pci_disable_device(pdev); + } + hdev->online_queues = 0; +} + +static void hiraid_disable_admin_queue(struct hiraid_dev *hdev, bool shutdown) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + hiraid_shutdown_control(hdev); + else + hiraid_disable_control(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "err, admin queue has been delete\n"); + return; + } + + spin_lock_irq(&adminq->cq_lock); + hiraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + hiraid_complete_cqes(adminq, start, end); +} + +static int hiraid_create_prp_pools(struct hiraid_dev *hdev) +{ + int i; + char poolname[20] = { 0 }; + + hdev->prp_page_pool = dma_pool_create("prp list page", hdev->dev, + PAGE_SIZE, PAGE_SIZE, 0); + + if (!hdev->prp_page_pool) { + dev_err(hdev->dev, "create prp_page_pool failed\n"); + return -ENOMEM; + } + + for (i = 0; i < extra_pool_num; i++) { + sprintf(poolname, "prp_list_256_%d", i); + hdev->prp_extra_pool[i] = dma_pool_create(poolname, hdev->dev, EXTRA_POOL_SIZE, + EXTRA_POOL_SIZE, 0); + + if (!hdev->prp_extra_pool[i]) { + dev_err(hdev->dev, "create prp extra pool[%d] failed\n", i); + goto destroy_prp_extra_pool; + } + } + + return 0; + +destroy_prp_extra_pool: + while (i > 0) + dma_pool_destroy(hdev->prp_extra_pool[--i]); + dma_pool_destroy(hdev->prp_page_pool); + + return -ENOMEM; +} + +static void hiraid_free_prp_pools(struct hiraid_dev *hdev) +{ + int i; + + for (i = 0; i < extra_pool_num; i++) + dma_pool_destroy(hdev->prp_extra_pool[i]); + dma_pool_destroy(hdev->prp_page_pool); +} + +static int hiraid_request_devices(struct hiraid_dev *hdev, struct hiraid_dev_info *dev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + struct hiraid_admin_command admin_cmd; + struct hiraid_dev_list *list_buf; + dma_addr_t buffer_phy = 0; + u32 i, idx, hdid, ndev; + int ret = 0; + + list_buf = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!list_buf) + return -ENOMEM; + + for (idx = 0; idx < nd;) { + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = HIRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = HIRAID_GET_DEVLIST_INFO; + admin_cmd.get_info.cdw11 = cpu_to_le32(idx); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + + if (ret) { + dev_err(hdev->dev, "get device list failed, nd[%u] idx[%u] ret[%d]\n", + nd, idx, ret); + goto out; + } + ndev = le32_to_cpu(list_buf->dev_num); + + dev_info(hdev->dev, "get dev list ndev num[%u]\n", ndev); + + for (i = 0; i < ndev; i++) { + hdid = le32_to_cpu(list_buf->devinfo[i].hdid); + dev_info(hdev->dev, "devices[%d], hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + i, hdid, le16_to_cpu(list_buf->devinfo[i].target), + list_buf->devinfo[i].channel, + list_buf->devinfo[i].lun, + list_buf->devinfo[i].attr); + if (hdid > nd || hdid == 0) { + dev_err(hdev->dev, "err, hdid[%d] invalid\n", hdid); + continue; + } + memcpy(&dev[hdid - 1], &list_buf->devinfo[i], + sizeof(struct hiraid_dev_info)); + } + idx += ndev; + + if (ndev < MAX_DEV_ENTRY_PER_PAGE_4K) + break; + } + +out: + dma_free_coherent(hdev->dev, PAGE_SIZE, list_buf, buffer_phy); + return ret; +} + +static void hiraid_send_async_event(struct hiraid_dev *hdev, u16 cid) +{ + struct hiraid_queue *adminq = &hdev->queues[0]; + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = HIRAID_ADMIN_ASYNC_EVENT; + admin_cmd.common.cmd_id = cpu_to_le16(cid); + + hiraid_submit_cmd(adminq, &admin_cmd); + dev_info(hdev->dev, "send async event to controller, cid[%d]\n", cid); +} + +static inline void hiraid_init_async_event(struct hiraid_dev *hdev) +{ + u16 i; + + for (i = 0; i < hdev->ctrl_info->asynevent; i++) + hiraid_send_async_event(hdev, i + HIRAID_AQ_BLK_MQ_DEPTH); +} + +static int hiraid_add_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "add device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (sdev) { + dev_warn(hdev->dev, "device is already exist, channel[%d] targetid[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + scsi_device_put(sdev); + return -EEXIST; + } + scsi_add_device(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + return 0; +} + +static int hiraid_rescan_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "rescan device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit rescan it, channel[%d] target_id[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + return -ENODEV; + } + + scsi_rescan_device(&sdev->sdev_gendev); + scsi_device_put(sdev); + return 0; +} + +static int hiraid_delete_device(struct hiraid_dev *hdev, struct hiraid_dev_info *devinfo) +{ + struct Scsi_Host *shost = hdev->shost; + struct scsi_device *sdev; + + dev_info(hdev->dev, "remove device, hdid[%u] target[%d] channel[%d] lun[%d] attr[0x%x]\n", + le32_to_cpu(devinfo->hdid), le16_to_cpu(devinfo->target), + devinfo->channel, devinfo->lun, devinfo->attr); + + sdev = scsi_device_lookup(shost, devinfo->channel, le16_to_cpu(devinfo->target), 0); + if (!sdev) { + dev_warn(hdev->dev, "device is not exit remove it, channel[%d] target_id[%d] lun[%d]\n", + devinfo->channel, le16_to_cpu(devinfo->target), 0); + return -ENODEV; + } + + scsi_remove_device(sdev); + scsi_device_put(sdev); + return 0; +} + +static int hiraid_dev_list_init(struct hiraid_dev *hdev) +{ + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + + hdev->dev_info = kzalloc_node(nd * sizeof(struct hiraid_dev_info), + GFP_KERNEL, hdev->numa_node); + if (!hdev->dev_info) + return -ENOMEM; + + return 0; +} + +static int hiraid_luntarget_sort(const void *l, const void *r) +{ + const struct hiraid_dev_info *ln = l; + const struct hiraid_dev_info *rn = r; + int l_attr = HIRAID_DEV_INFO_ATTR_BOOT(ln->attr); + int r_attr = HIRAID_DEV_INFO_ATTR_BOOT(rn->attr); + + /* boot first */ + if (l_attr != r_attr) + return (r_attr - l_attr); + + if (ln->channel == rn->channel) + return le16_to_cpu(ln->target) - le16_to_cpu(rn->target); + + return ln->channel - rn->channel; +} + +static void hiraid_scan_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = + container_of(work, struct hiraid_dev, scan_work); + struct hiraid_dev_info *dev, *old_dev, *new_dev; + u32 nd = le32_to_cpu(hdev->ctrl_info->nd); + u8 flag, org_flag; + int i, ret; + int count = 0; + + dev = kcalloc(nd, sizeof(struct hiraid_dev_info), GFP_KERNEL); + if (!dev) + return; + + new_dev = kcalloc(nd, sizeof(struct hiraid_dev_info), GFP_KERNEL); + if (!new_dev) + goto free_list; + + ret = hiraid_request_devices(hdev, dev); + if (ret) + goto free_all; + old_dev = hdev->dev_info; + for (i = 0; i < nd; i++) { + org_flag = old_dev[i].flag; + flag = dev[i].flag; + + dev_log_dbg(hdev->dev, "i[%d] org_flag[0x%x] flag[0x%x]\n", i, org_flag, flag); + + if (HIRAID_DEV_INFO_FLAG_VALID(flag)) { + if (!HIRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->dev_rwsem); + memcpy(&old_dev[i], &dev[i], + sizeof(struct hiraid_dev_info)); + memcpy(&new_dev[count++], &dev[i], + sizeof(struct hiraid_dev_info)); + up_write(&hdev->dev_rwsem); + } else if (HIRAID_DEV_INFO_FLAG_CHANGE(flag)) { + hiraid_rescan_device(hdev, &dev[i]); + } + } else { + if (HIRAID_DEV_INFO_FLAG_VALID(org_flag)) { + down_write(&hdev->dev_rwsem); + old_dev[i].flag &= 0xfe; + up_write(&hdev->dev_rwsem); + hiraid_delete_device(hdev, &old_dev[i]); + } + } + } + + dev_info(hdev->dev, "scan work add device num[%d]\n", count); + + sort(new_dev, count, sizeof(new_dev[0]), hiraid_luntarget_sort, NULL); + + for (i = 0; i < count; i++) + hiraid_add_device(hdev, &new_dev[i]); + +free_all: + kfree(new_dev); +free_list: + kfree(dev); +} + +static void hiraid_timesyn_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = + container_of(work, struct hiraid_dev, timesyn_work); + + hiraid_configure_timestamp(hdev); +} + +static int hiraid_init_control_info(struct hiraid_dev *hdev); +static void hiraid_fwactive_work(struct work_struct *work) +{ + struct hiraid_dev *hdev = container_of(work, struct hiraid_dev, fwact_work); + + if (hiraid_init_control_info(hdev)) + dev_err(hdev->dev, "get controller info failed after fw activation\n"); +} + +static void