- Kernel - mailweb.openeuler.org

[PATCH OLK-5.10] roh/hns3: Fix the processing flow of ROH CMDq during the reset process.
by Ke Chen 21 Dec '23

21 Dec '23

driver inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8PS3D ----------------------------------------------------------------------- Fix the sequence issue in the ROH CMDq processing flow during the FLR reset process, which may occasionally cause calltrace. In this patch, we added a flag in the reset process to indicate the availability of the ROH CMDq, in order to fix this issue. Fixes:1cb9fff33035("roh/hns3: Add ROH cmdq interface support") Signed-off-by: Ke Chen <chenke54(a)huawei.com> Reviewed-by: Gang Zhang <gang.zhang(a)huawei.com> --- drivers/roh/hw/hns3/hns3_cmdq.c | 5 +++++ drivers/roh/hw/hns3/hns3_device.h | 3 ++- drivers/roh/hw/hns3/hns3_main.c | 2 ++ 3 files changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/roh/hw/hns3/hns3_cmdq.c b/drivers/roh/hw/hns3/hns3_cmdq.c index 75fa2e852152..96c0286779eb 100644 --- a/drivers/roh/hw/hns3/hns3_cmdq.c +++ b/drivers/roh/hw/hns3/hns3_cmdq.c @@ -141,6 +141,8 @@ int hns3_roh_cmdq_init(struct hns3_roh_device *hroh_dev) /* Init CRQ REG */ hns3_roh_cmdq_init_regs(hroh_dev, HNS3_ROH_CMDQ_CRQ); + clear_bit(HNS3_ROH_STATE_CMD_DISABLE, &priv->handle->rohinfo.reset_state); + return 0; err_crq: @@ -340,6 +342,9 @@ int hns3_roh_cmdq_send(struct hns3_roh_device *hroh_dev, struct hns3_roh_desc *d int ntc = 0; int ret = 0; + if (test_bit(HNS3_ROH_STATE_CMD_DISABLE, &priv->handle->rohinfo.reset_state)) + return -EIO; + spin_lock_bh(&csq->lock); ret = hns3_roh_cmdq_build(hroh_dev, desc, num, &ntc); if (ret) { diff --git a/drivers/roh/hw/hns3/hns3_device.h b/drivers/roh/hw/hns3/hns3_device.h index 07e46879a4b9..4cc5f6aeffb8 100644 --- a/drivers/roh/hw/hns3/hns3_device.h +++ b/drivers/roh/hw/hns3/hns3_device.h @@ -8,7 +8,8 @@ enum hns3_roh_state { HNS3_ROH_STATE_RESETTING, HNS3_ROH_STATE_INIT, HNS3_ROH_STATE_INITED, - HNS3_ROH_STATE_DOWN + HNS3_ROH_STATE_DOWN, + HNS3_ROH_STATE_CMD_DISABLE, }; enum { HNS3_ROH_RST_DIRECT_RETURN = 0 }; diff --git a/drivers/roh/hw/hns3/hns3_main.c b/drivers/roh/hw/hns3/hns3_main.c index 64f9a13e5d83..78e57a3d8a95 100644 --- a/drivers/roh/hw/hns3/hns3_main.c +++ b/drivers/roh/hw/hns3/hns3_main.c @@ -388,6 +388,8 @@ static int hns3_roh_reset_notify(struct hnae3_handle *handle, case HNAE3_UNINIT_CLIENT: ret = hns3_roh_reset_notify_uninit(handle); break; + case HNAE3_DOWN_CLIENT: + set_bit(HNS3_ROH_STATE_CMD_DISABLE, &handle->rohinfo.reset_state); default: break; } -- 2.30.0

2 1

[PATCH OLK-6.6 v3 0/3] Revert "compiler: remove CONFIG_OPTIMIZE_INLINING entirely"
by Yuntao Liu 21 Dec '23

21 Dec '23

*** BLURB HERE *** Guo Xuenan (3): Revert "compiler: remove CONFIG_OPTIMIZE_INLINING entirely" make OPTIMIZE_INLINING config editable disable OPTIMIZE_INLINING by default arch/arm64/kvm/sys_regs.h | 5 +++++ arch/x86/configs/i386_defconfig | 1 + arch/x86/configs/x86_64_defconfig | 1 + drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h | 2 +- .../pci/hive_isp_css_include/print_support.h | 4 ++++ include/linux/compiler_types.h | 8 ++++++++ include/trace/trace_events.h | 2 +- kernel/configs/tiny.config | 1 + lib/Kconfig.debug | 13 +++++++++++++ 9 files changed, 35 insertions(+), 2 deletions(-) -- 2.34.1

2 4

[PATCH OLK-6.6 0/3] xfs: fix hung and warning
by Long Li 21 Dec '23

21 Dec '23

This patch set fix some hung task and warning bug: Long Li (1): xfs: fix hung when transaction commit fail in xfs_inactive_ifree Ye Bin (2): xfs: fix dead loop when do mount with IO fault injection xfs: fix warning in xfs_vm_writepages() fs/xfs/xfs_buf_item.c | 20 ++++++++++++++++++++ fs/xfs/xfs_icache.c | 6 ++++++ fs/xfs/xfs_inode.c | 3 ++- fs/xfs/xfs_inode_item.c | 3 ++- fs/xfs/xfs_log.c | 4 +++- 5 files changed, 33 insertions(+), 3 deletions(-) -- 2.31.1

2 4

[PATCH OLK-6.6 v2 0/3] Revert "compiler: remove CONFIG_OPTIMIZE_INLINING entirely"
by Yuntao Liu 20 Dec '23

20 Dec '23

*** BLURB HERE *** Guo Xuenan (3): Revert "compiler: remove CONFIG_OPTIMIZE_INLINING entirely" make OPTIMIZE_INLINING config editable disable OPTIMIZE_INLINING by default arch/arm64/kvm/sys_regs.h | 5 +++++ arch/x86/configs/i386_defconfig | 1 + arch/x86/configs/x86_64_defconfig | 1 + drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h | 2 +- .../pci/hive_isp_css_include/print_support.h | 4 ++++ include/linux/compiler_types.h | 8 ++++++++ include/trace/trace_events.h | 2 +- kernel/configs/tiny.config | 1 + lib/Kconfig.debug | 13 +++++++++++++ 9 files changed, 35 insertions(+), 2 deletions(-) -- 2.34.1

2 4

[PATCH OLK-6.6 0/6] kabi: net: reserve space for net
by Zhengchao Shao 20 Dec '23

20 Dec '23

2 7

[PATCH OLK-5.10 0/2] fix CVE-2023-6817 and CVE-2023-4244
by Dong Chenchen 20 Dec '23

20 Dec '23

fix CVE-2023-6817 and CVE-2023-4244 Florian Westphal (2): netfilter: nf_tables: don't skip expired elements during walk netfilter: nft_set_pipapo: skip inactive elements during set walk net/netfilter/nf_tables_api.c | 4 ++++ net/netfilter/nft_set_hash.c | 2 -- net/netfilter/nft_set_pipapo.c | 19 ++++++++++++++----- net/netfilter/nft_set_rbtree.c | 2 -- 4 files changed, 18 insertions(+), 9 deletions(-) -- 2.25.1

2 3

[PATCH OLK-6.6 0/8] adapt kabi reserve patch to OLK6.6
by Lin Yujun 20 Dec '23

20 Dec '23

*** BLURB HERE *** Chen Jiahao (2): KABI: reserve space for several i2c structures KABI: reserve space for struct input_dev Lin Ruizhe (6): bootparam: Add kabi_reserve in bootparam interrupt: Add kabi_reserve in interrupt.h irq: Add kabi_reserve in irq irq_desc: Add kabi_reserve in irq_desc irqdomain: Add kabi_reserve in irqdomain msi: Add kabi_reserve in msi.h arch/x86/include/uapi/asm/bootparam.h | 2 ++ include/linux/i2c.h | 11 +++++++++++ include/linux/input.h | 4 ++++ include/linux/interrupt.h | 4 ++++ include/linux/irq.h | 6 ++++++ include/linux/irqdesc.h | 3 ++- include/linux/irqdomain.h | 2 ++ include/linux/msi.h | 3 +++ 8 files changed, 34 insertions(+), 1 deletion(-) -- 2.34.1

2 9

[PATCH OLK-6.6 0/5] ubi: Enhanced fault injection capability for the UBI driver
by ZhaoLong Wang 20 Dec '23

20 Dec '23

CONFIG Dependency CONFIG_MTD=y CONFIG_MTD_NAND_NANDSIM=m CONFIG_MTD_UBI=m CONFIG_FAULT_INJECTION_DEBUG_FS=y CONFIG_MTD_PARTITIONED_MASTER The existing fault injection capability of UBI is too simple. It uses hard-coded fault probability values and lacks other configurable options. As a result, these interfaces are difficult to use when digging defects in the abnormal path of code and reproducing some problems. The kernel provides a powerful fault injection framework, which provides rich configurable fault injection attributes during runtime. So it can be used to improve the fault injection capability of the UBI driver. This series of patches refactor the existing fault injection interface and add some fault injection types to help testers and developers find potential problems in the code. This series of patches enhance the existing fault injection interface and retain the old debugfs interface, and add some fault injection types to help testers and developers Look for potential problems in the code. ZhaoLong Wang (5): ubi: Use the fault injection framework to enhance the fault injection capability ubi: Split io_failures into write_failure and erase_failure ubi: Add six fault injection type for testing ubi: Reserve sufficient buffer length for the input mask mtd: Add several functions to the fail_function list drivers/mtd/mtdcore.c | 5 + drivers/mtd/ubi/Kconfig | 9 ++ drivers/mtd/ubi/debug.c | 107 ++++++++++++-- drivers/mtd/ubi/debug.h | 304 +++++++++++++++++++++++++++++++++++++--- drivers/mtd/ubi/io.c | 86 +++++++++++- drivers/mtd/ubi/ubi.h | 45 +++--- 6 files changed, 497 insertions(+), 59 deletions(-) -- 2.39.2

2 6

[PATCH OLK-6.6 v5] ascend: export interfaces required by ascend drivers
by Yuan Can 20 Dec '23

20 Dec '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8OX3Q CVE: NA -------------------------------- Export oom_type_notifier_call and __get_vm_area_caller. Export pm_autosleep_set_state and __vmalloc_node_range. Export alloc_workqueue_attrs, free_workqueue_attrs and apply_workqueue_attrs. Export find_get_task_by_vpid. Signed-off-by: Yuan Can <yuancan(a)huawei.com> --- kernel/pid.c | 1 + kernel/power/autosleep.c | 1 + kernel/workqueue.c | 3 +++ mm/oom_kill.c | 1 + mm/vmalloc.c | 2 ++ 5 files changed, 8 insertions(+) diff --git a/kernel/pid.c b/kernel/pid.c index 6500ef956f2f..383abde0c208 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -437,6 +437,7 @@ struct task_struct *find_get_task_by_vpid(pid_t nr) return task; } +EXPORT_SYMBOL_GPL(find_get_task_by_vpid); struct pid *get_task_pid(struct task_struct *task, enum pid_type type) { diff --git a/kernel/power/autosleep.c b/kernel/power/autosleep.c index b29c8aca7486..80ba474daa40 100644 --- a/kernel/power/autosleep.c +++ b/kernel/power/autosleep.c @@ -113,6 +113,7 @@ int pm_autosleep_set_state(suspend_state_t state) mutex_unlock(&autosleep_lock); return 0; } +EXPORT_SYMBOL_GPL(pm_autosleep_set_state); int __init pm_autosleep_init(void) { diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 0f682da96e1c..3eabf97c4e9a 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -3713,6 +3713,7 @@ void free_workqueue_attrs(struct workqueue_attrs *attrs) kfree(attrs); } } +EXPORT_SYMBOL_GPL(free_workqueue_attrs); /** * alloc_workqueue_attrs - allocate a workqueue_attrs @@ -3741,6 +3742,7 @@ struct workqueue_attrs *alloc_workqueue_attrs(void) free_workqueue_attrs(attrs); return NULL; } +EXPORT_SYMBOL_GPL(alloc_workqueue_attrs); static void copy_workqueue_attrs(struct workqueue_attrs *to, const struct workqueue_attrs *from) @@ -4482,6 +4484,7 @@ int apply_workqueue_attrs(struct workqueue_struct *wq, return ret; } +EXPORT_SYMBOL_GPL(apply_workqueue_attrs); /** * wq_update_pod - update pod affinity of a wq for CPU hot[un]plug diff --git a/mm/oom_kill.c b/mm/oom_kill.c index 601ee56cc7d7..81b79f82f601 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -1141,6 +1141,7 @@ int oom_type_notifier_call(unsigned int type, struct oom_control *oc) return blocking_notifier_call_chain(&oom_type_notify_list, type, NULL); } +EXPORT_SYMBOL_GPL(oom_type_notifier_call); #endif /** diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 30665fb33589..31761425d7a8 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -2622,6 +2622,7 @@ struct vm_struct *__get_vm_area_caller(unsigned long size, unsigned long flags, return __get_vm_area_node(size, 1, PAGE_SHIFT, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, caller); } +EXPORT_SYMBOL(__get_vm_area_caller); /** * get_vm_area - reserve a contiguous kernel virtual area @@ -3362,6 +3363,7 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align, return NULL; } +EXPORT_SYMBOL(__vmalloc_node_range); /** * __vmalloc_node - allocate virtually contiguous memory -- 2.17.1

