- Kernel - mailweb.openeuler.org

[PATCH openEuler-22.03-LTS-SP1] ALSA: usb-audio: Stop parsing channels bits when all channels are found.
by Yuan Can 26 Sep '24

26 Sep '24

From: Johan Carlsson <johan.carlsson(a)teenage.engineering> stable inclusion from stable-v5.10.214 commit 5cd466673b34bac369334f66cbe14bb77b7d7827 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9Q91N CVE: CVE-2024-27436 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit a39d51ff1f52cd0b6fe7d379ac93bd8b4237d1b7 ] If a usb audio device sets more bits than the amount of channels it could write outside of the map array. Signed-off-by: Johan Carlsson <johan.carlsson(a)teenage.engineering> Fixes: 04324ccc75f9 ("ALSA: usb-audio: add channel map support") Message-ID: <20240313081509.9801-1-johan.carlsson(a)teenage.engineering> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Yuan Can <yuancan(a)huawei.com> --- sound/usb/stream.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/sound/usb/stream.c b/sound/usb/stream.c index c4f4585f9b85..f51e901a9689 100644 --- a/sound/usb/stream.c +++ b/sound/usb/stream.c @@ -301,9 +301,12 @@ static struct snd_pcm_chmap_elem *convert_chmap(int channels, unsigned int bits, c = 0; if (bits) { - for (; bits && *maps; maps++, bits >>= 1) + for (; bits && *maps; maps++, bits >>= 1) { if (bits & 1) chmap->map[c++] = *maps; + if (c == chmap->channels) + break; + } } else { /* If we're missing wChannelConfig, then guess something to make sure the channel map is not skipped entirely */ -- 2.17.1

2 1

[PATCH] fix: fix version showed in ethtool
by s00851154 26 Sep '24

26 Sep '24

Fix the version in the output of ethtool commands from empty to the kernel version. Use the macro HINIC3_NIC_DRV_VERSION to record driver changes for developers. The last number 1.0.X grow as minor changes, like this revision which does not affect the main functions. The middle number 1.X.1 iterates with changes to drivers functionality and fixing of serious bugs, and the first one X.0.1 means the number of releases of the driver. Signed-off-by: Shi Jing <shijing34(a)huawei.com> --- drivers/net/ethernet/huawei/hinic3/hinic3_ethtool.c | 1 - drivers/net/ethernet/huawei/hinic3/hinic3_nic_dev.h | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/net/ethernet/huawei/hinic3/hinic3_ethtool.c b/drivers/net/ethernet/huawei/hinic3/hinic3_ethtool.c index 5a1995956a555..4ea526329d155 100644 --- a/drivers/net/ethernet/huawei/hinic3/hinic3_ethtool.c +++ b/drivers/net/ethernet/huawei/hinic3/hinic3_ethtool.c @@ -45,7 +45,6 @@ static void hinic3_get_drvinfo(struct net_device *netdev, int err; strscpy(info->driver, HINIC3_NIC_DRV_NAME, sizeof(info->driver)); - strscpy(info->version, HINIC3_NIC_DRV_VERSION, sizeof(info->version)); strscpy(info->bus_info, pci_name(pdev), sizeof(info->bus_info)); err = hinic3_get_mgmt_version(nic_dev->hwdev, mgmt_ver, diff --git a/drivers/net/ethernet/huawei/hinic3/hinic3_nic_dev.h b/drivers/net/ethernet/huawei/hinic3/hinic3_nic_dev.h index 2967311aab76c..5868187eafa99 100644 --- a/drivers/net/ethernet/huawei/hinic3/hinic3_nic_dev.h +++ b/drivers/net/ethernet/huawei/hinic3/hinic3_nic_dev.h @@ -17,7 +17,7 @@ #include "hinic3_dcb.h" #define HINIC3_NIC_DRV_NAME "hinic3" -#define HINIC3_NIC_DRV_VERSION "" +#define HINIC3_NIC_DRV_VERSION "1.0.1" #define HINIC3_FUNC_IS_VF(hwdev) (hinic3_func_type(hwdev) == TYPE_VF) -- 2.28.0.windows.1

1 0

[PATCH openEuler-22.03-LTS-SP1 v2] drm/msm/dpu: cleanup FB if dpu_format_populate_layout fails
by Wang Hai 26 Sep '24