hiraid_queue_scan(struct hiraid_dev *hdev) +{ + queue_work(work_queue, &hdev->scan_work); +} + +static void hiraid_handle_async_notice(struct hiraid_dev *hdev, u32 result) +{ + switch ((result & 0xff00) >> 8) { + case HIRAID_ASYN_DEV_CHANGED: + hiraid_queue_scan(hdev); + break; + case HIRAID_ASYN_FW_ACT_START: + dev_info(hdev->dev, "fw activation starting\n"); + break; + case HIRAID_ASYN_HOST_PROBING: + break; + default: + dev_warn(hdev->dev, "async event result[%08x]\n", result); + } +} + +static void hiraid_handle_async_vs(struct hiraid_dev *hdev, u32 result, u32 result1) +{ + switch ((result & 0xff00) >> 8) { + case HIRAID_ASYN_TIMESYN: + queue_work(work_queue, &hdev->timesyn_work); + break; + case HIRAID_ASYN_FW_ACT_FINISH: + dev_info(hdev->dev, "fw activation finish\n"); + queue_work(work_queue, &hdev->fwact_work); + break; + case HIRAID_ASYN_EVENT_MIN ... HIRAID_ASYN_EVENT_MAX: + dev_info(hdev->dev, "recv card event[%d] param1[0x%x] param2[0x%x]\n", + (result & 0xff00) >> 8, result, result1); + break; + default: + dev_warn(hdev->dev, "async event result[0x%x]\n", result); + } +} + +static int hiraid_alloc_resources(struct hiraid_dev *hdev) +{ + int ret, nqueue; + + hdev->ctrl_info = kzalloc_node(sizeof(*hdev->ctrl_info), GFP_KERNEL, hdev->numa_node); + if (!hdev->ctrl_info) + return -ENOMEM; + + ret = hiraid_create_prp_pools(hdev); + if (ret) + goto free_ctrl_info; + nqueue = min(num_possible_cpus(), max_hwq_num) + 1; + hdev->queues = kcalloc_node(nqueue, sizeof(struct hiraid_queue), + GFP_KERNEL, hdev->numa_node); + if (!hdev->queues) { + ret = -ENOMEM; + goto destroy_dma_pools; + } + + ret = hiraid_create_admin_cmds(hdev); + if (ret) + goto free_queues; + + dev_info(hdev->dev, "total queues num[%d]\n", nqueue); + + return 0; + +free_queues: + kfree(hdev->queues); +destroy_dma_pools: + hiraid_free_prp_pools(hdev); +free_ctrl_info: + kfree(hdev->ctrl_info); + + return ret; +} + +static void hiraid_free_resources(struct hiraid_dev *hdev) +{ + hiraid_free_admin_cmds(hdev); + kfree(hdev->queues); + hiraid_free_prp_pools(hdev); + kfree(hdev->ctrl_info); +} + +static void hiraid_bsg_buf_unmap(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_mapmange *mapbuf = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + + if (mapbuf->sge_cnt) + dma_unmap_sg(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir); + + hiraid_free_mapbuf(hdev, mapbuf); +} + +static int hiraid_bsg_buf_map(struct hiraid_dev *hdev, struct bsg_job *job, + struct hiraid_admin_command *cmd) +{ + struct hiraid_bsg_request *bsg_req = job->request; + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_mapmange *mapbuf = job->dd_data; + enum dma_data_direction dma_dir = rq_data_dir(rq) ? DMA_TO_DEVICE : DMA_FROM_DEVICE; + int ret = 0; + + /* No data to DMA, it may be scsi no-rw command */ + mapbuf->sge_cnt = job->request_payload.sg_cnt; + mapbuf->sgl = job->request_payload.sg_list; + mapbuf->len = job->request_payload.payload_len; + mapbuf->page_cnt = -1; + if (unlikely(mapbuf->sge_cnt == 0)) + goto out; + + mapbuf->use_sgl = !hiraid_is_prp(hdev, mapbuf->sgl, mapbuf->sge_cnt); + + ret = dma_map_sg_attrs(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir, DMA_ATTR_NO_WARN); + if (!ret) + goto out; + + if ((mapbuf->use_sgl == (bool)true) && (bsg_req->msgcode == HIRAID_BSG_IOPTHRU) && + (hdev->ctrl_info->pt_use_sgl != (bool)false)) { + ret = hiraid_build_passthru_sgl(hdev, cmd, mapbuf); + } else { + mapbuf->use_sgl = false; + + ret = hiraid_build_passthru_prp(hdev, mapbuf); + cmd->common.dptr.prp1 = cpu_to_le64(sg_dma_address(mapbuf->sgl)); + cmd->common.dptr.prp2 = cpu_to_le64(mapbuf->first_dma); + } + + if (ret) + goto unmap; + + return 0; + +unmap: + dma_unmap_sg(hdev->dev, mapbuf->sgl, mapbuf->sge_cnt, dma_dir); +out: + return ret; +} + +static int hiraid_get_control_info(struct hiraid_dev *hdev, struct hiraid_ctrl_info *ctrl_info) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.get_info.opcode = HIRAID_ADMIN_GET_INFO; + admin_cmd.get_info.type = HIRAID_GET_CTRL_INFO; + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); + if (!ret) + memcpy(ctrl_info, data_ptr, sizeof(struct hiraid_ctrl_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_init_control_info(struct hiraid_dev *hdev) +{ + int ret; + + hdev->ctrl_info->nd = cpu_to_le32(240); + hdev->ctrl_info->mdts = 8; + hdev->ctrl_info->max_cmds = cpu_to_le16(4096); + hdev->ctrl_info->max_num_sge = cpu_to_le16(128); + hdev->ctrl_info->max_channel = cpu_to_le16(4); + hdev->ctrl_info->max_tgt_id = cpu_to_le32(3239); + hdev->ctrl_info->max_lun = cpu_to_le16(2); + + ret = hiraid_get_control_info(hdev, hdev->ctrl_info); + if (ret) + dev_err(hdev->dev, "get controller info failed[%d]\n", ret); + + dev_info(hdev->dev, "device_num = %d\n", hdev->ctrl_info->nd); + dev_info(hdev->dev, "max_cmd = %d\n", hdev->ctrl_info->max_cmds); + dev_info(hdev->dev, "max_channel = %d\n", hdev->ctrl_info->max_channel); + dev_info(hdev->dev, "max_tgt_id = %d\n", hdev->ctrl_info->max_tgt_id); + dev_info(hdev->dev, "max_lun = %d\n", hdev->ctrl_info->max_lun); + dev_info(hdev->dev, "max_num_sge = %d\n", hdev->ctrl_info->max_num_sge); + dev_info(hdev->dev, "lun_num_boot = %d\n", hdev->ctrl_info->lun_num_boot); + dev_info(hdev->dev, "max_data_transfer_size = %d\n", hdev->ctrl_info->mdts); + dev_info(hdev->dev, "abort_cmd_limit = %d\n", hdev->ctrl_info->acl); + dev_info(hdev->dev, "asyn_event_num = %d\n", hdev->ctrl_info->asynevent); + dev_info(hdev->dev, "card_type = %d\n", hdev->ctrl_info->card_type); + dev_info(hdev->dev, "pt_use_sgl = %d\n", hdev->ctrl_info->pt_use_sgl); + dev_info(hdev->dev, "rtd3e = %d\n", hdev->ctrl_info->rtd3e); + dev_info(hdev->dev, "serial_num = %s\n", hdev->ctrl_info->sn); + dev_info(hdev->dev, "fw_verion = %s\n", hdev->ctrl_info->fw_version); + + if (!hdev->ctrl_info->asynevent) + hdev->ctrl_info->asynevent = 1; + if (hdev->ctrl_info->asynevent > HIRAID_ASYN_COMMANDS) + hdev->ctrl_info->asynevent = HIRAID_ASYN_COMMANDS; + + hdev->scsi_qd = work_mode ? + le16_to_cpu(hdev->ctrl_info->max_cmds) : (hdev->ioq_depth - HIRAID_PTHRU_CMDS_PERQ); + + return 0; +} + +static int hiraid_user_send_admcmd(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct hiraid_bsg_request *bsg_req = job->request; + struct hiraid_passthru_common_cmd *ptcmd = &(bsg_req->admcmd); + struct hiraid_admin_command admin_cmd; + u32 timeout = msecs_to_jiffies(ptcmd->timeout_ms); + u32 result[2] = {0}; + int status; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", + hdev->state); + return -EBUSY; + } + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.common.opcode = ptcmd->opcode; + admin_cmd.common.flags = ptcmd->flags; + admin_cmd.common.hdid = cpu_to_le32(ptcmd->nsid); + admin_cmd.common.cdw2[0] = cpu_to_le32(ptcmd->cdw2); + admin_cmd.common.cdw2[1] = cpu_to_le32(ptcmd->cdw3); + admin_cmd.common.cdw10 = cpu_to_le32(ptcmd->cdw10); + admin_cmd.common.cdw11 = cpu_to_le32(ptcmd->cdw11); + admin_cmd.common.cdw12 = cpu_to_le32(ptcmd->cdw12); + admin_cmd.common.cdw13 = cpu_to_le32(ptcmd->cdw13); + admin_cmd.common.cdw14 = cpu_to_le32(ptcmd->cdw14); + admin_cmd.common.cdw15 = cpu_to_le32(ptcmd->cdw15); + + status = hiraid_bsg_buf_map(hdev, job, &admin_cmd); + if (status) { + dev_err(hdev->dev, "err, map data failed\n"); + return status; + } + + status = hiraid_put_admin_sync_request(hdev, &admin_cmd, &result[0], &result[1], timeout); + if (status >= 0) { + job->reply_len = sizeof(result); + memcpy(job->reply, result, sizeof(result)); + } + if (status) + dev_info(hdev->dev, "opcode[0x%x] subopcode[0x%x] status[0x%x] result0[0x%x];" + "result1[0x%x]\n", ptcmd->opcode, ptcmd->info_0.subopcode, status, + result[0], result[1]); + + hiraid_bsg_buf_unmap(hdev, job); + + return status; +} + +static int hiraid_alloc_io_ptcmds(struct hiraid_dev *hdev) +{ + u32 i; + u32 ptnum = HIRAID_TOTAL_PTCMDS(hdev->online_queues - 1); + + INIT_LIST_HEAD(&hdev->io_pt_list); + spin_lock_init(&hdev->io_pt_lock); + + hdev->io_ptcmds = kcalloc_node(ptnum, sizeof(struct hiraid_cmd), + GFP_KERNEL, hdev->numa_node); + + if (!hdev->io_ptcmds) { + dev_err(hdev->dev, "alloc io pthrunum failed\n"); + return -ENOMEM; + } + + for (i = 0; i < ptnum; i++) { + hdev->io_ptcmds[i].qid = i / HIRAID_PTHRU_CMDS_PERQ + 1; + hdev->io_ptcmds[i].cid = i % HIRAID_PTHRU_CMDS_PERQ + hdev->scsi_qd; + list_add_tail(&(hdev->io_ptcmds[i].list), &hdev->io_pt_list); + } + + dev_info(hdev->dev, "alloc io pthru cmd success, pthrunum[%d]\n", ptnum); + + return 0; +} + +static void hiraid_free_io_ptcmds(struct hiraid_dev *hdev) +{ + kfree(hdev->io_ptcmds); + hdev->io_ptcmds = NULL; + + INIT_LIST_HEAD(&hdev->io_pt_list); +} + +static int hiraid_put_io_sync_request(struct hiraid_dev *hdev, struct hiraid_scsi_io_cmd *io_cmd, + u32 *result, u32 *reslen, u32 timeout) +{ + int ret; + dma_addr_t buffer_phy; + struct hiraid_queue *ioq; + void *sense_addr = NULL; + struct hiraid_cmd *pt_cmd = hiraid_get_cmd(hdev, HIRAID_CMD_PTHRU); + + if (!pt_cmd) { + dev_err(hdev->dev, "err, get ioq cmd failed\n"); + return -EFAULT; + } + + timeout = timeout ? timeout : ADMIN_TIMEOUT; + + init_completion(&pt_cmd->cmd_done); + + ioq = &hdev->queues[pt_cmd->qid]; + if (work_mode) { + ret = ((pt_cmd->qid - 1) * HIRAID_PTHRU_CMDS_PERQ + pt_cmd->cid) * + SCSI_SENSE_BUFFERSIZE; + sense_addr = hdev->sense_buffer_virt + ret; + buffer_phy = hdev->sense_buffer_phy + ret; + } else { + ret = pt_cmd->cid * SCSI_SENSE_BUFFERSIZE; + sense_addr = ioq->sense_buffer_virt + ret; + buffer_phy = ioq->sense_buffer_phy + ret; + } + + io_cmd->common.sense_addr = cpu_to_le64(buffer_phy); + io_cmd->common.sense_len = cpu_to_le16(SCSI_SENSE_BUFFERSIZE); + io_cmd->common.cmd_id = cpu_to_le16(pt_cmd->cid); + + hiraid_submit_cmd(ioq, io_cmd); + + if (!wait_for_completion_timeout(&pt_cmd->cmd_done, timeout)) { + dev_err(hdev->dev, "cid[%d] qid[%d] timeout, opcode[0x%x] subopcode[0x%x]\n", + pt_cmd->cid, pt_cmd->qid, io_cmd->common.opcode, + (le32_to_cpu(io_cmd->common.cdw3[0]) & 0xffff)); + + hiraid_admin_timeout(hdev, pt_cmd); + + hiraid_put_cmd(hdev, pt_cmd, HIRAID_CMD_PTHRU); + return -ETIME; + } + + if (result && reslen) { + if ((pt_cmd->status & 0x17f) == 0x101) { + memcpy(result, sense_addr, SCSI_SENSE_BUFFERSIZE); + *reslen = SCSI_SENSE_BUFFERSIZE; + } + } + + hiraid_put_cmd(hdev, pt_cmd, HIRAID_CMD_PTHRU); + + return pt_cmd->status; +} + +static int hiraid_user_send_ptcmd(struct hiraid_dev *hdev, struct bsg_job *job) +{ + struct hiraid_bsg_request *bsg_req = (struct hiraid_bsg_request *)(job->request); + struct hiraid_passthru_io_cmd *cmd = &(bsg_req->pthrucmd); + struct hiraid_scsi_io_cmd pthru_cmd; + int status = 0; + u32 timeout = msecs_to_jiffies(cmd->timeout_ms); + // data len is 4k before use sgl, now len is 1M + u32 io_pt_data_len = (hdev->ctrl_info->pt_use_sgl == (bool)true) ? + IOQ_PT_SGL_DATA_LEN : IOQ_PT_DATA_LEN; + + if (cmd->data_len > io_pt_data_len) { + dev_err(hdev->dev, "data len bigger than %d\n", io_pt_data_len); + return -EFAULT; + } + + if (hdev->state != DEV_LIVE) { + dev_err(hdev->dev, "err, host state[%d] is not live\n", hdev->state); + return -EBUSY; + } + + memset(&pthru_cmd, 0, sizeof(pthru_cmd)); + pthru_cmd.common.opcode = cmd->opcode; + pthru_cmd.common.flags = cmd->flags; + pthru_cmd.common.hdid = cpu_to_le32(cmd->nsid); + pthru_cmd.common.sense_len = cpu_to_le16(cmd->info_0.res_sense_len); + pthru_cmd.common.cdb_len = cmd->info_0.cdb_len; + pthru_cmd.common.rsvd2 = cmd->info_0.rsvd0; + pthru_cmd.common.cdw3[0] = cpu_to_le32(cmd->cdw3); + pthru_cmd.common.cdw3[1] = cpu_to_le32(cmd->cdw4); + pthru_cmd.common.cdw3[2] = cpu_to_le32(cmd->cdw5); + + pthru_cmd.common.cdw10[0] = cpu_to_le32(cmd->cdw10); + pthru_cmd.common.cdw10[1] = cpu_to_le32(cmd->cdw11); + pthru_cmd.common.cdw10[2] = cpu_to_le32(cmd->cdw12); + pthru_cmd.common.cdw10[3] = cpu_to_le32(cmd->cdw13); + pthru_cmd.common.cdw10[4] = cpu_to_le32(cmd->cdw14); + pthru_cmd.common.cdw10[5] = cpu_to_le32(cmd->data_len); + + memcpy(pthru_cmd.common.cdb, &cmd->cdw16, cmd->info_0.cdb_len); + + pthru_cmd.common.cdw26[0] = cpu_to_le32(cmd->cdw26[0]); + pthru_cmd.common.cdw26[1] = cpu_to_le32(cmd->cdw26[1]); + pthru_cmd.common.cdw26[2] = cpu_to_le32(cmd->cdw26[2]); + pthru_cmd.common.cdw26[3] = cpu_to_le32(cmd->cdw26[3]); + + status = hiraid_bsg_buf_map(hdev, job, (struct hiraid_admin_command *)&pthru_cmd); + if (status) { + dev_err(hdev->dev, "err, map data failed\n"); + return status; + } + + status = hiraid_put_io_sync_request(hdev, &pthru_cmd, job->reply, &job->reply_len, timeout); + + if (status) + dev_info(hdev->dev, "opcode[0x%x] subopcode[0x%x] status[0x%x] replylen[%d]\n", + cmd->opcode, cmd->info_1.subopcode, status, job->reply_len); + + hiraid_bsg_buf_unmap(hdev, job); + + return status; +} + +static bool hiraid_check_scmd_finished(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_queue *hiraidq; + + hiraidq = mapbuf->hiraidq; + if (!hiraidq) + return false; + if (READ_ONCE(mapbuf->state) == CMD_COMPLETE || hiraid_poll_cq(hiraidq, mapbuf->cid)) { + dev_warn(hdev->dev, "cid[%d] qid[%d] has been completed\n", + mapbuf->cid, hiraidq->qid); + return true; + } + return false; +} + +static enum blk_eh_timer_return hiraid_timed_out(struct scsi_cmnd *scmd) +{ + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + unsigned int timeout = scmd->device->request_queue->rq_timeout; + + if (hiraid_check_scmd_finished(scmd)) + goto out; + + if (time_after(jiffies, scmd->jiffies_at_alloc + timeout)) { + if (cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_TIMEOUT) == CMD_FLIGHT) + return BLK_EH_DONE; + } +out: + return BLK_EH_RESET_TIMER; +} + +/* send abort command by admin queue temporary */ +static int hiraid_send_abort_cmd(struct hiraid_dev *hdev, u32 hdid, u16 qid, u16 cid) +{ + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.abort.opcode = HIRAID_ADMIN_ABORT_CMD; + admin_cmd.abort.hdid = cpu_to_le32(hdid); + admin_cmd.abort.sqid = cpu_to_le16(qid); + admin_cmd.abort.cid = cpu_to_le16(cid); + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +/* send reset command by admin quueue temporary */ +static int hiraid_send_reset_cmd(struct hiraid_dev *hdev, u8 type, u32 hdid) +{ + struct hiraid_admin_command admin_cmd; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.reset.opcode = HIRAID_ADMIN_RESET; + admin_cmd.reset.hdid = cpu_to_le32(hdid); + admin_cmd.reset.type = type; + + return hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, 0); +} + +static bool hiraid_dev_state_trans(struct hiraid_dev *hdev, enum hiraid_dev_state new_state) +{ + unsigned long flags; + enum hiraid_dev_state old_state; + bool change = false; + + spin_lock_irqsave(&hdev->state_lock, flags); + + old_state = hdev->state; + switch (new_state) { + case DEV_LIVE: + switch (old_state) { + case DEV_NEW: + case DEV_RESETTING: + change = true; + break; + default: + break; + } + break; + case DEV_RESETTING: + switch (old_state) { + case DEV_LIVE: + change = true; + break; + default: + break; + } + break; + case DEV_DELETING: + if (old_state != DEV_DELETING) + change = true; + break; + case DEV_DEAD: + switch (old_state) { + case DEV_NEW: + case DEV_LIVE: + case DEV_RESETTING: + change = true; + break; + default: + break; + } + break; + default: + break; + } + if (change) + hdev->state = new_state; + spin_unlock_irqrestore(&hdev->state_lock, flags); + + dev_info(hdev->dev, "oldstate[%d]->newstate[%d], change[%d]\n", + old_state, new_state, change); + + return change; +} + +static void hiraid_drain_pending_ios(struct hiraid_dev *hdev); + +static void hiraid_flush_running_cmds(struct hiraid_dev *hdev) +{ + int i, j; + + scsi_block_requests(hdev->shost); + hiraid_drain_pending_ios(hdev); + scsi_unblock_requests(hdev->shost); + + j = HIRAID_AQ_BLK_MQ_DEPTH; + for (i = 0; i < j; i++) { + if (READ_ONCE(hdev->adm_cmds[i].state) == CMD_FLIGHT) { + dev_info(hdev->dev, "flush admin, cid[%d]\n", i); + hdev->adm_cmds[i].status = 