March 2021 - Kernel - mailweb.openeuler.org

[PATCH v1 openEuler-21.03] fix a bug of memory leakage of openEuler-21.03
by Zhang ming 17 Mar '21

17 Mar '21

From: Zhang Ming <154842638(a)qq.com> [PATCH v1 openEuler-21.03]arch/arm64/kernel/mpam/mpam_ctrlmon.c: fix a bug of memory leakage openEuler inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I3BPPX CVE: NA ----------------------------------------------------------- The default branch in switch will not run at present, but there may be related extensions in the future, which may lead to memory leakage Signed-off-by: Zhang Ming <154842638(a)qq.com> Reported-by: Wang ShaoBo <bobo.shaobowang(a)huawei.com> Suggested-by: Jian Cheng <cj.chengjian(a)huawei.com> --- arch/arm64/kernel/mpam/mpam_ctrlmon.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm64/kernel/mpam/mpam_ctrlmon.c b/arch/arm64/kernel/mpam/mpam_ctrlmon.c index 4acf9234c3a5..b1d32d432556 100644 --- a/arch/arm64/kernel/mpam/mpam_ctrlmon.c +++ b/arch/arm64/kernel/mpam/mpam_ctrlmon.c @@ -74,6 +74,7 @@ static int add_schema(enum resctrl_conf_type t, struct resctrl_resource *r) suffix = ""; break; default: + kfree(s); return -EINVAL; } -- 2.17.1

1 0

[PATCH v1 kernel-4.19] fix a bug of memory leakage
by Zhang ming 17 Mar '21

17 Mar '21

From: Zhang Ming <154842638(a)qq.com> [PATCH kernel-4.19]arch/arm64/kernel/mpam/mpam_ctrlmon.c: fix a bug of memory leakage openEuler inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I3BPPX CVE: NA ----------------------------------------------------------- The default branch in switch will not run at present, but there may be related extensions in the future, which may lead to memory leakage Signed-off-by: Zhang Ming <154842638(a)qq.com> Reported-by: Wang ShaoBo <bobo.shaobowang(a)huawei.com> Suggested-by: Jian Cheng <cj.chengjian(a)huawei.com> --- arch/arm64/kernel/mpam/mpam_ctrlmon.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm64/kernel/mpam/mpam_ctrlmon.c b/arch/arm64/kernel/mpam/mpam_ctrlmon.c index aae585e7d7df..a4a298a455e0 100644 --- a/arch/arm64/kernel/mpam/mpam_ctrlmon.c +++ b/arch/arm64/kernel/mpam/mpam_ctrlmon.c @@ -78,6 +78,7 @@ static int add_schema(enum resctrl_conf_type t, struct resctrl_resource *r) suffix = ""; break; default: + kfree(s); return -EINVAL; } -- 2.17.1

1 0

openEuler kernel sig meeting
by openEuler conference 17 Mar '21

17 Mar '21

1 0

[PATCH openEuler-20.09 00/23] CVE for openEuler 20.09
by Cheng Jian 16 Mar '21

16 Mar '21

CVE for openEuler 20.09 Chris Leech (2): scsi: iscsi: Ensure sysfs attributes are limited to PAGE_SIZE scsi: iscsi: Verify lengths on passthrough PDUs Jan Beulich (10): Xen/x86: don't bail early from clear_foreign_p2m_mapping() Xen/x86: also check kernel mapping in set_foreign_p2m_mapping() Xen/gntdev: correct dev_bus_addr handling in gntdev_map_grant_pages() Xen/gntdev: correct error checking in gntdev_map_grant_pages() xen-blkback: don't "handle" error by BUG() xen-netback: don't "handle" error by BUG() xen-scsiback: don't "handle" error by BUG() xen-blkback: fix error handling in xen_blkbk_map() Xen/gnttab: handle p2m update errors on a per-slot basis xen-netback: respect gnttab_map_refs()'s return value Joe Perches (1): sysfs: Add sysfs_emit and sysfs_emit_at to format sysfs output Lee Duncan (1): scsi: iscsi: Restrict sessions and handles to admin capabilities Miklos Szeredi (6): ovl: pass correct flags for opening real directory ovl: switch to mounter creds in readdir ovl: verify permissions in ovl_path_open() ovl: call secutiry hook in ovl_real_ioctl() ovl: check permission to open real file ovl: do not fail because of O_NOATIME Stefano Stabellini (1): xen/arm: don't ignore return errors from set_phys_to_machine Wenchao Hao (1): virtio-blk: modernize sysfs attribute creation Yang Yingliang (1): sysfs: fix kabi broken when add sysfs_emit and sysfs_emit_at Documentation/filesystems/sysfs.txt | 8 +- arch/arm/xen/p2m.c | 33 ++++++- arch/x86/xen/p2m.c | 59 ++++++++--- drivers/block/virtio_blk.c | 67 +++++++------ drivers/block/xen-blkback/blkback.c | 30 +++--- drivers/net/xen-netback/netback.c | 10 +- drivers/scsi/libiscsi.c | 148 ++++++++++++++-------------- drivers/scsi/scsi_transport_iscsi.c | 38 +++++-- drivers/xen/gntdev.c | 37 +++---- drivers/xen/xen-scsiback.c | 4 +- fs/overlayfs/file.c | 28 ++++-- fs/overlayfs/readdir.c | 37 +++++-- fs/overlayfs/util.c | 27 ++++- fs/sysfs/file.c | 55 +++++++++++ include/linux/sysfs.h | 16 +++ include/xen/grant_table.h | 1 + security/security.c | 1 + 17 files changed, 419 insertions(+), 180 deletions(-) -- 2.25.1