2 1

[PATCH OLK-6.6 v4] ascend: export interfaces required by ascend drivers
by Yuan Can 20 Dec '23

20 Dec '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8OX3Q CVE: NA -------------------------------- Export oom_type_notifier_call and __get_vm_area_caller. Export pm_autosleep_set_state and __vmalloc_node_range. Export alloc_workqueue_attrs, free_workqueue_attrs and apply_workqueue_attrs. Export find_get_task_by_vpid. Signed-off-by: Yuan Can <yuancan(a)huawei.com> --- kernel/pid.c | 1 + kernel/power/autosleep.c | 1 + kernel/workqueue.c | 3 +++ mm/oom_kill.c | 1 + mm/vmalloc.c | 2 ++ 5 files changed, 8 insertions(+) diff --git a/kernel/pid.c b/kernel/pid.c index 6500ef956f2f..383abde0c208 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -437,6 +437,7 @@ struct task_struct *find_get_task_by_vpid(pid_t nr) return task; } +EXPORT_SYMBOL_GPL(find_get_task_by_vpid); struct pid *get_task_pid(struct task_struct *task, enum pid_type type) { diff --git a/kernel/power/autosleep.c b/kernel/power/autosleep.c index b29c8aca7486..80ba474daa40 100644 --- a/kernel/power/autosleep.c +++ b/kernel/power/autosleep.c @@ -113,6 +113,7 @@ int pm_autosleep_set_state(suspend_state_t state) mutex_unlock(&autosleep_lock); return 0; } +EXPORT_SYMBOL_GPL(pm_autosleep_set_state); int __init pm_autosleep_init(void) { diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 0f682da96e1c..3eabf97c4e9a 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -3713,6 +3713,7 @@ void free_workqueue_attrs(struct workqueue_attrs *attrs) kfree(attrs); } } +EXPORT_SYMBOL_GPL(free_workqueue_attrs); /** * alloc_workqueue_attrs - allocate a workqueue_attrs @@ -3741,6 +3742,7 @@ struct workqueue_attrs *alloc_workqueue_attrs(void) free_workqueue_attrs(attrs); return NULL; } +EXPORT_SYMBOL_GPL(alloc_workqueue_attrs); static void copy_workqueue_attrs(struct workqueue_attrs *to, const struct workqueue_attrs *from) @@ -4482,6 +4484,7 @@ int apply_workqueue_attrs(struct workqueue_struct *wq, return ret; } +EXPORT_SYMBOL_GPL(apply_workqueue_attrs); /** * wq_update_pod - update pod affinity of a wq for CPU hot[un]plug diff --git a/mm/oom_kill.c b/mm/oom_kill.c index 601ee56cc7d7..81b79f82f601 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -1141,6 +1141,7 @@ int oom_type_notifier_call(unsigned int type, struct oom_control *oc) return blocking_notifier_call_chain(&oom_type_notify_list, type, NULL); } +EXPORT_SYMBOL_GPL(oom_type_notifier_call); #endif /** diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 30665fb33589..719539b32488 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -2622,6 +2622,7 @@ struct vm_struct *__get_vm_area_caller(unsigned long size, unsigned long flags, return __get_vm_area_node(size, 1, PAGE_SHIFT, flags, start, end, NUMA_NO_NODE, GFP_KERNEL, caller); } +EXPORT_SYMBOL(__get_vm_area_caller); /** * get_vm_area - reserve a contiguous kernel virtual area @@ -3362,6 +3363,7 @@ void *__vmalloc_node_range(unsigned long size, unsigned long align, return NULL; } +EXPORT_SYMBOL_GPL(__vmalloc_node_range); /** * __vmalloc_node - allocate virtually contiguous memory -- 2.17.1

2 1

[PATCH OLK-6.6 0/1] Support priority load balance for qos scheduler
by Song Zhang 20 Dec '23

20 Dec '23

Song Zhang (1): sched: Introduce priority load balance for qos scheduler arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + init/Kconfig | 9 +++ kernel/sched/core.c | 3 + kernel/sched/fair.c | 90 +++++++++++++++++++++++++- kernel/sched/sched.h | 3 + 6 files changed, 106 insertions(+), 1 deletion(-) -- 2.33.0

2 2

[PATCH V3 OLK-6.6 0/3] sync smmu patches for olk-6.6
by Zhang Zekun 19 Dec '23

19 Dec '23

Make the rache depot scale better. The origin depot size is fixed, may cause problem in some condition. Add another patch to fix smmu problem. v2->v3: - Add a smmu bugfix patch v1->v2: - Change the commit messge Robin Murphy (2): iommu/iova: Make the rcache depot scale better iommu/iova: Manage the depot list size Wenkai Lin (1): iommu/arm-smmu-v3: disable stall for quiet_cd drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 3 + drivers/iommu/iova.c | 95 ++++++++++++++------- 2 files changed, 68 insertions(+), 30 deletions(-) -- 2.17.1

2 4

[PATCH OLK-6.6 0/8] adapt kabi reserve patch to OLK6.6
by Lin Yujun 19 Dec '23

19 Dec '23

*** BLURB HERE *** Chen Jiahao (2): KABI: reserve space for several i2c structures KABI: reserve space for struct input_dev Lin Ruizhe (6): bootparam: Add kabi_reserve in bootparam interrupt: Add kabi_reserve in interrupt.h irq: Add kabi_reserve in irq irq_desc: Add kabi_reserve in irq_desc irqdomain: Add kabi_reserve in irqdomain msi: Add kabi_reserve in msi.h arch/x86/include/uapi/asm/bootparam.h | 2 ++ include/linux/i2c.h | 11 +++++++++++ include/linux/input.h | 4 ++++ include/linux/interrupt.h | 4 ++++ include/linux/irq.h | 6 ++++++ include/linux/irqdesc.h | 3 ++- include/linux/irqdomain.h | 2 ++ include/linux/msi.h | 3 +++ 8 files changed, 34 insertions(+), 1 deletion(-) -- 2.34.1

2 9

[PATCH OLK-6.6 v2 0/5] Terrace Service Acceleration
by Zhengchao Shao 19 Dec '23

19 Dec '23

Terrace Service Acceleration. Liu Jian (3): bpf: Add bpf_get_sockops_uid_gid helper function bpf: Add new bpf helper to get SO_ORIGINAL_DST/REPLY_SRC bpf: Access bpf_sock's src_ip4 and sorc_port in BPF_CGROUP_INET_SOCK_RELEASE hook Lu Wei (1): net: core: Add a GID field to struct sock. Wang Yufen (1): bpf, sockmap: Add sk_rmem_alloc check for sockmap include/net/netfilter/nf_conntrack.h | 6 ++ include/net/sock.h | 18 ++++++ include/uapi/linux/bpf.h | 13 +++++ include/uapi/linux/netfilter_ipv4.h | 2 + net/Kconfig | 7 +++ net/core/filter.c | 84 ++++++++++++++++++++++++++++ net/core/skmsg.c | 10 +++- net/core/sock.c | 6 ++ net/ipv4/tcp_bpf.c | 10 ++++ net/netfilter/nf_conntrack_proto.c | 71 +++++++++++++++++++++++ net/socket.c | 8 ++- tools/include/uapi/linux/bpf.h | 13 +++++ 12 files changed, 245 insertions(+), 3 deletions(-) -- 2.34.1

2 6

[PATCH openEuler-1.0-LTS v2] net: Remove acked SYN flag from packet in the transmit queue correctly
by Dong Chenchen 19 Dec '23

19 Dec '23

mainline inclusion from mainline-v6.7-rc4 commit f99cd56230f56c8b6b33713c5be4da5d6766be1f category: bugfix bugzilla: 187254, https://gitee.com/openeuler/kernel/issues/I8PRTS Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- syzkaller report: kernel BUG at net/core/skbuff.c:3452! invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI CPU: 0 PID: 0 Comm: swapper/0 Not tainted 6.7.0-rc4-00009-gbee0e7762ad2-dirty #135 RIP: 0010:skb_copy_and_csum_bits (net/core/skbuff.c:3452) Call Trace: icmp_glue_bits (net/ipv4/icmp.c:357) __ip_append_data.isra.0 (net/ipv4/ip_output.c:1165) ip_append_data (net/ipv4/ip_output.c:1362 net/ipv4/ip_output.c:1341) icmp_push_reply (net/ipv4/icmp.c:370) __icmp_send (./include/net/route.h:252 net/ipv4/icmp.c:772) ip_fragment.constprop.0 (./include/linux/skbuff.h:1234 net/ipv4/ip_output.c:592 net/ipv4/ip_output.c:577) __ip_finish_output (net/ipv4/ip_output.c:311 net/ipv4/ip_output.c:295) ip_output (net/ipv4/ip_output.c:427) __ip_queue_xmit (net/ipv4/ip_output.c:535) __tcp_transmit_skb (net/ipv4/tcp_output.c:1462) __tcp_retransmit_skb (net/ipv4/tcp_output.c:3387) tcp_retransmit_skb (net/ipv4/tcp_output.c:3404) tcp_retransmit_timer (net/ipv4/tcp_timer.c:604) tcp_write_timer (./include/linux/spinlock.h:391 net/ipv4/tcp_timer.c:716) The panic issue was trigered by tcp simultaneous initiation. The initiation process is as follows: TCP A TCP B 1. CLOSED CLOSED 2. SYN-SENT --> <SEQ=100><CTL=SYN> ... 3. SYN-RECEIVED <-- <SEQ=300><CTL=SYN> <-- SYN-SENT 4. ... <SEQ=100><CTL=SYN> --> SYN-RECEIVED 5. SYN-RECEIVED --> <SEQ=100><ACK=301><CTL=SYN,ACK> ... // TCP B: not send challenge ack for ack limit or packet loss // TCP A: close tcp_close tcp_send_fin if (!tskb && tcp_under_memory_pressure(sk)) tskb = skb_rb_last(&sk->tcp_rtx_queue); //pick SYN_ACK packet TCP_SKB_CB(tskb)->tcp_flags |= TCPHDR_FIN; // set FIN flag 6. FIN_WAIT_1 --> <SEQ=100><ACK=301><END_SEQ=102><CTL=SYN,FIN,ACK> ... // TCP B: send challenge ack to SYN_FIN_ACK 7. ... <SEQ=301><ACK=101><CTL=ACK> <-- SYN-RECEIVED //challenge ack // TCP A: <SND.UNA=101> 8. FIN_WAIT_1 --> <SEQ=101><ACK=301><END_SEQ=102><CTL=SYN,FIN,ACK> ... // retransmit panic __tcp_retransmit_skb //skb->len=0 tcp_trim_head len = tp->snd_una - TCP_SKB_CB(skb)->seq // len=101-100 __pskb_trim_head skb->data_len -= len // skb->len=-1, wrap around ... ... ip_fragment icmp_glue_bits //BUG_ON If we use tcp_trim_head() to remove acked SYN from packet that contains data or other flags, skb->len will be incorrectly decremented. We can remove SYN flag that has been acked from rtx_queue earlier than tcp_trim_head(), which can fix the problem mentioned above. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Co-developed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> Link: https://lore.kernel.org/r/20231210020200.1539875-1-dongchenchen2@huawei.com Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- net/ipv4/tcp_output.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 8cd673d75f68..90e0041ae9f7 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2930,7 +2930,13 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs) if (skb_still_in_host_queue(sk, skb)) return -EBUSY; +start: if (before(TCP_SKB_CB(skb)->seq, tp->snd_una)) { + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_SYN; + TCP_SKB_CB(skb)->seq++; + goto start; + } if (unlikely(before(TCP_SKB_CB(skb)->end_seq, tp->snd_una))) { WARN_ON_ONCE(1); return -EINVAL; -- 2.25.1

2 1

[PATCH v3 OLK-6.6 0/2] Make the rcache depot scale better
by Zhang Zekun 19 Dec '23

19 Dec '23

Make the rache depot scale better. The origin depot size is fixed, may cause problem in some condition. Robin Murphy (2): [Backport] iommu/iova: Make the rcache depot scale better [Backport] iommu/iova: Manage the depot list size drivers/iommu/iova.c | 95 ++++++++++++++++++++++++++++++-------------- 1 file changed, 65 insertions(+), 30 deletions(-) -- 2.17.1