0xFFFF; + WRITE_ONCE(hdev->adm_cmds[i].state, CMD_COMPLETE); + complete(&(hdev->adm_cmds[i].cmd_done)); + } + } + + j = HIRAID_TOTAL_PTCMDS(hdev->online_queues - 1); + for (i = 0; i < j; i++) { + if (READ_ONCE(hdev->io_ptcmds[i].state) == CMD_FLIGHT) { + hdev->io_ptcmds[i].status = 0xFFFF; + WRITE_ONCE(hdev->io_ptcmds[i].state, CMD_COMPLETE); + complete(&(hdev->io_ptcmds[i].cmd_done)); + } + } +} + +static int hiraid_dev_disable(struct hiraid_dev *hdev, bool shutdown) +{ + int ret = -ENODEV; + struct hiraid_queue *adminq = &hdev->queues[0]; + u16 start, end; + + if (pci_device_is_present(hdev->pdev)) { + if (shutdown) + hiraid_shutdown_control(hdev); + else + ret = hiraid_disable_control(hdev); + } + + if (hdev->queue_count == 0) { + dev_err(hdev->dev, "warn: queue has been delete\n"); + return ret; + } + + spin_lock_irq(&adminq->cq_lock); + hiraid_process_cq(adminq, &start, &end, -1); + spin_unlock_irq(&adminq->cq_lock); + hiraid_complete_cqes(adminq, start, end); + + hiraid_pci_disable(hdev); + + hiraid_flush_running_cmds(hdev); + + return ret; +} + +static void hiraid_reset_work(struct work_struct *work) +{ + int ret = 0; + struct hiraid_dev *hdev = container_of(work, struct hiraid_dev, reset_work); + + if (hdev->state != DEV_RESETTING) { + dev_err(hdev->dev, "err, host is not reset state\n"); + return; + } + + dev_info(hdev->dev, "enter host reset\n"); + + if (hdev->ctrl_config & HIRAID_CC_ENABLE) { + dev_info(hdev->dev, "start dev_disable\n"); + ret = hiraid_dev_disable(hdev, false); + } + + if (ret) + goto out; + + ret = hiraid_pci_enable(hdev); + if (ret) + goto out; + + ret = hiraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = hiraid_setup_io_queues(hdev); + if (ret || hdev->online_queues != hdev->last_qcnt) + goto pci_disable; + + hiraid_dev_state_trans(hdev, DEV_LIVE); + + hiraid_init_async_event(hdev); + + hiraid_queue_scan(hdev); + + return; + +pci_disable: + hiraid_pci_disable(hdev); +out: + hiraid_dev_state_trans(hdev, DEV_DEAD); + dev_err(hdev->dev, "err, host reset failed\n"); +} + +static int hiraid_reset_work_sync(struct hiraid_dev *hdev) +{ + if (!hiraid_dev_state_trans(hdev, DEV_RESETTING)) { + dev_info(hdev->dev, "can't change to reset state\n"); + return -EBUSY; + } + + if (!queue_work(work_queue, &hdev->reset_work)) { + dev_err(hdev->dev, "err, host is already in reset state\n"); + return -EBUSY; + } + + flush_work(&hdev->reset_work); + if (hdev->state != DEV_LIVE) + return -ENODEV; + + return 0; +} + +static int hiraid_wait_io_completion(struct hiraid_mapmange *mapbuf) +{ + u16 times = 0; + + do { + if (READ_ONCE(mapbuf->state) == CMD_TMO_COMPLETE) + break; + msleep(500); + times++; + } while (times <= HIRAID_WAIT_ABNL_CMD_TIMEOUT); + + /* wait command completion timeout after abort/reset success */ + if (times >= HIRAID_WAIT_ABNL_CMD_TIMEOUT) + return -ETIMEDOUT; + + return 0; +} + +static bool hiraid_tgt_rst_pending_io_count(struct request *rq, void *data, bool reserved) +{ + unsigned int id = *(unsigned int *)data; + struct scsi_cmnd *scmd = blk_mq_rq_to_pdu(rq); + struct hiraid_mapmange *mapbuf; + struct hiraid_sdev_hostdata *hostdata; + + if (scmd) { + mapbuf = scsi_cmd_priv(scmd); + if ((mapbuf->state == CMD_FLIGHT) || (mapbuf->state == CMD_TIMEOUT)) { + if ((scmd->device) && (scmd->device->id == id)) { + hostdata = scmd->device->hostdata; + hostdata->pend_count++; + } + } + } + return true; +} +static bool hiraid_clean_pending_io(struct request *rq, void *data, bool reserved) +{ + struct hiraid_dev *hdev = data; + struct scsi_cmnd *scmd; + struct hiraid_mapmange *mapbuf; + + if (unlikely(!rq || !blk_mq_request_started(rq))) + return true; + + scmd = blk_mq_rq_to_pdu(rq); + mapbuf = scsi_cmd_priv(scmd); + + if ((cmpxchg(&mapbuf->state, CMD_FLIGHT, CMD_COMPLETE) != CMD_FLIGHT) && + (cmpxchg(&mapbuf->state, CMD_TIMEOUT, CMD_COMPLETE) != CMD_TIMEOUT)) + return true; + + set_host_byte(scmd, DID_NO_CONNECT); + if (mapbuf->sge_cnt) + scsi_dma_unmap(scmd); + hiraid_free_mapbuf(hdev, mapbuf); + dev_warn_ratelimited(hdev->dev, "back unfinished CQE, cid[%d] qid[%d]\n", + mapbuf->cid, mapbuf->hiraidq->qid); + scmd->scsi_done(scmd); + + return true; +} + +static void hiraid_drain_pending_ios(struct hiraid_dev *hdev) +{ + blk_mq_tagset_busy_iter(&hdev->shost->tag_set, hiraid_clean_pending_io, (void *)(hdev)); +} + +static int wait_tgt_reset_io_done(struct scsi_cmnd *scmd) +{ + u16 timeout = 0; + struct hiraid_sdev_hostdata *hostdata; + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + + hostdata = scmd->device->hostdata; + + do { + hostdata->pend_count = 0; + blk_mq_tagset_busy_iter(&hdev->shost->tag_set, hiraid_tgt_rst_pending_io_count, + (void *)(&scmd->device->id)); + + if (!hostdata->pend_count) + return 0; + + msleep(500); + timeout++; + } while (timeout <= HIRAID_WAIT_RST_IO_TIMEOUT); + + return -ETIMEDOUT; +} + +static int hiraid_abort(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_mapmange *mapbuf = scsi_cmd_priv(scmd); + struct hiraid_sdev_hostdata *hostdata; + u16 hwq, cid; + int ret; + + scsi_print_command(scmd); + + if (hdev->state != DEV_LIVE || !hiraid_wait_io_completion(mapbuf) || + hiraid_check_scmd_finished(scmd)) + return SUCCESS; + + hostdata = scmd->device->hostdata; + cid = mapbuf->cid; + hwq = mapbuf->hiraidq->qid; + + dev_warn(hdev->dev, "cid[%d] qid[%d] timeout, send abort\n", cid, hwq); + ret = hiraid_send_abort_cmd(hdev, hostdata->hdid, hwq, cid); + if (ret != -ETIME) { + ret = hiraid_wait_io_completion(mapbuf); + if (ret) { + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, not found\n", cid, hwq); + return FAILED; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort succ\n", cid, hwq); + return SUCCESS; + } + dev_warn(hdev->dev, "cid[%d] qid[%d] abort failed, timeout\n", cid, hwq); + return FAILED; +} + +static int hiraid_scsi_reset(struct scsi_cmnd *scmd, enum hiraid_rst_type rst) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + struct hiraid_sdev_hostdata *hostdata; + int ret; + + if (hdev->state != DEV_LIVE) + return SUCCESS; + + hostdata = scmd->device->hostdata; + + dev_warn(hdev->dev, "sdev[%d:%d] send %s reset\n", scmd->device->channel, scmd->device->id, + rst ? "bus" : "target"); + ret = hiraid_send_reset_cmd(hdev, rst, hostdata->hdid); + if ((ret == 0) || (ret == FW_EH_DEV_NONE && rst == HIRAID_RESET_TARGET)) { + if (rst == HIRAID_RESET_TARGET) { + ret = wait_tgt_reset_io_done(scmd); + if (ret) { + dev_warn(hdev->dev, "sdev[%d:%d] target has %d peding cmd, target reset failed\n", + scmd->device->channel, scmd->device->id, + hostdata->pend_count); + return FAILED; + } + } + dev_warn(hdev->dev, "sdev[%d:%d] %s reset success\n", + scmd->device->channel, scmd->device->id, rst ? "bus" : "target"); + return SUCCESS; + } + + dev_warn(hdev->dev, "sdev[%d:%d] %s reset failed\n", + scmd->device->channel, scmd->device->id, rst ? "bus" : "target"); + return FAILED; +} + +static int hiraid_target_reset(struct scsi_cmnd *scmd) +{ + return hiraid_scsi_reset(scmd, HIRAID_RESET_TARGET); +} + +static int hiraid_bus_reset(struct scsi_cmnd *scmd) +{ + return hiraid_scsi_reset(scmd, HIRAID_RESET_BUS); +} + +static int hiraid_host_reset(struct scsi_cmnd *scmd) +{ + struct hiraid_dev *hdev = shost_priv(scmd->device->host); + + if (hdev->state != DEV_LIVE) + return SUCCESS; + + dev_warn(hdev->dev, "sdev[%d:%d] send host reset\n", + scmd->device->channel, scmd->device->id); + if (hiraid_reset_work_sync(hdev) == -EBUSY) + flush_work(&hdev->reset_work); + + if (hdev->state != DEV_LIVE) { + dev_warn(hdev->dev, "sdev[%d:%d] host reset failed\n", + scmd->device->channel, scmd->device->id); + return FAILED; + } + + dev_warn(hdev->dev, "sdev[%d:%d] host reset success\n", + scmd->device->channel, scmd->device->id); + + return SUCCESS; +} + +static pci_ers_result_t hiraid_pci_error_detected(struct pci_dev *pdev, + pci_channel_state_t state) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "pci error detected, state[%d]\n", state); + + switch (state) { + case pci_channel_io_normal: + dev_warn(hdev->dev, "channel