2 24

[PATCH openEuler-1.0-LTS 00/32] CVE in March for 20.03
by Cheng Jian 15 Mar '21

15 Mar '21

CVE-2021-27365 CVE-2021-27363 CVE-2021-27364 Chris Leech (2): scsi: iscsi: Ensure sysfs attributes are limited to PAGE_SIZE scsi: iscsi: Verify lengths on passthrough PDUs Christoph Hellwig (1): mm/swapfile.c: fix a comment in sys_swapon() Darrick J. Wong (2): mm: set S_SWAPFILE on blockdev swap devices vfs: don't allow writes to swap files Domenico Andreoli (1): hibernate: Allow uswsusp to write to swap Jan Beulich (10): Xen/x86: don't bail early from clear_foreign_p2m_mapping() Xen/x86: also check kernel mapping in set_foreign_p2m_mapping() Xen/gntdev: correct dev_bus_addr handling in gntdev_map_grant_pages() Xen/gntdev: correct error checking in gntdev_map_grant_pages() xen-blkback: don't "handle" error by BUG() xen-netback: don't "handle" error by BUG() xen-scsiback: don't "handle" error by BUG() xen-blkback: fix error handling in xen_blkbk_map() Xen/gnttab: handle p2m update errors on a per-slot basis xen-netback: respect gnttab_map_refs()'s return value Joe Perches (1): sysfs: Add sysfs_emit and sysfs_emit_at to format sysfs output Lee Duncan (1): scsi: iscsi: Restrict sessions and handles to admin capabilities Miaohe Lin (1): mm/swapfile.c: fix potential memory leak in sys_swapon Miklos Szeredi (6): ovl: pass correct flags for opening real directory ovl: switch to mounter creds in readdir ovl: verify permissions in ovl_path_open() ovl: call secutiry hook in ovl_real_ioctl() ovl: check permission to open real file ovl: do not fail because of O_NOATIME Naohiro Aota (1): mm/swapfile.c: move inode_lock out of claim_swapfile Stefano Stabellini (1): xen/arm: don't ignore return errors from set_phys_to_machine Wenchao Hao (2): nvme: register ns_id attributes as default sysfs groups virtio-blk: modernize sysfs attribute creation Yang Yingliang (1): sysfs: fix kabi broken when add sysfs_emit and sysfs_emit_at Ye Bin (1): ext4: Fix not report exception message when mount with errors=continue zhangyi (F) (1): block_dump: remove block_dump feature when dirting inode Documentation/filesystems/sysfs.txt | 8 +- arch/arm/xen/p2m.c | 33 ++++++- arch/x86/xen/p2m.c | 59 ++++++++--- drivers/block/virtio_blk.c | 67 +++++++------ drivers/block/xen-blkback/blkback.c | 30 +++--- drivers/net/xen-netback/netback.c | 10 +- drivers/nvme/host/core.c | 20 ++-- drivers/nvme/host/lightnvm.c | 105 +++++++++----------- drivers/nvme/host/multipath.c | 11 +-- drivers/nvme/host/nvme.h | 10 +- drivers/scsi/libiscsi.c | 148 ++++++++++++++-------------- drivers/scsi/scsi_transport_iscsi.c | 38 +++++-- drivers/xen/gntdev.c | 37 +++---- drivers/xen/xen-scsiback.c | 4 +- fs/block_dev.c | 5 + fs/ext4/super.c | 6 +- fs/fs-writeback.c | 25 ----- fs/overlayfs/file.c | 28 ++++-- fs/overlayfs/readdir.c | 37 +++++-- fs/overlayfs/util.c | 27 ++++- fs/sysfs/file.c | 55 +++++++++++ include/linux/fs.h | 11 +++ include/linux/sysfs.h | 16 +++ include/xen/grant_table.h | 1 + mm/filemap.c | 3 + mm/memory.c | 4 + mm/mmap.c | 8 +- mm/swapfile.c | 72 ++++++++------ security/security.c | 1 + 29 files changed, 552 insertions(+), 327 deletions(-) -- 2.25.1

1 32

[PATCH openEuler-21.03 v1] arm64: Uninstall cpu park after cpu up
by sangyan＠huawei.com 12 Mar '21

12 Mar '21

From: Sang Yan <sangyan(a)huawei.com> hulk inclusion category: feature bugzilla: 48159 CVE: N/A ------------------------------ One cpu in PARK state could not come up in this case: CPU 0 | CPU 1 boot_secondary(cpu 1) | --> write_park_exit(cpu 1) | | cpu uping from PARK | ... uninstall_cpu_park() | --> memset to 0 park text | | ... | Exception in memory !! wait for cpu up | Cpu 1 uping from PARK may trap into exception while cpu 0 clear cpu 1's park text memory. This uninstall_cpu_park should be after waiting for cpu up. Signed-off-by: Sang Yan <sangyan(a)huawei.com> --- arch/arm64/kernel/smp.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c index d7b750a..fb6007d 100644 --- a/arch/arm64/kernel/smp.c +++ b/arch/arm64/kernel/smp.c @@ -300,15 +300,15 @@ int __cpu_up(unsigned int cpu, struct task_struct *idle) return ret; } -#ifdef CONFIG_ARM64_CPU_PARK - uninstall_cpu_park(cpu); -#endif /* * CPU was successfully started, wait for it to come online or * time out. */ wait_for_completion_timeout(&cpu_running, msecs_to_jiffies(5000)); +#ifdef CONFIG_ARM64_CPU_PARK + uninstall_cpu_park(cpu); +#endif if (cpu_online(cpu)) return 0; -- 2.9.5

2 1

[PATCH OLK-5.10 v1] arm64: Uninstall cpu park after cpu up
by sangyan＠huawei.com 12 Mar '21