2 3

[PATCH OLK-6.6 0/5] Terrace Service Acceleration
by Zhengchao Shao 19 Dec '23

19 Dec '23

Terrace Service Acceleration. Liu Jian (3): bpf: Add bpf_get_sockops_uid_gid helper function bpf: Add new bpf helper to get SO_ORIGINAL_DST/REPLY_SRC bpf: Access bpf_sock's src_ip4 and sorc_port in BPF_CGROUP_INET_SOCK_RELEASE hook Lu Wei (1): net: core: Add a GID field to struct sock. Wang Yufen (1): bpf, sockmap: Add sk_rmem_alloc check for sockmap include/net/netfilter/nf_conntrack.h | 6 ++ include/net/sock.h | 18 ++++++ include/uapi/linux/bpf.h | 13 +++++ include/uapi/linux/netfilter_ipv4.h | 2 + net/Kconfig | 7 +++ net/core/filter.c | 84 ++++++++++++++++++++++++++++ net/core/skmsg.c | 10 +++- net/core/sock.c | 6 ++ net/ipv4/tcp_bpf.c | 10 ++++ net/netfilter/nf_conntrack_proto.c | 71 +++++++++++++++++++++++ net/socket.c | 8 ++- tools/include/uapi/linux/bpf.h | 13 +++++ 12 files changed, 245 insertions(+), 3 deletions(-) -- 2.34.1

2 6

[PATCH OLK-6.6 v4] cache: Workaround HiSilicon Linxicore DC CVAU
by Yuan Can 19 Dec '23

19 Dec '23

From: Weilong Chen <chenweilong(a)huawei.com> ascend inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8OX47 CVE: NA ------------------------------------- Linxicore's L1/L2 cache is inclusive, and the data is consistent. Any change of L1 does not require DC operation to brush CL in L1 to L2. It's safe that don't clean data cache by address to point of unification. Some of the Linxicore report wrong value to kernel. Without reporting this IDC featrue, the kernel needs to flush icache as well as dcache, causes performance degradation. Signed-off-by: Weilong Chen <chenweilong(a)huawei.com> --- Documentation/arch/arm64/silicon-errata.rst | 2 ++ arch/arm64/Kconfig | 10 +++++++ arch/arm64/include/asm/cache.h | 9 ++++++ arch/arm64/kernel/cpu_errata.c | 33 +++++++++++++++++++++ arch/arm64/tools/cpucaps | 1 + 5 files changed, 55 insertions(+) diff --git a/Documentation/arch/arm64/silicon-errata.rst b/Documentation/arch/arm64/silicon-errata.rst index bb62e0dbd546..30d3091452f9 100644 --- a/Documentation/arch/arm64/silicon-errata.rst +++ b/Documentation/arch/arm64/silicon-errata.rst @@ -205,6 +205,8 @@ stable kernels. +----------------+-----------------+-----------------+-----------------------------+ | Hisilicon | LINXICORE9100 | #162100125 | HISILICON_ERRATUM_162100125 | +----------------+-----------------+-----------------+-----------------------------+ +| Hisilicon | LINXICORE9100 | #1980005 | HISILICON_ERRATUM_1980005 | ++----------------+-----------------+-----------------+-----------------------------+ +----------------+-----------------+-----------------+-----------------------------+ | Qualcomm Tech. | Kryo/Falkor v1 | E1003 | QCOM_FALKOR_ERRATUM_1003 | +----------------+-----------------+-----------------+-----------------------------+ diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index 73333648a98c..2aca373a7038 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -1170,6 +1170,16 @@ config HISILICON_ERRATUM_162100125 If unsure, say Y. +config HISILICON_ERRATUM_1980005 + bool "Hisilicon erratum 1980005: IDC support" + default n + help + Some of the Hisilicon linxicore support idc but report wrong value to + the kernel, which result to the unnecessary cache flush, leading to + performance degradation. + + If unsure, say N. + config QCOM_FALKOR_ERRATUM_1003 bool "Falkor E1003: Incorrect translation due to ASID change" default y diff --git a/arch/arm64/include/asm/cache.h b/arch/arm64/include/asm/cache.h index ceb368d33bf4..1613779be63a 100644 --- a/arch/arm64/include/asm/cache.h +++ b/arch/arm64/include/asm/cache.h @@ -112,6 +112,15 @@ int cache_line_size(void); static inline u32 __attribute_const__ read_cpuid_effective_cachetype(void) { u32 ctr = read_cpuid_cachetype(); +#ifdef CONFIG_HISILICON_ERRATUM_1980005 + static const struct midr_range idc_support_list[] = { + MIDR_ALL_VERSIONS(MIDR_HISI_TSV110), + MIDR_REV(MIDR_HISI_LINXICORE9100, 1, 0), + { /* sentinel */ } + }; + if (is_midr_in_range_list(read_cpuid_id(), idc_support_list)) + ctr |= BIT(CTR_EL0_IDC_SHIFT); +#endif if (!(ctr & BIT(CTR_EL0_IDC_SHIFT))) { u64 clidr = read_sysreg(clidr_el1); diff --git a/arch/arm64/kernel/cpu_errata.c b/arch/arm64/kernel/cpu_errata.c index e40b76ac4711..a686a96d966a 100644 --- a/arch/arm64/kernel/cpu_errata.c +++ b/arch/arm64/kernel/cpu_errata.c @@ -60,6 +60,30 @@ is_kryo_midr(const struct arm64_cpu_capabilities *entry, int scope) return model == entry->midr_range.model; } +#ifdef CONFIG_HISILICON_ERRATUM_1980005 +static bool +hisilicon_1980005_match(const struct arm64_cpu_capabilities *entry, + int scope) +{ + static const struct midr_range idc_support_list[] = { + MIDR_ALL_VERSIONS(MIDR_HISI_TSV110), + MIDR_REV(MIDR_HISI_LINXICORE9100, 1, 0), + { /* sentinel */ } + }; + + return is_midr_in_range_list(read_cpuid_id(), idc_support_list); +} + +static void +hisilicon_1980005_enable(const struct arm64_cpu_capabilities *__unused) +{ + __set_bit(ARM64_HAS_CACHE_IDC, system_cpucaps); + arm64_ftr_reg_ctrel0.sys_val |= BIT(CTR_EL0_IDC_SHIFT); + arm64_ftr_reg_ctrel0.strict_mask &= ~BIT(CTR_EL0_IDC_SHIFT); + sysreg_clear_set(sctlr_el1, SCTLR_EL1_UCT, 0); +} +#endif + static bool has_mismatched_cache_type(const struct arm64_cpu_capabilities *entry, int scope) @@ -567,6 +591,15 @@ const struct arm64_cpu_capabilities arm64_errata[] = { ERRATA_MIDR_RANGE_LIST(hisilicon_erratum_162100125_cpus), }, #endif +#ifdef CONFIG_HISILICON_ERRATUM_1980005 + { + .desc = "Hisilicon erratum 1980005 (IDC)", + .capability = ARM64_WORKAROUND_HISILICON_1980005, + .matches = hisilicon_1980005_match, + .type = ARM64_CPUCAP_SYSTEM_FEATURE, + .cpu_enable = hisilicon_1980005_enable, + }, +#endif #ifdef CONFIG_QCOM_FALKOR_ERRATUM_1003 { .desc = "Qualcomm Technologies Falkor/Kryo erratum 1003", diff --git a/arch/arm64/tools/cpucaps b/arch/arm64/tools/cpucaps index cb4996edc09f..0b62edb91876 100644 --- a/arch/arm64/tools/cpucaps +++ b/arch/arm64/tools/cpucaps @@ -102,3 +102,4 @@ WORKAROUND_REPEAT_TLBI WORKAROUND_SPECULATIVE_AT WORKAROUND_HISILICON_ERRATUM_162100125 WORKAROUND_HISI_HIP08_RU_PREFETCH +WORKAROUND_HISILICON_1980005 -- 2.17.1

2 1

[PATCH OLK-6.6 0/3] memcg: support OOM priority for memcg
by Jinjiang Tu 19 Dec '23

19 Dec '23

support OOM priority for memcg. Jing Xiangfeng (2): memcg: support priority for oom memcg: Add sysctl memcg_qos_enable Jinjiang Tu (1): memcg: enable CONFIG_MEMCG_OOM_PROIRITY by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/memcontrol.h | 26 +++ init/Kconfig | 14 ++ mm/memcontrol.c | 230 +++++++++++++++++++++++++ mm/oom_kill.c | 59 ++++++- 6 files changed, 326 insertions(+), 5 deletions(-) -- 2.25.1

2 4

[PATCH OLK-6.6 0/3] memcg: support OOM priority for memcg
by Jinjiang Tu 19 Dec '23

19 Dec '23

Support OOM priority for memcg. Jing Xiangfeng (2): memcg: support priority for oom memcg: Add sysctl memcg_qos_enable memcg: enable CONFIG_MEMCG_OOM_PROIRITY by default arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/memcontrol.h | 26 +++ init/Kconfig | 14 ++ mm/memcontrol.c | 230 +++++++++++++++++++++++++ mm/oom_kill.c | 59 ++++++- 6 files changed, 326 insertions(+), 5 deletions(-) -- 2.25.1

2 4

[PATCH OLK-6.6 0/2] Make the rcache depot scale better
by Zhang Zekun 19 Dec '23

19 Dec '23

Make the rache depot scale better. The origin depot size is fixed, may cause problem in some condition. Robin Murphy (2): [Backport] iommu/iova: Make the rcache depot scale better [Backport] iommu/iova: Manage the depot list size drivers/iommu/iova.c | 95 ++++++++++++++++++++++++++++++-------------- 1 file changed, 65 insertions(+), 30 deletions(-) -- 2.17.1

2 3

[PATCH OLK-6.6 0/9] Support dynamic affinity scheduler
by Hui Tang 19 Dec '23

19 Dec '23

patch 1->7: Introduce dynamic affinity for cfs scheduler. patch 8->9: Tiny bugfix. Hui Tang (7): sched: Introduce dynamic affinity for cfs scheduler cpuset: Introduce new interface for scheduler dynamic affinity sched: Adjust wakeup cpu range according CPU util dynamicly sched: Adjust cpu allowed in load balance dynamicly sched: Add statistics for scheduler dynamic affinity sched: Add cmdline for dynamic affinity config: enable CONFIG_QOS_SCHED_DYNAMIC_AFFINITY by default zhangwei123171 (2): sched/fair: Remove invalid cpu selection logic in dynamic affinity sched/fair: Modify idle cpu judgment in dynamic affinity arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + fs/proc/base.c | 73 +++++++++++ include/linux/sched.h | 22 ++++ include/linux/sched/sysctl.h | 3 + init/Kconfig | 10 ++ init/init_task.c | 3 + kernel/cgroup/cpuset.c | 153 +++++++++++++++++++++- kernel/fork.c | 16 +++ kernel/sched/core.c | 98 ++++++++++++++ kernel/sched/debug.c | 6 + kernel/sched/fair.c | 170 +++++++++++++++++++++++++ kernel/sched/features.h | 7 + kernel/sysctl.c | 11 ++ 14 files changed, 573 insertions(+), 1 deletion(-) -- 2.34.1

2 10

[PATCH 1/2] [Backport] iommu/iova: Make the rcache depot scale better
by Zhang Zekun 19 Dec '23