is normal, do nothing\n"); + + return PCI_ERS_RESULT_CAN_RECOVER; + case pci_channel_io_frozen: + dev_warn(hdev->dev, "channel io frozen, need reset controller\n"); + + scsi_block_requests(hdev->shost); + + hiraid_dev_state_trans(hdev, DEV_RESETTING); + + return PCI_ERS_RESULT_NEED_RESET; + case pci_channel_io_perm_failure: + dev_warn(hdev->dev, "channel io failure, disconnect\n"); + + return PCI_ERS_RESULT_DISCONNECT; + } + + return PCI_ERS_RESULT_NEED_RESET; +} + +static pci_ers_result_t hiraid_pci_slot_reset(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "restart after slot reset\n"); + + pci_restore_state(pdev); + + if (!queue_work(work_queue, &hdev->reset_work)) { + dev_err(hdev->dev, "err, the device is resetting state\n"); + return PCI_ERS_RESULT_NONE; + } + + flush_work(&hdev->reset_work); + + scsi_unblock_requests(hdev->shost); + + return PCI_ERS_RESULT_RECOVERED; +} + +static void hiraid_reset_pci_finish(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + dev_info(hdev->dev, "enter hiraid reset finish\n"); +} + +static ssize_t csts_pp_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_PP_MASK); + ret >>= HIRAID_CSTS_PP_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_shst_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_SHST_MASK); + ret >>= HIRAID_CSTS_SHST_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_cfs_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) { + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_CFS_MASK); + ret >>= HIRAID_CSTS_CFS_SHIFT; + } + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t csts_rdy_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + int ret = -1; + + if (pci_device_is_present(hdev->pdev)) + ret = (readl(hdev->bar + HIRAID_REG_CSTS) & HIRAID_CSTS_RDY); + + return snprintf(buf, PAGE_SIZE, "%d\n", ret); +} + +static ssize_t fw_version_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%s\n", hdev->ctrl_info->fw_version); +} + +static ssize_t hdd_dispatch_store(struct device *cdev, struct device_attribute *attr, + const char *buf, size_t count) +{ + int val = 0; + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + if (kstrtoint(buf, 0, &val) != 0) + return -EINVAL; + if (val < DISPATCH_BY_CPU || val > DISPATCH_BY_DISK) + return -EINVAL; + hdev->hdd_dispatch = val; + + return strlen(buf); +} +static ssize_t hdd_dispatch_show(struct device *cdev, struct device_attribute *attr, char *buf) +{ + struct Scsi_Host *shost = class_to_shost(cdev); + struct hiraid_dev *hdev = shost_priv(shost); + + return snprintf(buf, PAGE_SIZE, "%d\n", hdev->hdd_dispatch); +} + +static DEVICE_ATTR_RO(csts_pp); +static DEVICE_ATTR_RO(csts_shst); +static DEVICE_ATTR_RO(csts_cfs); +static DEVICE_ATTR_RO(csts_rdy); +static DEVICE_ATTR_RO(fw_version); +static DEVICE_ATTR_RW(hdd_dispatch); + +static struct device_attribute *hiraid_host_attrs[] = { + &dev_attr_csts_rdy, + &dev_attr_csts_pp, + &dev_attr_csts_cfs, + &dev_attr_fw_version, + &dev_attr_csts_shst, + &dev_attr_hdd_dispatch, + NULL, +}; + +static int hiraid_get_vd_info(struct hiraid_dev *hdev, struct hiraid_vd_info *vd_info, u16 vid) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", hdev->state); + return -EBUSY; + } + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_VDINFO); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.usr_cmd.info_1.param_len = cpu_to_le16(VDINFO_PARAM_LEN); + admin_cmd.usr_cmd.cdw10 = cpu_to_le32(vid); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, USRCMD_TIMEOUT); + if (!ret) + memcpy(vd_info, data_ptr, sizeof(struct hiraid_vd_info)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static int hiraid_get_bgtask(struct hiraid_dev *hdev, struct hiraid_bgtask *bgtask) +{ + struct hiraid_admin_command admin_cmd; + u8 *data_ptr = NULL; + dma_addr_t buffer_phy = 0; + int ret; + + if (hdev->state >= DEV_RESETTING) { + dev_err(hdev->dev, "err, host state[%d] is not right\n", hdev->state); + return -EBUSY; + } + + data_ptr = dma_alloc_coherent(hdev->dev, PAGE_SIZE, &buffer_phy, GFP_KERNEL); + if (!data_ptr) + return -ENOMEM; + + memset(&admin_cmd, 0, sizeof(admin_cmd)); + admin_cmd.usr_cmd.opcode = USR_CMD_READ; + admin_cmd.usr_cmd.info_0.subopcode = cpu_to_le16(USR_CMD_BGTASK); + admin_cmd.usr_cmd.info_1.data_len = cpu_to_le16(USR_CMD_RDLEN); + admin_cmd.common.dptr.prp1 = cpu_to_le64(buffer_phy); + + ret = hiraid_put_admin_sync_request(hdev, &admin_cmd, NULL, NULL, USRCMD_TIMEOUT); + if (!ret) + memcpy(bgtask, data_ptr, sizeof(struct hiraid_bgtask)); + + dma_free_coherent(hdev->dev, PAGE_SIZE, data_ptr, buffer_phy); + + return ret; +} + +static ssize_t raid_level_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + vd_info->rg_level = ARRAY_SIZE(raid_levels) - 1; + + ret = (vd_info->rg_level < ARRAY_SIZE(raid_levels)) ? + vd_info->rg_level : (ARRAY_SIZE(raid_levels) - 1); + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "RAID-%s\n", raid_levels[ret]); +} + +static ssize_t raid_state_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_sdev_hostdata *hostdata; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) { + vd_info->vd_status = 0; + vd_info->rg_id = 0xff; + } + + ret = (vd_info->vd_status < ARRAY_SIZE(raid_states)) ? vd_info->vd_status : 0; + + kfree(vd_info); + + return snprintf(buf, PAGE_SIZE, "%s\n", raid_states[ret]); +} + +static ssize_t raid_resync_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct scsi_device *sdev; + struct hiraid_dev *hdev; + struct hiraid_vd_info *vd_info; + struct hiraid_bgtask *bgtask; + struct hiraid_sdev_hostdata *hostdata; + u8 rg_id, i, progress = 0; + int ret; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + vd_info = kmalloc(sizeof(*vd_info), GFP_KERNEL); + if (!vd_info || !HIRAID_DEV_INFO_ATTR_VD(hostdata->attr)) + return snprintf(buf, PAGE_SIZE, "NA\n"); + + ret = hiraid_get_vd_info(hdev, vd_info, sdev->id); + if (ret) + goto out; + + rg_id = vd_info->rg_id; + + bgtask = (struct hiraid_bgtask *)vd_info; + ret = hiraid_get_bgtask(hdev, bgtask); + if (ret) + goto out; + for (i = 0; i < bgtask->task_num; i++) { + if ((bgtask->bgtask[i].type == BGTASK_TYPE_REBUILD) && + (le16_to_cpu(bgtask->bgtask[i].vd_id) == rg_id)) + progress = bgtask->bgtask[i].progress; + } + +out: + kfree(vd_info); + return snprintf(buf, PAGE_SIZE, "%d\n", progress); +} + +static ssize_t dispatch_hwq_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + struct hiraid_sdev_hostdata *hostdata; + + hostdata = to_scsi_device(dev)->hostdata; + return snprintf(buf, PAGE_SIZE, "%d\n", hostdata->hwq); +} + +static ssize_t dispatch_hwq_store(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + int val; + struct hiraid_dev *hdev; + struct scsi_device *sdev; + struct hiraid_sdev_hostdata *hostdata; + + sdev = to_scsi_device(dev); + hdev = shost_priv(sdev->host); + hostdata = sdev->hostdata; + + if (kstrtoint(buf, 0, &val) != 0) + return -EINVAL; + if (val <= 0 || val >= hdev->online_queues) + return -EINVAL; + if (!hiraid_disk_is_hdd(hostdata->attr)) + return -EINVAL; + + hostdata->hwq = val; + return strlen(buf); +} + +static DEVICE_ATTR_RO(raid_level); +static DEVICE_ATTR_RO(raid_state); +static DEVICE_ATTR_RO(raid_resync); +static DEVICE_ATTR_RW(dispatch_hwq); + +static struct device_attribute *hiraid_dev_attrs[] = { + &dev_attr_raid_state, + &dev_attr_raid_level, + &dev_attr_raid_resync, + &dev_attr_dispatch_hwq, + NULL, +}; + +static struct pci_error_handlers hiraid_err_handler = { + .error_detected = hiraid_pci_error_detected, + .slot_reset = hiraid_pci_slot_reset, + .reset_done = hiraid_reset_pci_finish, +}; + +static int hiraid_sysfs_host_reset(struct Scsi_Host *shost, int reset_type) +{ + int ret; + struct hiraid_dev *hdev = shost_priv(shost); + + dev_info(hdev->dev, "start sysfs host reset cmd\n"); + ret = hiraid_reset_work_sync(hdev); + dev_info(hdev->dev, "stop sysfs host reset cmd[%d]\n", ret); + + return ret; +} + +static int hiraid_scan_finished(struct Scsi_Host *shost, unsigned long time) +{ + struct