12 Mar '21

From: Sang Yan <sangyan(a)huawei.com> hulk inclusion category: feature bugzilla: 48159 CVE: N/A ------------------------------ One cpu in PARK state could not come up in this case: CPU 0 | CPU 1 boot_secondary(cpu 1) | --> write_park_exit(cpu 1) | | cpu uping from PARK | ... uninstall_cpu_park() | --> memset to 0 park text | | ... | Exception in memory !! wait for cpu up | Cpu 1 uping from PARK may trap into exception while cpu 0 clear cpu 1's park text memory. This uninstall_cpu_park should be after waiting for cpu up. Signed-off-by: Sang Yan <sangyan(a)huawei.com> --- arch/arm64/kernel/smp.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c index d7b750a..fb6007d 100644 --- a/arch/arm64/kernel/smp.c +++ b/arch/arm64/kernel/smp.c @@ -300,15 +300,15 @@ int __cpu_up(unsigned int cpu, struct task_struct *idle) return ret; } -#ifdef CONFIG_ARM64_CPU_PARK - uninstall_cpu_park(cpu); -#endif /* * CPU was successfully started, wait for it to come online or * time out. */ wait_for_completion_timeout(&cpu_running, msecs_to_jiffies(5000)); +#ifdef CONFIG_ARM64_CPU_PARK + uninstall_cpu_park(cpu); +#endif if (cpu_online(cpu)) return 0; -- 2.9.5

1 0

[PATCH openEuler-21.03 v2] park: Reserve park mem before kexec reserved
by sangyan＠huawei.com 12 Mar '21

12 Mar '21

From: Sang Yan <sangyan(a)huawei.com> hulk inclusion category: feature bugzilla: 48159 CVE: N/A ------------------------------ reserve_crashkernel or reserve_quick_kexec may find one sutiable memory region and reserves it, which address of the region is not fixed. As a result, cpu park reserves memory could be failed while specified address used by crashkernel or quickkexec. So, move reserve_park_mem before reserve_crashkernel and reserve_quick_kexec. Signed-off-by: Sang Yan <sangyan(a)huawei.com> Reviewed-by: Jing Xiangfeng <jingxiangfeng(a)huawei.com> --- arch/arm64/mm/init.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index b343744..dbcc801 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -497,16 +497,25 @@ void __init arm64_memblock_init(void) else arm64_dma32_phys_limit = PHYS_MASK + 1; + /* + * Reserve park memory before crashkernel and quick kexec. + * Because park memory must be specified by address, but + * crashkernel and quickkexec may be specified by memory length, + * then find one sutiable memory region to reserve. + * + * So reserve park memory firstly is better, but it may cause + * crashkernel or quickkexec reserving failed. + */ +#ifdef CONFIG_ARM64_CPU_PARK + reserve_park_mem(); +#endif + reserve_crashkernel(); #ifdef CONFIG_QUICK_KEXEC reserve_quick_kexec(); #endif -#ifdef CONFIG_ARM64_CPU_PARK - reserve_park_mem(); -#endif - reserve_pin_memory_res(); reserve_elfcorehdr(); -- 2.9.5

1 0

[PATCH openEuler-21.03 v4 1/2] arm64: Add memmap parameter and register pmem
by Zhuling 12 Mar '21