19 Dec '23

From: Robin Murphy <robin.murphy(a)arm.com> mainline inclusion from mainline-v6.6-rc4 commit 911aa1245da83ff5e76d33bb612d8b5a3f2ec4a5 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8PJ8L Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------------- The algorithm in the original paper specifies the storage of full magazines in the depot as an unbounded list rather than a fixed-size array. It turns out to be pretty straightforward to do this in our implementation with no significant loss of efficiency. This allows the depot to scale up to the working set sizes of larger systems, while also potentially saving some memory on smaller ones too. Since this involves touching struct iova_magazine with the requisite care, we may as well reinforce the comment with a proper assertion too. Reviewed-by: John Garry <john.g.garry(a)oracle.com> Reviewed-by: Jerry Snitselaar <jsnitsel(a)redhat.com> Signed-off-by: Robin Murphy <robin.murphy(a)arm.com> Link: https://lore.kernel.org/r/f597aa72fc3e1d315bc4574af0ce0ebe5c31cd22.16945355… Signed-off-by: Joerg Roedel <jroedel(a)suse.de> Signed-off-by: Zhang Zekun <zhangzekun11(a)huawei.com> --- drivers/iommu/iova.c | 65 ++++++++++++++++++++++++-------------------- 1 file changed, 35 insertions(+), 30 deletions(-) diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c index 10b964600948..dd2309e9a6c5 100644 --- a/drivers/iommu/iova.c +++ b/drivers/iommu/iova.c @@ -622,15 +622,19 @@ EXPORT_SYMBOL_GPL(reserve_iova); /* * As kmalloc's buffer size is fixed to power of 2, 127 is chosen to * assure size of 'iova_magazine' to be 1024 bytes, so that no memory - * will be wasted. + * will be wasted. Since only full magazines are inserted into the depot, + * we don't need to waste PFN capacity on a separate list head either. */ #define IOVA_MAG_SIZE 127 -#define MAX_GLOBAL_MAGS 32 /* magazines per bin */ struct iova_magazine { - unsigned long size; + union { + unsigned long size; + struct iova_magazine *next; + }; unsigned long pfns[IOVA_MAG_SIZE]; }; +static_assert(!(sizeof(struct iova_magazine) & (sizeof(struct iova_magazine) - 1))); struct iova_cpu_rcache { spinlock_t lock; @@ -640,8 +644,7 @@ struct iova_cpu_rcache { struct iova_rcache { spinlock_t lock; - unsigned long depot_size; - struct iova_magazine *depot[MAX_GLOBAL_MAGS]; + struct iova_magazine *depot; struct iova_cpu_rcache __percpu *cpu_rcaches; }; @@ -717,6 +720,21 @@ static void iova_magazine_push(struct iova_magazine *mag, unsigned long pfn) mag->pfns[mag->size++] = pfn; } +static struct iova_magazine *iova_depot_pop(struct iova_rcache *rcache) +{ + struct iova_magazine *mag = rcache->depot; + + rcache->depot = mag->next; + mag->size = IOVA_MAG_SIZE; + return mag; +} + +static void iova_depot_push(struct iova_rcache *rcache, struct iova_magazine *mag) +{ + mag->next = rcache->depot; + rcache->depot = mag; +} + int iova_domain_init_rcaches(struct iova_domain *iovad) { unsigned int cpu; @@ -734,7 +752,6 @@ int iova_domain_init_rcaches(struct iova_domain *iovad) rcache = &iovad->rcaches[i]; spin_lock_init(&rcache->lock); - rcache->depot_size = 0; rcache->cpu_rcaches = __alloc_percpu(sizeof(*cpu_rcache), cache_line_size()); if (!rcache->cpu_rcaches) { @@ -776,7 +793,6 @@ static bool __iova_rcache_insert(struct iova_domain *iovad, struct iova_rcache *rcache, unsigned long iova_pfn) { - struct iova_magazine *mag_to_free = NULL; struct iova_cpu_rcache *cpu_rcache; bool can_insert = false; unsigned long flags; @@ -794,12 +810,7 @@ static bool __iova_rcache_insert(struct iova_domain *iovad, if (new_mag) { spin_lock(&rcache->lock); - if (rcache->depot_size < MAX_GLOBAL_MAGS) { - rcache->depot[rcache->depot_size++] = - cpu_rcache->loaded; - } else { - mag_to_free = cpu_rcache->loaded; - } + iova_depot_push(rcache, cpu_rcache->loaded); spin_unlock(&rcache->lock); cpu_rcache->loaded = new_mag; @@ -812,11 +823,6 @@ static bool __iova_rcache_insert(struct iova_domain *iovad, spin_unlock_irqrestore(&cpu_rcache->lock, flags); - if (mag_to_free) { - iova_magazine_free_pfns(mag_to_free, iovad); - iova_magazine_free(mag_to_free); - } - return can_insert; } @@ -854,9 +860,9 @@ static unsigned long __iova_rcache_get(struct iova_rcache *rcache, has_pfn = true; } else { spin_lock(&rcache->lock); - if (rcache->depot_size > 0) { + if (rcache->depot) { iova_magazine_free(cpu_rcache->loaded); - cpu_rcache->loaded = rcache->depot[--rcache->depot_size]; + cpu_rcache->loaded = iova_depot_pop(rcache); has_pfn = true; } spin_unlock(&rcache->lock); @@ -895,9 +901,8 @@ static void free_iova_rcaches(struct iova_domain *iovad) struct iova_rcache *rcache; struct iova_cpu_rcache *cpu_rcache; unsigned int cpu; - int i, j; - for (i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) { + for (int i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) { rcache = &iovad->rcaches[i]; if (!rcache->cpu_rcaches) break; @@ -907,8 +912,8 @@ static void free_iova_rcaches(struct iova_domain *iovad) iova_magazine_free(cpu_rcache->prev); } free_percpu(rcache->cpu_rcaches); - for (j = 0; j < rcache->depot_size; ++j) - iova_magazine_free(rcache->depot[j]); + while (rcache->depot) + iova_magazine_free(iova_depot_pop(rcache)); } kfree(iovad->rcaches); @@ -942,16 +947,16 @@ static void free_global_cached_iovas(struct iova_domain *iovad) { struct iova_rcache *rcache; unsigned long flags; - int i, j; - for (i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) { + for (int i = 0; i < IOVA_RANGE_CACHE_MAX_SIZE; ++i) { rcache = &iovad->rcaches[i]; spin_lock_irqsave(&rcache->lock, flags); - for (j = 0; j < rcache->depot_size; ++j) { - iova_magazine_free_pfns(rcache->depot[j], iovad); - iova_magazine_free(rcache->depot[j]); + while (rcache->depot) { + struct iova_magazine *mag = iova_depot_pop(rcache); + + iova_magazine_free_pfns(mag, iovad); + iova_magazine_free(mag); } - rcache->depot_size = 0; spin_unlock_irqrestore(&rcache->lock, flags); } } -- 2.17.1

1 1

[PATCH OLK-6.6 v3 0/8] kaslr: ppc64: Introduce KASLR for PPC64
by GUO Zihua 19 Dec '23

19 Dec '23

This patchset introduces KASLR for PowerPC64 chips. v3: Fixed one more commit message. v2: Replaced bugzilla in commit messages. GUO Zihua (2): powerpc/fsl_booke/kaslr: Provide correct r5 value for relocated kernel powerpc/fsl_booke/kaslr: Fix preserved memory size for int-vectors issue Jason Yan (6): powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and kaslr_early_init() powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper powerpc/fsl_booke/64: implement KASLR for fsl_booke64 powerpc/fsl_booke/64: do not clear the BSS for the second pass powerpc/fsl_booke/64: clear the original kernel if randomized powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst and add 64bit part Documentation/powerpc/index.rst | 2 +- .../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++- arch/powerpc/Kconfig | 5 +- arch/powerpc/kernel/exceptions-64e.S | 27 ++++++ arch/powerpc/kernel/head_64.S | 22 +++++ arch/powerpc/kernel/prom.c | 8 +- arch/powerpc/kernel/setup_64.c | 3 + arch/powerpc/mm/mmu_decl.h | 25 ++--- arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------ 9 files changed, 169 insertions(+), 49 deletions(-) rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%) -- 2.34.1

2 9

[PATCH OLK-6.6 00/19] Support basic functions of kernel livepatch
by Zheng Yejian 19 Dec '23

19 Dec '23

Port patches from v5.10 to support basic functions of kernel livepatch on arch x86_64/arm64/arm32/ppc32. Zheng Yejian (19): livepatch/core: Allow implementation without ftrace livepatch/core: Reuse common codes in the solution without ftrace Revert "x86/insn: Make insn_complete() static" livepatch/x86: Support livepatch without ftrace livepatch/core: Disable support for replacing livepatch/core: Restrict livepatch patched/unpatched when plant kprobe livepatch/core: Support load and unload hooks livepatch: samples: Adapt livepatch-sample for solution without ftrace livepatch/core: Support jump_label livepatch: Fix crash when access the global variable in hook livepatch: Fix patching functions which have static_call livepatch/core: Avoid conflict with static {call,key} livepatch/arm64: Support livepatch without ftrace livepatch/core: Revert module_enable_ro and module_disable_ro livepatch: Enable livepatch configs in openeuler_defconfig arm/module: Use plt section indices for relocations livepatch/core: Add support for arm for klp relocation livepatch/arm: Support livepatch without ftrace livepatch/ppc32: Support livepatch without ftrace Documentation/filesystems/proc.rst | 2 +- arch/arm/Kconfig | 3 + arch/arm/include/asm/livepatch.h | 51 + arch/arm/include/asm/module.h | 4 +- arch/arm/kernel/Makefile | 1 + arch/arm/kernel/ftrace.c | 4 +- arch/arm/kernel/livepatch.c | 256 ++++ arch/arm/kernel/module-plts.c | 22 +- arch/arm/kernel/module.c | 4 +- arch/arm64/Kconfig | 3 + arch/arm64/configs/openeuler_defconfig | 12 + arch/arm64/include/asm/livepatch.h | 48 + arch/arm64/kernel/Makefile | 1 + arch/arm64/kernel/livepatch.c | 225 ++++ arch/powerpc/Kconfig | 5 +- arch/powerpc/include/asm/livepatch.h | 24 + arch/powerpc/kernel/Makefile | 1 + arch/powerpc/kernel/livepatch.c | 282 +++++ arch/powerpc/kernel/livepatch_32.c | 124 ++ arch/s390/Kconfig | 2 +- arch/s390/configs/debug_defconfig | 1 + arch/s390/configs/defconfig | 1 + arch/x86/Kconfig | 3 +- arch/x86/configs/openeuler_defconfig | 12 +- arch/x86/include/asm/insn.h | 7 + arch/x86/include/asm/livepatch.h | 33 + arch/x86/kernel/Makefile | 1 + arch/x86/kernel/livepatch.c | 289 +++++ arch/x86/kernel/module.c | 2 +- arch/x86/lib/insn.c | 7 - include/linux/jump_label.h | 10 + include/linux/livepatch.h | 99 +- include/linux/livepatch_sched.h | 6 +- include/linux/module.h | 33 + include/linux/moduleloader.h | 4 +- include/linux/static_call.h | 6 + kernel/jump_label.c | 22 + kernel/livepatch/Kconfig | 78 +- kernel/livepatch/Makefile | 3 +- kernel/livepatch/core.c | 1242 +++++++++++++++++++- kernel/livepatch/core.h | 4 + kernel/module/main.c | 10 +- kernel/module/strict_rwx.c | 17 + kernel/static_call_inline.c | 20 + lib/Kconfig.debug | 2 +- samples/livepatch/Makefile | 2 + samples/livepatch/livepatch-sample.c | 36 + tools/arch/x86/include/asm/insn.h | 7 + tools/arch/x86/lib/insn.c | 7 - tools/testing/selftests/bpf/config.aarch64 | 1 + tools/testing/selftests/bpf/config.s390x | 1 + tools/testing/selftests/livepatch/README | 1 + tools/testing/selftests/livepatch/config | 1 + 53 files changed, 2985 insertions(+), 57 deletions(-) create mode 100644 arch/arm/include/asm/livepatch.h create mode 100644 arch/arm/kernel/livepatch.c create mode 100644 arch/arm64/include/asm/livepatch.h create mode 100644 arch/arm64/kernel/livepatch.c create mode 100644 arch/powerpc/kernel/livepatch.c create mode 100644 arch/powerpc/kernel/livepatch_32.c create mode 100644 arch/x86/include/asm/livepatch.h create mode 100644 arch/x86/kernel/livepatch.c -- 2.25.1

2 20

[PATCH OLK-6.6 v2 0/8] kaslr: ppc64: Introduce KASLR for PPC64
by GUO Zihua 19 Dec '23

19 Dec '23

This patchset introduces KASLR for PowerPC64 chips. v2: Replaced bugzilla in commit messages. GUO Zihua (2): powerpc/fsl_booke/kaslr: Provide correct r5 value for relocated kernel powerpc/fsl_booke/kaslr: Fix preserved memory size for int-vectors issue Jason Yan (6): powerpc/fsl_booke/kaslr: refactor kaslr_legal_offset() and kaslr_early_init() powerpc/fsl_booke/64: introduce reloc_kernel_entry() helper powerpc/fsl_booke/64: implement KASLR for fsl_booke64 powerpc/fsl_booke/64: do not clear the BSS for the second pass powerpc/fsl_booke/64: clear the original kernel if randomized powerpc/fsl_booke/kaslr: rename kaslr-booke32.rst to kaslr-booke.rst and add 64bit part Documentation/powerpc/index.rst | 2 +- .../{kaslr-booke32.rst => kaslr-booke.rst} | 35 ++++++- arch/powerpc/Kconfig | 5 +- arch/powerpc/kernel/exceptions-64e.S | 27 ++++++ arch/powerpc/kernel/head_64.S | 22 +++++ arch/powerpc/kernel/prom.c | 8 +- arch/powerpc/kernel/setup_64.c | 3 + arch/powerpc/mm/mmu_decl.h | 25 ++--- arch/powerpc/mm/nohash/kaslr_booke.c | 91 +++++++++++++------ 9 files changed, 169 insertions(+), 49 deletions(-) rename Documentation/powerpc/{kaslr-booke32.rst => kaslr-booke.rst} (59%) -- 2.34.1

