From 1e106aa3509b86738769775969822ffc1ec21bf4 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 6 Nov 2020 11:52:05 +0300 Subject: futex: Don't enable IRQs unconditionally in put_pi_state() The exit_pi_state_list() function calls put_pi_state() with IRQs disabled and is not expecting that IRQs will be enabled inside the function. Use the _irqsave() variant so that IRQs are restored to the original state instead of being enabled unconditionally. Fixes: 153fbd1226fb ("futex: Fix more put_pi_state() vs. exit_pi_state_list() races") Signed-off-by: Dan Carpenter Signed-off-by: Thomas Gleixner Acked-by: Peter Zijlstra (Intel) Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/20201106085205.GA1159983@mwanda --- kernel/futex.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/futex.c b/kernel/futex.c index ac328874f6e5..00259c7e288e 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -788,8 +788,9 @@ static void put_pi_state(struct futex_pi_state *pi_state) */ if (pi_state->owner) { struct task_struct *owner; + unsigned long flags; - raw_spin_lock_irq(&pi_state->pi_mutex.wait_lock); + raw_spin_lock_irqsave(&pi_state->pi_mutex.wait_lock, flags); owner = pi_state->owner; if (owner) { raw_spin_lock(&owner->pi_lock); @@ -797,7 +798,7 @@ static void put_pi_state(struct futex_pi_state *pi_state) raw_spin_unlock(&owner->pi_lock); } rt_mutex_proxy_unlock(&pi_state->pi_mutex, owner); - raw_spin_unlock_irq(&pi_state->pi_mutex.wait_lock); + raw_spin_unlock_irqrestore(&pi_state->pi_mutex.wait_lock, flags); } if (current->pi_state_cache) { -- cgit v1.2.3 From 267fb27352b6fc9fdbad753127a239f75618ecbc Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 30 Oct 2020 15:50:32 +0100 Subject: perf: Reduce stack usage of perf_output_begin() __perf_output_begin() has an on-stack struct perf_sample_data in the unlikely case it needs to generate a LOST record. However, every call to perf_output_begin() must already have a perf_sample_data on-stack. Reported-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20201030151954.985416146@infradead.org --- arch/powerpc/perf/imc-pmu.c | 2 +- arch/s390/kernel/perf_cpum_sf.c | 2 +- arch/x86/events/intel/ds.c | 4 ++-- include/linux/perf_event.h | 7 +++++-- kernel/events/core.c | 32 +++++++++++++++++--------------- kernel/events/ring_buffer.c | 20 +++++++++++--------- 6 files changed, 37 insertions(+), 30 deletions(-) (limited to 'kernel') diff --git a/arch/powerpc/perf/imc-pmu.c b/arch/powerpc/perf/imc-pmu.c index 9ed4fcccf8a9..7b25548ec42b 100644 --- a/arch/powerpc/perf/imc-pmu.c +++ b/arch/powerpc/perf/imc-pmu.c @@ -1336,7 +1336,7 @@ static void dump_trace_imc_data(struct perf_event *event) /* If this is a valid record, create the sample */ struct perf_output_handle handle; - if (perf_output_begin(&handle, event, header.size)) + if (perf_output_begin(&handle, &data, event, header.size)) return; perf_output_sample(&handle, &header, &data, event); diff --git a/arch/s390/kernel/perf_cpum_sf.c b/arch/s390/kernel/perf_cpum_sf.c index 4f9e4626df55..00255ae3979d 100644 --- a/arch/s390/kernel/perf_cpum_sf.c +++ b/arch/s390/kernel/perf_cpum_sf.c @@ -672,7 +672,7 @@ static void cpumsf_output_event_pid(struct perf_event *event, rcu_read_lock(); perf_prepare_sample(&header, data, event, regs); - if (perf_output_begin(&handle, event, header.size)) + if (perf_output_begin(&handle, data, event, header.size)) goto out; /* Update the process ID (see also kernel/events/core.c) */ diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index 404315df1e16..cd2ae14a0a98 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -642,8 +642,8 @@ int intel_pmu_drain_bts_buffer(void) rcu_read_lock(); perf_prepare_sample(&header, &data, event, ®s); - if (perf_output_begin(&handle, event, header.size * - (top - base - skip))) + if (perf_output_begin(&handle, &data, event, + header.size * (top - base - skip))) goto unlock; for (at = base; at < top; at++) { diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index 0c19d279b97f..b775ae0a8c87 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -1400,11 +1400,14 @@ perf_event_addr_filters(struct perf_event *event) extern void perf_event_addr_filters_sync(struct perf_event *event); extern int perf_output_begin(struct perf_output_handle *handle, + struct perf_sample_data *data, struct perf_event *event, unsigned int size); extern int perf_output_begin_forward(struct perf_output_handle *handle, - struct perf_event *event, - unsigned int size); + struct perf_sample_data *data, + struct perf_event *event, + unsigned int size); extern int perf_output_begin_backward(struct perf_output_handle *handle, + struct perf_sample_data *data, struct perf_event *event, unsigned int size); diff --git a/kernel/events/core.c b/kernel/events/core.c index 5a29ab09e72d..fc681c7c1e03 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -7186,6 +7186,7 @@ __perf_event_output(struct perf_event *event, struct perf_sample_data *data, struct pt_regs *regs, int (*output_begin)(struct perf_output_handle *, + struct perf_sample_data *, struct perf_event *, unsigned int)) { @@ -7198,7 +7199,7 @@ __perf_event_output(struct perf_event *event, perf_prepare_sample(&header, data, event, regs); - err = output_begin(&handle, event, header.size); + err = output_begin(&handle, data, event, header.size); if (err) goto exit; @@ -7264,7 +7265,7 @@ perf_event_read_event(struct perf_event *event, int ret; perf_event_header__init_id(&read_event.header, &sample, event); - ret = perf_output_begin(&handle, event, read_event.header.size); + ret = perf_output_begin(&handle, &sample, event, read_event.header.size); if (ret) return; @@ -7533,7 +7534,7 @@ static void perf_event_task_output(struct perf_event *event, perf_event_header__init_id(&task_event->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, &sample, event, task_event->event_id.header.size); if (ret) goto out; @@ -7636,7 +7637,7 @@ static void perf_event_comm_output(struct perf_event *event, return; perf_event_header__init_id(&comm_event->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, &sample, event, comm_event->event_id.header.size); if (ret) @@ -7736,7 +7737,7 @@ static void perf_event_namespaces_output(struct perf_event *event, perf_event_header__init_id(&namespaces_event->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, &sample, event, namespaces_event->event_id.header.size); if (ret) goto out; @@ -7863,7 +7864,7 @@ static void perf_event_cgroup_output(struct perf_event *event, void *data) perf_event_header__init_id(&cgroup_event->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, &sample, event, cgroup_event->event_id.header.size); if (ret) goto out; @@ -7989,7 +7990,7 @@ static void perf_event_mmap_output(struct perf_event *event, } perf_event_header__init_id(&mmap_event->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, &sample, event, mmap_event->event_id.header.size); if (ret) goto out; @@ -8299,7 +8300,7 @@ void perf_event_aux_event(struct perf_event *event, unsigned long head, int ret; perf_event_header__init_id(&rec.header, &sample, event); - ret = perf_output_begin(&handle, event, rec.header.size); + ret = perf_output_begin(&handle, &sample, event, rec.header.size); if (ret) return; @@ -8333,7 +8334,7 @@ void perf_log_lost_samples(struct perf_event *event, u64 lost) perf_event_header__init_id(&lost_samples_event.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, &sample, event, lost_samples_event.header.size); if (ret) return; @@ -8388,7 +8389,7 @@ static void perf_event_switch_output(struct perf_event *event, void *data) perf_event_header__init_id(&se->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, se->event_id.header.size); + ret = perf_output_begin(&handle, &sample, event, se->event_id.header.size); if (ret) return; @@ -8463,7 +8464,7 @@ static void perf_log_throttle(struct perf_event *event, int enable) perf_event_header__init_id(&throttle_event.