hiraid_dev *hdev = shost_priv(shost); + + hiraid_scan_work(&hdev->scan_work); + + return 1; +} + +static struct scsi_host_template hiraid_driver_template = { + .module = THIS_MODULE, + .name = "hiraid", + .proc_name = "hiraid", + .queuecommand = hiraid_queue_command, + .slave_alloc = hiraid_slave_alloc, + .slave_destroy = hiraid_slave_destroy, + .slave_configure = hiraid_slave_configure, + .scan_finished = hiraid_scan_finished, + .eh_timed_out = hiraid_timed_out, + .eh_abort_handler = hiraid_abort, + .eh_target_reset_handler = hiraid_target_reset, + .eh_bus_reset_handler = hiraid_bus_reset, + .eh_host_reset_handler = hiraid_host_reset, + .change_queue_depth = scsi_change_queue_depth, + .this_id = -1, + .unchecked_isa_dma = 0, + .shost_attrs = hiraid_host_attrs, + .sdev_attrs = hiraid_dev_attrs, + .host_reset = hiraid_sysfs_host_reset, +}; + +static void hiraid_shutdown(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + + hiraid_delete_io_queues(hdev); + hiraid_disable_admin_queue(hdev, true); +} + +static bool hiraid_bsg_is_valid(struct bsg_job *job) +{ + u64 timeout = 0; + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_bsg_request *bsg_req = job->request; + struct hiraid_dev *hdev = shost_priv(dev_to_shost(job->dev)); + + if (bsg_req == NULL || job->request_len != sizeof(struct hiraid_bsg_request)) + return false; + + switch (bsg_req->msgcode) { + case HIRAID_BSG_ADMIN: + timeout = msecs_to_jiffies(bsg_req->admcmd.timeout_ms); + break; + case HIRAID_BSG_IOPTHRU: + timeout = msecs_to_jiffies(bsg_req->pthrucmd.timeout_ms); + break; + default: + dev_info(hdev->dev, "bsg unsupport msgcode[%d]\n", bsg_req->msgcode); + return false; + } + + if ((timeout + CTL_RST_TIME) > rq->timeout) { + dev_err(hdev->dev, "bsg invalid time\n"); + return false; + } + + return true; +} + +/* bsg dispatch user command */ +static int hiraid_bsg_dispatch(struct bsg_job *job) +{ + struct Scsi_Host *shost = dev_to_shost(job->dev); + struct hiraid_dev *hdev = shost_priv(shost); + struct request *rq = blk_mq_rq_from_pdu(job); + struct hiraid_bsg_request *bsg_req = job->request; + int ret = -ENOMSG; + + job->reply_len = 0; + + if (!hiraid_bsg_is_valid(job)) { + bsg_job_done(job, ret, 0); + return 0; + } + + dev_log_dbg(hdev->dev, "bsg msgcode[%d] msglen[%d] timeout[%d];" + "reqnsge[%d], reqlen[%d]\n", + bsg_req->msgcode, job->request_len, rq->timeout, + job->request_payload.sg_cnt, job->request_payload.payload_len); + + switch (bsg_req->msgcode) { + case HIRAID_BSG_ADMIN: + ret = hiraid_user_send_admcmd(hdev, job); + break; + case HIRAID_BSG_IOPTHRU: + ret = hiraid_user_send_ptcmd(hdev, job); + break; + default: + break; + } + + if (ret > 0) + ret = ret | (ret << 8); + + bsg_job_done(job, ret, 0); + return 0; +} + +static inline void hiraid_unregist_bsg(struct hiraid_dev *hdev) +{ + if (hdev->bsg_queue) { + bsg_unregister_queue(hdev->bsg_queue); + blk_cleanup_queue(hdev->bsg_queue); + } +} +static int hiraid_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct hiraid_dev *hdev; + struct Scsi_Host *shost; + int node, ret; + char bsg_name[15]; + + shost = scsi_host_alloc(&hiraid_driver_template, sizeof(*hdev)); + if (!shost) { + dev_err(&pdev->dev, "failed to allocate scsi host\n"); + return -ENOMEM; + } + hdev = shost_priv(shost); + hdev->pdev = pdev; + hdev->dev = get_device(&pdev->dev); + + node = dev_to_node(hdev->dev); + if (node == NUMA_NO_NODE) { + node = first_memory_node; + set_dev_node(hdev->dev, node); + } + hdev->numa_node = node; + hdev->shost = shost; + hdev->instance = shost->host_no; + pci_set_drvdata(pdev, hdev); + + ret = hiraid_dev_map(hdev); + if (ret) + goto put_dev; + + init_rwsem(&hdev->dev_rwsem); + INIT_WORK(&hdev->scan_work, hiraid_scan_work); + INIT_WORK(&hdev->timesyn_work, hiraid_timesyn_work); + INIT_WORK(&hdev->reset_work, hiraid_reset_work); + INIT_WORK(&hdev->fwact_work, hiraid_fwactive_work); + spin_lock_init(&hdev->state_lock); + + ret = hiraid_alloc_resources(hdev); + if (ret) + goto dev_unmap; + + ret = hiraid_pci_enable(hdev); + if (ret) + goto resources_free; + + ret = hiraid_setup_admin_queue(hdev); + if (ret) + goto pci_disable; + + ret = hiraid_init_control_info(hdev); + if (ret) + goto disable_admin_q; + + ret = hiraid_setup_io_queues(hdev); + if (ret) + goto disable_admin_q; + + hiraid_shost_init(hdev); + + ret = scsi_add_host(hdev->shost, hdev->dev); + if (ret) { + dev_err(hdev->dev, "add shost to system failed, ret[%d]\n", ret); + goto remove_io_queues; + } + + snprintf(bsg_name, sizeof(bsg_name), "hiraid%d", shost->host_no); + hdev->bsg_queue = bsg_setup_queue(&shost->shost_gendev, bsg_name, hiraid_bsg_dispatch, + NULL, hiraid_get_max_cmd_size(hdev)); + if (IS_ERR(hdev->bsg_queue)) { + dev_err(hdev->dev, "err, setup bsg failed\n"); + hdev->bsg_queue = NULL; + goto remove_io_queues; + } + + if (hdev->online_queues == HIRAID_ADMIN_QUEUE_NUM) { + dev_warn(hdev->dev, "warn: only admin queue can be used\n"); + return 0; + } + + hdev->state = DEV_LIVE; + + hiraid_init_async_event(hdev); + + ret = hiraid_dev_list_init(hdev); + if (ret) + goto unregist_bsg; + + ret = hiraid_configure_timestamp(hdev); + if (ret) + dev_warn(hdev->dev, "time synchronization failed\n"); + + ret = hiraid_alloc_io_ptcmds(hdev); + if (ret) + goto unregist_bsg; + + scsi_scan_host(hdev->shost); + + return 0; + +unregist_bsg: + hiraid_unregist_bsg(hdev); +remove_io_queues: + hiraid_delete_io_queues(hdev); +disable_admin_q: + hiraid_free_sense_buffer(hdev); + hiraid_disable_admin_queue(hdev, false); +pci_disable: + hiraid_free_all_queues(hdev); + hiraid_pci_disable(hdev); +resources_free: + hiraid_free_resources(hdev); +dev_unmap: + hiraid_dev_unmap(hdev); +put_dev: + put_device(hdev->dev); + scsi_host_put(shost); + + return -ENODEV; +} + +static void hiraid_remove(struct pci_dev *pdev) +{ + struct hiraid_dev *hdev = pci_get_drvdata(pdev); + struct Scsi_Host *shost = hdev->shost; + + dev_info(hdev->dev, "enter hiraid remove\n"); + + hiraid_dev_state_trans(hdev, DEV_DELETING); + flush_work(&hdev->reset_work); + + if (!pci_device_is_present(pdev)) + hiraid_flush_running_cmds(hdev); + + hiraid_unregist_bsg(hdev); + scsi_remove_host(shost); + hiraid_free_io_ptcmds(hdev); + kfree(hdev->dev_info); + hiraid_delete_io_queues(hdev); + hiraid_free_sense_buffer(hdev); + hiraid_disable_admin_queue(hdev, false); + hiraid_free_all_queues(hdev); + hiraid_pci_disable(hdev); + hiraid_free_resources(hdev); + hiraid_dev_unmap(hdev); + put_device(hdev->dev); + scsi_host_put(shost); + + dev_info(hdev->dev, "exit hiraid remove\n"); +} + +static const struct pci_device_id hiraid_hw_card_ids[] = { + { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI_LOGIC, HIRAID_SERVER_DEVICE_HBA_DID) }, + { PCI_DEVICE(PCI_VENDOR_ID_HUAWEI_LOGIC, HIRAID_SERVER_DEVICE_RAID_DID) }, + { 0, } +}; +MODULE_DEVICE_TABLE(pci, hiraid_hw_card_ids); + +static struct pci_driver hiraid_driver = { + .name = "hiraid", + .id_table = hiraid_hw_card_ids, + .probe = hiraid_probe, + .remove = hiraid_remove, + .shutdown = hiraid_shutdown, + .err_handler = &hiraid_err_handler, +}; + +static int __init hiraid_init(void) +{ + int ret; + + work_queue = alloc_workqueue("hiraid-wq", WQ_UNBOUND | WQ_MEM_RECLAIM | WQ_SYSFS, 0); + if (!work_queue) + return -ENOMEM; + + hiraid_class = class_create(THIS_MODULE, "hiraid"); + if (IS_ERR(hiraid_class)) { + ret = PTR_ERR(hiraid_class); + goto destroy_wq; + } + + ret = pci_register_driver(&hiraid_driver); + if (ret < 0) + goto destroy_class; + + return 0; + +destroy_class: + class_destroy(hiraid_class); +destroy_wq: + destroy_workqueue(work_queue); + + return ret; +} + +static void __exit hiraid_exit(void) +{ + pci_unregister_driver(&hiraid_driver); + class_destroy(hiraid_class); + destroy_workqueue(work_queue); +} + +MODULE_AUTHOR("Huawei Technologies CO., Ltd"); +MODULE_DESCRIPTION("Huawei RAID driver"); +MODULE_LICENSE("GPL"); +MODULE_VERSION(HIRAID_DRV_VERSION); +module_init(hiraid_init); +module_exit(hiraid_exit); -- 2.22.0.windows.1

