From 7bb83f6fc4ee84e95d0ac0d14452c2619fb3fe70 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 8 Jan 2021 13:19:38 +0900 Subject: tracing/kprobes: Do the notrace functions check without kprobes on ftrace Enable the notrace function check on the architecture which doesn't support kprobes on ftrace but support dynamic ftrace. This notrace function check is not only for the kprobes on ftrace but also sw-breakpoint based kprobes. Thus there is no reason to limit this check for the arch which supports kprobes on ftrace. This also changes the dependency of Kconfig. Because kprobe event uses the function tracer's address list for identifying notrace function, if the CONFIG_DYNAMIC_FTRACE=n, it can not check whether the target function is notrace or not. Link: https://lkml.kernel.org/r/20210105065730.2634785-1-naveen.n.rao@linux.vnet.ibm.com Link: https://lkml.kernel.org/r/161007957862.114704.4512260007555399463.stgit@devnote2 Cc: stable@vger.kernel.org Fixes: 45408c4f92506 ("tracing: kprobes: Prohibit probing on notrace function") Acked-by: Naveen N. Rao Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/Kconfig | 2 +- kernel/trace/trace_kprobe.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index d5a19413d4f8..c1a62ae7e812 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -538,7 +538,7 @@ config KPROBE_EVENTS config KPROBE_EVENTS_ON_NOTRACE bool "Do NOT protect notrace function from kprobe events" depends on KPROBE_EVENTS - depends on KPROBES_ON_FTRACE + depends on DYNAMIC_FTRACE default n help This is only for the developers who want to debug ftrace itself diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 9c31f42245e9..e6fba1798771 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -434,7 +434,7 @@ static int disable_trace_kprobe(struct trace_event_call *call, return 0; } -#if defined(CONFIG_KPROBES_ON_FTRACE) && \ +#if defined(CONFIG_DYNAMIC_FTRACE) && \ !defined(CONFIG_KPROBE_EVENTS_ON_NOTRACE) static bool __within_notrace_func(unsigned long addr) { -- cgit v1.2.3 From 7e0a9220467dbcfdc5bc62825724f3e52e50ab31 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 29 Jan 2021 10:13:53 -0500 Subject: fgraph: Initialize tracing_graph_pause at task creation On some archs, the idle task can call into cpu_suspend(). The cpu_suspend() will disable or pause function graph tracing, as there's some paths in bringing down the CPU that can have issues with its return address being modified. The task_struct structure has a "tracing_graph_pause" atomic counter, that when set to something other than zero, the function graph tracer will not modify the return address. The problem is that the tracing_graph_pause counter is initialized when the function graph tracer is enabled. This can corrupt the counter for the idle task if it is suspended in these architectures. CPU 1 CPU 2 ----- ----- do_idle() cpu_suspend() pause_graph_tracing() task_struct->tracing_graph_pause++ (0 -> 1) start_graph_tracing() for_each_online_cpu(cpu) { ftrace_graph_init_idle_task(cpu) task-struct->tracing_graph_pause = 0 (1 -> 0) unpause_graph_tracing() task_struct->tracing_graph_pause-- (0 -> -1) The above should have gone from 1 to zero, and enabled function graph tracing again. But instead, it is set to -1, which keeps it disabled. There's no reason that the field tracing_graph_pause on the task_struct can not be initialized at boot up. Cc: stable@vger.kernel.org Fixes: 380c4b1411ccd ("tracing/function-graph-tracer: append the tracing_graph_flag") Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=211339 Reported-by: pierre.gondois@arm.com Signed-off-by: Steven Rostedt (VMware) --- init/init_task.c | 3 ++- kernel/trace/fgraph.c | 2 -- 2 files changed, 2 insertions(+), 3 deletions(-) (limited to 'kernel/trace') diff --git a/init/init_task.c b/init/init_task.c index 8a992d73e6fb..3711cdaafed2 100644 --- a/init/init_task.c +++ b/init/init_task.c @@ -198,7 +198,8 @@ struct task_struct init_task .lockdep_recursion = 0, #endif #ifdef CONFIG_FUNCTION_GRAPH_TRACER - .ret_stack = NULL, + .ret_stack = NULL, + .tracing_graph_pause = ATOMIC_INIT(0), #endif #if defined(CONFIG_TRACING) && defined(CONFIG_PREEMPTION) .trace_recursion = 0, diff --git a/kernel/trace/fgraph.c b/kernel/trace/fgraph.c index 73edb9e4f354..29a6ebeebc9e 100644 --- a/kernel/trace/fgraph.c +++ b/kernel/trace/fgraph.c @@ -394,7 +394,6 @@ static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list) } if (t->ret_stack == NULL) { - atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); t->curr_ret_stack = -1; t->curr_ret_depth = -1; @@ -489,7 +488,6 @@ static DEFINE_PER_CPU(struct ftrace_ret_stack *, idle_ret_stack); static void graph_init_task(struct task_struct *t, struct ftrace_ret_stack *ret_stack) { - atomic_set(&t->tracing_graph_pause, 0); atomic_set(&t->trace_overrun, 0); t->ftrace_timestamp = 0; /* make curr_ret_stack visible before we add the ret_stack */ -- cgit v1.2.3 From da7f84cdf02fd5f66864041f45018b328911b722 Mon Sep 17 00:00:00 2001 From: Viktor Rosendahl Date: Tue, 19 Jan 2021 17:43:43 +0100 Subject: tracing: Use pause-on-trace with the latency tracers Eaerlier, tracing was disabled when reading the trace file. This behavior was changed with: commit 06e0a548bad0 ("tracing: Do not disable tracing when reading the trace file"). This doesn't seem to work with the latency tracers. The above mentioned commit dit not only change the behavior but also added an option to emulate the old behavior. The idea with this patch is to enable this pause-on-trace option when the latency tracers are used. Link: https://lkml.kernel.org/r/20210119164344.37500-2-Viktor.Rosendahl@bmw.de Cc: stable@vger.kernel.org Fixes: 06e0a548bad0 ("tracing: Do not disable tracing when reading the trace file") Signed-off-by: Viktor Rosendahl Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_irqsoff.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c index d06aab4dcbb8..6756379b661f 100644 --- a/kernel/trace/trace_irqsoff.c +++ b/kernel/trace/trace_irqsoff.c @@ -562,6 +562,8 @@ static int __irqsoff_tracer_init(struct trace_array *tr) /* non overwrite screws up the latency tracers */ set_tracer_flag(tr, TRACE_ITER_OVERWRITE, 1); set_tracer_flag(tr, TRACE_ITER_LATENCY_FMT, 1); + /* without pause, we will produce garbage if another latency occurs */ + set_tracer_flag(tr, TRACE_ITER_PAUSE_ON_TRACE, 1); tr->max_latency = 0; irqsoff_trace = tr; @@ -583,11 +585,13 @@ static void __irqsoff_tracer_reset(struct trace_array *tr) { int lat_flag = save_flags & TRACE_ITER_LATENCY_FMT; int overwrite_flag = save_flags & TRACE_ITER_OVERWRITE; + int pause_flag = save_flags & TRACE_ITER_PAUSE_ON_TRACE; stop_irqsoff_tracer(tr, is_graph(tr)); set_tracer_flag(tr, TRACE_ITER_LATENCY_FMT, lat_flag); set_tracer_flag(tr, TRACE_ITER_OVERWRITE, overwrite_flag); + set_tracer_flag(tr, TRACE_ITER_PAUSE_ON_TRACE, pause_flag); ftrace_reset_array_ops(tr); irqsoff_busy = false; -- cgit v1.2.3 From 97c753e62e6c31a404183898d950d8c08d752dbd Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Thu, 28 Jan 2021 00:37:51 +0900 Subject: tracing/kprobe: Fix to support kretprobe events on unloaded modules Fix kprobe_on_func_entry() returns error code instead of false so that register_kretprobe() can return an appropriate error code. append_trace_kprobe() expects the kprobe registration returns -ENOENT when the target symbol is not found, and it checks whether the target module is unloaded or not. If the target module doesn't exist, it defers to probe the target symbol until the module is loaded. However, since register_kretprobe() returns -EINVAL instead of -ENOENT in that case, it always fail on putting the kretprobe event on unloaded modules. e.g. Kprobe event: /sys/kernel/debug/tracing # echo p xfs:xfs_end_io >> kprobe_events [ 16.515574] trace_kprobe: This probe might be able to register after target module is loaded. Continue. Kretprobe event: (p -> r) /sys/kernel/debug/tracing # echo r xfs:xfs_end_io >> kprobe_events sh: write error: Invalid argument /sys/kernel/debug/tracing # cat error_log [ 41.122514] trace_kprobe: error: Failed to register probe event Command: r xfs:xfs_end_io ^ To fix this bug, change kprobe_on_func_entry() to detect symbol lookup failure and return -ENOENT in that case. Otherwise it returns -EINVAL or 0 (succeeded, given address is on the entry). Link: https://lkml.kernel.org/r/161176187132.1067016.8118042342894378981.stgit@devnote2 Cc: stable@vger.kernel.org Fixes: 59158ec4aef7 ("tracing/kprobes: Check the probe on unloaded module correctly") Reported-by: Jianlin Lv Signed-off-by: Masami Hiramatsu Signed-off-by: Steven Rostedt (VMware) --- include/linux/kprobes.h | 2 +- kernel/kprobes.c | 34 +++++++++++++++++++++++++--------- kernel/trace/trace_kprobe.c | 10 ++++++---- 3 files changed, 32 insertions(+), 14 deletions(-) (limited to 'kernel/trace') diff --git a/include/linux/kprobes.h b/include/linux/kprobes.h index b3a36b0cfc81..1883a4a9f16a 