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, &sample, event, throttle_event.header.size); if (ret) return; @@ -8506,7 +8507,7 @@ static void perf_event_ksymbol_output(struct perf_event *event, void *data) perf_event_header__init_id(&ksymbol_event->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, &sample, event, ksymbol_event->event_id.header.size); if (ret) return; @@ -8596,7 +8597,7 @@ static void perf_event_bpf_output(struct perf_event *event, void *data) perf_event_header__init_id(&bpf_event->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, + ret = perf_output_begin(&handle, data, event, bpf_event->event_id.header.size); if (ret) return; @@ -8705,7 +8706,8 @@ static void perf_event_text_poke_output(struct perf_event *event, void *data) perf_event_header__init_id(&text_poke_event->event_id.header, &sample, event); - ret = perf_output_begin(&handle, event, text_poke_event->event_id.header.size); + ret = perf_output_begin(&handle, &sample, event, + text_poke_event->event_id.header.size); if (ret) return; @@ -8786,7 +8788,7 @@ static void perf_log_itrace_start(struct perf_event *event) rec.tid = perf_event_tid(event, current); perf_event_header__init_id(&rec.header, &sample, event); - ret = perf_output_begin(&handle, event, rec.header.size); + ret = perf_output_begin(&handle, &sample, event, rec.header.size); if (ret) return; diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c index 192b8abc6330..ef91ae75ca56 100644 --- a/kernel/events/ring_buffer.c +++ b/kernel/events/ring_buffer.c @@ -147,6 +147,7 @@ ring_buffer_has_space(unsigned long head, unsigned long tail, static __always_inline int __perf_output_begin(struct perf_output_handle *handle, + struct perf_sample_data *data, struct perf_event *event, unsigned int size, bool backward) { @@ -237,18 +238,16 @@ __perf_output_begin(struct perf_output_handle *handle, handle->size = (1UL << page_shift) - offset; if (unlikely(have_lost)) { - struct perf_sample_data sample_data; - lost_event.header.size = sizeof(lost_event); lost_event.header.type = PERF_RECORD_LOST; lost_event.header.misc = 0; lost_event.id = event->id; lost_event.lost = local_xchg(&rb->lost, 0); - perf_event_header__init_id(&lost_event.header, - &sample_data, event); + /* XXX mostly redundant; @data is already fully initializes */ + perf_event_header__init_id(&lost_event.header, data, event); perf_output_put(handle, lost_event); - perf_event__output_id_sample(event, handle, &sample_data); + perf_event__output_id_sample(event, handle, data); } return 0; @@ -263,22 +262,25 @@ out: } int perf_output_begin_forward(struct perf_output_handle *handle, - struct perf_event *event, unsigned int size) + struct perf_sample_data *data, + struct perf_event *event, unsigned int size) { - return __perf_output_begin(handle, event, size, false); + return __perf_output_begin(handle, data, event, size, false); } int perf_output_begin_backward(struct perf_output_handle *handle, + struct perf_sample_data *data, struct perf_event *event, unsigned int size) { - return __perf_output_begin(handle, event, size, true); + return __perf_output_begin(handle, data, event, size, true); } int perf_output_begin(struct perf_output_handle *handle, + struct perf_sample_data *data, struct perf_event *event, unsigned int size) { - return __perf_output_begin(handle, event, size, + return __perf_output_begin(handle, data, event, size, unlikely(is_write_backward(event))); } -- cgit v1.2.3 From ce0f17fc93f63ee91428af10b7b2ddef38cd19e5 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 30 Oct 2020 12:49:45 +0100 Subject: perf: Fix get_recursion_context() One should use in_serving_softirq() to detect SoftIRQ context. Fixes: 96f6d4444302 ("perf_counter: avoid recursion") Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20201030151955.120572175@infradead.org --- kernel/events/internal.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/events/internal.h b/kernel/events/internal.h index fcbf5616a441..402054e755f2 100644 --- a/kernel/events/internal.h +++ b/kernel/events/internal.h @@ -211,7 +211,7 @@ static inline int get_recursion_context(int *recursion) rctx = 3; else if (in_irq()) rctx = 2; - else if (in_softirq()) + else if (in_serving_softirq()) rctx = 1; else rctx = 0; -- cgit v1.2.3 From 09da9c81253dd8e43e0d2d7cea02de6f9f19499d Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 30 Oct 2020 13:43:16 +0100 Subject: perf: Optimize get_recursion_context() "Look ma, no branches!" Signed-off-by: Peter Zijlstra (Intel) Acked-by: Jesper Dangaard Brouer Link: https://lkml.kernel.org/r/20201030151955.187580298@infradead.org --- kernel/events/internal.h | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/events/internal.h b/kernel/events/internal.h index 402054e755f2..228801e20788 100644 --- a/kernel/events/internal.h +++ b/kernel/events/internal.h @@ -205,16 +205,12 @@ DEFINE_OUTPUT_COPY(__output_copy_user, arch_perf_out_copy_user) static inline int get_recursion_context(int *recursion) { - int rctx; - - if (unlikely(in_nmi())) - rctx = 3; - else if (in_irq()) - rctx = 2; - else if (in_serving_softirq()) - rctx = 1; - else - rctx = 0; + unsigned int pc = preempt_count(); + unsigned char rctx = 0; + + rctx += !!(pc & (NMI_MASK)); + rctx += !!(pc & (NMI_MASK | HARDIRQ_MASK)); + rctx += !!(pc & (NMI_MASK | HARDIRQ_MASK | SOFTIRQ_OFFSET)); if (recursion[rctx]) return -1; -- cgit v1.2.3 From 76a4efa80900fc40e0fdf243b42aec9fb8c35d24 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 30 Oct 2020 12:14:21 +0100 Subject: perf/arch: Remove perf_sample_data::regs_user_copy struct perf_sample_data lives on-stack, we should be careful about it's size. Furthermore, the pt_regs copy in there is only because x86_64 is a trainwreck, solve it differently. Reported-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Tested-by: Steven Rostedt Link: https://lkml.kernel.org/r/20201030151955.258178461@infradead.org --- arch/arm/kernel/perf_regs.c | 3 +-- arch/arm64/kernel/perf_regs.c | 3 +-- arch/csky/kernel/perf_regs.c | 3 +-- arch/powerpc/perf/perf_regs.c | 3 +-- arch/riscv/kernel/perf_regs.c | 3 +-- arch/s390/kernel/perf_regs.c | 3 +-- arch/x86/kernel/perf_regs.c | 15 +++++++++++---- include/linux/perf_event.h | 6 ------ include/linux/perf_regs.h | 6 ++---- kernel/events/core.c | 8 +++----- 10 files changed, 22 insertions(+), 31 deletions(-) (limited to 'kernel') diff --git a/arch/arm/kernel/perf_regs.c b/arch/arm/kernel/perf_regs.c index 05fe92aa7d98..0529f90395c9 100644 --- a/arch/arm/kernel/perf_regs.c +++ b/arch/arm/kernel/perf_regs.c @@ -32,8 +32,7 @@ u64 perf_reg_abi(struct task_struct *task) } void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { regs_user->regs = task_pt_regs(current); regs_user->abi = perf_reg_abi(current); diff --git a/arch/arm64/kernel/perf_regs.c b/arch/arm64/kernel/perf_regs.c index 94e8718e7229..f6f58e6265df 100644 --- a/arch/arm64/kernel/perf_regs.c +++ b/arch/arm64/kernel/perf_regs.c @@ -73,8 +73,7 @@ u64 perf_reg_abi(struct task_struct *task) } void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { regs_user->regs = task_pt_regs(current); regs_user->abi = perf_reg_abi(current); diff --git a/arch/csky/kernel/perf_regs.c b/arch/csky/kernel/perf_regs.c index eb32838b8210..09b7f88a2d6a 100644 --- a/arch/csky/kernel/perf_regs.c +++ b/arch/csky/kernel/perf_regs.c @@ -32,8 +32,7 @@ u64 perf_reg_abi(struct task_struct *task) } void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { regs_user->regs = task_pt_regs(current); regs_user->abi = perf_reg_abi(current); diff --git a/arch/powerpc/perf/perf_regs.c b/arch/powerpc/perf/perf_regs.c index 8e53f2fc3fe0..6f681b105eec 100644 --- a/arch/powerpc/perf/perf_regs.c +++ b/arch/powerpc/perf/perf_regs.c @@ -144,8 +144,7 @@ u64 perf_reg_abi(struct task_struct *task) } void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { regs_user->regs = task_pt_regs(current); regs_user->abi = (regs_user->regs) ? perf_reg_abi(current) : diff --git a/arch/riscv/kernel/perf_regs.c b/arch/riscv/kernel/perf_regs.c index 04a38fbeb9c7..fd304a248de6 100644 --- a/arch/riscv/kernel/perf_regs.c +++ b/arch/riscv/kernel/perf_regs.c @@ -36,8 +36,7 @@ u64 perf_reg_abi(struct task_struct *task) } void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { regs_user->regs = task_pt_regs(current); regs_user->abi = perf_reg_abi(current); diff --git a/arch/s390/kernel/perf_regs.c b/arch/s390/kernel/perf_regs.c index 4352a504f235..6e9e5d5e927e 100644 --- a/arch/s390/kernel/perf_regs.c +++ b/arch/s390/kernel/perf_regs.c @@ -53,8 +53,7 @@ u64 perf_reg_abi(struct task_struct *task) } void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { /* * Use the regs from the first interruption and let diff --git a/arch/x86/kernel/perf_regs.c b/arch/x86/kernel/perf_regs.c index bb7e1132290b..f9e5352b3bef 100644 --- a/arch/x86/kernel/perf_regs.c +++ b/arch/x86/kernel/perf_regs.c @@ -101,8 +101,7 @@ u64 perf_reg_abi(struct task_struct *task) } void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { regs_user->regs = task_pt_regs(current); regs_user->abi = perf_reg_abi(current); @@ -129,12 +128,20 @@ u64 perf_reg_abi(struct task_struct *task) return PERF_SAMPLE_REGS_ABI_64; } +static DEFINE_PER_CPU(struct pt_regs, nmi_user_regs); + void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { + struct pt_regs *regs_user_copy = this_cpu_ptr(&nmi_user_regs); struct pt_regs *user_regs = task_pt_regs(current); + if (!in_nmi()) { + regs_user->regs = user_regs; + regs_user->abi = perf_reg_abi(current); + return; + } + /* * If we're in an NMI that interrupted task_pt_regs setup, then * we can't sample user regs at all. This check isn't really diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index b775ae0a8c87..96450f6fb1de 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -1022,13 +1022,7 @@ struct perf_sample_data { struct perf_callchain_entry *callchain; u64 aux_size; - /* - * regs_user may point to task_pt_regs or to regs_user_copy, depending - * on arch details. - */ struct perf_regs regs_user; - struct pt_regs regs_user_copy; - struct perf_regs regs_intr; u64 stack_user_size; diff --git a/include/linux/perf_regs.h b/include/linux/perf_regs.h index 2d12e97d5e7b..f632c5725f16 100644 --- a/include/linux/perf_regs.h +++ b/include/linux/perf_regs.h @@ -20,8 +20,7 @@ u64 perf_reg_value(struct pt_regs *regs, int idx); int perf_reg_validate(u64 mask); u64 perf_reg_abi(struct task_struct *task); void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy); + struct pt_regs *regs); #else #define PERF_REG_EXTENDED_MASK 0 @@ -42,8 +41,7 @@ static inline u64 perf_reg_abi(struct task_struct *task) } static inline void perf_get_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { regs_user->regs = task_pt_regs(current); regs_user->abi = perf_reg_abi(current); diff --git a/kernel/events/core.c b/kernel/events/core.c index fc681c7c1e03..d67c9cbb0f6a 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -6374,14 +6374,13 @@ perf_output_sample_regs(struct perf_output_handle *handle, } static void perf_sample_regs_user(struct perf_regs *regs_user, - struct pt_regs *regs, - struct pt_regs *regs_user_copy) + struct pt_regs *regs) { if (user_mode(regs)) { regs_user->abi = perf_reg_abi(current); regs_user->regs = regs; } else if (!(current->flags & PF_KTHREAD)) { - perf_get_regs_user(regs_user, regs, regs_user_copy); + perf_get_regs_user(regs_user, regs); } else { regs_user->abi = PERF_SAMPLE_REGS_ABI_NONE; regs_user->regs = NULL; @@ -7083,8 +7082,7 @@ void perf_prepare_sample(struct perf_event_header *header, } if (sample_type & (PERF_SAMPLE_REGS_USER | PERF_SAMPLE_STACK_USER)) - perf_sample_regs_user(&data->regs_user, regs, - &data->regs_user_copy); + perf_sample_regs_user(&data->regs_user, regs); if (sample_type & PERF_SAMPLE_REGS_USER) { /* regs dump ABI info */ -- cgit v1.2.3 From 8c7855d82933bab7fa5e96f0e568fc125c2e1ab4 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 29 Oct 2020 16:28:25 +0100 Subject: perf: Simplify group_sched_out() Since event_sched_out() clears cpuctx->exclusive upon removal of an exclusive event (and only group leaders can be exclusive), there is no point in group_sched_out() trying to do it too. It is impossible for cpuctx->exclusive to still be set here. Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20201029162901.904060564@infradead.org --- kernel/events/core.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index d67c9cbb0f6a..9a5736617a82 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -2312,9 +2312,6 @@ group_sched_out(struct perf_event *group_event, event_sched_out(event, cpuctx, ctx); perf_pmu_enable(ctx->pmu); - - if (group_event->attr.exclusive) - cpuctx->exclusive = 0; } #define DETACH_GROUP 0x01UL -- cgit v1.2.3 From 251ff2d49347793d348babcff745289b11910e96 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 29 Oct 2020 16:29:15 +0100 Subject: perf: Simplify group_sched_in() Collate the error paths. Code duplication only leads to divergence and extra bugs. Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20201029162901.972161394@infradead.org --- kernel/events/core.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 9a5736617a82..f0e526866a1c 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -2580,11 +2580,8 @@ group_sched_in(struct perf_event *group_event, pmu->start_txn(pmu, PERF_PMU_TXN_ADD); - if (event_sched_in(group_event, cpuctx, ctx)) { - pmu->cancel_txn(pmu); - perf_mux_hrtimer_restart(cpuctx); - return -EAGAIN; - } + if (event_sched_in(group_event, cpuctx, ctx)) + goto error; /* * Schedule in siblings as one group (if any): @@ -2613,10 +2610,9 @@ group_error: } event_sched_out(group_event, cpuctx, ctx); +error: pmu->cancel_txn(pmu); - perf_mux_hrtimer_restart(cpuctx); - return -EAGAIN; } -- cgit v1.2.3 From 2714c3962f304d031d5016c963c4b459337b0749 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 29 Oct 2020 16:29:53 +0100 Subject: perf: Fix event multiplexing for exclusive groups Commit 9e6302056f80 ("perf: Use hrtimers for event multiplexing") placed the hrtimer (re)start call in the wrong place. Instead of capturing all scheduling failures, it only considered the PMU failure. The result is that groups using perf_event_attr::exclusive are no longer rotated. Fixes: 9e6302056f80 ("perf: Use hrtimers for event multiplexing") Reported-by: Andi Kleen Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20201029162902.038667689@infradead.org --- kernel/events/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index f0e526866a1c..00be48acdc36 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -2612,7 +2612,6 @@ group_error: error: pmu->cancel_txn(pmu); - perf_mux_hrtimer_restart(cpuctx); return -EAGAIN; } @@ -3672,6 +3671,7 @@ static int merge_sched_in(struct perf_event *event, void *data) *can_add_hw = 0; ctx->rotate_necessary = 1; + perf_mux_hrtimer_restart(cpuctx); } return 0; -- cgit v1.2.3 From 1908dc911792067287458fdb0800f036f4f4e0f6 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 29 Oct 2020 16:32:22 +0100 Subject: perf: Tweak perf_event_attr::exclusive semantics Currently perf_event_attr::exclusive can be used to ensure an event(group) is the sole group scheduled on the PMU. One consequence is that when you have a pinned event (say the watchdog) you can no longer have regular exclusive event(group)s. Inspired by the fact that !pinned events are considered less strict, allow !pinned,exclusive events to share the PMU with pinned,!exclusive events. Pinned,exclusive is still fully exclusive. Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20201029162902.105962225@infradead.org --- kernel/events/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/events/core.c b/kernel/events/core.c index 00be48acdc36..dc568ca295bd 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -2637,7 +2637,7 @@ static int group_can_go_on(struct perf_event *event, * If this group is exclusive and there are already * events on the CPU, it can't go on. */ - if (event->attr.exclusive && cpuctx->active_oncpu) + if (event->attr.exclusive && !list_empty(get_event_list(event))) return 0; /* * Otherwise, try to add it if all previous groups were able -- cgit v1.2.3 From abbaa433de07076fb8ef524b77ce55d94bad5fc5 Mon Sep 17 00:00:00 2001 From: Wang Qing Date: Sat, 7 Nov 2020 15:45:44 +0800 Subject: bpf: Fix passing zero to PTR_ERR() in bpf_btf_printf_prepare There is a bug when passing zero to PTR_ERR() and return. Fix the smatch error. Fixes: c4d0bfb45068 ("bpf: Add bpf_snprintf_btf helper") Signed-off-by: Wang Qing Signed-off-by: Daniel Borkmann Acked-by: Yonghong Song Acked-by: John Fastabend Link: https://lore.kernel.org/bpf/1604735144-686-1-git-send-email-wangqing@vivo.com --- kernel/trace/bpf_trace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 4517c8b66518..5113fd423cdf 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -1198,7 +1198,7 @@ static int bpf_btf_printf_prepare(struct btf_ptr *ptr, u32 btf_ptr_size, *btf = bpf_get_btf_vmlinux(); if (IS_ERR_OR_NULL(*btf)) - return PTR_ERR(*btf); + return IS_ERR(*btf) ? PTR_ERR(*btf) : -EINVAL; if (ptr->type_id > 0) *btf_id = ptr->type_id; -- cgit v1.2.3 From 04e613ded8c26489b3e0f9101b44462f780d1a35 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Fri, 6 Nov 2020 10:25:49 +0000 Subject: arm64: smp: Tell RCU about CPUs that fail to come online Commit ce3d31ad3cac ("arm64/smp: Move rcu_cpu_starting() earlier") ensured that RCU is informed early about incoming CPUs that might end up calling into printk() before they are online. However, if such a CPU fails the early CPU feature compatibility checks in check_local_cpu_capabilities(), then it will be powered off or parked without informing RCU, leading to an endless stream of stalls: | rcu: INFO: rcu_preempt detected stalls on CPUs/tasks: | rcu: 2-O...: (0 ticks this GP) idle=002/1/0x4000000000000000 softirq=0/0 fqs=2593 | (detected by 0, t=5252 jiffies, g=9317, q=136) | Task dump for CPU 2: | task:swapper/2 state:R running task stack: 0 pid: 0 ppid: 1 flags:0x00000028 | Call trace: | ret_from_fork+0x0/0x30 Ensure that the dying CPU invokes rcu_report_dead() prior to being powered off or parked. Cc: Qian Cai Cc: "Paul E. McKenney" Reviewed-by: Paul E. McKenney Suggested-by: Qian Cai Link: https://lore.kernel.org/r/20201105222242.GA8842@willie-the-truck Link: https://lore.kernel.org/r/20201106103602.9849-3-will@kernel.org Signed-off-by: Will Deacon --- arch/arm64/kernel/smp.c | 1 + kernel/rcu/tree.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c index 09c96f57818c..18e9727d3f64 100644 --- a/arch/arm64/kernel/smp.c +++ b/arch/arm64/kernel/smp.c @@ -413,6 +413,7 @@ void cpu_die_early(void) /* Mark this CPU absent */ set_cpu_present(cpu, 0); + rcu_report_dead(cpu); if (IS_ENABLED(CONFIG_HOTPLUG_CPU)) { update_cpu_boot_status(CPU_KILL_ME); diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index 06895ef85d69..946e7c0c4cf7 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -4077,7 +4077,6 @@ void rcu_cpu_starting(unsigned int cpu) smp_mb(); /* Ensure RCU read-side usage follows above initialization. */ } -#ifdef CONFIG_HOTPLUG_CPU /* * The outgoing function has no further need of RCU, so remove it from * the rcu_node tree's ->qsmaskinitnext bit masks. @@ -4117,6 +4116,7 @@ void rcu_report_dead(unsigned int cpu) rdp->cpu_started = false; } +#ifdef CONFIG_HOTPLUG_CPU /* * The outgoing CPU has just passed through the dying-idle state, and we * are being invoked from the CPU that was IPIed to continue the offline -- cgit v1.2.3 From d61fc96a37603384cd531622c1e89de1096b5123 Mon Sep 17 00:00:00 2001 From: Boqun Feng Date: Mon, 2 Nov 2020 13:37:41 +0800 Subject: lockdep: Avoid to modify chain keys in validate_chain() Chris Wilson reported a problem spotted by check_chain_key(): a chain key got changed in validate_chain() because we modify the ->read in validate_chain() to skip checks for dependency adding, and ->read is taken into calculation for chain key since commit f611e8cf98ec ("lockdep: Take read/write status in consideration when generate chainkey"). Fix this by avoiding to modify ->read in validate_chain() based on two facts: a) since we now support recursive read lock detection, there is no need to skip checks for dependency adding for recursive readers, b) since we have a), there is only one case left (nest_lock) where we want to skip checks in validate_chain(), we simply remove the modification for ->read and rely on the return value of check_deadlock() to skip the dependency adding. Reported-by: Chris Wilson Signed-off-by: Boqun Feng Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/20201102053743.450459-1-boqun.feng@gmail.com --- kernel/locking/lockdep.c | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c index b71ad8d9f1c9..d9fb9e19d2ed 100644 --- a/kernel/locking/lockdep.c +++ b/kernel/locking/lockdep.c @@ -2765,7 +2765,9 @@ print_deadlock_bug(struct task_struct *curr, struct held_lock *prev, * (Note that this has to be done separately, because the graph cannot * detect such classes of deadlocks.) * - * Returns: 0 on deadlock detected, 1 on OK, 2 on recursive read + * Returns: 0 on deadlock detected, 1 on OK, 2 if another lock with the same + * lock class is held but nest_lock is also held, i.e. we rely on the + * nest_lock to avoid the deadlock. */ static int check_deadlock(struct task_struct *curr, struct held_lock *next) @@ -2788,7 +2790,7 @@ check_deadlock(struct task_struct *curr, struct held_lock *next) * lock class (i.e. read_lock(lock)+read_lock(lock)): */ if ((next->read == 2) && prev->read) - return 2; + continue; /* * We're holding the nest_lock, which serializes this lock's @@ -3592,16 +3594,13 @@ static int validate_chain(struct task_struct *curr, if (!ret) return 0; - /* - * Mark recursive read, as we jump over it when - * building dependencies (just like we jump over - * trylock entries): - */ - if (ret == 2) - hlock->read = 2; /* * Add dependency only if this lock is not the head - * of the chain, and if it's not a secondary read-lock: + * of the chain, and if the new lock introduces no more + * lock dependency (because we already hold a lock with the + * same lock class) nor deadlock (because the nest_lock + * serializes nesting locks), see the comments for + * check_deadlock(). */ if (!chain_head && ret != 2) { if (!check_prevs_add(curr, hlock)) -- cgit v1.2.3 From 16b0a7a1a0af9db6e008fecd195fe4d6cb366d83 Mon Sep 17 00:00:00 2001 From: Vincent Guittot Date: Mon, 2 Nov 2020 11:24:57 +0100 Subject: sched/fair: Ensure tasks spreading in LLC during LB schbench shows latency increase for 95 percentile above since: commit 0b0695f2b34a ("sched/fair: Rework load_balance()") Align the behavior of the load balancer with the wake up path, which tries to select an idle CPU which belongs to the LLC for a waking task. calculate_imbalance() will use nr_running instead of the spare capacity when CPUs share resources (ie cache) at the domain level. This will ensure a better spread of tasks on idle CPUs. Running schbench on a hikey (8cores arm64) shows the problem: tip/sched/core : schbench -m 2 -t 4 -s 10000 -c 1000000 -r 10 Latency percentiles (usec) 50.0th: 33 75.0th: 45 90.0th: 51 95.0th: 4152 *99.0th: 14288 99.5th: 14288 99.9th: 14288 min=0, max=14276 tip/sched/core + patch : schbench -m 2 -t 4 -s 10000 -c 1000000 -r 10 Latency percentiles (usec) 50.0th: 34 75.0th: 47 90.0th: 52 95.0th: 78 *99.0th: 94 99.5th: 94 99.9th: 94 min=0, max=94 Fixes: 0b0695f2b34a ("sched/fair: Rework load_balance()") Reported-by: Chris Mason Suggested-by: Rik van Riel Signed-off-by: Vincent Guittot Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Rik van Riel Tested-by: Rik van Riel Link: https://lkml.kernel.org/r/20201102102457.28808-1-vincent.guittot@linaro.org --- kernel/sched/fair.