26 Sep '24

From: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> mainline inclusion from mainline-v6.11-rc5 commit bfa1a6283be390947d3649c482e5167186a37016 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IAOXZV CVE: CVE-2024-44982 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- If the dpu_format_populate_layout() fails, then FB is prepared, but not cleaned up. This ends up leaking the pin_count on the GEM object and causes a splat during DRM file closure: msm_obj->pin_count WARNING: CPU: 2 PID: 569 at drivers/gpu/drm/msm/msm_gem.c:121 update_lru_locked+0xc4/0xcc [...] Call trace: update_lru_locked+0xc4/0xcc put_pages+0xac/0x100 msm_gem_free_object+0x138/0x180 drm_gem_object_free+0x1c/0x30 drm_gem_object_handle_put_unlocked+0x108/0x10c drm_gem_object_release_handle+0x58/0x70 idr_for_each+0x68/0xec drm_gem_release+0x28/0x40 drm_file_free+0x174/0x234 drm_release+0xb0/0x160 __fput+0xc0/0x2c8 __fput_sync+0x50/0x5c __arm64_sys_close+0x38/0x7c invoke_syscall+0x48/0x118 el0_svc_common.constprop.0+0x40/0xe0 do_el0_svc+0x1c/0x28 el0_svc+0x4c/0x120 el0t_64_sync_handler+0x100/0x12c el0t_64_sync+0x190/0x194 irq event stamp: 129818 hardirqs last enabled at (129817): [<ffffa5f6d953fcc0>] console_unlock+0x118/0x124 hardirqs last disabled at (129818): [<ffffa5f6da7dcf04>] el1_dbg+0x24/0x8c softirqs last enabled at (129808): [<ffffa5f6d94afc18>] handle_softirqs+0x4c8/0x4e8 softirqs last disabled at (129785): [<ffffa5f6d94105e4>] __do_softirq+0x14/0x20 Fixes: 25fdd5933e4c ("drm/msm: Add SDM845 DPU support") Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> Reviewed-by: Abhinav Kumar <quic_abhinavk(a)quicinc.com> Patchwork: https://patchwork.freedesktop.org/patch/600714/ Link: https://lore.kernel.org/r/20240625-dpu-mode-config-width-v5-1-501d984d634f@… Signed-off-by: Abhinav Kumar <quic_abhinavk(a)quicinc.com> Conflicts: drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c [Conflict due to commit 9e4dde28e9cd ("drm/msm: Avoid dirtyfb stalls on video mode displays (v2)") didn't merge.] Signed-off-by: Wang Hai <wanghai38(a)huawei.com> --- v1->v2: fix build failed drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c index 8aa9f2335f57..e9a9522f3904 100644 --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c @@ -891,6 +891,8 @@ static int dpu_plane_prepare_fb(struct drm_plane *plane, new_state->fb, &layout); if (ret) { DPU_ERROR_PLANE(pdpu, "failed to get format layout, %d\n", ret); + if (pstate->aspace) + msm_framebuffer_cleanup(new_state->fb, pstate->aspace); return ret; } -- 2.17.1

2 1

[PATCH OLK-5.10 v2] drm/msm/dpu: cleanup FB if dpu_format_populate_layout fails
by Wang Hai 26 Sep '24