2 9

[PATCH OLK-6.6 v6 0/3] add redis sockmap sample code
by Zhengchao Shao 19 Dec '23

19 Dec '23

add redis sockmap sample code. Liu Jian (3): net: let sockops can use bpf_get_current_comm() net: add local_skb parameter to identify local tcp connection tools: add sample sockmap code for redis include/linux/filter.h | 3 + include/linux/skbuff.h | 5 + include/uapi/linux/bpf.h | 1 + net/Kconfig | 6 + net/core/filter.c | 18 ++ net/ipv4/tcp_input.c | 6 +- net/ipv4/tcp_output.c | 9 + tools/include/uapi/linux/bpf.h | 1 + tools/netacc/Makefile | 24 +++ tools/netacc/bpf_sockmap.h | 167 +++++++++++++++++++ tools/netacc/net-acc | 45 +++++ tools/netacc/netacc.c | 296 +++++++++++++++++++++++++++++++++ tools/netacc/netaccsockmap.c | 160 ++++++++++++++++++ 13 files changed, 740 insertions(+), 1 deletion(-) create mode 100644 tools/netacc/Makefile create mode 100644 tools/netacc/bpf_sockmap.h create mode 100755 tools/netacc/net-acc create mode 100644 tools/netacc/netacc.c create mode 100644 tools/netacc/netaccsockmap.c -- 2.34.1

2 4

[PATCH OLK-5.10] tick/broadcast-hrtimer: Prevent the timer device on broadcast duty CPU from being disabled
by Yu Liao 19 Dec '23

19 Dec '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8PL17 CVE: NA ---------------------------------------- It was found that running the LTP hotplug stress test on a aarch64 system could produce rcu_sched stall warnings. The issue is the following: CPU1 (owns the broadcast hrtimer) CPU2 tick_broadcast_enter() //shut down local timer device ... tick_broadcast_exit() //exits with tick_broadcast_force_mask set, timer device remains disabled initiates offlining of CPU1 take_cpu_down() //CPU1 shuts down and does not send broadcast IPI anymore takedown_cpu() hotplug_cpu__broadcast_tick_pull() //move broadcast hrtimer to this CPU clockevents_program_event() bc_set_next() hrtimer_start() //does not call hrtimer_reprogram() to program timer device if expires equals dev->next_event, so the timer device remains disabled. CPU2 takes over the broadcast duty but local timer device is disabled, causing many CPUs to become stuck. Fix this by calling tick_program_event() to reprogram the local timer device in this scenario. Signed-off-by: Yu Liao <liaoyu15(a)huawei.com> --- kernel/time/tick-broadcast-hrtimer.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/kernel/time/tick-broadcast-hrtimer.c b/kernel/time/tick-broadcast-hrtimer.c index b5a65e212df2..9cc9e6703694 100644 --- a/kernel/time/tick-broadcast-hrtimer.c +++ b/kernel/time/tick-broadcast-hrtimer.c @@ -42,6 +42,8 @@ static int bc_shutdown(struct clock_event_device *evt) */ static int bc_set_next(ktime_t expires, struct clock_event_device *bc) { + ktime_t next_event; + /* * This is called either from enter/exit idle code or from the * broadcast handler. In all cases tick_broadcast_lock is held. @@ -62,6 +64,19 @@ static int bc_set_next(ktime_t expires, struct clock_event_device *bc) * hrtimer_start() can call into tracing. */ RCU_NONIDLE( { + next_event = this_cpu_ptr(&tick_cpu_device)->evtdev->next_event; + + /* + * This can be called from CPU offline operation to move broadcast + * assignment. If tick_broadcast_force_mask is set, the CPU local + * timer device may be disabled. And hrtimer_reprogram() will not + * called if the timer is not the first expiring timer. Reprogram + * the cpu local timer device to ensure we can take over the + * broadcast duty. + */ + if (tick_check_broadcast_expired() && expires >= next_event) + tick_program_event(next_event, 1); + hrtimer_start(&bctimer, expires, HRTIMER_MODE_ABS_PINNED_HARD); /* * The core tick broadcast mode expects bc->bound_on to be set -- 2.33.0

2 1

[PATCH OLK-6.6 0/1] Support priority load balance for qos scheduler
by Song Zhang 19 Dec '23

19 Dec '23

Song Zhang (1): sched: Introduce priority load balance for qos scheduler arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/sched/sysctl.h | 4 ++ init/Kconfig | 9 +++ kernel/sched/core.c | 3 + kernel/sched/fair.c | 90 +++++++++++++++++++++++++- kernel/sched/sched.h | 3 + 7 files changed, 110 insertions(+), 1 deletion(-) -- 2.33.0

2 2

[PATCH OLK-6.6 0/1] Support priority load balance for qos scheduler
by Song Zhang 19 Dec '23

19 Dec '23

Song Zhang (1): sched: Introduce priority load balance for qos scheduler arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/sched/sysctl.h | 4 ++ init/Kconfig | 9 +++ kernel/sched/core.c | 3 + kernel/sched/fair.c | 90 +++++++++++++++++++++++++- kernel/sched/sched.h | 3 + 7 files changed, 110 insertions(+), 1 deletion(-) -- 2.33.0

2 2

[PATCH OLK-6.6 0/3] Revert "compiler: remove CONFIG_OPTIMIZE_INLINING entirely"
by Yuntao Liu 19 Dec '23

19 Dec '23

*** BLURB HERE *** Guo Xuenan (3): Revert "compiler: remove CONFIG_OPTIMIZE_INLINING entirely" disable OPTIMIZE_INLINING by default make OPTIMIZE_INLINING config editable arch/arm64/kvm/sys_regs.h | 5 +++++ arch/x86/configs/i386_defconfig | 1 + arch/x86/configs/x86_64_defconfig | 1 + .../pci/hive_isp_css_include/print_support.h | 4 ++++ include/linux/compiler_types.h | 9 ++++++++- kernel/configs/tiny.config | 1 + lib/Kconfig.debug | 13 +++++++++++++ 7 files changed, 33 insertions(+), 1 deletion(-) -- 2.34.1

2 4

[PATCH OLK-5.10] livepatch/core: Disable support for replacing
by Zheng Yejian 19 Dec '23

19 Dec '23

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8ORHJ -------------------------------- Replacing is an optimization for livepatching based on ftrace, the wo_ftrace solution do not support it. Signed-off-by: Zheng Yejian <zhengyejian1(a)huawei.com> --- kernel/livepatch/core.c | 10 ++++++++++ kernel/livepatch/patch.c | 2 ++ 2 files changed, 12 insertions(+) diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index 8768ec1bddf3..7288456906b6 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -1156,6 +1156,9 @@ static void klp_init_func_early(struct klp_object *obj, kobject_init(&func->kobj, &klp_ktype_func); list_add_tail(&func->node, &obj->func_list); func->func_node = NULL; +#ifdef CONFIG_LIVEPATCH_STOP_MACHINE_CONSISTENCY + func->nop = false; +#endif } static void klp_init_object_early(struct klp_patch *patch, @@ -1166,6 +1169,7 @@ static void klp_init_object_early(struct klp_patch *patch, list_add_tail(&obj->node, &patch->obj_list); #ifdef CONFIG_LIVEPATCH_STOP_MACHINE_CONSISTENCY obj->mod = NULL; + obj->dynamic = false; #endif } @@ -1245,6 +1249,12 @@ static int klp_init_patch(struct klp_patch *patch) if (ret) return ret; +#ifdef CONFIG_LIVEPATCH_STOP_MACHINE_CONSISTENCY + if (patch->replace) { + pr_err("Replacing is not supported\n"); + return -EINVAL; + } +#endif if (patch->replace) { ret = klp_add_nops(patch); if (ret) diff --git a/kernel/livepatch/patch.c b/kernel/livepatch/patch.c index bea6c5d0af94..df0c487a04fd 100644 --- a/kernel/livepatch/patch.c +++ b/kernel/livepatch/patch.c @@ -315,7 +315,9 @@ int klp_patch_object(struct klp_object *obj, bool rollback) if (obj->patched) return 0; + WARN_ON(obj->dynamic); klp_for_each_func(obj, func) { + WARN_ON(func->nop); ret = klp_patch_func(func); if (ret && klp_need_rollback(ret, rollback)) { klp_unpatch_object(obj); -- 2.25.1

2 1

[PATCH OLK-6.6 0/1] Support priority load balance for qos scheduler
by Song Zhang 19 Dec '23

19 Dec '23

Song Zhang (1): sched: Introduce priority load balance for qos scheduler arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/sched/sysctl.h | 4 ++ init/Kconfig | 9 +++ kernel/sched/core.c | 3 + kernel/sched/fair.c | 90 +++++++++++++++++++++++++- kernel/sched/sched.h | 3 + 7 files changed, 110 insertions(+), 1 deletion(-) -- 2.33.0

2 2

[PATCH OLK-6.6 0/1] Support priority load balance for qos scheduler
by Song Zhang 19 Dec '23

19 Dec '23

Song Zhang (1): sched: Introduce priority load balance for qos scheduler arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/sched/sysctl.h | 4 ++ init/Kconfig | 9 +++ kernel/sched/core.c | 3 + kernel/sched/fair.c | 90 +++++++++++++++++++++++++- kernel/sched/sched.h | 3 + 7 files changed, 110 insertions(+), 1 deletion(-) -- 2.33.0

2 2

[PATCH OLK-5.10 0/2] Support node_reclaim_distance adjust
by Hui Tang 19 Dec '23

19 Dec '23

Allow arch to adjust node_reclaim_distance. Hui Tang (2): arm64/numa: Support node_reclaim_distance adjust for arch config: enable COBFIG_ARCH_CUSTOM_NUMA_DISTANCE for arm64 arch/arm64/Kconfig | 13 +++++ arch/arm64/configs/openeuler_defconfig | 2 + arch/arm64/mm/numa.c | 79 ++++++++++++++++++++++++++ 3 files changed, 94 insertions(+) -- 2.34.1

2 3

[PATCH OLK-5.10 0/2] Support node_reclaim_distance adjust
by Hui Tang 19 Dec '23

19 Dec '23

Allow arch to adjust node_reclaim_distance. Hui Tang (2): arm64/numa: Support node_reclaim_distance adjust for arch config: enable COBFIG_ARCH_CUSTOM_NUMA_DISTANCE for arm64 arch/arm64/Kconfig | 13 +++++ arch/arm64/configs/openeuler_defconfig | 2 + arch/arm64/mm/numa.c | 80 ++++++++++++++++++++++++++ 3 files changed, 95 insertions(+) -- 2.34.1

2 3

[PATCH v3 OLK-5.10 0/1] Revert "mm/mempolicy: don't handle MPOL_LOCA
by Wupeng Ma 18 Dec '23

18 Dec '23

From: Ma Wupeng <mawupeng1(a)huawei.com> Revert to bring MPOL_LOCAL back. changelog since v2: - introduce mpol_rebind_preferred changelog since v1: - bugfix in __mpol_equal Ma Wupeng (1): Revert "mm/mempolicy: don't handle MPOL_LOCAL like a fake MPOL_PREFERRED policy" include/uapi/linux/mempolicy.h | 1 + mm/mempolicy.c | 142 ++++++++++++++++++++------------- 2 files changed, 87 insertions(+), 56 deletions(-) -- 2.25.1

2 2

[PATCH v2 OLK-5.10 0/1] Revert "mm/mempolicy: don't handle MPOL_LOCA
by Wupeng Ma 18 Dec '23

18 Dec '23

From: Ma Wupeng <mawupeng1(a)huawei.com> Revert to bring MPOL_LOCAL back. changelog since v1: - bugfix in __mpol_equal Ma Wupeng (1): Revert "mm/mempolicy: don't handle MPOL_LOCAL like a fake MPOL_PREFERRED policy" include/uapi/linux/mempolicy.h | 1 + mm/mempolicy.c | 138 +++++++++++++++++++-------------- 2 files changed, 81 insertions(+), 58 deletions(-) -- 2.25.1

2 2

[PATCH v2 OLK-5.10 0/1] Revert "mm/mempolicy: don't handle MPOL_LOCA
by Wupeng Ma 18 Dec '23

18 Dec '23

From: Ma Wupeng <mawupeng1(a)huawei.com> Revert to bring MPOL_LOCAL back. changelog since v1: - bugfix in __mpol_equal Ma Wupeng (1): Revert "mm/mempolicy: don't handle MPOL_LOCAL like a fake MPOL_PREFERRED policy" include/uapi/linux/mempolicy.h | 1 + mm/mempolicy.c | 138 +++++++++++++++++++-------------- 2 files changed, 81 insertions(+), 58 deletions(-) -- 2.25.1

