pm24.git/kernel/trace/trace_uprobe.c, branch v6.11-rc2

uprobes: prevent mutex_lock() under rcu_read_lock()

2024-05-23T22:44:44Z

Recent changes made uprobe_cpu_buffer preparation lazy, and moved it deeper into __uprobe_trace_func(). This is problematic because __uprobe_trace_func() is called inside rcu_read_lock()/rcu_read_unlock() block, which then calls prepare_uprobe_buffer() -> uprobe_buffer_get() -> mutex_lock(&ucb->mutex), leading to a splat about using mutex under non-sleepable RCU: BUG: sleeping function called from invalid context at kernel/locking/mutex.c:585 in_atomic(): 0, irqs_disabled(): 0, non_block: 0, pid: 98231, name: stress-ng-sigq preempt_count: 0, expected: 0 RCU nest depth: 1, expected: 0 ... Call Trace: dump_stack_lvl+0x3d/0xe0 __might_resched+0x24c/0x270 ? prepare_uprobe_buffer+0xd5/0x1d0 __mutex_lock+0x41/0x820 ? ___perf_sw_event+0x206/0x290 ? __perf_event_task_sched_in+0x54/0x660 ? __perf_event_task_sched_in+0x54/0x660 prepare_uprobe_buffer+0xd5/0x1d0 __uprobe_trace_func+0x4a/0x140 uprobe_dispatcher+0x135/0x280 ? uprobe_dispatcher+0x94/0x280 uprobe_notify_resume+0x650/0xec0 ? atomic_notifier_call_chain+0x21/0x110 ? atomic_notifier_call_chain+0xf8/0x110 irqentry_exit_to_user_mode+0xe2/0x1e0 asm_exc_int3+0x35/0x40 RIP: 0033:0x7f7e1d4da390 Code: 33 04 00 0f 1f 80 00 00 00 00 f3 0f 1e fa b9 01 00 00 00 e9 b2 fc ff ff 66 90 f3 0f 1e fa 31 c9 e9 a5 fc ff ff 0f 1f 44 00 00 0f 1e fa b8 27 00 00 00 0f 05 c3 0f 1f 40 00 f3 0f 1e fa b8 6e RSP: 002b:00007ffd2abc3608 EFLAGS: 00000246 RAX: 0000000000000000 RBX: 0000000076d325f1 RCX: 0000000000000000 RDX: 0000000076d325f1 RSI: 000000000000000a RDI: 00007ffd2abc3690 RBP: 000000000000000a R08: 00017fb700000000 R09: 00017fb700000000 R10: 00017fb700000000 R11: 0000000000000246 R12: 0000000000017ff2 R13: 00007ffd2abc3610 R14: 0000000000000000 R15: 00007ffd2abc3780 Luckily, it's easy to fix by moving prepare_uprobe_buffer() to be called slightly earlier: into uprobe_trace_func() and uretprobe_trace_func(), outside of RCU locked section. This still keeps this buffer preparation lazy and helps avoid the overhead when it's not needed. E.g., if there is only BPF uprobe handler installed on a given uprobe, buffer won't be initialized. Note, the other user of prepare_uprobe_buffer(), __uprobe_perf_func(), is not affected, as it doesn't prepare buffer under RCU read lock. Link: https://lore.kernel.org/all/20240521053017.3708530-1-andrii@kernel.org/ Fixes: 1b8f85defbc8 ("uprobes: prepare uprobe args buffer lazily") Reported-by: Breno Leitao Signed-off-by: Andrii Nakryiko Signed-off-by: Masami Hiramatsu (Google)