2 1

[PATCH openEuler-1.0-LTS v2] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 16 Nov '23

16 Nov '23

From: Florian Westphal <fw(a)strlen.de> mainline inclusion from mainline-v6.5-rc1 commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Conflicts: net/netfilter/nf_conntrack_proto_dccp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 50 ++++++++++++++++++++++++- 1 file changed, 48 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 3ba1f4d9934f..cd7a51a0f100 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -433,17 +433,47 @@ static u64 dccp_ack_seq(const struct dccp_hdr *dh) ntohl(dhack->dccph_ack_nr_low); } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + static int dccp_packet(struct nf_conn *ct, const struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); BUG_ON(dh == NULL); + + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (type == DCCP_PKT_RESET && @@ -526,10 +556,20 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, struct sk_buff *skb, unsigned int dataoff, u_int8_t pf, unsigned int hooknum) { + static const unsigned long require_seq48 = 1 << DCCP_PKT_REQUEST | + 1 << DCCP_PKT_RESPONSE | + 1 << DCCP_PKT_CLOSEREQ | + 1 << DCCP_PKT_CLOSE | + 1 << DCCP_PKT_RESET | + 1 << DCCP_PKT_SYNC | + 1 << DCCP_PKT_SYNCACK; struct dccp_hdr _dh, *dh; unsigned int dccp_len = skb->len - dataoff; unsigned int cscov; const char *msg; + u8 type; + + BUILD_BUG_ON(DCCP_PKT_INVALID >= BITS_PER_LONG); dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); if (dh == NULL) { @@ -559,11 +599,17 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, goto out_invalid; } - if (dh->dccph_type >= DCCP_PKT_INVALID) { + type = dh->dccph_type; + if (type >= DCCP_PKT_INVALID) { msg = "nf_ct_dccp: reserved packet type "; goto out_invalid; } + if (test_bit(type, &require_seq48) && !dh->dccph_x) { + msg = "nf_ct_dccp: type lacks 48bit sequence numbers"; + goto out_invalid; + } + return NF_ACCEPT; out_invalid: -- 2.34.1

2 1

[OLK-5.10 0/2] Support SRQ Context tracing by debugfs
by Chengchang Tang 16 Nov '23

16 Nov '23

From: Juan Zhou <zhoujuan51(a)h-partners.com> The first patch refactors the hns debugfs function, and the second patch supports querying SRQ context by debugfs. Junxian Huang (1): RDMA/hns: Refactor hns RoCE debugfs wenglianfa (1): RDMA/hns: Support SRQ Context tracing by debugfs drivers/infiniband/hw/hns/hns_roce_debugfs.c | 318 +++++++++---------- drivers/infiniband/hw/hns/hns_roce_debugfs.h | 57 ++++ drivers/infiniband/hw/hns/hns_roce_device.h | 7 +- drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 26 +- drivers/infiniband/hw/hns/hns_roce_main.c | 1 - drivers/infiniband/hw/hns/hns_roce_srq.c | 12 + 6 files changed, 242 insertions(+), 179 deletions(-) -- 2.30.0

1 2

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH openEuler-1.0-LTS,v2] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 16 Nov '23

16 Nov '23

From: Florian Westphal <fw(a)strlen.de> mainline inclusion from mainline-v6.5-rc1 commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Conflicts: net/netfilter/nf_conntrack_proto_dccp.c Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 50 ++++++++++++++++++++++++- 1 file changed, 48 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 3ba1f4d9934f..cd7a51a0f100 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -433,17 +433,47 @@ static u64 dccp_ack_seq(const struct dccp_hdr *dh) ntohl(dhack->dccph_ack_nr_low); } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + static int dccp_packet(struct nf_conn *ct, const struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); BUG_ON(dh == NULL); + + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (type == DCCP_PKT_RESET && @@ -526,10 +556,20 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, struct sk_buff *skb, unsigned int dataoff, u_int8_t pf, unsigned int hooknum) { + static const unsigned long require_seq48 = 1 << DCCP_PKT_REQUEST | + 1 << DCCP_PKT_RESPONSE | + 1 << DCCP_PKT_CLOSEREQ | + 1 << DCCP_PKT_CLOSE | + 1 << DCCP_PKT_RESET | + 1 << DCCP_PKT_SYNC | + 1 << DCCP_PKT_SYNCACK; struct dccp_hdr _dh, *dh; unsigned int dccp_len = skb->len - dataoff; unsigned int cscov; const char *msg; + u8 type; + + BUILD_BUG_ON(DCCP_PKT_INVALID >= BITS_PER_LONG); dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); if (dh == NULL) { @@ -559,11 +599,17 @@ static int dccp_error(struct net *net, struct nf_conn *tmpl, goto out_invalid; } - if (dh->dccph_type >= DCCP_PKT_INVALID) { + type = dh->dccph_type; + if (type >= DCCP_PKT_INVALID) { msg = "nf_ct_dccp: reserved packet type "; goto out_invalid; } + if (test_bit(type, &require_seq48) && !dh->dccph_x) { + msg = "nf_ct_dccp: type lacks 48bit sequence numbers"; + goto out_invalid; + } + return NF_ACCEPT; out_invalid: -- 2.34.1

1 0

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH OLK-5.10 v5 00/19] introduce smart_grid zone
by Yipeng Zou 16 Nov '23