12 Mar '21

From: ZhuLing <zhuling8(a)huawei.com> hulk inclusion category: feature bugzilla: 48159 CVE: NA Register pmem in arm64: Use memmap(memmap=nn[KMG]!ss[KMG]) reserve memory and e820(driver/nvdimm/e820.c) function to register persistent memory in arm64. when the kernel restart or update, the data in PMEM will not be lost and can be loaded faster. this is a general features. driver/nvdimm/e820.c: The function of this file is scan "iomem_resource" and take advantage of nvdimm resource discovery mechanism by registering a resource named "Persistent Memory (legacy)", this function doesn't depend on architecture. We will push the feature to linux kernel community and discuss to modify the file name. because people have a mistaken notion that the e820.c is depend on x86. If you want use this features, you need do as follows: 1.Reserve memory: add memmap to reserve memory in grub.cfg memmap=nn[KMG]!ss[KMG] exp:memmap=100K!0x1a0000000. 2.Insmod nd_e820.ko: modprobe nd_e820. 3.Check pmem device in /dev exp: /dev/pmem0 Signed-off-by: ZhuLing <zhuling8(a)huawei.com> Signed-off-by: Sang Yan <sangyan(a)huawei.com> Acked-by: Hanjun Guo <guohanjun(a)huawei.com> --- arch/arm64/Kconfig | 21 +++++++++ arch/arm64/kernel/Makefile | 1 + arch/arm64/kernel/pmem.c | 35 ++++++++++++++ arch/arm64/kernel/setup.c | 10 ++++ arch/arm64/mm/init.c | 97 ++++++++++++++++++++++++++++++++++++++ drivers/nvdimm/Kconfig | 5 ++ drivers/nvdimm/Makefile | 2 +- 7 files changed, 170 insertions(+), 1 deletion(-) create mode 100644 arch/arm64/kernel/pmem.c diff --git a/arch/arm64/Kconfig b/arch/arm64/Kconfig index c451137ab..326f26d40 100644 --- a/arch/arm64/Kconfig +++ b/arch/arm64/Kconfig @@ -360,6 +360,27 @@ config ARM64_CPU_PARK config ARCH_HAS_CPU_RELAX def_bool y +config ARM64_PMEM_RESERVE + bool "Reserve memory for persistent storage" + default n + help + Use memmap=nn[KMG]!ss[KMG](memmap=100K!0x1a0000000) reserve + memory for persistent storage. + + Say y here to enable this feature. + +config ARM64_PMEM_LEGACY_DEVICE + bool "Create persistent storage" + depends on BLK_DEV + depends on LIBNVDIMM + select ARM64_PMEM_RESERVE + help + Use reserved memory for persistent storage when the kernel + restart or update. the data in PMEM will not be lost and + can be loaded faster. + + Say y if unsure. + source "arch/arm64/Kconfig.platforms" menu "Kernel Features" diff --git a/arch/arm64/kernel/Makefile b/arch/arm64/kernel/Makefile index 967cb3c6d..be996f3c1 100644 --- a/arch/arm64/kernel/Makefile +++ b/arch/arm64/kernel/Makefile @@ -67,6 +67,7 @@ obj-$(CONFIG_ARM64_PTR_AUTH) += pointer_auth.o obj-$(CONFIG_SHADOW_CALL_STACK) += scs.o obj-$(CONFIG_ARM64_MTE) += mte.o obj-$(CONFIG_MPAM) += mpam/ +obj-$(CONFIG_ARM64_PMEM_LEGACY_DEVICE) += pmem.o obj-y += vdso/ probes/ obj-$(CONFIG_COMPAT_VDSO) += vdso32/ diff --git a/arch/arm64/kernel/pmem.c b/arch/arm64/kernel/pmem.c new file mode 100644 index 000000000..16eaf706f --- /dev/null +++ b/arch/arm64/kernel/pmem.c @@ -0,0 +1,35 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright(c) 2021 Huawei Technologies Co., Ltd + * + * Derived from x86 and arm64 implement PMEM. + */ +#include <linux/platform_device.h> +#include <linux/init.h> +#include <linux/ioport.h> +#include <linux/module.h> + +static int found(struct resource *res, void *data) +{ + return 1; +} + +static int __init register_e820_pmem(void) +{ + struct platform_device *pdev; + int rc; + + rc = walk_iomem_res_desc(IORES_DESC_PERSISTENT_MEMORY_LEGACY, + IORESOURCE_MEM, 0, -1, NULL, found); + if (rc <= 0) + return 0; + + /* + * See drivers/nvdimm/e820.c for the implementation, this is + * simply here to trigger the module to load on demand. + */ + pdev = platform_device_alloc("e820_pmem", -1); + + return platform_device_add(pdev); +} +device_initcall(register_e820_pmem); diff --git a/arch/arm64/kernel/setup.c b/arch/arm64/kernel/setup.c index 5e282d31a..84c71c88d 100644 --- a/arch/arm64/kernel/setup.c +++ b/arch/arm64/kernel/setup.c @@ -57,6 +57,10 @@ static int num_standard_resources; static struct resource *standard_resources; +#ifdef