- Kernel - mailweb.openeuler.org

[PATCH openEuler-23.09 v2 0/2] configs: open drivers' config
by Liu Chao 04 Sep '23

04 Sep '23

From: root <root(a)localhost.localdomain> openeuler inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/I7VEEV CVE: NA -------------------------------- Changes in v2: Enable more driver configurations, such as crypto, Intel platform, virtio, vdpa, and so on. Liu Chao (2): configs: aarch64: open drivers' config configs: open drivers' config arch/arm64/configs/openeuler_defconfig | 4525 ++++++++++++++++++----- arch/x86/configs/openeuler_defconfig | 4617 ++++++++++++++++-------- 2 files changed, 6594 insertions(+), 2548 deletions(-) -- 2.33.0

2 3

[PATCH openEuler-23.09 0/2] configs: open drivers' config
by Liu Chao 04 Sep '23

04 Sep '23

The number of drivers opened by openEuler is less than other distributions. In order to narrow the gap with other distributions and support more hardware devices, especially various PC hardware commonly used by developers and personal users, more hardware driver modules need to be enabled. Reference:https://gitee.com/openeuler/TC/blob/master/oEEP/oEEP-0006%20openE… Liu Chao (2): configs: aarch64: open drivers' config configs: x86_64: open drivers' config arch/arm64/configs/openeuler_defconfig | 3107 +++++++++++++++++++----- arch/x86/configs/openeuler_defconfig | 2694 ++++++++++++-------- 2 files changed, 4142 insertions(+), 1659 deletions(-) -- 2.33.0

1 3

[PATCH OLK-5.10] sched/smt: fix unbalance sched_smt_present dec/inc
by Yu Liao 04 Sep '23

04 Sep '23

From: Yang Yingliang <yangyingliang(a)huawei.com> hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7YBSW CVE: NA -------------------------------- If cpuset_cpu_inactive() fails in sched_cpu_deactivate(), the cpu offline failed, but sched_smt_present is decreased before calling sched_cpu_deactivate, it leads unbalance dec/inc, so fix it by increasing sched_smt_present in the error path. Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Signed-off-by: Yu Liao <liaoyu15(a)huawei.com> --- kernel/sched/core.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 920a5af0e346..669a62d16ba6 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -8060,6 +8060,10 @@ int sched_cpu_deactivate(unsigned int cpu) ret = cpuset_cpu_inactive(cpu); if (ret) { +#ifdef CONFIG_SCHED_SMT + if (cpumask_weight(cpu_smt_mask(cpu)) == 2) + static_branch_inc_cpuslocked(&sched_smt_present); +#endif set_cpu_active(cpu, true); return ret; } -- 2.25.1

2 4

[PATCH openEuler-23.09 0/9] add kabi macros helpers
by Yu Liao 04 Sep '23

04 Sep '23

Add kabi macros helpers. Li Lingfeng (1): kabi: deduplication friendly structs Wei Li (2): kabi: fix build error when CONFIG_KABI_RESERVE=n kabi: cleanup config entries of kabi Xie XiuQi (3): kabi: add kabi helper macros kabi: add KABI_SIZE_ALIGN_CHECKS for more stringent kabi checks kabi: enables more stringent kabi checks Zheng Zengkai (3): KABI: Add CONFIG_KABI_RESERVE to control KABI padding reserve openeuler_defconfig: Enable CONFIG_KABI_RESERVE for x86 and arm64 kabi: Generalize naming of kabi helper macros arch/arm64/configs/openeuler_defconfig | 2 + arch/x86/configs/openeuler_defconfig | 2 + include/linux/kabi.h | 509 +++++++++++++++++++++++++ init/Kconfig | 16 + 4 files changed, 529 insertions(+) create mode 100644 include/linux/kabi.h -- 2.33.0