2 4

[PATCH OLK-5.10 0/1] Revert "mm/mempolicy: don't handle MPOL_LOCA
by Wupeng Ma 18 Dec '23

18 Dec '23

From: Ma Wupeng <mawupeng1(a)huawei.com> Revert to bring MPOL_LOCAL back. Ma Wupeng (1): Revert "mm/mempolicy: don't handle MPOL_LOCAL like a fake MPOL_PREFERRED policy" include/uapi/linux/mempolicy.h | 1 + mm/mempolicy.c | 134 +++++++++++++++++++-------------- 2 files changed, 80 insertions(+), 55 deletions(-) -- 2.25.1

2 2

[PATCH OLK-6.6 0/3] xfs: fix attr inactive problems
by Long Li 18 Dec '23

18 Dec '23

This patch set fix two attr inactive problems: Guo Xuenan (1): xfs: force shutdown xfs when xfs_attr_inactive fails Zhang Yi (2): xfs: factor out __xfs_da3_node_read() xfs: atomic drop extent entries when inactiving attr fs/xfs/libxfs/xfs_da_btree.c | 5 +-- fs/xfs/libxfs/xfs_da_btree.h | 15 ++++++-- fs/xfs/xfs_attr_inactive.c | 66 ++++++++++++++++++++++++++---------- 3 files changed, 65 insertions(+), 21 deletions(-) -- 2.31.1

2 4

[PATCH OLK-5.10] sched: Support custom numa distance for schedluler
by Hui Tang 18 Dec '23

18 Dec '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8PG0C CVE: NA -------------------------------- Support arch redefine far numa distance, which affects the initialization of sched domain, such as SD_WAKE_AFFINE. Signed-off-by: Hui Tang <tanghui20(a)huawei.com> --- arch/arm64/Kconfig | 5 +++++ arch/arm64/include/asm/numa.h | 1 + arch/arm64/mm/numa.c | 40 +++++++++++++++++++++++++++++++++++ kernel/sched/topology.c | 9 +++++++- 4 files changed, 54 insertions(+), 1 deletion(-) diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c5eca65c27fc..88d9862aa139 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -2102,6 +2102,11 @@ config STACKPROTECTOR_PER_TASK def_bool y depends on STACKPROTECTOR && CC_HAVE_STACKPROTECTOR_SYSREG +config ARCH_CUSTOM_NUMA_DISTANCE + bool "Support custom far numa distance" + depends on ARCH_HISI + default n + menuconfig ASCEND_FEATURES bool "Support Ascend Features" depends on ARM64 diff --git a/arch/arm64/include/asm/numa.h b/arch/arm64/include/asm/numa.h index 43bfff72a32f..3af06fa3e969 100644 --- a/arch/arm64/include/asm/numa.h +++ b/arch/arm64/include/asm/numa.h @@ -43,6 +43,7 @@ void __init early_map_cpu_to_node(unsigned int cpu, int nid); void numa_store_cpu_info(unsigned int cpu); void numa_add_cpu(unsigned int cpu); void numa_remove_cpu(unsigned int cpu); +bool arch_numa_distance_is_far(int distance, int numa_levels); #else /* CONFIG_NUMA */ diff --git a/arch/arm64/mm/numa.c b/arch/arm64/mm/numa.c index 480fcfb1b917..871394038ae0 100644 --- a/arch/arm64/mm/numa.c +++ b/arch/arm64/mm/numa.c @@ -619,6 +619,46 @@ static int __init dummy_numa_init(void) return 0; } +#ifdef CONFIG_ARCH_CUSTOM_NUMA_DISTANCE +static int arch_custom_numa_distance_flag = 1; +static int __init arch_custom_numa_distance_flag_setup(char *str) +{ + int val; + + if (kstrtoint(str, 0, &val)) + pr_warn("Unable to set arch_custom_numa_distance\n"); + + if (val != 0 || val != 1) + return 1; + + arch_custom_numa_distance_flag = val; + return 1; +} +early_param("arch_custom_numa_distance_flag=", arch_custom_numa_distance_flag_setup); + +bool arch_numa_distance_is_far(int distance, int numa_levels) +{ + unsigned int model = read_cpuid_id() & MIDR_CPU_MODEL_MASK; + int far_numa_distance = node_reclaim_distance; + + if (arch_custom_numa_distance_flag == 0) + return distance > node_reclaim_distance; + + switch (model) { + case MIDR_HISI_TSV200: + if (nr_node_ids != 4 || numa_levels <= 3) + break; + + far_numa_distance = 40; + break; + default: + break; + } + + return distance > far_numa_distance; +} +#endif + /** * arm64_numa_init() - Initialize NUMA * diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index e7413d6dd75b..e0fd4e68616c 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -1529,6 +1529,12 @@ int sched_max_numa_distance; static int *sched_domains_numa_distance; static struct cpumask ***sched_domains_numa_masks; int __read_mostly node_reclaim_distance = RECLAIM_DISTANCE; + +#ifndef CONFIG_ARCH_CUSTOM_NUMA_DISTANCE +bool arch_numa_distance_is_far(int distance, int numa_levels) { + return distance > node_reclaim_distance; +} +#endif #endif /* @@ -1640,7 +1646,8 @@ sd_init(struct sched_domain_topology_level *tl, sd->flags &= ~SD_PREFER_SIBLING; sd->flags |= SD_SERIALIZE; - if (sched_domains_numa_distance[tl->numa_level] > node_reclaim_distance) { + if (arch_numa_distance_is_far(tl->numa_level, + sched_domains_numa_levels)) { sd->flags &= ~(SD_BALANCE_EXEC | SD_BALANCE_FORK | SD_WAKE_AFFINE); -- 2.34.1

2 1

[PATCH] sched: Support custom numa distance for schedluler
by Hui Tang 18 Dec '23

18 Dec '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8PG0C CVE: NA -------------------------------- Support arch redefine far numa distance, which affects the initialization of sched domain, such as SD_WAKE_AFFINE. Signed-off-by: Hui Tang <tanghui20(a)huawei.com> --- arch/arm64/Kconfig | 5 +++++ arch/arm64/include/asm/numa.h | 1 + arch/arm64/mm/numa.c | 40 +++++++++++++++++++++++++++++++++++ kernel/sched/topology.c | 9 +++++++- 4 files changed, 54 insertions(+), 1 deletion(-) diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c5eca65c27fc..88d9862aa139 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -2102,6 +2102,11 @@ config STACKPROTECTOR_PER_TASK def_bool y depends on STACKPROTECTOR && CC_HAVE_STACKPROTECTOR_SYSREG +config ARCH_CUSTOM_NUMA_DISTANCE + bool "Support custom far numa distance" + depends on ARCH_HISI + default n + menuconfig ASCEND_FEATURES bool "Support Ascend Features" depends on ARM64 diff --git a/arch/arm64/include/asm/numa.h b/arch/arm64/include/asm/numa.h index 43bfff72a32f..3af06fa3e969 100644 --- a/arch/arm64/include/asm/numa.h +++ b/arch/arm64/include/asm/numa.h @@ -43,6 +43,7 @@ void __init early_map_cpu_to_node(unsigned int cpu, int nid); void numa_store_cpu_info(unsigned int cpu); void numa_add_cpu(unsigned int cpu); void numa_remove_cpu(unsigned int cpu); +bool arch_numa_distance_is_far(int distance, int numa_levels); #else /* CONFIG_NUMA */ diff --git a/arch/arm64/mm/numa.c b/arch/arm64/mm/numa.c index 480fcfb1b917..871394038ae0 100644 --- a/arch/arm64/mm/numa.c +++ b/arch/arm64/mm/numa.c @@ -619,6 +619,46 @@ static int __init dummy_numa_init(void) return 0; } +#ifdef CONFIG_ARCH_CUSTOM_NUMA_DISTANCE +static int arch_custom_numa_distance_flag = 1; +static int __init arch_custom_numa_distance_flag_setup(char *str) +{ + int val; + + if (kstrtoint(str, 0, &val)) + pr_warn("Unable to set arch_custom_numa_distance\n"); + + if (val != 0 || val != 1) + return 1; + + arch_custom_numa_distance_flag = val; + return 1; +} +early_param("arch_custom_numa_distance_flag=", arch_custom_numa_distance_flag_setup); + +bool arch_numa_distance_is_far(int distance, int numa_levels) +{ + unsigned int model = read_cpuid_id() & MIDR_CPU_MODEL_MASK; + int far_numa_distance = node_reclaim_distance; + + if (arch_custom_numa_distance_flag == 0) + return distance > node_reclaim_distance; + + switch (model) { + case MIDR_HISI_TSV200: + if (nr_node_ids != 4 || numa_levels <= 3) + break; + + far_numa_distance = 40; + break; + default: + break; + } + + return distance > far_numa_distance; +} +#endif + /** * arm64_numa_init() - Initialize NUMA * diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index e7413d6dd75b..e0fd4e68616c 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -1529,6 +1529,12 @@ int sched_max_numa_distance; static int *sched_domains_numa_distance; static struct cpumask ***sched_domains_numa_masks; int __read_mostly node_reclaim_distance = RECLAIM_DISTANCE; + +#ifndef CONFIG_ARCH_CUSTOM_NUMA_DISTANCE +bool arch_numa_distance_is_far(int distance, int numa_levels) { + return distance > node_reclaim_distance; +} +#endif #endif /* @@ -1640,7 +1646,8 @@ sd_init(struct sched_domain_topology_level *tl, sd->flags &= ~SD_PREFER_SIBLING; sd->flags |= SD_SERIALIZE; - if (sched_domains_numa_distance[tl->numa_level] > node_reclaim_distance) { + if (arch_numa_distance_is_far(tl->numa_level, + sched_domains_numa_levels)) { sd->flags &= ~(SD_BALANCE_EXEC | SD_BALANCE_FORK | SD_WAKE_AFFINE); -- 2.34.1

1 0

[PATCH OLK-5.10] sched: Support custom numa distance for schedluler
by Hui Tang 18 Dec '23

18 Dec '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8PG0C CVE: NA -------------------------------- Support arch redefine far numa distance, which affects the initialization of sched domain, such as SD_WAKE_AFFINE. Signed-off-by: Hui Tang <tanghui20(a)huawei.com> --- arch/arm64/Kconfig | 5 +++++ arch/arm64/include/asm/numa.h | 1 + arch/arm64/mm/numa.c | 40 +++++++++++++++++++++++++++++++++++ kernel/sched/topology.c | 9 +++++++- 4 files changed, 54 insertions(+), 1 deletion(-) diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c5eca65c27fc..88d9862aa139 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -2102,6 +2102,11 @@ config STACKPROTECTOR_PER_TASK def_bool y depends on STACKPROTECTOR && CC_HAVE_STACKPROTECTOR_SYSREG +config ARCH_CUSTOM_NUMA_DISTANCE + bool "Support custom far numa distance" + depends on ARCH_HISI + default n + menuconfig ASCEND_FEATURES bool "Support Ascend Features" depends on ARM64 diff --git a/arch/arm64/include/asm/numa.h b/arch/arm64/include/asm/numa.h index 43bfff72a32f..3af06fa3e969 100644 --- a/arch/arm64/include/asm/numa.h +++ b/arch/arm64/include/asm/numa.h @@ -43,6 +43,7 @@ void __init early_map_cpu_to_node(unsigned int cpu, int nid); void numa_store_cpu_info(unsigned int cpu); void numa_add_cpu(unsigned int cpu); void numa_remove_cpu(unsigned int cpu); +bool arch_numa_distance_is_far(int distance, int numa_levels); #else /* CONFIG_NUMA */ diff --git a/arch/arm64/mm/numa.c b/arch/arm64/mm/numa.c index 480fcfb1b917..871394038ae0 100644 --- a/arch/arm64/mm/numa.c +++ b/arch/arm64/mm/numa.c @@ -619,6 +619,46 @@ static int __init dummy_numa_init(void) return 0; } +#ifdef CONFIG_ARCH_CUSTOM_NUMA_DISTANCE +static int arch_custom_numa_distance_flag = 1; +static int __init arch_custom_numa_distance_flag_setup(char *str) +{ + int val; + + if (kstrtoint(str, 0, &val)) + pr_warn("Unable to set arch_custom_numa_distance\n"); + + if (val != 0 || val != 1) + return 1; + + arch_custom_numa_distance_flag = val; + return 1; +} +early_param("arch_custom_numa_distance_flag=", arch_custom_numa_distance_flag_setup); + +bool arch_numa_distance_is_far(int distance, int numa_levels) +{ + unsigned int model = read_cpuid_id() & MIDR_CPU_MODEL_MASK; + int far_numa_distance = node_reclaim_distance; + + if (arch_custom_numa_distance_flag == 0) + return distance > node_reclaim_distance; + + switch (model) { + case MIDR_HISI_TSV200: + if (nr_node_ids != 4 || numa_levels <= 3) + break; + + far_numa_distance = 40; + break; + default: + break; + } + + return distance > far_numa_distance; +} +#endif + /** * arm64_numa_init() - Initialize NUMA * diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index e7413d6dd75b..fb6f20956ee7 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -1529,6 +1529,12 @@ int sched_max_numa_distance; static int *sched_domains_numa_distance; static struct cpumask ***sched_domains_numa_masks; int __read_mostly node_reclaim_distance = RECLAIM_DISTANCE; + +#ifdef CONFIG_ARCH_CUSTOM_NUMA_DISTANCE +static inline bool arch_numa_distance_is_far(int distance, int numa_levels) { + return distance > node_reclaim_distance; +} +#endif #endif /* @@ -1640,7 +1646,8 @@ sd_init(struct sched_domain_topology_level *tl, sd->flags &= ~SD_PREFER_SIBLING; sd->flags |= SD_SERIALIZE; - if (sched_domains_numa_distance[tl->numa_level] > node_reclaim_distance) { + if (arch_numa_distance_is_far(tl->numa_level, + sched_domains_numa_levels)) { sd->flags &= ~(SD_BALANCE_EXEC | SD_BALANCE_FORK | SD_WAKE_AFFINE); -- 2.34.1