26 Sep '24

From: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> mainline inclusion from mainline-v6.11-rc5 commit bfa1a6283be390947d3649c482e5167186a37016 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IAOXZV CVE: CVE-2024-44982 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- If the dpu_format_populate_layout() fails, then FB is prepared, but not cleaned up. This ends up leaking the pin_count on the GEM object and causes a splat during DRM file closure: msm_obj->pin_count WARNING: CPU: 2 PID: 569 at drivers/gpu/drm/msm/msm_gem.c:121 update_lru_locked+0xc4/0xcc [...] Call trace: update_lru_locked+0xc4/0xcc put_pages+0xac/0x100 msm_gem_free_object+0x138/0x180 drm_gem_object_free+0x1c/0x30 drm_gem_object_handle_put_unlocked+0x108/0x10c drm_gem_object_release_handle+0x58/0x70 idr_for_each+0x68/0xec drm_gem_release+0x28/0x40 drm_file_free+0x174/0x234 drm_release+0xb0/0x160 __fput+0xc0/0x2c8 __fput_sync+0x50/0x5c __arm64_sys_close+0x38/0x7c invoke_syscall+0x48/0x118 el0_svc_common.constprop.0+0x40/0xe0 do_el0_svc+0x1c/0x28 el0_svc+0x4c/0x120 el0t_64_sync_handler+0x100/0x12c el0t_64_sync+0x190/0x194 irq event stamp: 129818 hardirqs last enabled at (129817): [<ffffa5f6d953fcc0>] console_unlock+0x118/0x124 hardirqs last disabled at (129818): [<ffffa5f6da7dcf04>] el1_dbg+0x24/0x8c softirqs last enabled at (129808): [<ffffa5f6d94afc18>] handle_softirqs+0x4c8/0x4e8 softirqs last disabled at (129785): [<ffffa5f6d94105e4>] __do_softirq+0x14/0x20 Fixes: 25fdd5933e4c ("drm/msm: Add SDM845 DPU support") Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> Reviewed-by: Abhinav Kumar <quic_abhinavk(a)quicinc.com> Patchwork: https://patchwork.freedesktop.org/patch/600714/ Link: https://lore.kernel.org/r/20240625-dpu-mode-config-width-v5-1-501d984d634f@… Signed-off-by: Abhinav Kumar <quic_abhinavk(a)quicinc.com> Conflicts: drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c [Conflict due to commit 9e4dde28e9cd ("drm/msm: Avoid dirtyfb stalls on video mode displays (v2)") didn't merge.] Signed-off-by: Wang Hai <wanghai38(a)huawei.com> --- v1->v2: fix build failed drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c index 8aa9f2335f57..e9a9522f3904 100644 --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c @@ -891,6 +891,8 @@ static int dpu_plane_prepare_fb(struct drm_plane *plane, new_state->fb, &layout); if (ret) { DPU_ERROR_PLANE(pdpu, "failed to get format layout, %d\n", ret); + if (pstate->aspace) + msm_framebuffer_cleanup(new_state->fb, pstate->aspace); return ret; } -- 2.17.1

2 1

[openeuler:OLK-6.6 2385/14122] kernel/cgroup/cgroup.c:3968:29: error: 'cgroup_psi_stat_show' undeclared here (not in a function); did you mean 'cgroup_stat_show'?
by kernel test robot 26 Sep '24

26 Sep '24

Hi Lu, FYI, the error/warning still remains. tree: https://gitee.com/openeuler/kernel.git OLK-6.6 head: 81a41d2ac1de43215c014bc71d907a026042e55b commit: 4a69b56540f40550585b9f93c35e415ac90bdecc [2385/14122] sched/psi: Introduce pressure.stat in psi config: loongarch-randconfig-r061-20240926 (https://download.01.org/0day-ci/archive/20240926/202409260923.Yh7ALQqg-lkp@…) compiler: loongarch64-linux-gcc (GCC) 14.1.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240926/202409260923.Yh7ALQqg-lkp@…) If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags | Reported-by: kernel test robot <lkp(a)intel.com> | Closes: https://lore.kernel.org/oe-kbuild-all/202409260923.Yh7ALQqg-lkp@intel.com/ All errors (new ones prefixed by >>): >> kernel/cgroup/cgroup.c:3968:29: error: 'cgroup_psi_stat_show' undeclared here (not in a function); did you mean 'cgroup_stat_show'? 3968 | .seq_show = cgroup_psi_stat_show, | ^~~~~~~~~~~~~~~~~~~~ | cgroup_stat_show vim +3968 kernel/cgroup/cgroup.c 3929 3930 struct cftype cgroup_v1_psi_files[] = { 3931 { 3932 .name = "io.pressure", 3933 .flags = CFTYPE_NO_PREFIX, 3934 .seq_show = cgroup_io_pressure_show, 3935 .write = cgroup_io_pressure_write, 3936 .poll = cgroup_pressure_poll, 3937 .release = cgroup_pressure_release, 3938 }, 3939 { 3940 .name = "memory.pressure", 3941 .flags = CFTYPE_NO_PREFIX, 3942 .seq_show = cgroup_memory_pressure_show, 3943 .write = cgroup_memory_pressure_write, 3944 .poll = cgroup_pressure_poll, 3945 .release = cgroup_pressure_release, 3946 }, 3947 { 3948 .name = "cpu.pressure", 3949 .flags = CFTYPE_NO_PREFIX, 3950 .seq_show = cgroup_cpu_pressure_show, 3951 .write = cgroup_cpu_pressure_write, 3952 .poll = cgroup_pressure_poll, 3953 .release = cgroup_pressure_release, 3954 }, 3955 #ifdef CONFIG_IRQ_TIME_ACCOUNTING 3956 { 3957 .name = "irq.pressure", 3958 .flags = CFTYPE_NO_PREFIX, 3959 .seq_show = cgroup_irq_pressure_show, 3960 .write = cgroup_irq_pressure_write, 3961 .poll = cgroup_pressure_poll, 3962 .release = cgroup_pressure_release, 3963 }, 3964 #endif 3965 { 3966 .name = "pressure.stat", 3967 .flags = CFTYPE_NO_PREFIX, > 3968 .seq_show = cgroup_psi_stat_show, 3969 }, 3970 { } /* terminate */ 3971 }; 3972 #endif 3973 #else /* CONFIG_PSI */ 3974 bool cgroup_psi_enabled(void) 3975 { 3976 return false; 3977 } 3978 -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0