CONFIG_ARM64_PMEM_RESERVE +extern struct resource pmem_res; +#endif + phys_addr_t __fdt_pointer __initdata; /* @@ -270,6 +274,12 @@ static void __init request_standard_resources(void) request_resource(res, &pin_memory_resource); #endif } + +#ifdef CONFIG_ARM64_PMEM_RESERVE + if (pmem_res.end && pmem_res.start) + request_resource(&iomem_resource, &pmem_res); +#endif + } static int __init reserve_memblock_reserved_regions(void) diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c index b3437440d..f22faea1a 100644 --- a/arch/arm64/mm/init.c +++ b/arch/arm64/mm/init.c @@ -66,6 +66,18 @@ EXPORT_SYMBOL(memstart_addr); phys_addr_t arm64_dma_phys_limit __ro_after_init; phys_addr_t arm64_dma32_phys_limit __ro_after_init; +static unsigned long long pmem_size, pmem_start; + +#ifdef CONFIG_ARM64_PMEM_RESERVE +struct resource pmem_res = { + .name = "Persistent Memory (legacy)", + .start = 0, + .end = 0, + .flags = IORESOURCE_MEM, + .desc = IORES_DESC_PERSISTENT_MEMORY_LEGACY +}; +#endif + #ifndef CONFIG_KEXEC_CORE static void __init reserve_crashkernel(void) { @@ -378,6 +390,87 @@ static int __init reserve_park_mem(void) } #endif +static int __init is_mem_valid(unsigned long long mem_size, unsigned long long mem_start) +{ + if (!memblock_is_region_memory(mem_start, mem_size)) { + pr_warn("cannot reserve mem: region is not memory!\n"); + return -EINVAL; + } + + if (memblock_is_region_reserved(mem_start, mem_size)) { + pr_warn("cannot reserve mem: region overlaps reserved memory!\n"); + return -EINVAL; + } + + if (!IS_ALIGNED(mem_start, SZ_2M)) { + pr_warn("cannot reserve mem: base address is not 2MB aligned!\n"); + return -EINVAL; + } + + return 0; +} + +static int __init parse_memmap_one(char *p) +{ + char *oldp; + phys_addr_t start_at, mem_size; + + if (!p) + return -EINVAL; + + oldp = p; + mem_size = memparse(p, &p); + if (p == oldp) + return -EINVAL; + + if (!mem_size) + return -EINVAL; + + mem_size = PAGE_ALIGN(mem_size); + + if (*p == '!') { + start_at = memparse(p+1, &p); + + if (is_mem_valid(mem_size, start_at) != 0) + return -EINVAL; + + pr_info("pmem reserved: 0x%016llx - 0x%016llx (%lld MB)\n", + start_at, start_at + mem_size, mem_size >> 20); + pmem_start = start_at; + pmem_size = mem_size; + } else + pr_info("Unrecognized memmap option, please check the parameter.\n"); + + return *p == '\0' ? 0 : -EINVAL; +} + +static int __init parse_memmap_opt(char *str) +{ + while (str) { + char *k = strchr(str, ','); + + if (k) + *k++ = 0; + + parse_memmap_one(str); + str = k; + } + + return 0; +} +early_param("memmap", parse_memmap_opt); + +#ifdef CONFIG_ARM64_PMEM_RESERVE +static void __init reserve_pmem(void) +{ + memblock_remove(pmem_start, pmem_size); + pr_info("pmem reserved: 0x%016llx - 0x%016llx (%lld MB)\n", + pmem_start, pmem_start + pmem_size, pmem_size >> 20); + pmem_res.start = pmem_start; + pmem_res.end = pmem_start + pmem_size - 1; +} +#endif + void __init arm64_memblock_init(void) { const s64 linear_region_size = BIT(vabits_actual - 1); @@ -511,6 +604,10 @@ void __init arm64_memblock_init(void) reserve_elfcorehdr(); +#ifdef CONFIG_ARM64_PMEM_RESERVE + reserve_pmem(); +#endif + high_memory = __va(memblock_end_of_DRAM() - 1) + 1; dma_contiguous_reserve(arm64_dma32_phys_limit); diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig index b7d1eb38b..ce4de7526 100644 --- a/drivers/nvdimm/Kconfig +++ b/drivers/nvdimm/Kconfig @@ -132,3 +132,8 @@ config NVDIMM_TEST_BUILD infrastructure. endif + +config PMEM_LEGACY + tristate "Pmem_legacy" + select X86_PMEM_LEGACY if X86 + select ARM64_PMEM_LEGACY_DEVICE if ARM64 diff --git a/drivers/nvdimm/Makefile b/drivers/nvdimm/Makefile index 29203f3d3..6f8dc9242 100644 --- a/drivers/nvdimm/Makefile +++ b/drivers/nvdimm/Makefile @@ -3,7 +3,7 @@ obj-$(CONFIG_LIBNVDIMM) += libnvdimm.o obj-$(CONFIG_BLK_DEV_PMEM) += nd_pmem.o obj-$(CONFIG_ND_BTT) += nd_btt.o obj-$(CONFIG_ND_BLK) += nd_blk.o -obj-$(CONFIG_X86_PMEM_LEGACY) += nd_e820.o +obj-$(CONFIG_PMEM_LEGACY) += nd_e820.o obj-$(CONFIG_OF_PMEM) += of_pmem.o obj-$(CONFIG_VIRTIO_PMEM) += virtio_pmem.o nd_virtio.o -- 2.19.1