2 1

[PATCH] sched: Support custom numa distance for schedluler
by Hui Tang 18 Dec '23

18 Dec '23

hulk inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8PG0C CVE: NA -------------------------------- Support arch redefine far numa distance, which affects the initialization of sched domain, such as SD_WAKE_AFFINE. Signed-off-by: Hui Tang <tanghui20(a)huawei.com> --- arch/arm64/Kconfig | 5 +++++ arch/arm64/include/asm/numa.h | 1 + arch/arm64/mm/numa.c | 40 +++++++++++++++++++++++++++++++++++ kernel/sched/topology.c | 9 +++++++- 4 files changed, 54 insertions(+), 1 deletion(-) diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c5eca65c27fc..88d9862aa139 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -2102,6 +2102,11 @@ config STACKPROTECTOR_PER_TASK def_bool y depends on STACKPROTECTOR && CC_HAVE_STACKPROTECTOR_SYSREG +config ARCH_CUSTOM_NUMA_DISTANCE + bool "Support custom far numa distance" + depends on ARCH_HISI + default n + menuconfig ASCEND_FEATURES bool "Support Ascend Features" depends on ARM64 diff --git a/arch/arm64/include/asm/numa.h b/arch/arm64/include/asm/numa.h index 43bfff72a32f..3af06fa3e969 100644 --- a/arch/arm64/include/asm/numa.h +++ b/arch/arm64/include/asm/numa.h @@ -43,6 +43,7 @@ void __init early_map_cpu_to_node(unsigned int cpu, int nid); void numa_store_cpu_info(unsigned int cpu); void numa_add_cpu(unsigned int cpu); void numa_remove_cpu(unsigned int cpu); +bool arch_numa_distance_is_far(int distance, int numa_levels); #else /* CONFIG_NUMA */ diff --git a/arch/arm64/mm/numa.c b/arch/arm64/mm/numa.c index 480fcfb1b917..871394038ae0 100644 --- a/arch/arm64/mm/numa.c +++ b/arch/arm64/mm/numa.c @@ -619,6 +619,46 @@ static int __init dummy_numa_init(void) return 0; } +#ifdef CONFIG_ARCH_CUSTOM_NUMA_DISTANCE +static int arch_custom_numa_distance_flag = 1; +static int __init arch_custom_numa_distance_flag_setup(char *str) +{ + int val; + + if (kstrtoint(str, 0, &val)) + pr_warn("Unable to set arch_custom_numa_distance\n"); + + if (val != 0 || val != 1) + return 1; + + arch_custom_numa_distance_flag = val; + return 1; +} +early_param("arch_custom_numa_distance_flag=", arch_custom_numa_distance_flag_setup); + +bool arch_numa_distance_is_far(int distance, int numa_levels) +{ + unsigned int model = read_cpuid_id() & MIDR_CPU_MODEL_MASK; + int far_numa_distance = node_reclaim_distance; + + if (arch_custom_numa_distance_flag == 0) + return distance > node_reclaim_distance; + + switch (model) { + case MIDR_HISI_TSV200: + if (nr_node_ids != 4 || numa_levels <= 3) + break; + + far_numa_distance = 40; + break; + default: + break; + } + + return distance > far_numa_distance; +} +#endif + /** * arm64_numa_init() - Initialize NUMA * diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c index e7413d6dd75b..fb6f20956ee7 100644 --- a/kernel/sched/topology.c +++ b/kernel/sched/topology.c @@ -1529,6 +1529,12 @@ int sched_max_numa_distance; static int *sched_domains_numa_distance; static struct cpumask ***sched_domains_numa_masks; int __read_mostly node_reclaim_distance = RECLAIM_DISTANCE; + +#ifdef CONFIG_ARCH_CUSTOM_NUMA_DISTANCE +static inline bool arch_numa_distance_is_far(int distance, int numa_levels) { + return distance > node_reclaim_distance; +} +#endif #endif /* @@ -1640,7 +1646,8 @@ sd_init(struct sched_domain_topology_level *tl, sd->flags &= ~SD_PREFER_SIBLING; sd->flags |= SD_SERIALIZE; - if (sched_domains_numa_distance[tl->numa_level] > node_reclaim_distance) { + if (arch_numa_distance_is_far(tl->numa_level, + sched_domains_numa_levels)) { sd->flags &= ~(SD_BALANCE_EXEC | SD_BALANCE_FORK | SD_WAKE_AFFINE); -- 2.34.1

1 0

[PATCH OLK-6.6 v5 0/4] add redis sockmap sample code
by Zhengchao Shao 18 Dec '23

18 Dec '23

add redis sockmap sample code. Liu Jian (4): cgroup: make cgroup_bpf_prog_attach work when cgroup2 is not mounted net: let sockops can use bpf_get_current_comm() net: add local_skb parameter to identify local tcp connection tools: add sample sockmap code for redis include/linux/filter.h | 3 + include/linux/skbuff.h | 5 + include/uapi/linux/bpf.h | 1 + kernel/bpf/cgroup.c | 16 ++ net/Kconfig | 5 + net/core/filter.c | 18 ++ net/ipv4/tcp_input.c | 6 +- net/ipv4/tcp_output.c | 9 + tools/include/uapi/linux/bpf.h | 1 + tools/netacc/Makefile | 24 +++ tools/netacc/bpf_sockmap.h | 167 +++++++++++++++++++ tools/netacc/net-acc | 35 ++++ tools/netacc/netacc.c | 296 +++++++++++++++++++++++++++++++++ tools/netacc/netaccsockmap.c | 160 ++++++++++++++++++ 14 files changed, 745 insertions(+), 1 deletion(-) create mode 100644 tools/netacc/Makefile create mode 100644 tools/netacc/bpf_sockmap.h create mode 100755 tools/netacc/net-acc create mode 100644 tools/netacc/netacc.c create mode 100644 tools/netacc/netaccsockmap.c -- 2.34.1

2 5

[PATCH OLK-5.10] livepatch/core: Disable support for replacing
by Zheng Yejian 18 Dec '23

18 Dec '23

Offering: HULK hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I8ORHJ -------------------------------- Replacing is an optimization for livepatching based on ftrace, the wo_ftrace solution do not support it. Signed-off-by: Zheng Yejian <zhengyejian1(a)huawei.com> --- kernel/livepatch/core.c | 10 ++++++++++ kernel/livepatch/patch.c | 2 ++ 2 files changed, 12 insertions(+) diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index 8768ec1bddf3..7288456906b6 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -1156,6 +1156,9 @@ static void klp_init_func_early(struct klp_object *obj, kobject_init(&func->kobj, &klp_ktype_func); list_add_tail(&func->node, &obj->func_list); func->func_node = NULL; +#ifdef CONFIG_LIVEPATCH_STOP_MACHINE_CONSISTENCY + func->nop = false; +#endif } static void klp_init_object_early(struct klp_patch *patch, @@ -1166,6 +1169,7 @@ static void klp_init_object_early(struct klp_patch *patch, list_add_tail(&obj->node, &patch->obj_list); #ifdef CONFIG_LIVEPATCH_STOP_MACHINE_CONSISTENCY obj->mod = NULL; + obj->dynamic = false; #endif } @@ -1245,6 +1249,12 @@ static int klp_init_patch(struct klp_patch *patch) if (ret) return ret; +#ifdef CONFIG_LIVEPATCH_STOP_MACHINE_CONSISTENCY + if (patch->replace) { + pr_err("Replacing is not supported\n"); + return -EINVAL; + } +#endif if (patch->replace) { ret = klp_add_nops(patch); if (ret) diff --git a/kernel/livepatch/patch.c b/kernel/livepatch/patch.c index bea6c5d0af94..df0c487a04fd 100644 --- a/kernel/livepatch/patch.c +++ b/kernel/livepatch/patch.c @@ -315,7 +315,9 @@ int klp_patch_object(struct klp_object *obj, bool rollback) if (obj->patched) return 0; + WARN_ON(obj->dynamic); klp_for_each_func(obj, func) { + WARN_ON(func->nop); ret = klp_patch_func(func); if (ret && klp_need_rollback(ret, rollback)) { klp_unpatch_object(obj); -- 2.25.1

2 1

[PATCH openEuler-1.0-LTS] net: Remove acked SYN flag from packet in the transmit queue correctly
by Dong Chenchen 18 Dec '23

18 Dec '23

mainline inclusion from mainline-v6.7-rc4 commit f99cd56230f56c8b6b33713c5be4da5d6766be1f category: bugfix bugzilla: 187254 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… --------------------------- syzkaller report: kernel BUG at net/core/skbuff.c:3452! invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI CPU: 0 PID: 0 Comm: swapper/0 Not tainted 6.7.0-rc4-00009-gbee0e7762ad2-dirty #135 RIP: 0010:skb_copy_and_csum_bits (net/core/skbuff.c:3452) Call Trace: icmp_glue_bits (net/ipv4/icmp.c:357) __ip_append_data.isra.0 (net/ipv4/ip_output.c:1165) ip_append_data (net/ipv4/ip_output.c:1362 net/ipv4/ip_output.c:1341) icmp_push_reply (net/ipv4/icmp.c:370) __icmp_send (./include/net/route.h:252 net/ipv4/icmp.c:772) ip_fragment.constprop.0 (./include/linux/skbuff.h:1234 net/ipv4/ip_output.c:592 net/ipv4/ip_output.c:577) __ip_finish_output (net/ipv4/ip_output.c:311 net/ipv4/ip_output.c:295) ip_output (net/ipv4/ip_output.c:427) __ip_queue_xmit (net/ipv4/ip_output.c:535) __tcp_transmit_skb (net/ipv4/tcp_output.c:1462) __tcp_retransmit_skb (net/ipv4/tcp_output.c:3387) tcp_retransmit_skb (net/ipv4/tcp_output.c:3404) tcp_retransmit_timer (net/ipv4/tcp_timer.c:604) tcp_write_timer (./include/linux/spinlock.h:391 net/ipv4/tcp_timer.c:716) The panic issue was trigered by tcp simultaneous initiation. The initiation process is as follows: TCP A TCP B 1. CLOSED CLOSED 2. SYN-SENT --> <SEQ=100><CTL=SYN> ... 3. SYN-RECEIVED <-- <SEQ=300><CTL=SYN> <-- SYN-SENT 4. ... <SEQ=100><CTL=SYN> --> SYN-RECEIVED 5. SYN-RECEIVED --> <SEQ=100><ACK=301><CTL=SYN,ACK> ... // TCP B: not send challenge ack for ack limit or packet loss // TCP A: close tcp_close tcp_send_fin if (!tskb && tcp_under_memory_pressure(sk)) tskb = skb_rb_last(&sk->tcp_rtx_queue); //pick SYN_ACK packet TCP_SKB_CB(tskb)->tcp_flags |= TCPHDR_FIN; // set FIN flag 6. FIN_WAIT_1 --> <SEQ=100><ACK=301><END_SEQ=102><CTL=SYN,FIN,ACK> ... // TCP B: send challenge ack to SYN_FIN_ACK 7. ... <SEQ=301><ACK=101><CTL=ACK> <-- SYN-RECEIVED //challenge ack // TCP A: <SND.UNA=101> 8. FIN_WAIT_1 --> <SEQ=101><ACK=301><END_SEQ=102><CTL=SYN,FIN,ACK> ... // retransmit panic __tcp_retransmit_skb //skb->len=0 tcp_trim_head len = tp->snd_una - TCP_SKB_CB(skb)->seq // len=101-100 __pskb_trim_head skb->data_len -= len // skb->len=-1, wrap around ... ... ip_fragment icmp_glue_bits //BUG_ON If we use tcp_trim_head() to remove acked SYN from packet that contains data or other flags, skb->len will be incorrectly decremented. We can remove SYN flag that has been acked from rtx_queue earlier than tcp_trim_head(), which can fix the problem mentioned above. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Co-developed-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Eric Dumazet <edumazet(a)google.com> Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> Link: https://lore.kernel.org/r/20231210020200.1539875-1-dongchenchen2@huawei.com Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- net/ipv4/tcp_output.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 8cd673d75f68..90e0041ae9f7 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2930,7 +2930,13 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs) if (skb_still_in_host_queue(sk, skb)) return -EBUSY; +start: if (before(TCP_SKB_CB(skb)->seq, tp->snd_una)) { + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_SYN; + TCP_SKB_CB(skb)->seq++; + goto start; + } if (unlikely(before(TCP_SKB_CB(skb)->end_seq, tp->snd_una))) { WARN_ON_ONCE(1); return -EINVAL; -- 2.25.1