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index aa4c6227cd6d..210b15f068a6 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -9031,7 +9031,8 @@ static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *s * emptying busiest. */ if (local->group_type == group_has_spare) { - if (busiest->group_type > group_fully_busy) { + if ((busiest->group_type > group_fully_busy) && + !(env->sd->flags & SD_SHARE_PKG_RESOURCES)) { /* * If busiest is overloaded, try to fill spare * capacity. This might end up creating spare capacity -- cgit v1.2.3 From b4c9c9f15649c98a5b45408919d1ff4fd7f5531c Mon Sep 17 00:00:00 2001 From: Vincent Guittot Date: Thu, 29 Oct 2020 17:18:24 +0100 Subject: sched/fair: Prefer prev cpu in asymmetric wakeup path During fast wakeup path, scheduler always check whether local or prev cpus are good candidates for the task before looking for other cpus in the domain. With commit b7a331615d25 ("sched/fair: Add asymmetric CPU capacity wakeup scan") the heterogenous system gains a dedicated path but doesn't try to reuse prev cpu whenever possible. If the previous cpu is idle and belong to the LLC domain, we should check it 1st before looking for another cpu because it stays one of the best candidate and this also stabilizes task placement on the system. This change aligns asymmetric path behavior with symmetric one and reduces cases where the task migrates across all cpus of the sd_asym_cpucapacity domains at wakeup. This change does not impact normal EAS mode but only the overloaded case or when EAS is not used. - On hikey960 with performance governor (EAS disable) ./perf bench sched pipe -T -l 50000 mainline w/ patch # migrations 999364 0 ops/sec 149313(+/-0.28%) 182587(+/- 0.40) +22% - On hikey with performance governor ./perf bench sched pipe -T -l 50000 mainline w/ patch # migrations 0 0 ops/sec 47721(+/-0.76%) 47899(+/- 0.56) +0.4% According to test on hikey, the patch doesn't impact symmetric system compared to current implementation (only tested on arm64) Also read the uclamped value of task's utilization at most twice instead instead each time we compare task's utilization with cpu's capacity. Fixes: b7a331615d25 ("sched/fair: Add asymmetric CPU capacity wakeup scan") Signed-off-by: Vincent Guittot Signed-off-by: Peter Zijlstra (Intel) Tested-by: Dietmar Eggemann Reviewed-by: Valentin Schneider Link: https://lkml.kernel.org/r/20201029161824.26389-1-vincent.guittot@linaro.org --- kernel/sched/fair.c | 67 ++++++++++++++++++++++++++++++++++------------------- 1 file changed, 43 insertions(+), 24 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 210b15f068a6..8e563cf2b5e7 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -6172,21 +6172,21 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t static int select_idle_capacity(struct task_struct *p, struct sched_domain *sd, int target) { - unsigned long best_cap = 0; + unsigned long task_util, best_cap = 0; int cpu, best_cpu = -1; struct cpumask *cpus; - sync_entity_load_avg(&p->se); - cpus = this_cpu_cpumask_var_ptr(select_idle_mask); cpumask_and(cpus, sched_domain_span(sd), p->cpus_ptr); + task_util = uclamp_task_util(p); + for_each_cpu_wrap(cpu, cpus, target) { unsigned long cpu_cap = capacity_of(cpu); if (!available_idle_cpu(cpu) && !sched_idle_cpu(cpu)) continue; - if (task_fits_capacity(p, cpu_cap)) + if (fits_capacity(task_util, cpu_cap)) return cpu; if (cpu_cap > best_cap) { @@ -6198,44 +6198,42 @@ select_idle_capacity(struct task_struct *p, struct sched_domain *sd, int target) return best_cpu; } +static inline bool asym_fits_capacity(int task_util, int cpu) +{ + if (static_branch_unlikely(&sched_asym_cpucapacity)) + return fits_capacity(task_util, capacity_of(cpu)); + + return true; +} + /* * Try and locate an idle core/thread in the LLC cache domain. */ static int select_idle_sibling(struct task_struct *p, int prev, int target) { struct sched_domain *sd; + unsigned long task_util; int i, recent_used_cpu; /* - * For asymmetric CPU capacity systems, our domain of interest is - * sd_asym_cpucapacity rather than sd_llc. + * On asymmetric system, update task utilization because we will check + * that the task fits with cpu's capacity. */ if (static_branch_unlikely(&sched_asym_cpucapacity)) { - sd = rcu_dereference(per_cpu(sd_asym_cpucapacity, target)); - /* - * On an asymmetric CPU capacity system where an exclusive - * cpuset defines a symmetric island (i.e. one unique - * capacity_orig value through the cpuset), the key will be set - * but the CPUs within that cpuset will not have a domain with - * SD_ASYM_CPUCAPACITY. These should follow the usual symmetric - * capacity path. - */ - if (!sd) - goto symmetric; - - i = select_idle_capacity(p, sd, target); - return ((unsigned)i < nr_cpumask_bits) ? i : target; + sync_entity_load_avg(&p->se); + task_util = uclamp_task_util(p); } -symmetric: - if (available_idle_cpu(target) || sched_idle_cpu(target)) + if ((available_idle_cpu(target) || sched_idle_cpu(target)) && + asym_fits_capacity(task_util, target)) return target; /* * If the previous CPU is cache affine and idle, don't be stupid: */ if (prev != target && cpus_share_cache(prev, target) && - (available_idle_cpu(prev) || sched_idle_cpu(prev))) + (available_idle_cpu(prev) || sched_idle_cpu(prev)) && + asym_fits_capacity(task_util, prev)) return prev; /* @@ -6258,7 +6256,8 @@ symmetric: recent_used_cpu != target && cpus_share_cache(recent_used_cpu, target) && (available_idle_cpu(recent_used_cpu) || sched_idle_cpu(recent_used_cpu)) && - cpumask_test_cpu(p->recent_used_cpu, p->cpus_ptr)) { + cpumask_test_cpu(p->recent_used_cpu, p->cpus_ptr) && + asym_fits_capacity(task_util, recent_used_cpu)) { /* * Replace recent_used_cpu with prev as it is a potential * candidate for the next wake: @@ -6267,6 +6266,26 @@ symmetric: return recent_used_cpu; } + /* + * For asymmetric CPU capacity systems, our domain of interest is + * sd_asym_cpucapacity rather than sd_llc. + */ + if (static_branch_unlikely(&sched_asym_cpucapacity)) { + sd = rcu_dereference(per_cpu(sd_asym_cpucapacity, target)); + /* + * On an asymmetric CPU capacity system where an exclusive + * cpuset defines a symmetric island (i.e. one unique + * capacity_orig value through the cpuset), the key will be set + * but the CPUs within that cpuset will not have a domain with + * SD_ASYM_CPUCAPACITY. These should follow the usual symmetric + * capacity path. + */ + if (sd) { + i = select_idle_capacity(p, sd, target); + return ((unsigned)i < nr_cpumask_bits) ? i : target; + } + } + sd = rcu_dereference(per_cpu(sd_llc, target)); if (!sd) return target; -- cgit v1.2.3 From 8d4d9c7b4333abccb3bf310d76ef7ea2edb9828f Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Thu, 29 Oct 2020 15:11:03 +0000 Subject: sched/debug: Fix memory corruption caused by multiple small reads of flags Reading /proc/sys/kernel/sched_domain/cpu*/domain0/flags mutliple times with small reads causes oopses with slub corruption issues because the kfree is free'ing an offset from a previous allocation. Fix this by adding in a new pointer 'buf' for the allocation and kfree and use the temporary pointer tmp to handle memory copies of the buf offsets. Fixes: 5b9f8ff7b320 ("sched/debug: Output SD flag names rather than their values") Reported-by: Jeff Bastian Signed-off-by: Colin Ian King Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Valentin Schneider Link: https://lkml.kernel.org/r/20201029151103.373410-1-colin.king@canonical.com --- kernel/sched/debug.