2 10

[PATCH openEuler-23.09 0/8] support Android vendor hooks
by Yu Liao 04 Sep '23

04 Sep '23

support android vendor hooks for openEuler-23.09 Jialin Zhang (2): vendor_hooks: make android vendor hooks feature generic. openeuler_defconfig: enable CONFIG_VENDOR_HOOKS for x86 and arm64 Nick Desaulniers (1): ANDROID: vendor_hooks: fix __section macro Todd Kjos (5): ANDROID: add support for vendor hooks ANDROID: fix redefinition error for restricted vendor hooks ANDROID: use static_call() for restricted hooks ANDROID: simplify vendor hooks for non-GKI builds ANDROID: fixup restricted hooks after tracepont refactoring arch/arm64/configs/openeuler_defconfig | 6 ++ arch/x86/configs/openeuler_defconfig | 6 ++ drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/hooks/Kconfig | 13 +++ drivers/hooks/Makefile | 4 + drivers/hooks/vendor_hooks.c | 16 ++++ include/trace/hooks/vendor_hooks.h | 119 +++++++++++++++++++++++++ 8 files changed, 167 insertions(+) create mode 100644 drivers/hooks/Kconfig create mode 100644 drivers/hooks/Makefile create mode 100644 drivers/hooks/vendor_hooks.c create mode 100644 include/trace/hooks/vendor_hooks.h -- 2.33.0

2 9

[PATCH openEuler-23.09 0/8] support Android vendor hooks
by Yu Liao 04 Sep '23

04 Sep '23

support android vendor hooks for openEuler-23.09 Jialin Zhang (2): vendor_hooks: make android vendor hooks feature generic. openeuler_defconfig: enable CONFIG_VENDOR_HOOKS for x86 and arm64 Nick Desaulniers (1): ANDROID: vendor_hooks: fix __section macro Todd Kjos (5): ANDROID: add support for vendor hooks ANDROID: fix redefinition error for restricted vendor hooks ANDROID: use static_call() for restricted hooks ANDROID: simplify vendor hooks for non-GKI builds ANDROID: fixup restricted hooks after tracepont refactoring arch/arm64/configs/openeuler_defconfig | 6 ++ arch/x86/configs/openeuler_defconfig | 6 ++ drivers/Kconfig | 2 + drivers/Makefile | 1 + drivers/hooks/Kconfig | 13 +++ drivers/hooks/Makefile | 4 + drivers/hooks/vendor_hooks.c | 16 ++++ include/trace/hooks/vendor_hooks.h | 119 +++++++++++++++++++++++++ 8 files changed, 167 insertions(+) create mode 100644 drivers/hooks/Kconfig create mode 100644 drivers/hooks/Makefile create mode 100644 drivers/hooks/vendor_hooks.c create mode 100644 include/trace/hooks/vendor_hooks.h -- 2.33.0

1 8

[PATCH openEuler-23.09 0/1] Fix wrong initialization of rpg_kmsg_cbftns
by Chen Jun 04 Sep '23

04 Sep '23

*** BLURB HERE *** Chen Jun (1): remote_pager: Fix wrong initialization of rpg_kmsg_cbftns drivers/remote_pager/main.c | 4 ++-- drivers/remote_pager/msg_handler_comm.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) -- 2.17.1

2 2

[PATCH openEuler-23.09 1/1] remote_pager: Fix wrong initialization of rpg_kmsg_cbftns
by Chen Jun 04 Sep '23