uprobes: add speculative lockless system-wide uprobe filter check

2024-05-01T14:18:46Z

It's very common with BPF-based uprobe/uretprobe use cases to have a system-wide (not PID specific) probes used. In this case uprobe's trace_uprobe_filter->nr_systemwide counter is bumped at registration time, and actual filtering is short circuited at the time when uprobe/uretprobe is triggered. This is a great optimization, and the only issue with it is that to even get to checking this counter uprobe subsystem is taking read-side trace_uprobe_filter->rwlock. This is actually noticeable in profiles and is just another point of contention when uprobe is triggered on multiple CPUs simultaneously. This patch moves this nr_systemwide check outside of filter list's rwlock scope, as rwlock is meant to protect list modification, while nr_systemwide-based check is speculative and racy already, despite the lock (as discussed in [0]). trace_uprobe_filter_remove() and trace_uprobe_filter_add() already check for filter->nr_systewide explicitly outside of __uprobe_perf_filter, so no modifications are required there. Confirming with BPF selftests's based benchmarks. BEFORE (based on changes in previous patch) =========================================== uprobe-nop : 2.732 ± 0.022M/s uprobe-push : 2.621 ± 0.016M/s uprobe-ret : 1.105 ± 0.007M/s uretprobe-nop : 1.396 ± 0.007M/s uretprobe-push : 1.347 ± 0.008M/s uretprobe-ret : 0.800 ± 0.006M/s AFTER ===== uprobe-nop : 2.878 ± 0.017M/s (+5.5%, total +8.3%) uprobe-push : 2.753 ± 0.013M/s (+5.3%, total +10.2%) uprobe-ret : 1.142 ± 0.010M/s (+3.8%, total +3.8%) uretprobe-nop : 1.444 ± 0.008M/s (+3.5%, total +6.5%) uretprobe-push : 1.410 ± 0.010M/s (+4.8%, total +7.1%) uretprobe-ret : 0.816 ± 0.002M/s (+2.0%, total +3.9%) In the above, first percentage value is based on top of previous patch (lazy uprobe buffer optimization), while the "total" percentage is based on kernel without any of the changes in this patch set. As can be seen, we get about 4% - 10% speed up, in total, with both lazy uprobe buffer and speculative filter check optimizations. [0] https://lore.kernel.org/bpf/20240313131926.GA19986@redhat.com/ Reviewed-by: Jiri Olsa Link: https://lore.kernel.org/all/20240318181728.2795838-4-andrii@kernel.org/ Signed-off-by: Andrii Nakryiko Acked-by: Masami Hiramatsu (Google) Signed-off-by: Masami Hiramatsu (Google)

uprobes: prepare uprobe args buffer lazily

2024-05-01T14:18:46Z

uprobe_cpu_buffer and corresponding logic to store uprobe args into it are used for uprobes/uretprobes that are created through tracefs or perf events. BPF is yet another user of uprobe/uretprobe infrastructure, but doesn't need uprobe_cpu_buffer and associated data. For BPF-only use cases this buffer handling and preparation is a pure overhead. At the same time, BPF-only uprobe/uretprobe usage is very common in practice. Also, for a lot of cases applications are very senstivie to performance overheads, as they might be tracing a very high frequency functions like malloc()/free(), so every bit of performance improvement matters. All that is to say that this uprobe_cpu_buffer preparation is an unnecessary overhead that each BPF user of uprobes/uretprobe has to pay. This patch is changing this by making uprobe_cpu_buffer preparation optional. It will happen only if either tracefs-based or perf event-based uprobe/uretprobe consumer is registered for given uprobe/uretprobe. For BPF-only use cases this step will be skipped. We used uprobe/uretprobe benchmark which is part of BPF selftests (see [0]) to estimate the improvements. We have 3 uprobe and 3 uretprobe scenarios, which vary an instruction that is replaced by uprobe: nop (fastest uprobe case), `push rbp` (typical case), and non-simulated `ret` instruction (slowest case). Benchmark thread is constantly calling user space function in a tight loop. User space function has attached BPF uprobe or uretprobe program doing nothing but atomic counter increments to count number of triggering calls. Benchmark emits throughput in millions of executions per second. BEFORE these changes ==================== uprobe-nop : 2.657 ± 0.024M/s uprobe-push : 2.499 ± 0.018M/s uprobe-ret : 1.100 ± 0.006M/s uretprobe-nop : 1.356 ± 0.004M/s uretprobe-push : 1.317 ± 0.019M/s uretprobe-ret : 0.785 ± 0.007M/s AFTER these changes =================== uprobe-nop : 2.732 ± 0.022M/s (+2.8%) uprobe-push : 2.621 ± 0.016M/s (+4.9%) uprobe-ret : 1.105 ± 0.007M/s (+0.5%) uretprobe-nop : 1.396 ± 0.007M/s (+2.9%) uretprobe-push : 1.347 ± 0.008M/s (+2.3%) uretprobe-ret : 0.800 ± 0.006M/s (+1.9) So the improvements on this particular machine seems to be between 2% and 5%. [0] https://github.com/torvalds/linux/blob/master/tools/testing/selftests/bpf/benchs/bench_trigger.c Reviewed-by: Jiri Olsa Link: https://lore.kernel.org/all/20240318181728.2795838-3-andrii@kernel.org/ Signed-off-by: Andrii Nakryiko Acked-by: Masami Hiramatsu (Google) Signed-off-by: Masami Hiramatsu (Google)