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c index 0655524700d2..2357921580f9 100644 --- a/kernel/sched/debug.c +++ b/kernel/sched/debug.c @@ -251,7 +251,7 @@ static int sd_ctl_doflags(struct ctl_table *table, int write, unsigned long flags = *(unsigned long *)table->data; size_t data_size = 0; size_t len = 0; - char *tmp; + char *tmp, *buf; int idx; if (write) @@ -269,17 +269,17 @@ static int sd_ctl_doflags(struct ctl_table *table, int write, return 0; } - tmp = kcalloc(data_size + 1, sizeof(*tmp), GFP_KERNEL); - if (!tmp) + buf = kcalloc(data_size + 1, sizeof(*buf), GFP_KERNEL); + if (!buf) return -ENOMEM; for_each_set_bit(idx, &flags, __SD_FLAG_CNT) { char *name = sd_flag_debug[idx].name; - len += snprintf(tmp + len, strlen(name) + 2, "%s ", name); + len += snprintf(buf + len, strlen(name) + 2, "%s ", name); } - tmp += *ppos; + tmp = buf + *ppos; len -= *ppos; if (len > *lenp) @@ -294,7 +294,7 @@ static int sd_ctl_doflags(struct ctl_table *table, int write, *lenp = len; *ppos += len; - kfree(tmp); + kfree(buf); return 0; } -- cgit v1.2.3 From c583bcb8f5edd48c1798798e341f78afb9bf4f6f Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 24 Sep 2020 15:11:55 -0700 Subject: rcu: Don't invoke try_invoke_on_locked_down_task() with irqs disabled The try_invoke_on_locked_down_task() function requires that interrupts be enabled, but it is called with interrupts disabled from rcu_print_task_stall(), resulting in an "IRQs not enabled as expected" diagnostic. This commit therefore updates rcu_print_task_stall() to accumulate a list of the first few tasks while holding the current leaf rcu_node structure's ->lock, then releases that lock and only then uses try_invoke_on_locked_down_task() to attempt to obtain per-task detailed information. Of course, as soon as ->lock is released, the task might exit, so the get_task_struct() function is used to prevent the task structure from going away in the meantime. Link: https://lore.kernel.org/lkml/000000000000903d5805ab908fc4@google.com/ Fixes: 5bef8da66a9c ("rcu: Add per-task state to RCU CPU stall warnings") Reported-by: syzbot+cb3b69ae80afd6535b0e@syzkaller.appspotmail.com Reported-by: syzbot+f04854e1c5c9e913cc27@syzkaller.appspotmail.com Tested-by: Sebastian Andrzej Siewior Signed-off-by: Paul E. McKenney --- kernel/rcu/tree_stall.h | 22 +++++++++++++++++----- 1 file changed, 17 insertions(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/rcu/tree_stall.h b/kernel/rcu/tree_stall.h index 0fde39b8daab..ca21d28a0f98 100644 --- a/kernel/rcu/tree_stall.h +++ b/kernel/rcu/tree_stall.h @@ -249,13 +249,16 @@ static bool check_slow_task(struct task_struct *t, void *arg) /* * Scan the current list of tasks blocked within RCU read-side critical - * sections, printing out the tid of each. + * sections, printing out the tid of each of the first few of them. */ -static int rcu_print_task_stall(struct rcu_node *rnp) +static int rcu_print_task_stall(struct rcu_node *rnp, unsigned long flags) + __releases(rnp->lock) { + int i = 0; int ndetected = 0; struct rcu_stall_chk_rdr rscr; struct task_struct *t; + struct task_struct *ts[8]; if (!rcu_preempt_blocked_readers_cgp(rnp)) return 0; @@ -264,6 +267,14 @@ static int rcu_print_task_stall(struct rcu_node *rnp) t = list_entry(rnp->gp_tasks->prev, struct task_struct, rcu_node_entry); list_for_each_entry_continue(t, &rnp->blkd_tasks, rcu_node_entry) { + get_task_struct(t); + ts[i++] = t; + if (i >= ARRAY_SIZE(ts)) + break; + } + raw_spin_unlock_irqrestore_rcu_node(rnp, flags); + for (i--; i; i--) { + t = ts[i]; if (!try_invoke_on_locked_down_task(t, check_slow_task, &rscr)) pr_cont(" P%d", t->pid); else @@ -273,6 +284,7 @@ static int rcu_print_task_stall(struct rcu_node *rnp) ".q"[rscr.rs.b.need_qs], ".e"[rscr.rs.b.exp_hint], ".l"[rscr.on_blkd_list]); + put_task_struct(t); ndetected++; } pr_cont("\n"); @@ -293,8 +305,9 @@ static void rcu_print_detail_task_stall_rnp(struct rcu_node *rnp) * Because preemptible RCU does not exist, we never have to check for * tasks blocked within RCU read-side critical sections. */ -static int rcu_print_task_stall(struct rcu_node *rnp) +static int rcu_print_task_stall(struct rcu_node *rnp, unsigned long flags) { + raw_spin_unlock_irqrestore_rcu_node(rnp, flags); return 0; } #endif /* #else #ifdef CONFIG_PREEMPT_RCU */ @@ -472,7 +485,6 @@ static void print_other_cpu_stall(unsigned long gp_seq, unsigned long gps) pr_err("INFO: %s detected stalls on CPUs/tasks:\n", rcu_state.name); rcu_for_each_leaf_node(rnp) { raw_spin_lock_irqsave_rcu_node(rnp, flags); - ndetected += rcu_print_task_stall(rnp); if (rnp->qsmask != 0) { for_each_leaf_node_possible_cpu(rnp, cpu) if (rnp->qsmask & leaf_node_cpu_bit(rnp, cpu)) { @@ -480,7 +492,7 @@ static void print_other_cpu_stall(unsigned long gp_seq, unsigned long gps) ndetected++; } } - raw_spin_unlock_irqrestore_rcu_node(rnp, flags); + ndetected += rcu_print_task_stall(rnp, flags); // Releases rnp->lock. } for_each_possible_cpu(cpu) -- cgit v1.2.3 From f16e631333a8f12ae8128826e695db4b2a528407 Mon Sep 17 00:00:00 2001 From: Kaixu Xia Date: Wed, 11 Nov 2020 13:03:46 +0800 Subject: bpf: Fix unsigned 'datasec_id' compared with zero in check_pseudo_btf_id The unsigned variable datasec_id is assigned a return value from the call to check_pseudo_btf_id(), which may return negative error code. This fixes the following coccicheck warning: ./kernel/bpf/verifier.c:9616:5-15: WARNING: Unsigned expression compared with zero: datasec_id > 0 Fixes: eaa6bcb71ef6 ("bpf: Introduce bpf_per_cpu_ptr()") Reported-by: Tosk Robot Signed-off-by: Kaixu Xia Signed-off-by: Daniel Borkmann Acked-by: Andrii Nakryiko Acked-by: John Fastabend Cc: Hao Luo Link: https://lore.kernel.org/bpf/1605071026-25906-1-git-send-email-kaixuxia@tencent.com --- kernel/bpf/verifier.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 6200519582a6..6204ec705d80 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -9572,12 +9572,13 @@ static int check_pseudo_btf_id(struct bpf_verifier_env *env, struct bpf_insn *insn, struct bpf_insn_aux_data *aux) { - u32 datasec_id, type, id = insn->imm; const struct btf_var_secinfo *vsi; const struct btf_type *datasec; const struct btf_type *t; const char *sym_name; bool percpu = false; + u32 type, id = insn->imm; + s32 datasec_id; u64 addr; int i; -- cgit v1.2.3 From f782e2c300a717233b64697affda3ea7aac00b2b Mon Sep 17 00:00:00 2001 From: Dmitrii Banshchikov Date: Fri, 13 Nov 2020 17:17:56 +0000 Subject: bpf: Relax return code check for subprograms Currently verifier enforces return code checks for subprograms in the same manner as it does for program entry points. This prevents returning arbitrary scalar values from subprograms. Scalar type of returned values is checked by btf_prepare_func_args() and hence it should be safe to allow only scalars for now. Relax return code checks for subprograms and allow any correct scalar values. Fixes: 51c39bb1d5d10 (bpf: Introduce function-by-function verification) Signed-off-by: Dmitrii Banshchikov Signed-off-by: Alexei Starovoitov Acked-by: Andrii Nakryiko Link: https://lore.kernel.org/bpf/20201113171756.90594-1-me@ubique.spb.ru --- kernel/bpf/verifier.c | 15 +++++++++++++-- .../selftests/bpf/prog_tests/test_global_funcs.c | 1 + tools/testing/selftests/bpf/progs/test_global_func8.c | 19 +++++++++++++++++++ 3 files changed, 33 insertions(+), 2 deletions(-) create mode 100644 tools/testing/selftests/bpf/progs/test_global_func8.c (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 6204ec705d80..1388bf733071 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -7786,9 +7786,11 @@ static int check_return_code(struct bpf_verifier_env *env) struct tnum range = tnum_range(0, 1); enum bpf_prog_type prog_type = resolve_prog_type(env->prog); int err; + const bool is_subprog = env->cur_state->frame[0]->subprogno; /* LSM and struct_ops func-ptr's return type could be "void" */ - if ((prog_type == BPF_PROG_TYPE_STRUCT_OPS || + if (!is_subprog && + (prog_type == BPF_PROG_TYPE_STRUCT_OPS || prog_type == BPF_PROG_TYPE_LSM) && !prog->aux->attach_func_proto->type) return 0; @@ -7808,6 +7810,16 @@ static int check_return_code(struct bpf_verifier_env *env) return -EACCES; } + reg = cur_regs(env) + BPF_REG_0; + if (is_subprog) { + if (reg->type != SCALAR_VALUE) { + verbose(env, "At subprogram exit the register R0 is not a scalar value (%s)\n", + reg_type_str[reg->type]); + return -EINVAL; + } + return 0; + } + switch (prog_type) { case BPF_PROG_TYPE_CGROUP_SOCK_ADDR: if (env->prog->expected_attach_type == BPF_CGROUP_UDP4_RECVMSG || @@ -7861,7 +7873,6 @@ static int check_return_code(struct bpf_verifier_env *env) return 0; } - reg = cur_regs(env) + BPF_REG_0; if (reg->type != SCALAR_VALUE) { verbose(env, "At program exit the register R0 is not a known value (%s)\n", reg_type_str[reg->type]); diff --git a/tools/testing/selftests/bpf/prog_tests/test_global_funcs.c b/tools/testing/selftests/bpf/prog_tests/test_global_funcs.c index 193002b14d7f..32e4348b714b 100644 --- a/tools/testing/selftests/bpf/prog_tests/test_global_funcs.c +++ b/tools/testing/selftests/bpf/prog_tests/test_global_funcs.c @@ -60,6 +60,7 @@ void test_test_global_funcs(void) { "test_global_func5.o" , "expected pointer to ctx, but got PTR" }, { "test_global_func6.o" , "modified ctx ptr R2" }, { "test_global_func7.o" , "foo() doesn't return scalar" }, + { "test_global_func8.o" }, }; libbpf_print_fn_t old_print_fn = NULL; int err, i, duration = 0; diff --git a/tools/testing/selftests/bpf/progs/test_global_func8.c b/tools/testing/selftests/bpf/progs/test_global_func8.c new file mode 100644 index 000000000000..d55a6544b1ab --- /dev/null +++ b/tools/testing/selftests/bpf/progs/test_global_func8.c @@ -0,0 +1,19 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* Copyright (c) 2020 Facebook */ +#include +#include +#include + +__noinline int foo(struct __sk_buff *skb) +{ + return bpf_get_prandom_u32(); +} + +SEC("cgroup_skb/ingress") +int test_cls(struct __sk_buff *skb) +{ + if (!foo(skb)) + return 0; + + return 1; +} -- cgit v1.2.3 From 8b92c4ff4423aa9900cf838d3294fcade4dbda35 Mon Sep 17 00:00:00 2001 From: Matteo Croce Date: Fri, 13 Nov 2020 22:52:02 -0800 Subject: Revert "kernel/reboot.c: convert simple_strtoul to kstrtoint" Patch series "fix parsing of reboot= cmdline", v3. The parsing of the reboot= cmdline has two major errors: - a missing bound check can crash the system on reboot - parsing of the cpu number only works if specified last Fix both. This patch (of 2): This reverts commit 616feab753972b97. kstrtoint() and simple_strtoul() have a subtle difference which makes them non interchangeable: if a non digit character is found amid the parsing, the former will return an error, while the latter will just stop parsing, e.g. simple_strtoul("123xyx") = 123. The kernel cmdline reboot= argument allows to specify the CPU used for rebooting, with the syntax `s####` among the other flags, e.g. "reboot=warm,s31,force", so if this flag is not the last given, it's silently ignored as well as the subsequent ones. Fixes: 616feab75397 ("kernel/reboot.c: convert simple_strtoul to kstrtoint") Signed-off-by: Matteo Croce Signed-off-by: Andrew Morton Cc: Guenter Roeck Cc: Petr Mladek Cc: Arnd Bergmann Cc: Mike Rapoport Cc: Kees Cook Cc: Pavel Tatashin Cc: Robin Holt Cc: Fabian Frederick Cc: Greg Kroah-Hartman Cc: Link: https://lkml.kernel.org/r/20201103214025.116799-2-mcroce@linux.microsoft.com Signed-off-by: Linus Torvalds --- kernel/reboot.c | 21 +++++++-------------- 1 file changed, 7 insertions(+), 14 deletions(-) (limited to 'kernel') diff --git a/kernel/reboot.c b/kernel/reboot.c index e7b78d5ae1ab..8fbba433725e 100644 --- a/kernel/reboot.c +++ b/kernel/reboot.c @@ -551,22 +551,15 @@ static int __init reboot_setup(char *str) break; case 's': - { - int rc; - - if (isdigit(*(str+1))) { - rc = kstrtoint(str+1, 0, &reboot_cpu); - if (rc) - return rc; - } else if (str[1] == 'm' && str[2] == 'p' && - isdigit(*(str+3))) { - rc = kstrtoint(str+3, 0, &reboot_cpu); - if (rc) - return rc; - } else + if (isdigit(*(str+1))) + reboot_cpu = simple_strtoul(str+1, NULL, 0); + else if (str[1] == 'm' && str[2] == 'p' && + isdigit(*(str+3))) + reboot_cpu = simple_strtoul(str+3, NULL, 0); + else *mode = REBOOT_SOFT; break; - } + case 'g': *mode = REBOOT_GPIO; break; -- cgit v1.2.3 From df5b0ab3e08a156701b537809914b339b0daa526 Mon Sep 17 00:00:00 2001 From: Matteo Croce Date: Fri, 13 Nov 2020 22:52:07 -0800 Subject: reboot: fix overflow parsing reboot cpu number Limit the CPU number to num_possible_cpus(), because setting it to a value lower than INT_MAX but higher than NR_CPUS produces the following error on reboot and shutdown: BUG: unable to handle page fault for address: ffffffff90ab1bb0 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page PGD 1c09067 P4D 1c09067 PUD 1c0a063 PMD 0 Oops: 0000 [#1] SMP CPU: 1 PID: 1 Comm: systemd-shutdow Not tainted 5.9.0-rc8-kvm #110 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014 RIP: 0010:migrate_to_reboot_cpu+0xe/0x60 Code: ea ea 00 48 89 fa 48 c7 c7 30 57 f1 81 e9 fa ef ff ff 66 2e 0f 1f 84 00 00 00 00 00 53 8b 1d d5 ea ea 00 e8 14 33 fe ff 89 da <48> 0f a3 15 ea fc bd 00 48 89 d0 73 29 89 c2 c1 e8 06 65 48 8b 3c RSP: 0018:ffffc90000013e08 EFLAGS: 00010246 RAX: ffff88801f0a0000 RBX: 0000000077359400 RCX: 0000000000000000 RDX: 0000000077359400 RSI: 0000000000000002 RDI: ffffffff81c199e0 RBP: ffffffff81c1e3c0 R08: ffff88801f41f000 R09: ffffffff81c1e348 R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 R13: 00007f32bedf8830 R14: 00000000fee1dead R15: 0000000000000000 FS: 00007f32bedf8980(0000) GS:ffff88801f480000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: ffffffff90ab1bb0 CR3: 000000001d057000 CR4: 00000000000006a0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: __do_sys_reboot.cold+0x34/0x5b do_syscall_64+0x2d/0x40 Fixes: 1b3a5d02ee07 ("reboot: move arch/x86 reboot= handling to generic kernel") Signed-off-by: Matteo Croce Signed-off-by: Andrew Morton Cc: Arnd Bergmann Cc: Fabian Frederick Cc: Greg Kroah-Hartman Cc: Guenter Roeck Cc: Kees Cook Cc: Mike Rapoport Cc: Pavel Tatashin Cc: Petr Mladek Cc: Robin Holt Cc: Link: https://lkml.kernel.org/r/20201103214025.116799-3-mcroce@linux.microsoft.com Signed-off-by: Linus Torvalds --- kernel/reboot.c | 7 +++++++ 1 file changed, 7 insertions(+) (limited to 'kernel') diff --git a/kernel/reboot.c b/kernel/reboot.c index 8fbba433725e..af6f23d8bea1 100644 --- a/kernel/reboot.c +++ b/kernel/reboot.c @@ -558,6 +558,13 @@ static int __init reboot_setup(char *str) reboot_cpu = simple_strtoul(str+3, NULL, 0); else *mode = REBOOT_SOFT; + if (reboot_cpu >= num_possible_cpus()) { + pr_err("Ignoring the CPU number in reboot= option. " + "CPU %d exceeds possible cpu number %d\n", + reboot_cpu, num_possible_cpus()); + reboot_cpu = 0; + break; + } break; case 'g': -- cgit v1.2.3 From e7e046155af04cdca5e1157f28b07e1651eb317b Mon Sep 17 00:00:00 2001 From: Santosh Sivaraj Date: Fri, 13 Nov 2020 22:52:10 -0800 Subject: kernel/watchdog: fix watchdog_allowed_mask not used warning Define watchdog_allowed_mask only when SOFTLOCKUP_DETECTOR is enabled. Fixes: 7feeb9cd4f5b ("watchdog/sysctl: Clean up sysctl variable name space") Signed-off-by: Santosh Sivaraj Signed-off-by: Andrew Morton Reviewed-by: Petr Mladek Cc: Thomas Gleixner Link: https://lkml.kernel.org/r/20201106015025.1281561-1-santosh@fossix.org Signed-off-by: Linus Torvalds --- kernel/watchdog.