100644 --- a/include/linux/kprobes.h +++ b/include/linux/kprobes.h @@ -266,7 +266,7 @@ extern void kprobes_inc_nmissed_count(struct kprobe *p); extern bool arch_within_kprobe_blacklist(unsigned long addr); extern int arch_populate_kprobe_blacklist(void); extern bool arch_kprobe_on_func_entry(unsigned long offset); -extern bool kprobe_on_func_entry(kprobe_opcode_t *addr, const char *sym, unsigned long offset); +extern int kprobe_on_func_entry(kprobe_opcode_t *addr, const char *sym, unsigned long offset); extern bool within_kprobe_blacklist(unsigned long addr); extern int kprobe_add_ksym_blacklist(unsigned long entry); diff --git a/kernel/kprobes.c b/kernel/kprobes.c index f7fb5d135930..1a5bc321e0a5 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -1954,29 +1954,45 @@ bool __weak arch_kprobe_on_func_entry(unsigned long offset) return !offset; } -bool kprobe_on_func_entry(kprobe_opcode_t *addr, const char *sym, unsigned long offset) +/** + * kprobe_on_func_entry() -- check whether given address is function entry + * @addr: Target address + * @sym: Target symbol name + * @offset: The offset from the symbol or the address + * + * This checks whether the given @addr+@offset or @sym+@offset is on the + * function entry address or not. + * This returns 0 if it is the function entry, or -EINVAL if it is not. + * And also it returns -ENOENT if it fails the symbol or address lookup. + * Caller must pass @addr or @sym (either one must be NULL), or this + * returns -EINVAL. + */ +int kprobe_on_func_entry(kprobe_opcode_t *addr, const char *sym, unsigned long offset) { kprobe_opcode_t *kp_addr = _kprobe_addr(addr, sym, offset); if (IS_ERR(kp_addr)) - return false; + return PTR_ERR(kp_addr); - if (!kallsyms_lookup_size_offset((unsigned long)kp_addr, NULL, &offset) || - !arch_kprobe_on_func_entry(offset)) - return false; + if (!kallsyms_lookup_size_offset((unsigned long)kp_addr, NULL, &offset)) + return -ENOENT; - return true; + if (!arch_kprobe_on_func_entry(offset)) + return -EINVAL; + + return 0; } int register_kretprobe(struct kretprobe *rp) { - int ret = 0; + int ret; struct kretprobe_instance *inst; int i; void *addr; - if (!kprobe_on_func_entry(rp->kp.addr, rp->kp.symbol_name, rp->kp.offset)) - return -EINVAL; + ret = kprobe_on_func_entry(rp->kp.addr, rp->kp.symbol_name, rp->kp.offset); + if (ret) + return ret; if (kretprobe_blacklist_size) { addr = kprobe_addr(&rp->kp); diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index e6fba1798771..56c7fbff7bd7 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -221,9 +221,9 @@ bool trace_kprobe_on_func_entry(struct trace_event_call *call) { struct trace_kprobe *tk = trace_kprobe_primary_from_call(call); - return tk ? kprobe_on_func_entry(tk->rp.kp.addr, + return tk ? (kprobe_on_func_entry(tk->rp.kp.addr, tk->rp.kp.addr ? NULL : tk->rp.kp.symbol_name, - tk->rp.kp.addr ? 0 : tk->rp.kp.offset) : false; + tk->rp.kp.addr ? 0 : tk->rp.kp.offset) == 0) : false; } bool trace_kprobe_error_injectable(struct trace_event_call *call) @@ -828,9 +828,11 @@ static int trace_kprobe_create(int argc, const char *argv[]) } if (is_return) flags |= TPARG_FL_RETURN; - if (kprobe_on_func_entry(NULL, symbol, offset)) + ret = kprobe_on_func_entry(NULL, symbol, offset); + if (ret == 0) flags |= TPARG_FL_FENTRY; - if (offset && is_return && !(flags & TPARG_FL_FENTRY)) { + /* Defer the ENOENT case until register kprobe */ + if (ret == -EINVAL && is_return) { trace_probe_log_err(0, BAD_RETPROBE); goto parse_error; } -- cgit v1.2.3 From 548f1191d86ccb9bde2a5305988877b7584c01eb Mon Sep 17 00:00:00 2001 From: Alexei Starovoitov Date: Tue, 2 Feb 2021 23:06:36 -0800 Subject: bpf: Unbreak BPF_PROG_TYPE_KPROBE when kprobe is called via do_int3 The commit 0d00449c7a28 ("x86: Replace ist_enter() with nmi_enter()") converted do_int3 handler to be "NMI-like". That made old if (in_nmi()) check abort execution of bpf programs attached to kprobe when kprobe is firing via int3 (For example when kprobe is placed in the middle of the function). Remove the check to restore user visible behavior. Fixes: 0d00449c7a28 ("x86: Replace ist_enter() with nmi_enter()") Reported-by: Nikolay Borisov Signed-off-by: Alexei Starovoitov Signed-off-by: Daniel Borkmann Tested-by: Nikolay Borisov Reviewed-by: Masami Hiramatsu Link: https://lore.kernel.org/bpf/20210203070636.70926-1-alexei.starovoitov@gmail.com --- kernel/trace/bpf_trace.c | 3 --- 1 file changed, 3 deletions(-) (limited