2 1

[PATCH] sched: Introduce priority load balance for qos scheduler
by Song Zhang 18 Dec '23

18 Dec '23

euleros inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I8MWDD CVE: NA Add new sysctl interface: /proc/sys/kernel/sched_prio_load_balance_enabled 0: default behavior 1: enable priority load balance for qos scheduler For tasks co-location with qos scheduler, when CFS do load balance, it is reasonable to prefer migrating online(Latency Sensitive) tasks. So the CFS load balance can be changed to below: cfs_tasks list is owned by online tasks. Add new cfs_offline_tasks list which is owned by offline tasks. Prefer to migrate the online tasks of cfs_tasks list to dst rq. In the scenario of hyperthread interference, if the smt expeller feature enabled, CPU A and CPU B are two hyperthreads on a physical core, CPU A runs online tasks while CPU B only has offline tasks, The offline tasks on CPU B are expelled by the online tasks on CPU A and cannot be scheduled. However, when load balance is triggered, before CPU B can migrate some online tasks from CPU A, the load on the two cpus is already balanced. As a result, CPU B cannot run online tasks and online tasks cannot be evenly distributed among different cpus. Signed-off-by: Song Zhang <zhangsong34(a)huawei.com> --- arch/arm64/configs/openeuler_defconfig | 1 + arch/x86/configs/openeuler_defconfig | 1 + include/linux/sched/sysctl.h | 4 ++ init/Kconfig | 9 +++ kernel/sched/core.c | 3 + kernel/sched/fair.c | 90 +++++++++++++++++++++++++- kernel/sched/sched.h | 3 + 7 files changed, 110 insertions(+), 1 deletion(-) diff --git a/arch/arm64/configs/openeuler_defconfig b/arch/arm64/configs/openeuler_defconfig index 039f3496af..296421c0d5 100644 --- a/arch/arm64/configs/openeuler_defconfig +++ b/arch/arm64/configs/openeuler_defconfig @@ -162,6 +162,7 @@ CONFIG_BLK_CGROUP=y CONFIG_CGROUP_WRITEBACK=y CONFIG_CGROUP_SCHED=y CONFIG_QOS_SCHED=y +CONFIG_QOS_SCHED_PRIO_LB=y CONFIG_FAIR_GROUP_SCHED=y CONFIG_CFS_BANDWIDTH=y CONFIG_RT_GROUP_SCHED=y diff --git a/arch/x86/configs/openeuler_defconfig b/arch/x86/configs/openeuler_defconfig index f481c1c8b4..efb9a1efa7 100644 --- a/arch/x86/configs/openeuler_defconfig +++ b/arch/x86/configs/openeuler_defconfig @@ -184,6 +184,7 @@ CONFIG_BLK_CGROUP=y CONFIG_CGROUP_WRITEBACK=y CONFIG_CGROUP_SCHED=y CONFIG_QOS_SCHED=y +CONFIG_QOS_SCHED_PRIO_LB=y CONFIG_FAIR_GROUP_SCHED=y CONFIG_CFS_BANDWIDTH=y CONFIG_RT_GROUP_SCHED=y diff --git a/include/linux/sched/sysctl.h b/include/linux/sched/sysctl.h index 5a64582b08..efee7213a0 100644 --- a/include/linux/sched/sysctl.h +++ b/include/linux/sched/sysctl.h @@ -29,4 +29,8 @@ extern int sysctl_numa_balancing_mode; #define sysctl_numa_balancing_mode 0 #endif +#ifdef CONFIG_QOS_SCHED_PRIO_LB +extern unsigned int sysctl_sched_prio_load_balance_enabled; +#endif + #endif /* _LINUX_SCHED_SYSCTL_H */ diff --git a/init/Kconfig b/init/Kconfig index b6952df34e..c177b7591a 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -997,6 +997,15 @@ config QOS_SCHED If in doubt, say N. +config QOS_SCHED_PRIO_LB + bool "Priority load balance for Qos scheduler" + depends on QOS_SCHED + default n + help + This feature enable priority load balance + for Qos scheduler, which prefer migrating online tasks + and migrating offline tasks secondly between CPUs. + config FAIR_GROUP_SCHED bool "Group scheduling for SCHED_OTHER" depends on CGROUP_SCHED diff --git a/kernel/sched/core.c b/kernel/sched/core.c index a1c73dea1f..d68ddb4d28 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -10045,6 +10045,9 @@ void __init sched_init(void) rq->max_idle_balance_cost = sysctl_sched_migration_cost; INIT_LIST_HEAD(&rq->cfs_tasks); +#ifdef CONFIG_QOS_SCHED_PRIO_LB + INIT_LIST_HEAD(&rq->cfs_offline_tasks); +#endif rq_attach_root(rq, &def_root_domain); #ifdef CONFIG_NO_HZ_COMMON diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 8ae0d65713..3b35448448 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -147,6 +147,10 @@ static int hundred_thousand = 100000; static int unthrottle_qos_cfs_rqs(int cpu); #endif +#ifdef CONFIG_QOS_SCHED_PRIO_LB +unsigned int sysctl_sched_prio_load_balance_enabled; +#endif + #ifdef CONFIG_CFS_BANDWIDTH /* * Amount of runtime to allocate from global (tg) to local (per-cfs_rq) pool @@ -214,6 +218,17 @@ static struct ctl_table sched_fair_sysctls[] = { .extra1 = SYSCTL_ONE_HUNDRED, .extra2 = &one_thousand, }, +#endif +#ifdef CONFIG_QOS_SCHED_PRIO_LB + { + .procname = "sched_prio_load_balance_enabled", + .data = &sysctl_sched_prio_load_balance_enabled, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = SYSCTL_ZERO, + .extra2 = SYSCTL_ONE, + }, #endif {} }; @@ -3565,6 +3580,21 @@ static inline void update_scan_period(struct task_struct *p, int new_cpu) #endif /* CONFIG_NUMA_BALANCING */ +#ifdef CONFIG_QOS_SCHED_PRIO_LB +static void +adjust_rq_cfs_tasks(void (*list_op)(struct list_head *, struct list_head *), + struct rq *rq, + struct sched_entity *se) +{ + struct task_group *tg = task_group(task_of(se)); + + if (sysctl_sched_prio_load_balance_enabled && tg->qos_level == -1) + (*list_op)(&se->group_node, &rq->cfs_offline_tasks); + else + (*list_op)(&se->group_node, &rq->cfs_tasks); +} +#endif + static void account_entity_enqueue(struct cfs_rq *cfs_rq, struct sched_entity *se) { @@ -3574,7 +3604,11 @@ account_entity_enqueue(struct cfs_rq *cfs_rq, struct sched_entity *se) struct rq *rq = rq_of(cfs_rq); account_numa_enqueue(rq, task_of(se)); +#ifdef CONFIG_QOS_SCHED_PRIO_LB + adjust_rq_cfs_tasks(list_add, rq, se); +#else list_add(&se->group_node, &rq->cfs_tasks); +#endif } #endif cfs_rq->nr_running++; @@ -8611,7 +8645,11 @@ done: __maybe_unused; * the list, so our cfs_tasks list becomes MRU * one. */ +#ifdef CONFIG_QOS_SCHED_PRIO_LB + adjust_rq_cfs_tasks(list_move, rq, &p->se); +#else list_move(&p->se.group_node, &rq->cfs_tasks); +#endif #endif if (hrtick_enabled_fair(rq)) @@ -8959,6 +8997,14 @@ static int task_hot(struct task_struct *p, struct lb_env *env) (&p->se == cfs_rq_of(&p->se)->next)) return 1; +#ifdef CONFIG_QOS_SCHED_PRIO_LB + /* Preempt sched idle cpu do not consider migration cost */ + if (sysctl_sched_prio_load_balance_enabled && + cpus_share_cache(env->src_cpu, env->dst_cpu) && + sched_idle_cpu(env->dst_cpu)) + return 0; +#endif + if (sysctl_sched_migration_cost == -1) return 1; @@ -9149,11 +9195,18 @@ static void detach_task(struct task_struct *p, struct lb_env *env) static struct task_struct *detach_one_task(struct lb_env *env) { struct task_struct *p; + struct list_head *tasks = &env->src_rq->cfs_tasks; +#ifdef CONFIG_QOS_SCHED_PRIO_LB + int loop = 0; +#endif lockdep_assert_rq_held(env->src_rq); +#ifdef CONFIG_QOS_SCHED_PRIO_LB +again: +#endif list_for_each_entry_reverse(p, - &env->src_rq->cfs_tasks, se.group_node) { + tasks, se.group_node) { if (!can_migrate_task(p, env)) continue; @@ -9168,6 +9221,15 @@ static struct task_struct *detach_one_task(struct lb_env *env) schedstat_inc(env->sd->lb_gained[env->idle]); return p; } +#ifdef CONFIG_QOS_SCHED_PRIO_LB + if (sysctl_sched_prio_load_balance_enabled) { + loop++; + if (loop == 1) { + tasks = &env->src_rq->cfs_offline_tasks; + goto again; + } + } +#endif return NULL; } @@ -9183,6 +9245,9 @@ static int detach_tasks(struct lb_env *env) unsigned long util, load; struct task_struct *p; int detached = 0; +#ifdef CONFIG_QOS_SCHED_PRIO_LB + int loop = 0; +#endif lockdep_assert_rq_held(env->src_rq); @@ -9198,6 +9263,9 @@ static int detach_tasks(struct lb_env *env) if (env->imbalance <= 0) return 0; +#ifdef CONFIG_QOS_SCHED_PRIO_LB +again: +#endif while (!list_empty(tasks)) { /* * We don't want to steal all, otherwise we may be treated likewise, @@ -9303,6 +9371,22 @@ static int detach_tasks(struct lb_env *env) list_move(&p->se.group_node, tasks); } +#ifdef CONFIG_QOS_SCHED_PRIO_LB + if (sysctl_sched_prio_load_balance_enabled && env->imbalance > 0) { + /* + * Avoid offline tasks starve to death if env->loop exceed + * env->loop_max, so we should set env->loop to 0 and detach + * offline tasks again. + */ + if (env->loop > env->loop_max) + env->loop = 0; + loop++; + if (loop == 1) { + tasks = &env->src_rq->cfs_offline_tasks; + goto again; + } + } +#endif /* * Right now, this is one of only two places we collect this stat * so we can safely collect detach_one_task() stats here rather @@ -12920,7 +13004,11 @@ static void set_next_task_fair(struct rq *rq, struct task_struct *p, bool first) * Move the next running task to the front of the list, so our * cfs_tasks list becomes MRU one. */ +#ifdef CONFIG_QOS_SCHED_PRIO_LB + adjust_rq_cfs_tasks(list_move, rq, se); +#else list_move(&se->group_node, &rq->cfs_tasks); +#endif } #endif diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h index 3de84e95ba..76951226b6 100644 --- a/kernel/sched/sched.h +++ b/kernel/sched/sched.h @@ -1075,6 +1075,9 @@ struct rq { int online; struct list_head cfs_tasks; +#ifdef CONFIG_QOS_SCHED_PRIO_LB + struct list_head cfs_offline_tasks; +#endif struct sched_avg avg_rt; struct sched_avg avg_dl; -- 2.33.0

1 0

[PATCH OLK-6.6 0/2] Support iocost for cgroup v1
by Li Nan 18 Dec '23

18 Dec '23

Li Nan (2): cgroup: Support iocost for cgroup v1 iocost: add config option for iocost in cgroup v1 block/blk-iocost.c | 27 +++++++++++++++++++++++++++ block/Kconfig | 8 ++++++++ 2 files changed, 35 insertions(+) -- 2.39.2

2 3