04 Sep '23

euleros inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7WLVX --------------------------------------------- The handlers for GMEM_PAGE_FAULT_REQUEST and GMEM_EVICT_PAGE_REQUEST are not set if CONFIG_REMOTE_PAGER_MASTER=m Signed-off-by: Chen Jun <chenjun102(a)huawei.com> --- drivers/remote_pager/main.c | 4 ++-- drivers/remote_pager/msg_handler_comm.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/remote_pager/main.c b/drivers/remote_pager/main.c index 66f303ce61c9..b7c8d8440bfa 100644 --- a/drivers/remote_pager/main.c +++ b/drivers/remote_pager/main.c @@ -10,7 +10,7 @@ #include "msg_chan/msg_layer/msg_layer.h" #include "msg_handler.h" -#ifdef CONFIG_REMOTE_PAGER_SLAVE +#if IS_ENABLED(CONFIG_REMOTE_PAGER_SLAVE) #include "swap/device/swap_manager.h" #endif @@ -18,7 +18,7 @@ static int __init remote_pager_init(void) { msg_handle_init(); -#ifdef CONFIG_REMOTE_PAGER_SLAVE +#if IS_ENABLED(CONFIG_REMOTE_PAGER_SLAVE) init_swap_manager("list_lru"); #endif return 0; diff --git a/drivers/remote_pager/msg_handler_comm.c b/drivers/remote_pager/msg_handler_comm.c index 129a4247305f..f195ad15e646 100644 --- a/drivers/remote_pager/msg_handler_comm.c +++ b/drivers/remote_pager/msg_handler_comm.c @@ -17,7 +17,7 @@ static struct workqueue_struct *remote_pager_wq; struct msg_handler_st rpg_kmsg_cbftns[GMEM_MSG_MAX_ID] = { -#ifdef CONFIG_REMOTE_PAGER_SLAVE +#if IS_ENABLED(CONFIG_REMOTE_PAGER_SLAVE) /* HOST TO REMOTE */ [GMEM_TASK_PAIRING_REQUEST] = { gmem_handle_task_pairing @@ -42,7 +42,7 @@ struct msg_handler_st rpg_kmsg_cbftns[GMEM_MSG_MAX_ID] = { }, #endif -#ifdef CONFIG_REMOTE_PAGER_MASTER +#if IS_ENABLED(CONFIG_REMOTE_PAGER_MASTER) /* REMOTE TO HOST */ [GMEM_PAGE_FAULT_REQUEST] = { gmem_handle_dev_fault -- 2.17.1

2 1

[PATCH openEuler-1.0-LTS] tracing: Fix race issue between cpu buffer write and swap
by Zheng Yejian 04 Sep '23