[PATCH openEuler-22.03-LTS-SP1 0/2] CVE-2024-46781
by Zeng Heng 26 Sep '24

26 Sep '24

Ryusuke Konishi (2): nilfs2: fix use-after-free of nilfs_root in dirtying inodes via iput nilfs2: fix missing cleanup on rollforward recovery error fs/nilfs2/inode.c | 8 ++++++++ fs/nilfs2/recovery.c | 35 +++++++++++++++++++++++++++++++++-- fs/nilfs2/segment.c | 2 ++ fs/nilfs2/the_nilfs.h | 2 ++ 4 files changed, 45 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH openEuler-22.03-LTS-SP1 0/2] CVE-2024-46781
by Zeng Heng 26 Sep '24

26 Sep '24

Ryusuke Konishi (2): nilfs2: fix use-after-free of nilfs_root in dirtying inodes via iput nilfs2: fix missing cleanup on rollforward recovery error fs/nilfs2/inode.c | 8 ++++++++ fs/nilfs2/recovery.c | 35 +++++++++++++++++++++++++++++++++-- fs/nilfs2/segment.c | 2 ++ fs/nilfs2/the_nilfs.h | 2 ++ 4 files changed, 45 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 0/2] CVE-2024-46781
by Zeng Heng 26 Sep '24

26 Sep '24

Ryusuke Konishi (2): nilfs2: fix use-after-free of nilfs_root in dirtying inodes via iput nilfs2: fix missing cleanup on rollforward recovery error fs/nilfs2/inode.c | 8 ++++++++ fs/nilfs2/recovery.c | 35 +++++++++++++++++++++++++++++++++-- fs/nilfs2/segment.c | 2 ++ fs/nilfs2/the_nilfs.h | 2 ++ 4 files changed, 45 insertions(+), 2 deletions(-) -- 2.25.1

2 3

[PATCH OLK-6.6] perf/x86/intel: Limit the period on Haswell
by Li Huafei 26 Sep '24