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/watchdog.c b/kernel/watchdog.c index 5abb5b22ad13..71109065bd8e 100644 --- a/kernel/watchdog.c +++ b/kernel/watchdog.c @@ -44,8 +44,6 @@ int __read_mostly soft_watchdog_user_enabled = 1; int __read_mostly watchdog_thresh = 10; static int __read_mostly nmi_watchdog_available; -static struct cpumask watchdog_allowed_mask __read_mostly; - struct cpumask watchdog_cpumask __read_mostly; unsigned long *watchdog_cpumask_bits = cpumask_bits(&watchdog_cpumask); @@ -162,6 +160,8 @@ static void lockup_detector_update_enable(void) int __read_mostly sysctl_softlockup_all_cpu_backtrace; #endif +static struct cpumask watchdog_allowed_mask __read_mostly; + /* Global variables, exported for sysctl */ unsigned int __read_mostly softlockup_panic = CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC_VALUE; -- cgit v1.2.3 From 2f31ad64a9cce8b2409d2d4563482adfb8664082 Mon Sep 17 00:00:00 2001 From: Christophe Leroy Date: Fri, 13 Nov 2020 22:52:20 -0800 Subject: panic: don't dump stack twice on warn Before commit 3f388f28639f ("panic: dump registers on panic_on_warn"), __warn() was calling show_regs() when regs was not NULL, and show_stack() otherwise. After that commit, show_stack() is called regardless of whether show_regs() has been called or not, leading to duplicated Call Trace: ------------[ cut here ]------------ WARNING: CPU: 0 PID: 1 at arch/powerpc/mm/nohash/8xx.c:186 mmu_mark_initmem_nx+0x24/0x94 CPU: 0 PID: 1 Comm: swapper Not tainted 5.10.0-rc2-s3k-dev-01375-gf46ec0d3ecbd-dirty #4092 NIP: c00128b4 LR: c0010228 CTR: 00000000 REGS: c9023e40 TRAP: 0700 Not tainted (5.10.0-rc2-s3k-dev-01375-gf46ec0d3ecbd-dirty) MSR: 00029032 CR: 24000424 XER: 00000000 GPR00: c0010228 c9023ef8 c2100000 0074c000 ffffffff 00000000 c2151000 c07b3880 GPR08: ff000900 0074c000 c8000000 c33b53a8 24000822 00000000 c0003a20 00000000 GPR16: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 GPR24: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00800000 NIP [c00128b4] mmu_mark_initmem_nx+0x24/0x94 LR [c0010228] free_initmem+0x20/0x58 Call Trace: free_initmem+0x20/0x58 kernel_init+0x1c/0x114 ret_from_kernel_thread+0x14/0x1c Instruction dump: 7d291850 7d234b78 4e800020 9421ffe0 7c0802a6 bfc10018 3fe0c060 3bff0000 3fff4080 3bffffff 90010024 57ff0010 <0fe00000> 392001cd 7c3e0b78 953e0008 CPU: 0 PID: 1 Comm: swapper Not tainted 5.10.0-rc2-s3k-dev-01375-gf46ec0d3ecbd-dirty #4092 Call Trace: __warn+0x8c/0xd8 (unreliable) report_bug+0x11c/0x154 program_check_exception+0x1dc/0x6e0 ret_from_except_full+0x0/0x4 --- interrupt: 700 at mmu_mark_initmem_nx+0x24/0x94 LR = free_initmem+0x20/0x58 free_initmem+0x20/0x58 kernel_init+0x1c/0x114 ret_from_kernel_thread+0x14/0x1c ---[ end trace 31702cd2a9570752 ]--- Only call show_stack() when regs is NULL. Fixes: 3f388f28639f ("panic: dump registers on panic_on_warn") Signed-off-by: Christophe Leroy Signed-off-by: Andrew Morton Cc: Alexey Kardashevskiy Cc: Kefeng Wang Link: https://lkml.kernel.org/r/e8c055458b080707f1bc1a98ff8bea79d0cec445.1604748361.git.christophe.leroy@csgroup.eu Signed-off-by: Linus Torvalds --- kernel/panic.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/panic.c b/kernel/panic.c index 396142ee43fd..332736a72a58 100644 --- a/kernel/panic.c +++ b/kernel/panic.c @@ -605,7 +605,8 @@ void __warn(const char *file, int line, void *caller, unsigned taint, panic("panic_on_warn set ...\n"); } - dump_stack(); + if (!regs) + dump_stack(); print_irqtrace_events(current); -- cgit v1.2.3 From 6fa6d28051e9fcaa1570e69648ea13a353a5d218 Mon Sep 17 00:00:00 2001 From: Daniel Xu Date: Tue, 17 Nov 2020 12:05:45 -0800 Subject: lib/strncpy_from_user.c: Mask out bytes after NUL terminator. do_strncpy_from_user() may copy some extra bytes after the NUL terminator into the destination buffer. This usually does not matter for normal string operations. However, when BPF programs key BPF maps with strings, this matters a lot. A BPF program may read strings from user memory by calling the bpf_probe_read_user_str() helper which eventually calls do_strncpy_from_user(). The program can then key a map with the destination buffer. BPF map keys are fixed-width and string-agnostic, meaning that map keys are treated as a set of bytes. The issue is when do_strncpy_from_user() overcopies bytes after the NUL terminator, it can result in seemingly identical strings occupying multiple slots in a BPF map. This behavior is subtle and totally unexpected by the user. This commit masks out the bytes following the NUL while preserving long-sized stride in the fast path. Fixes: 6ae08ae3dea2 ("bpf: Add probe_read_{user, kernel} and probe_read_{user, kernel}_str helpers") Signed-off-by: Daniel Xu Signed-off-by: Alexei Starovoitov Link: https://lore.kernel.org/bpf/21efc982b3e9f2f7b0379eed642294caaa0c27a7.1605642949.git.dxu@dxuuu.xyz --- kernel/trace/bpf_trace.c | 10 ++++++++++ lib/strncpy_from_user.c | 19 +++++++++++++++++-- 2 files changed, 27 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 5113fd423cdf..048c655315f1 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -181,6 +181,16 @@ bpf_probe_read_user_str_common(void *dst, u32 size, { int ret; + /* + * NB: We rely on strncpy_from_user() not copying junk past the NUL + * terminator into `dst`. + * + * strncpy_from_user() does long-sized strides in the fast path. If the + * strncpy does not mask out the bytes after the NUL in `unsafe_ptr`, + * then there could be junk after the NUL in `dst`. If user takes `dst` + * and keys a hash map with it, then semantically identical strings can + * occupy multiple entries in the map. + */ ret = strncpy_from_user_nofault(dst, unsafe_ptr, size); if (unlikely(ret < 0)) memset(dst, 0, size); diff --git a/lib/strncpy_from_user.c b/lib/strncpy_from_user.c index e6d5fcc2cdf3..122d8d0e253c 100644 --- a/lib/strncpy_from_user.c +++ b/lib/strncpy_from_user.c @@ -35,17 +35,32 @@ static inline long do_strncpy_from_user(char *dst, const char __user *src, goto byte_at_a_time; while (max >= sizeof(unsigned long)) { - unsigned long c, data; + unsigned long c, data, mask; /* Fall back to byte-at-a-time if we get a page fault */ unsafe_get_user(c, (unsigned long __user *)(src+res), byte_at_a_time); - *(unsigned long *)(dst+res) = c; + /* + * Note that we mask out the bytes following the NUL. This is + * important to do because string oblivious code may read past + * the NUL. For those routines, we don't want to give them + * potentially random bytes after the NUL in `src`. + * + * One example of such code is BPF map keys. BPF treats map keys + * as an opaque set of bytes. Without the post-NUL mask, any BPF + * maps keyed by strings returned from strncpy_from_user() may + * have multiple entries for semantically identical strings. + */ if (has_zero(c, &data, &constants)) { data = prep_zero_mask(c, data, &constants); data = create_zero_mask(data); + mask = zero_bytemask(data); + *(unsigned long *)(dst+res) = c & mask; return res + find_zero(data); } + + *(unsigned long *)(dst+res) = c; + res += sizeof(unsigned long); max -= sizeof(unsigned long); } -- cgit v1.2.3 From 2801a5da5b25b7af9dd2addd19b2315c02d17b64 Mon Sep 17 00:00:00 2001 From: Luo Meng Date: Wed, 18 Nov 2020 22:49:31 +0900 Subject: fail_function: Remove a redundant mutex unlock Fix a mutex_unlock() issue where before copy_from_user() is not called mutex_locked. Fixes: 4b1a29a7f542 ("error-injection: Support fault injection framework") Reported-by: Hulk Robot Signed-off-by: Luo Meng Signed-off-by: Masami Hiramatsu Signed-off-by: Alexei Starovoitov Acked-by: Masami Hiramatsu Link: https://lore.kernel.org/bpf/160570737118.263807.8358435412898356284.stgit@devnote2 --- kernel/fail_function.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/fail_function.c b/kernel/fail_function.c index 63b349168da7..b0b1ad93fa95 100644 --- a/kernel/fail_function.c +++ b/kernel/fail_function.c @@ -253,7 +253,7 @@ static ssize_t fei_write(struct file *file, const char __user *buffer, if (copy_from_user(buf, buffer, count)) { ret = -EFAULT; - goto out; + goto out_free; } buf[count] = '\0'; sym = strstrip(buf); @@ -307,8 +307,9 @@ static ssize_t fei_write(struct file *file, const char __user *buffer, ret = count; } out: - kfree(buf); mutex_unlock(&fei_lock); +out_free: + kfree(buf); return ret; } -- cgit v1.2.3