to 'kernel/trace') diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c index 6c0018abe68a..764400260eb6 100644 --- a/kernel/trace/bpf_trace.c +++ b/kernel/trace/bpf_trace.c @@ -96,9 +96,6 @@ unsigned int trace_call_bpf(struct trace_event_call *call, void *ctx) { unsigned int ret; - if (in_nmi()) /* not supported yet */ - return 1; - cant_sleep(); if (unlikely(__this_cpu_inc_return(bpf_prog_active) != 1)) { -- cgit v1.2.3 From 256cfdd6fdf70c6fcf0f7c8ddb0ebd73ce8f3bc9 Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Fri, 5 Feb 2021 15:40:04 -0500 Subject: tracing: Do not count ftrace events in top level enable output The file /sys/kernel/tracing/events/enable is used to enable all events by echoing in "1", or disabling all events when echoing in "0". To know if all events are enabled, disabled, or some are enabled but not all of them, cating the file should show either "1" (all enabled), "0" (all disabled), or "X" (some enabled but not all of them). This works the same as the "enable" files in the individule system directories (like tracing/events/sched/enable). But when all events are enabled, the top level "enable" file shows "X". The reason is that its checking the "ftrace" events, which are special events that only exist for their format files. These include the format for the function tracer events, that are enabled when the function tracer is enabled, but not by the "enable" file. The check includes these events, which will always be disabled, and even though all true events are enabled, the top level "enable" file will show "X" instead of "1". To fix this, have the check test the event's flags to see if it has the "IGNORE_ENABLE" flag set, and if so, not test it. Cc: stable@vger.kernel.org Fixes: 553552ce1796c ("tracing: Combine event filter_active and enable into single flags field") Reported-by: "Yordan Karadzhov (VMware)" Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace_events.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index e9d28eeccb7e..d387b774ceeb 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -1212,7 +1212,8 @@ system_enable_read(struct file *filp, char __user *ubuf, size_t cnt, mutex_lock(&event_mutex); list_for_each_entry(file, &tr->events, list) { call = file->event_call; - if (!trace_event_name(call) || !call->class || !call->class->reg) + if ((call->flags & TRACE_EVENT_FL_IGNORE_ENABLE) || + !trace_event_name(call) || !call->class || !call->class->reg) continue; if (system && strcmp(call->class->system, system->name) != 0) -- cgit v1.2.3 From b220c049d5196dd94d992dd2dc8cba1a5e6123bf Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (VMware)" Date: Wed, 10 Feb 2021 11:53:22 -0500 Subject: tracing: Check length before giving out the filter buffer When filters are used by trace events, a page is allocated on each CPU and used to copy the trace event fields to this page before writing to the ring buffer. The reason to use the filter and not write directly into the ring buffer is because a filter may discard the event and there's more overhead on discarding from the ring buffer than the extra copy. The problem here is that there is no check against the size being allocated when using this page. If an event asks for more than a page size while being filtered, it will get only a page, leading to the caller writing more that what was allocated. Check the length of the request, and if it is more than PAGE_SIZE minus the header default back to allocating from the ring buffer directly. The ring buffer may reject the event if its too big anyway, but it wont overflow. Link: https://lore.kernel.org/ath10k/1612839593-2308-1-git-send-email-wgong@codeaurora.org/ Cc: stable@vger.kernel.org Fixes: 0fc1b09ff1ff4 ("tracing: Use temp buffer when filtering events") Reported-by: Wen Gong Signed-off-by: Steven Rostedt (VMware) --- kernel/trace/trace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel/trace') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index b8a2d786b503..b5815a022ecc 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -2745,7 +2745,7 @@ trace_event_buffer_lock_reserve(struct trace_buffer **current_rb, (entry = this_cpu_read(trace_buffered_event))) { /* Try to use the per cpu buffer first */ val = this_cpu_inc_return(trace_buffered_event_cnt); - if (val == 1) { + if ((len < (PAGE_SIZE - sizeof(*entry))) && val == 1) { trace_event_setup(entry, type, flags, pc); entry->array[0] = len; return entry; -- cgit v1.2.3