26 Sep '24

From: Kan Liang <kan.liang(a)linux.intel.com> stable inclusion from stable-v6.6.51 commit 0eaf812aa1506704f3b78be87036860e5d0fe81d category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IR511 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 25dfc9e357af8aed1ca79b318a73f2c59c1f0b2b upstream. Running the ltp test cve-2015-3290 concurrently reports the following warnings. perfevents: irq loop stuck! WARNING: CPU: 31 PID: 32438 at arch/x86/events/intel/core.c:3174 intel_pmu_handle_irq+0x285/0x370 Call Trace: <NMI> ? __warn+0xa4/0x220 ? intel_pmu_handle_irq+0x285/0x370 ? __report_bug+0x123/0x130 ? intel_pmu_handle_irq+0x285/0x370 ? __report_bug+0x123/0x130 ? intel_pmu_handle_irq+0x285/0x370 ? report_bug+0x3e/0xa0 ? handle_bug+0x3c/0x70 ? exc_invalid_op+0x18/0x50 ? asm_exc_invalid_op+0x1a/0x20 ? irq_work_claim+0x1e/0x40 ? intel_pmu_handle_irq+0x285/0x370 perf_event_nmi_handler+0x3d/0x60 nmi_handle+0x104/0x330 Thanks to Thomas Gleixner's analysis, the issue is caused by the low initial period (1) of the frequency estimation algorithm, which triggers the defects of the HW, specifically erratum HSW11 and HSW143. (For the details, please refer https://lore.kernel.org/lkml/87plq9l5d2.ffs@tglx/) The HSW11 requires a period larger than 100 for the INST_RETIRED.ALL event, but the initial period in the freq mode is 1. The erratum is the same as the BDM11, which has been supported in the kernel. A minimum period of 128 is enforced as well on HSW. HSW143 is regarding that the fixed counter 1 may overcount 32 with the Hyper-Threading is enabled. However, based on the test, the hardware has more issues than it tells. Besides the fixed counter 1, the message 'interrupt took too long' can be observed on any counter which was armed with a period < 32 and two events expired in the same NMI. A minimum period of 32 is enforced for the rest of the events. The recommended workaround code of the HSW143 is not implemented. Because it only addresses the issue for the fixed counter. It brings extra overhead through extra MSR writing. No related overcounting issue has been reported so far. Fixes: 3a632cb229bf ("perf/x86/intel: Add simple Haswell PMU support") Reported-by: Li Huafei <lihuafei1(a)huawei.com> Suggested-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: Kan Liang <kan.liang(a)linux.intel.com> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: stable(a)vger.kernel.org Link: https://lore.kernel.org/all/20240819183004.3132920-1-kan.liang@linux.intel.… Closes: https://lore.kernel.org/lkml/20240729223328.327835-1-lihuafei1@huawei.com/ Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Conflicts: arch/x86/events/intel/core.c [ Context conflict due to the backported commit b0560bfd4b70 ("perf/x86/intel: Clean up the hybrid CPU type handling code"). ] Signed-off-by: Li Huafei <lihuafei1(a)huawei.com> --- arch/x86/events/intel/core.c | 23 +++++++++++++++++++++-- 1 file changed, 21 insertions(+), 2 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 8b8e9189fd41..e2c1c51d8a01 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -4520,6 +4520,25 @@ static enum hybrid_cpu_type adl_get_hybrid_cpu_type(void) return HYBRID_INTEL_CORE; } +static inline bool erratum_hsw11(struct perf_event *event) +{ + return (event->hw.config & INTEL_ARCH_EVENT_MASK) == + X86_CONFIG(.event=0xc0, .umask=0x01); +} + +/* + * The HSW11 requires a period larger than 100 which is the same as the BDM11. + * A minimum period of 128 is enforced as well for the INST_RETIRED.ALL. + * + * The message 'interrupt took too long' can be observed on any counter which + * was armed with a period < 32 and two events expired in the same NMI. + * A minimum period of 32 is enforced for the rest of the events. + */ +static void hsw_limit_period(struct perf_event *event, s64 *left) +{ + *left = max(*left, erratum_hsw11(event) ? 128 : 32); +} + /* * Broadwell: * @@ -4537,8 +4556,7 @@ static enum hybrid_cpu_type adl_get_hybrid_cpu_type(void) */ static void bdw_limit_period(struct perf_event *event, s64 *left) { - if ((event->hw.config & INTEL_ARCH_EVENT_MASK) == - X86_CONFIG(.event=0xc0, .umask=0x01)) { + if (erratum_hsw11(event)) { if (*left < 128) *left = 128; *left &= ~0x3fULL; @@ -6598,6 +6616,7 @@ __init int intel_pmu_init(void) x86_pmu.hw_config = hsw_hw_config; x86_pmu.get_event_constraints = hsw_get_event_constraints; + x86_pmu.limit_period = hsw_limit_period; x86_pmu.lbr_double_abort = true; extra_attr = boot_cpu_has(X86_FEATURE_RTM) ? hsw_format_attr : nhm_format_attr; -- 2.25.1

2 1

[PATCH OLK-5.10] perf/x86/intel: Limit the period on Haswell
by Li Huafei 26 Sep '24