1 1

[PATCH v2] arm: keep the original function for non-RPi
by Zheng Zengkai 11 Mar '21

11 Mar '21

raspberrypi inclusion category: feature bugzilla: 50432 ------------------------------ This patch adjusts following arch arm related patches for raspberry pi on non-Raspberry Pi platforms, using specific config CONFIG_OPENEULER_RASPBERRYPI to distinguish them: d5c13edbd8 Improve __copy_to_user and __copy_from_user performance 97145d2a6a Update vfpmodule.c bffc462cbd Main bcm2708/bcm2709 linux port 588cfce788 cache: export clean and invalidate 41cd350cca ARM: Activate FIQs to avoid __irq_startup warnings 90607c7aaf ARM: proc-v7: Force misalignment of early stmia ee46d0fadf reboot: Use power off rather than busy spinning when halt is requested Signed-off-by: Zheng Zengkai <zhengzengkai(a)huawei.com> --- arch/arm/include/asm/string.h | 2 ++ arch/arm/include/asm/uaccess.h | 2 ++ arch/arm/kernel/fiq.c | 4 ++++ arch/arm/kernel/reboot.c | 6 ++++++ arch/arm/lib/Makefile | 15 +++++++++++++++ arch/arm/lib/copy_from_user.S | 6 ++++++ arch/arm/lib/uaccess_with_memcpy.c | 22 +++++++++++++++++++++- arch/arm/mm/cache-v6.S | 8 ++++++++ arch/arm/mm/cache-v7.S | 8 ++++++++ arch/arm/mm/proc-v6.S | 8 ++++++++ arch/arm/mm/proc-v7.S | 8 ++++++++ arch/arm/vfp/vfpmodule.c | 26 ++++++++++++++++++++++++++ 12 files changed, 114 insertions(+), 1 deletion(-) diff --git a/arch/arm/include/asm/string.h b/arch/arm/include/asm/string.h index c22d5869e7b6..3c4ae6b3c3a6 100644 --- a/arch/arm/include/asm/string.h +++ b/arch/arm/include/asm/string.h @@ -45,10 +45,12 @@ static inline void *memset64(uint64_t *p, uint64_t v, __kernel_size_t n) return __memset64(p, v, n * 8, v >> 32); } +#ifdef CONFIG_OPENEULER_RASPBERRYPI #ifdef CONFIG_BCM2835_FAST_MEMCPY #define __HAVE_ARCH_MEMCMP extern int memcmp(const void *, const void *, size_t); #endif +#endif /* * For files that are not instrumented (e.g. mm/slub.c) we diff --git a/arch/arm/include/asm/uaccess.h b/arch/arm/include/asm/uaccess.h index f24d3fabccd6..0d8cba7a9dde 100644 --- a/arch/arm/include/asm/uaccess.h +++ b/arch/arm/include/asm/uaccess.h @@ -516,8 +516,10 @@ do { \ extern unsigned long __must_check arm_copy_from_user(void *to, const void __user *from, unsigned long n); +#ifdef CONFIG_OPENEULER_RASPBERRYPI extern unsigned long __must_check __copy_from_user_std(void *to, const void __user *from, unsigned long n); +#endif static inline unsigned long __must_check raw_copy_from_user(void *to, const void __user *from, unsigned long n) diff --git a/arch/arm/kernel/fiq.c b/arch/arm/kernel/fiq.c index c3fe7d3cf482..8116f1e52e1b 100644 --- a/arch/arm/kernel/fiq.c +++ b/arch/arm/kernel/fiq.c @@ -56,7 +56,9 @@ static unsigned long dfl_fiq_insn; static struct pt_regs dfl_fiq_regs; +#ifdef CONFIG_OPENEULER_RASPBERRYPI extern int irq_activate(struct irq_desc *desc); +#endif /* Default reacquire function * - we always relinquish FIQ control @@ -142,8 +144,10 @@ static int fiq_start; void enable_fiq(int fiq) { +#ifdef CONFIG_OPENEULER_RASPBERRYPI struct irq_desc *desc = irq_to_desc(fiq + fiq_start); irq_activate(desc); +#endif enable_irq(fiq + fiq_start); } diff --git a/arch/arm/kernel/reboot.c b/arch/arm/kernel/reboot.c index 63373adab475..ffb170568dcc 100644 --- a/arch/arm/kernel/reboot.c +++ b/arch/arm/kernel/reboot.c @@ -102,7 +102,13 @@ void machine_shutdown(void) */ void machine_halt(void) { +#ifdef CONFIG_OPENEULER_RASPBERRYPI machine_power_off(); +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ + local_irq_disable(); + smp_send_stop(); + while (1); +#endif } /* diff --git a/arch/arm/lib/Makefile b/arch/arm/lib/Makefile index 8271cde92dec..63820a487d4b 100644 --- a/arch/arm/lib/Makefile +++ b/arch/arm/lib/Makefile @@ -5,6 +5,7 @@ # Copyright (C) 1995-2000 Russell King # +ifeq ($(CONFIG_OPENEULER_RASPBERRYPI),y) lib-y := changebit.o csumipv6.o csumpartial.o \ csumpartialcopy.o csumpartialcopyuser.o clearbit.o \ delay.o delay-loop.o findbit.o memchr.o \ @@ -15,6 +16,18 @@ lib-y := changebit.o csumipv6.o csumpartial.o \ ucmpdi2.o lib1funcs.o div64.o \ io-readsb.o io-writesb.o io-readsl.o io-writesl.o \ call_with_stack.o bswapsdi2.o +else +lib-y := changebit.o csumipv6.o csumpartial.o \ + csumpartialcopy.o csumpartialcopyuser.o clearbit.o \ + delay.o delay-loop.o findbit.o memchr.o memcpy.o \ + memmove.o memset.o setbit.o \ + strchr.o strrchr.o \ + testchangebit.o testclearbit.o testsetbit.o \ + ashldi3.o ashrdi3.o lshrdi3.o muldi3.o \ + ucmpdi2.o lib1funcs.o div64.o \ + io-readsb.o io-writesb.o io-readsl.o io-writesl.o \ + call_with_stack.o bswapsdi2.o +endif mmu-y := clear_user.o copy_page.o getuser.o putuser.o \ copy_from_user.o copy_to_user.o @@ -25,6 +38,7 @@ else lib-y += backtrace.o endif +ifeq ($(CONFIG_OPENEULER_RASPBERRYPI),y) # Choose optimised implementations for Raspberry Pi ifeq ($(CONFIG_BCM2835_FAST_MEMCPY),y) CFLAGS_uaccess_with_memcpy.o += -DCOPY_FROM_USER_THRESHOLD=1600 @@ -34,6 +48,7 @@ ifeq ($(CONFIG_BCM2835_FAST_MEMCPY),y) else lib-y += memcpy.o memmove.o memset.o endif +endif # using lib_ here won't override already available weak symbols