04 Sep '23

mainline inclusion from mainline-v6.6-rc1 commit 3163f635b20e9e1fb4659e74f47918c9dddfe64e category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7YE3Y Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ------------------------------------------- Warning happened in rb_end_commit() at code: if (RB_WARN_ON(cpu_buffer, !local_read(&cpu_buffer->committing))) WARNING: CPU: 0 PID: 139 at kernel/trace/ring_buffer.c:3142 rb_commit+0x402/0x4a0 Call Trace: ring_buffer_unlock_commit+0x42/0x250 trace_buffer_unlock_commit_regs+0x3b/0x250 trace_event_buffer_commit+0xe5/0x440 trace_event_buffer_reserve+0x11c/0x150 trace_event_raw_event_sched_switch+0x23c/0x2c0 __traceiter_sched_switch+0x59/0x80 __schedule+0x72b/0x1580 schedule+0x92/0x120 worker_thread+0xa0/0x6f0 It is because the race between writing event into cpu buffer and swapping cpu buffer through file per_cpu/cpu0/snapshot: Write on CPU 0 Swap buffer by per_cpu/cpu0/snapshot on CPU 1 -------- -------- tracing_snapshot_write() [...] ring_buffer_lock_reserve() cpu_buffer = buffer->buffers[cpu]; // 1. Suppose find 'cpu_buffer_a'; [...] rb_reserve_next_event() [...] ring_buffer_swap_cpu() if (local_read(&cpu_buffer_a->committing)) goto out_dec; if (local_read(&cpu_buffer_b->committing)) goto out_dec; buffer_a->buffers[cpu] = cpu_buffer_b; buffer_b->buffers[cpu] = cpu_buffer_a; // 2. cpu_buffer has swapped here. rb_start_commit(cpu_buffer); if (unlikely(READ_ONCE(cpu_buffer->buffer) != buffer)) { // 3. This check passed due to 'cpu_buffer->buffer' [...] // has not changed here. return NULL; } cpu_buffer_b->buffer = buffer_a; cpu_buffer_a->buffer = buffer_b; [...] // 4. Reserve event from 'cpu_buffer_a'. ring_buffer_unlock_commit() [...] cpu_buffer = buffer->buffers[cpu]; // 5. Now find 'cpu_buffer_b' !!! rb_commit(cpu_buffer) rb_end_commit() // 6. WARN for the wrong 'committing' state !!! Based on above analysis, we can easily reproduce by following testcase: ``` bash #!/bin/bash dmesg -n 7 sysctl -w kernel.panic_on_warn=1 TR=/sys/kernel/tracing echo 7 > ${TR}/buffer_size_kb echo "sched:sched_switch" > ${TR}/set_event while [ true ]; do echo 1 > ${TR}/per_cpu/cpu0/snapshot done & while [ true ]; do echo 1 > ${TR}/per_cpu/cpu0/snapshot done & while [ true ]; do echo 1 > ${TR}/per_cpu/cpu0/snapshot done & ``` To fix it, IIUC, we can use smp_call_function_single() to do the swap on the target cpu where the buffer is located, so that above race would be avoided. Link: https://lore.kernel.org/linux-trace-kernel/20230831132739.4070878-1-zhengye… Cc: <mhiramat(a)kernel.org> Fixes: f1affcaaa861 ("tracing: Add snapshot in the per_cpu trace directories") Signed-off-by: Zheng Yejian <zhengyejian1(a)huawei.com> Signed-off-by: Steven Rostedt (Google) <rostedt(a)goodmis.org> Signed-off-by: Zheng Yejian <zhengyejian1(a)huawei.com> --- kernel/trace/trace.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-) diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 8b5ede27b38e..64fd3c3f84f2 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -6463,6 +6463,11 @@ static int tracing_snapshot_open(struct inode *inode, struct file *file) return ret; } +static void tracing_swap_cpu_buffer(void *tr) +{ + update_max_tr_single((struct trace_array *)tr, current, smp_processor_id()); +} + static ssize_t tracing_snapshot_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) @@ -6512,13 +6517,15 @@ tracing_snapshot_write(struct file *filp, const char __user *ubuf, size_t cnt, ret = tracing_alloc_snapshot_instance(tr); if (ret < 0) break; - local_irq_disable(); /* Now, we're going to swap */ - if (iter->cpu_file == RING_BUFFER_ALL_CPUS) + if (iter->cpu_file == RING_BUFFER_ALL_CPUS) { + local_irq_disable(); update_max_tr(tr, current, smp_processor_id()); - else - update_max_tr_single(tr, current, iter->cpu_file); - local_irq_enable(); + local_irq_enable(); + } else { + smp_call_function_single(iter->cpu_file, tracing_swap_cpu_buffer, + (void *)tr, 1); + } break; default: if (tr->allocated_snapshot) { -- 2.25.1

2 1

[PATCH OLK-5.10] tracing: Fix race issue between cpu buffer write and swap
by Zheng Yejian 04 Sep '23