26 Sep '24

From: Kan Liang <kan.liang(a)linux.intel.com> mainline inclusion from mainline-v6.11-rc7 commit 25dfc9e357af8aed1ca79b318a73f2c59c1f0b2b category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IAR511 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Running the ltp test cve-2015-3290 concurrently reports the following warnings. perfevents: irq loop stuck! WARNING: CPU: 31 PID: 32438 at arch/x86/events/intel/core.c:3174 intel_pmu_handle_irq+0x285/0x370 Call Trace: <NMI> ? __warn+0xa4/0x220 ? intel_pmu_handle_irq+0x285/0x370 ? __report_bug+0x123/0x130 ? intel_pmu_handle_irq+0x285/0x370 ? __report_bug+0x123/0x130 ? intel_pmu_handle_irq+0x285/0x370 ? report_bug+0x3e/0xa0 ? handle_bug+0x3c/0x70 ? exc_invalid_op+0x18/0x50 ? asm_exc_invalid_op+0x1a/0x20 ? irq_work_claim+0x1e/0x40 ? intel_pmu_handle_irq+0x285/0x370 perf_event_nmi_handler+0x3d/0x60 nmi_handle+0x104/0x330 Thanks to Thomas Gleixner's analysis, the issue is caused by the low initial period (1) of the frequency estimation algorithm, which triggers the defects of the HW, specifically erratum HSW11 and HSW143. (For the details, please refer https://lore.kernel.org/lkml/87plq9l5d2.ffs@tglx/) The HSW11 requires a period larger than 100 for the INST_RETIRED.ALL event, but the initial period in the freq mode is 1. The erratum is the same as the BDM11, which has been supported in the kernel. A minimum period of 128 is enforced as well on HSW. HSW143 is regarding that the fixed counter 1 may overcount 32 with the Hyper-Threading is enabled. However, based on the test, the hardware has more issues than it tells. Besides the fixed counter 1, the message 'interrupt took too long' can be observed on any counter which was armed with a period < 32 and two events expired in the same NMI. A minimum period of 32 is enforced for the rest of the events. The recommended workaround code of the HSW143 is not implemented. Because it only addresses the issue for the fixed counter. It brings extra overhead through extra MSR writing. No related overcounting issue has been reported so far. Fixes: 3a632cb229bf ("perf/x86/intel: Add simple Haswell PMU support") Reported-by: Li Huafei <lihuafei1(a)huawei.com> Suggested-by: Thomas Gleixner <tglx(a)linutronix.de> Signed-off-by: Kan Liang <kan.liang(a)linux.intel.com> Signed-off-by: Thomas Gleixner <tglx(a)linutronix.de> Cc: stable(a)vger.kernel.org Link: https://lore.kernel.org/all/20240819183004.3132920-1-kan.liang@linux.intel.… Closes: https://lore.kernel.org/lkml/20240729223328.327835-1-lihuafei1@huawei.com/ Conflicts: arch/x86/events/intel/core.c [ Adapted x86_pmu::limit_period signature due to commit 28f0f3c44b5c (“perf/x86: Change x86_pmu::limit_period signature”) not backported. ] Signed-off-by: Li Huafei <lihuafei1(a)huawei.com> --- arch/x86/events/intel/core.c | 23 +++++++++++++++++++++-- 1 file changed, 21 insertions(+), 2 deletions(-) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 4372ed2d1637..1786e8d85b6b 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -4414,6 +4414,25 @@ static u8 adl_get_hybrid_cpu_type(void) return hybrid_big; } +static inline bool erratum_hsw11(struct perf_event *event) +{ + return (event->hw.config & INTEL_ARCH_EVENT_MASK) == + X86_CONFIG(.event=0xc0, .umask=0x01); +} + +/* + * The HSW11 requires a period larger than 100 which is the same as the BDM11. + * A minimum period of 128 is enforced as well for the INST_RETIRED.ALL. + * + * The message 'interrupt took too long' can be observed on any counter which + * was armed with a period < 32 and two events expired in the same NMI. + * A minimum period of 32 is enforced for the rest of the events. + */ +static u64 hsw_limit_period(struct perf_event *event, u64 left) +{ + return max(left, erratum_hsw11(event) ? 128ULL : 32ULL); +} + /* * Broadwell: * @@ -4431,8 +4450,7 @@ static u8 adl_get_hybrid_cpu_type(void) */ static u64 bdw_limit_period(struct perf_event *event, u64 left) { - if ((event->hw.config & INTEL_ARCH_EVENT_MASK) == - X86_CONFIG(.event=0xc0, .umask=0x01)) { + if (erratum_hsw11(event)) { if (left < 128) left = 128; left &= ~0x3fULL; @@ -6406,6 +6424,7 @@ __init int intel_pmu_init(void) x86_pmu.hw_config = hsw_hw_config; x86_pmu.get_event_constraints = hsw_get_event_constraints; + x86_pmu.limit_period = hsw_limit_period; x86_pmu.lbr_double_abort = true; extra_attr = boot_cpu_has(X86_FEATURE_RTM) ? hsw_format_attr : nhm_format_attr; -- 2.25.1

2 1

2024

2023

2022

2021

2020

2019

Kernel