obj-$(CONFIG_UACCESS_WITH_MEMCPY) += uaccess_with_memcpy.o diff --git a/arch/arm/lib/copy_from_user.S b/arch/arm/lib/copy_from_user.S index ab7bf28dbec0..3f83d8b18b0d 100644 --- a/arch/arm/lib/copy_from_user.S +++ b/arch/arm/lib/copy_from_user.S @@ -107,8 +107,12 @@ .text +#ifdef CONFIG_OPENEULER_RASPBERRYPI ENTRY(__copy_from_user_std) WEAK(arm_copy_from_user) +#else +ENTRY(arm_copy_from_user) +#endif #ifdef CONFIG_CPU_SPECTRE get_thread_info r3 ldr r3, [r3, #TI_ADDR_LIMIT] @@ -118,7 +122,9 @@ WEAK(arm_copy_from_user) #include "copy_template.S" ENDPROC(arm_copy_from_user) +#ifdef CONFIG_OPENEULER_RASPBERRYPI ENDPROC(__copy_from_user_std) +#endif .pushsection .text.fixup,"ax" .align 0 diff --git a/arch/arm/lib/uaccess_with_memcpy.c b/arch/arm/lib/uaccess_with_memcpy.c index b483e5713039..ab15ed7f599a 100644 --- a/arch/arm/lib/uaccess_with_memcpy.c +++ b/arch/arm/lib/uaccess_with_memcpy.c @@ -19,6 +19,7 @@ #include <asm/current.h> #include <asm/page.h> +#ifdef CONFIG_OPENEULER_RASPBERRYPI #ifndef COPY_FROM_USER_THRESHOLD #define COPY_FROM_USER_THRESHOLD 64 #endif @@ -26,6 +27,7 @@ #ifndef COPY_TO_USER_THRESHOLD #define COPY_TO_USER_THRESHOLD 64 #endif +#endif static int pin_page_for_write(const void __user *_addr, pte_t **ptep, spinlock_t **ptlp) @@ -51,7 +53,11 @@ pin_page_for_write(const void __user *_addr, pte_t **ptep, spinlock_t **ptlp) return 0; pmd = pmd_offset(pud, addr); +#ifdef CONFIG_OPENEULER_RASPBERRYPI if (unlikely(pmd_none(*pmd) || pmd_bad(*pmd))) +#else + if (unlikely(pmd_none(*pmd))) +#endif return 0; /* @@ -94,6 +100,7 @@ pin_page_for_write(const void __user *_addr, pte_t **ptep, spinlock_t **ptlp) return 1; } +#ifdef CONFIG_OPENEULER_RASPBERRYPI static int pin_page_for_read(const void __user *_addr, pte_t **ptep, spinlock_t **ptlp) { @@ -132,8 +139,13 @@ pin_page_for_read(const void __user *_addr, pte_t **ptep, spinlock_t **ptlp) return 1; } +#endif +#ifdef CONFIG_OPENEULER_RASPBERRYPI unsigned long noinline +#else +static unsigned long noinline +#endif __copy_to_user_memcpy(void __user *to, const void *from, unsigned long n) { unsigned long ua_flags; @@ -186,6 +198,7 @@ __copy_to_user_memcpy(void __user *to, const void *from, unsigned long n) return n; } +#ifdef CONFIG_OPENEULER_RASPBERRYPI unsigned long noinline __copy_from_user_memcpy(void *to, const void __user *from, unsigned long n) { @@ -236,6 +249,7 @@ __copy_from_user_memcpy(void *to, const void __user *from, unsigned long n) out: return n; } +#endif unsigned long arm_copy_to_user(void __user *to, const void *from, unsigned long n) @@ -247,7 +261,11 @@ arm_copy_to_user(void __user *to, const void *from, unsigned long n) * With frame pointer disabled, tail call optimization kicks in * as well making this test almost invisible. */ +#ifdef CONFIG_OPENEULER_RASPBERRYPI if (n < COPY_TO_USER_THRESHOLD) { +#else + if (n < 64) { +#endif unsigned long ua_flags = uaccess_save_and_enable(); n = __copy_to_user_std(to, from, n); uaccess_restore(ua_flags); @@ -258,6 +276,7 @@ arm_copy_to_user(void __user *to, const void *from, unsigned long n) return n; } +#ifdef CONFIG_OPENEULER_RASPBERRYPI unsigned long __must_check arm_copy_from_user(void *to, const void __user *from, unsigned long n) { @@ -283,7 +302,8 @@ arm_copy_from_user(void *to, const void __user *from, unsigned long n) #endif return n; } - +#endif + static unsigned long noinline __clear_user_memset(void __user *addr, unsigned long n) { diff --git a/arch/arm/mm/cache-v6.S b/arch/arm/mm/cache-v6.S index 868011801521..614d4ff2a760 100644 --- a/arch/arm/mm/cache-v6.S +++ b/arch/arm/mm/cache-v6.S @@ -198,7 +198,11 @@ ENTRY(v6_flush_kern_dcache_area) * - start - virtual start address of region * - end - virtual end address of region */ +#ifdef CONFIG_OPENEULER_RASPBERRYPI ENTRY(v6_dma_inv_range) +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ +v6_dma_inv_range: +#endif #ifdef CONFIG_DMA_CACHE_RWFO ldrb r2, [r0] @ read for ownership strb r2, [r0] @ write for ownership @@ -243,7 +247,11 @@ ENTRY(v6_dma_inv_range) * - start - virtual start address of region * - end - virtual end address of region */ +#ifdef CONFIG_OPENEULER_RASPBERRYPI ENTRY(v6_dma_clean_range) +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ +v6_dma_clean_range: +#endif bic r0, r0, #D_CACHE_LINE_SIZE - 1 1: #ifdef CONFIG_DMA_CACHE_RWFO diff --git a/arch/arm/mm/cache-v7.S b/arch/arm/mm/cache-v7.S index 536df5db66e4..58df59734e40 100644 --- a/arch/arm/mm/cache-v7.S +++ b/arch/arm/mm/cache-v7.S @@ -363,8 +363,12 @@ ENDPROC(v7_flush_kern_dcache_area) * - start - virtual start address of region * - end - virtual end address of region */ +#ifdef CONFIG_OPENEULER_RASPBERRYPI ENTRY(b15_dma_inv_range) ENTRY(v7_dma_inv_range) +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ +v7_dma_inv_range: +#endif dcache_line_size r2, r3 sub r3, r2, #1 tst r0, r3 @@ -394,8 +398,12 @@ ENDPROC(v7_dma_inv_range) * - start - virtual start address of region * - end - virtual end address of region */ +#ifdef CONFIG_OPENEULER_RASPBERRYPI ENTRY(b15_dma_clean_range) ENTRY(v7_dma_clean_range) +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ +v7_dma_clean_range: +#endif dcache_line_size r2, r3 sub r3, r2, #1 bic r0, r0, r3 diff --git a/arch/arm/mm/proc-v6.S b/arch/arm/mm/proc-v6.S index b3a2fce22eac..b651eaa1ee40 100644 --- a/arch/arm/mm/proc-v6.S +++ b/arch/arm/mm/proc-v6.S @@ -71,6 +71,7 @@ ENDPROC(cpu_v6_reset) * IRQs are already disabled. */ +#ifdef CONFIG_OPENEULER_RASPBERRYPI /* See jira SW-5991 for details of this workaround */ ENTRY(cpu_v6_do_idle) .align 5 @@ -84,6 +85,13 @@ ENTRY(cpu_v6_do_idle) nop bne 1b ret lr +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ +ENTRY(cpu_v6_do_idle) + mov r1, #0 + mcr p15, 0, r1, c7, c10, 4 @ DWB - WFI may enter a low-power mode + mcr p15, 0, r1, c7, c0, 4 @ wait for interrupt + ret lr +#endif ENTRY(cpu_v6_dcache_clean_area) 1: mcr p15, 0, r0, c7, c10, 1 @ clean D entry diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S index 3e77e8982df3..ef0e00249515 100644 --- a/arch/arm/mm/proc-v7.S +++ b/arch/arm/mm/proc-v7.S @@ -287,8 +287,10 @@ __v7_ca17mp_setup: mov r10, #0 1: adr r0, __v7_setup_stack_ptr ldr r12, [r0] +#ifdef CONFIG_OPENEULER_RASPBERRYPI tst r12, #0x1f addeq r12, r12, #4 +#endif add r12, r12, r0 @ the local stack stmia r12, {r1-r6, lr} @ v7_invalidate_l1 touches r0-r6 bl v7_invalidate_l1 @@ -476,8 +478,10 @@ __v7_setup: adr r0, __v7_setup_stack_ptr ldr r12, [r0] add r12, r12, r0 @ the local stack +#ifdef CONFIG_OPENEULER_RASPBERRYPI tst r12, #0x1f addeq r12, r12, #4 +#endif stmia r12, {r1-r6, lr} @ v7_invalidate_l1 touches r0-r6 bl v7_invalidate_l1 ldmia r12, {r1-r6, lr} @@ -561,7 +565,11 @@ ENDPROC(__v7_setup) .bss .align 2 __v7_setup_stack: +#ifdef CONFIG_OPENEULER_RASPBERRYPI .space 4 * 8 @ 7 registers + 1 spare +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ + .space 4 * 7 @ 7 registers +#endif __INITDATA diff --git a/arch/arm/vfp/vfpmodule.c b/arch/arm/vfp/vfpmodule.c index 1e2dcf81aefa..f7238e00db7c 100644 --- a/arch/arm/vfp/vfpmodule.c +++ b/arch/arm/vfp/vfpmodule.c @@ -176,11 +176,16 @@ static int vfp_notifier(struct notifier_block *self, unsigned long cmd, void *v) * case the thread migrates to a different CPU. The * restoring is done lazily. */ +#ifdef CONFIG_OPENEULER_RASPBERRYPI if ((fpexc & FPEXC_EN) && vfp_current_hw_state[cpu]) { /* vfp_save_state oopses on VFP11 if EX bit set */ fmxr(FPEXC, fpexc & ~FPEXC_EX); vfp_save_state(vfp_current_hw_state[cpu], fpexc); } +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ + if ((fpexc & FPEXC_EN) && vfp_current_hw_state[cpu]) + vfp_save_state(vfp_current_hw_state[cpu], fpexc); +#endif #endif /* @@ -457,16 +462,22 @@ static int vfp_pm_suspend(void) /* if vfp is on, then save state for resumption */ if (fpexc & FPEXC_EN) { pr_debug("%s: saving vfp state\n", __func__); +#ifdef CONFIG_OPENEULER_RASPBERRYPI /* vfp_save_state oopses on VFP11 if EX bit set */ fmxr(FPEXC, fpexc & ~FPEXC_EX); +#endif vfp_save_state(&ti->vfpstate, fpexc); /* disable, just in case */ fmxr(FPEXC, fmrx(FPEXC) & ~FPEXC_EN); } else if (vfp_current_hw_state[ti->cpu]) { #ifndef CONFIG_SMP +#ifdef CONFIG_OPENEULER_RASPBERRYPI /* vfp_save_state oopses on VFP11 if EX bit set */ fmxr(FPEXC, (fpexc & ~FPEXC_EX) | FPEXC_EN); +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ + fmxr(FPEXC, fpexc | FPEXC_EN); +#endif vfp_save_state(vfp_current_hw_state[ti->cpu], fpexc); fmxr(FPEXC, fpexc); #endif @@ -529,8 +540,12 @@ void vfp_sync_hwstate(struct thread_info *thread) /* * Save the last VFP state on this CPU. */ +#ifdef CONFIG_OPENEULER_RASPBERRYPI /* vfp_save_state oopses on VFP11 if EX bit set */ fmxr(FPEXC, (fpexc & ~FPEXC_EX) | FPEXC_EN); +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ + fmxr(FPEXC, fpexc | FPEXC_EN); +#endif vfp_save_state(&thread->vfpstate, fpexc | FPEXC_EN); fmxr(FPEXC, fpexc); } @@ -596,7 +611,9 @@ int vfp_restore_user_hwstate(struct user_vfp *ufp, struct user_vfp_exc *ufp_exc) struct thread_info *thread = current_thread_info(); struct vfp_hard_struct *hwstate = &thread->vfpstate.hard; unsigned long fpexc; +#ifdef CONFIG_OPENEULER_RASPBERRYPI u32 fpsid = fmrx(FPSID); +#endif /* Disable VFP to avoid corrupting the new thread state. */ vfp_flush_hwstate(thread); @@ -619,11 +636,16 @@ int vfp_restore_user_hwstate(struct user_vfp *ufp, struct user_vfp_exc *ufp_exc) /* Ensure the VFP is enabled. */ fpexc |= FPEXC_EN; +#ifdef CONFIG_OPENEULER_RASPBERRYPI /* Mask FPXEC_EX and FPEXC_FP2V if not required by VFP arch */ if ((fpsid & FPSID_ARCH_MASK) != (1 << FPSID_ARCH_BIT)) { /* Ensure FPINST2 is invalid and the exception flag is cleared. */ fpexc &= ~(FPEXC_EX | FPEXC_FP2V); } +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ + /* Ensure FPINST2 is invalid and the exception flag is cleared. */ + fpexc &= ~(FPEXC_EX | FPEXC_FP2V); +#endif hwstate->fpexc = fpexc; @@ -738,8 +760,12 @@ void kernel_neon_begin(void) cpu = get_cpu(); fpexc = fmrx(FPEXC) | FPEXC_EN; +#ifdef CONFIG_OPENEULER_RASPBERRYPI /* vfp_save_state oopses on VFP11 if EX bit set */ fmxr(FPEXC, fpexc & ~FPEXC_EX); +#else /* !CONFIG_OPENEULER_RASPBERRYPI */ + fmxr(FPEXC, fpexc); +#endif /* * Save the userland NEON/VFP state. Under UP, -- 2.20.1

2 1