16 Nov '23

The patch sets include two parts: 1. patch 1~15: Rebase smart_grid from openeuler-1.0-LTS to OLK-5.10 2. patch 16~19: introduce smart_grid zone qos and cpufreq Since v4: 1. Place the highest level task in current domain level itself in sched_grid_prefer_cpus Since v3: 1. fix CI warning Since v2: 1. static alloc sg_zone cpumask. 2. fix some warning Hui Tang (13): sched: Introduce smart grid scheduling strategy for cfs sched: fix smart grid usage count sched: fix WARN found by deadlock detect sched: Fix possible deadlock in tg_set_dynamic_affinity_mode sched: Fix negative count for jump label sched: Fix timer storm for smart grid sched: fix dereference NULL pointers sched: Fix memory leak on error branch sched: clear credit count in error branch sched: Adjust few parameters range for smart grid sched: Delete redundant updates to p->prefer_cpus sched: Fix memory leak for smart grid sched: Fix null pointer derefrence for sd->span Wang ShaoBo (2): sched: smart grid: init sched_grid_qos structure on QOS purpose config: enable CONFIG_QOS_SCHED_SMART_GRID by default Yipeng Zou (4): sched: introduce smart grid qos zone smart_grid: introduce /proc/pid/smart_grid_level smart_grid: introduce smart_grid_strategy_ctrl sysctl smart_grid: cpufreq: introduce smart_grid cpufreq control arch/arm64/configs/openeuler_defconfig | 1 + drivers/cpufreq/cpufreq.c | 234 ++++++++++++ fs/proc/array.c | 13 + fs/proc/base.c | 76 ++++ include/linux/cpufreq.h | 11 + include/linux/sched.h | 22 ++ include/linux/sched/grid_qos.h | 135 +++++++ include/linux/sched/sysctl.h | 5 + init/Kconfig | 13 + kernel/fork.c | 15 +- kernel/sched/Makefile | 1 + kernel/sched/core.c | 160 +++++++- kernel/sched/fair.c | 496 ++++++++++++++++++++++++- kernel/sched/grid/Makefile | 2 + kernel/sched/grid/internal.h | 6 + kernel/sched/grid/power.c | 27 ++ kernel/sched/grid/qos.c | 273 ++++++++++++++ kernel/sched/grid/stat.c | 47 +++ kernel/sched/sched.h | 48 +++ kernel/sysctl.c | 22 +- mm/mempolicy.c | 12 +- 21 files changed, 1601 insertions(+), 18 deletions(-) create mode 100644 include/linux/sched/grid_qos.h create mode 100644 kernel/sched/grid/Makefile create mode 100644 kernel/sched/grid/internal.h create mode 100644 kernel/sched/grid/power.c create mode 100644 kernel/sched/grid/qos.c create mode 100644 kernel/sched/grid/stat.c -- 2.34.1

1 19

[PATCH openEuler-1.0-LTS] mm, memory_hotplug: update pcp lists everytime onlining a memory block
by Liu Shixin 15 Nov '23

15 Nov '23

From: Charan Teja Reddy <charante(a)codeaurora.org> mainline inclusion from mainline-v5.9-rc1 commit de1193f0be66f88e2c6d0fd965137668fc2ec4a5 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8GREH CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When onlining a first memory block in a zone, pcp lists are not updated thus pcp struct will have the default setting of ->high = 0,->batch = 1. This means till the second memory block in a zone(if it have) is onlined the pcp lists of this zone will not contain any pages because pcp's ->count is always greater than ->high thus free_pcppages_bulk() is called to free batch size(=1) pages every time system wants to add a page to the pcp list through free_unref_page(). To put this in a word, system is not using benefits offered by the pcp lists when there is a single onlineable memory block in a zone. Correct this by always updating the pcp lists when memory block is onlined. Fixes: 1f522509c77a ("mem-hotplug: avoid multiple zones sharing same boot strapping boot_pageset") Signed-off-by: Charan Teja Reddy <charante(a)codeaurora.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> Reviewed-by: David Hildenbrand <david(a)redhat.com> Acked-by: Vlastimil Babka <vbabka(a)suse.cz> Acked-by: Michal Hocko <mhocko(a)suse.com> Cc: Vinayak Menon <vinmenon(a)codeaurora.org> Link: http://lkml.kernel.org/r/1596372896-15336-1-git-send-email-charante@codeaur… Signed-off-by: Linus Torvalds <torvalds(a)linux-foundation.org> Conflicts: mm/memory_hotplug.c Signed-off-by: Liu Shixin <liushixin2(a)huawei.com> --- mm/memory_hotplug.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 3b961b0fffcc..cebfa881e516 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -878,8 +878,7 @@ int __ref online_pages(unsigned long pfn, unsigned long nr_pages, int online_typ node_states_set_node(nid, &arg); if (need_zonelists_rebuild) build_all_zonelists(NULL); - else - zone_pcp_update(zone); + zone_pcp_update(zone); } init_per_zone_wmark_min(); -- 2.25.1

2 1

[PATCH OLK-5.10] netfilter: conntrack: dccp: copy entire header to stack buffer, not just basic one
by Zhengchao Shao 15 Nov '23

15 Nov '23

From: Florian Westphal <fw(a)strlen.de> stable inclusion from stable-v5.10.188 commit 9bdcda7abaf22f6453e5b5efb7eb4e524095d5d8 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I8EZSO CVE: CVE-2023-39197 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit ff0a3a7d52ff7282dbd183e7fc29a1fe386b0c30 ] Eric Dumazet says: nf_conntrack_dccp_packet() has an unique: dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); And nothing more is 'pulled' from the packet, depending on the content. dh->dccph_doff, and/or dh->dccph_x ...) So dccp_ack_seq() is happily reading stuff past the _dh buffer. BUG: KASAN: stack-out-of-bounds in nf_conntrack_dccp_packet+0x1134/0x11c0 Read of size 4 at addr ffff000128f66e0c by task syz-executor.2/29371 [..] Fix this by increasing the stack buffer to also include room for the extra sequence numbers and all the known dccp packet type headers, then pull again after the initial validation of the basic header. While at it, mark packets invalid that lack 48bit sequence bit but where RFC says the type MUST use them. Compile tested only. v2: first skb_header_pointer() now needs to adjust the size to only pull the generic header. (Eric) Heads-up: I intend to remove dccp conntrack support later this year. Fixes: 2bc780499aa3 ("[NETFILTER]: nf_conntrack: add DCCP protocol support") Reported-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Florian Westphal <fw(a)strlen.de> Reviewed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Zhengchao Shao <shaozhengchao(a)huawei.com> --- net/netfilter/nf_conntrack_proto_dccp.c | 52 +++++++++++++++++++++++-- 1 file changed, 49 insertions(+), 3 deletions(-) diff --git a/net/netfilter/nf_conntrack_proto_dccp.c b/net/netfilter/nf_conntrack_proto_dccp.c index 94001eb51ffe..a9ae292e932a 100644 --- a/net/netfilter/nf_conntrack_proto_dccp.c +++ b/net/netfilter/nf_conntrack_proto_dccp.c @@ -431,9 +431,19 @@ static bool dccp_error(const struct dccp_hdr *dh, struct sk_buff *skb, unsigned int dataoff, const struct nf_hook_state *state) { + static const unsigned long require_seq48 = 1 << DCCP_PKT_REQUEST | + 1 << DCCP_PKT_RESPONSE | + 1 << DCCP_PKT_CLOSEREQ | + 1 << DCCP_PKT_CLOSE | + 1 << DCCP_PKT_RESET | + 1 << DCCP_PKT_SYNC | + 1 << DCCP_PKT_SYNCACK; unsigned int dccp_len = skb->len - dataoff; unsigned int cscov; const char *msg; + u8 type; + + BUILD_BUG_ON(DCCP_PKT_INVALID >= BITS_PER_LONG); if (dh->dccph_doff * 4 < sizeof(struct dccp_hdr) || dh->dccph_doff * 4 > dccp_len) { @@ -458,10 +468,17 @@ static bool dccp_error(const struct dccp_hdr *dh, goto out_invalid; } - if (dh->dccph_type >= DCCP_PKT_INVALID) { + type = dh->dccph_type; + if (type >= DCCP_PKT_INVALID) { msg = "nf_ct_dccp: reserved packet type "; goto out_invalid; } + + if (test_bit(type, &require_seq48) && !dh->dccph_x) { + msg = "nf_ct_dccp: type lacks 48bit sequence numbers"; + goto out_invalid; + } + return false; out_invalid: nf_l4proto_log_invalid(skb, state->net, state->pf, @@ -469,24 +486,53 @@ static bool dccp_error(const struct dccp_hdr *dh, return true; } +struct nf_conntrack_dccp_buf { + struct dccp_hdr dh; /* generic header part */ + struct dccp_hdr_ext ext; /* optional depending dh->dccph_x */ + union { /* depends on header type */ + struct dccp_hdr_ack_bits ack; + struct dccp_hdr_request req; + struct dccp_hdr_response response; + struct dccp_hdr_reset rst; + } u; +}; + +static struct dccp_hdr * +dccp_header_pointer(const struct sk_buff *skb, int offset, const struct dccp_hdr *dh, + struct nf_conntrack_dccp_buf *buf) +{ + unsigned int hdrlen = __dccp_hdr_len(dh); + + if (hdrlen > sizeof(*buf)) + return NULL; + + return skb_header_pointer(skb, offset, hdrlen, buf); +} + int nf_conntrack_dccp_packet(struct nf_conn *ct, struct sk_buff *skb, unsigned int dataoff, enum ip_conntrack_info ctinfo, const struct nf_hook_state *state) { enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - struct dccp_hdr _dh, *dh; + struct nf_conntrack_dccp_buf _dh; u_int8_t type, old_state, new_state; enum ct_dccp_roles role; unsigned int *timeouts; + struct dccp_hdr *dh; - dh = skb_header_pointer(skb, dataoff, sizeof(_dh), &_dh); + dh = skb_header_pointer(skb, dataoff, sizeof(*dh), &_dh.dh); if (!dh) return NF_DROP; if (dccp_error(dh, skb, dataoff, state)) return -NF_ACCEPT; + /* pull again, including possible 48 bit sequences and subtype header */ + dh = dccp_header_pointer(skb, dataoff, dh, &_dh); + if (!dh) + return NF_DROP; + type = dh->dccph_type; if (!nf_ct_is_confirmed(ct) && !dccp_new(ct, skb, dh)) return -NF_ACCEPT; -- 2.34.1

2 1