04 Sep '23

mainline inclusion from mainline-v6.6-rc1 commit 3163f635b20e9e1fb4659e74f47918c9dddfe64e category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I7YE3Y CVE: NA Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… ------------------------------------------- Warning happened in rb_end_commit() at code: if (RB_WARN_ON(cpu_buffer, !local_read(&cpu_buffer->committing))) WARNING: CPU: 0 PID: 139 at kernel/trace/ring_buffer.c:3142 rb_commit+0x402/0x4a0 Call Trace: ring_buffer_unlock_commit+0x42/0x250 trace_buffer_unlock_commit_regs+0x3b/0x250 trace_event_buffer_commit+0xe5/0x440 trace_event_buffer_reserve+0x11c/0x150 trace_event_raw_event_sched_switch+0x23c/0x2c0 __traceiter_sched_switch+0x59/0x80 __schedule+0x72b/0x1580 schedule+0x92/0x120 worker_thread+0xa0/0x6f0 It is because the race between writing event into cpu buffer and swapping cpu buffer through file per_cpu/cpu0/snapshot: Write on CPU 0 Swap buffer by per_cpu/cpu0/snapshot on CPU 1 -------- -------- tracing_snapshot_write() [...] ring_buffer_lock_reserve() cpu_buffer = buffer->buffers[cpu]; // 1. Suppose find 'cpu_buffer_a'; [...] rb_reserve_next_event() [...] ring_buffer_swap_cpu() if (local_read(&cpu_buffer_a->committing)) goto out_dec; if (local_read(&cpu_buffer_b->committing)) goto out_dec; buffer_a->buffers[cpu] = cpu_buffer_b; buffer_b->buffers[cpu] = cpu_buffer_a; // 2. cpu_buffer has swapped here. rb_start_commit(cpu_buffer); if (unlikely(READ_ONCE(cpu_buffer->buffer) != buffer)) { // 3. This check passed due to 'cpu_buffer->buffer' [...] // has not changed here. return NULL; } cpu_buffer_b->buffer = buffer_a; cpu_buffer_a->buffer = buffer_b; [...] // 4. Reserve event from 'cpu_buffer_a'. ring_buffer_unlock_commit() [...] cpu_buffer = buffer->buffers[cpu]; // 5. Now find 'cpu_buffer_b' !!! rb_commit(cpu_buffer) rb_end_commit() // 6. WARN for the wrong 'committing' state !!! Based on above analysis, we can easily reproduce by following testcase: ``` bash #!/bin/bash dmesg -n 7 sysctl -w kernel.panic_on_warn=1 TR=/sys/kernel/tracing echo 7 > ${TR}/buffer_size_kb echo "sched:sched_switch" > ${TR}/set_event while [ true ]; do echo 1 > ${TR}/per_cpu/cpu0/snapshot done & while [ true ]; do echo 1 > ${TR}/per_cpu/cpu0/snapshot done & while [ true ]; do echo 1 > ${TR}/per_cpu/cpu0/snapshot done & ``` To fix it, IIUC, we can use smp_call_function_single() to do the swap on the target cpu where the buffer is located, so that above race would be avoided. Link: https://lore.kernel.org/linux-trace-kernel/20230831132739.4070878-1-zhengye… Cc: <mhiramat(a)kernel.org> Fixes: f1affcaaa861 ("tracing: Add snapshot in the per_cpu trace directories") Signed-off-by: Zheng Yejian <zhengyejian1(a)huawei.com> Signed-off-by: Steven Rostedt (Google) <rostedt(a)goodmis.org> Signed-off-by: Zheng Yejian <zhengyejian1(a)huawei.com> --- kernel/trace/trace.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-) diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 2ba13e9d61bd..6eead8a61195 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -7068,6 +7068,11 @@ static int tracing_snapshot_open(struct inode *inode, struct file *file) return ret; } +static void tracing_swap_cpu_buffer(void *tr) +{ + update_max_tr_single((struct trace_array *)tr, current, smp_processor_id()); +} + static ssize_t tracing_snapshot_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) @@ -7126,13 +7131,15 @@ tracing_snapshot_write(struct file *filp, const char __user *ubuf, size_t cnt, ret = tracing_alloc_snapshot_instance(tr); if (ret < 0) break; - local_irq_disable(); /* Now, we're going to swap */ - if (iter->cpu_file == RING_BUFFER_ALL_CPUS) + if (iter->cpu_file == RING_BUFFER_ALL_CPUS) { + local_irq_disable(); update_max_tr(tr, current, smp_processor_id(), NULL); - else - update_max_tr_single(tr, current, iter->cpu_file); - local_irq_enable(); + local_irq_enable(); + } else { + smp_call_function_single(iter->cpu_file, tracing_swap_cpu_buffer, + (void *)tr, 1); + } break; default: if (tr->allocated_snapshot) { -- 2.25.1

2 1

2024

2023

2022

2021

2020

2019

Kernel