uprobes: encapsulate preparation of uprobe args buffer

2024-05-01T14:18:46Z

Move the logic of fetching temporary per-CPU uprobe buffer and storing uprobes args into it to a new helper function. Store data size as part of this buffer, simplifying interfaces a bit, as now we only pass single uprobe_cpu_buffer reference around, instead of pointer + dsize. This logic was duplicated across uprobe_dispatcher and uretprobe_dispatcher, and now will be centralized. All this is also in preparation to make this uprobe_cpu_buffer handling logic optional in the next patch. Link: https://lore.kernel.org/all/20240318181728.2795838-2-andrii@kernel.org/ [Masami: update for v6.9-rc3 kernel] Signed-off-by: Andrii Nakryiko Reviewed-by: Jiri Olsa Acked-by: Masami Hiramatsu (Google) Signed-off-by: Masami Hiramatsu (Google)

tracing/probes: Support $argN in return probe (kprobe and fprobe)

2024-03-06T15:27:34Z

Support accessing $argN in the return probe events. This will help users to record entry data in function return (exit) event for simplfing the function entry/exit information in one event, and record the result values (e.g. allocated object/initialized object) at function exit. For example, if we have a function `int init_foo(struct foo *obj, int param)` sometimes we want to check how `obj` is initialized. In such case, we can define a new return event like below; # echo 'r init_foo retval=$retval param=$arg2 field1=+0($arg1)' >> kprobe_events Thus it records the function parameter `param` and its result `obj->field1` (the dereference will be done in the function exit timing) value at once. This also support fprobe, BTF args and'$arg*'. So if CONFIG_DEBUG_INFO_BTF is enabled, we can trace both function parameters and the return value by following command. # echo 'f target_function%return $arg* $retval' >> dynamic_events Link: https://lore.kernel.org/all/170952365552.229804.224112990211602895.stgit@devnote2/ Signed-off-by: Masami Hiramatsu (Google)

tracing/probes: cleanup: Set trace_probe::nr_args at trace_probe_init

2024-03-06T15:27:15Z

Instead of incrementing the trace_probe::nr_args, init it at trace_probe_init(). Without this change, there is no way to get the number of trace_probe arguments while parsing it. This is a cleanup, so the behavior is not changed. Link: https://lore.kernel.org/all/170952363585.229804.13060759900346411951.stgit@devnote2/ Signed-off-by: Masami Hiramatsu (Google)

tracing/uprobe: Replace strlcpy() with strscpy()

2023-12-01T18:25:35Z

strlcpy() reads the entire source buffer first. This read may exceed the destination size limit. This is both inefficient and can lead to linear read overflows if a source string is not NUL-terminated[1]. Additionally, it returns the size of the source string, not the resulting size of the destination string. In an effort to remove strlcpy() completely[2], replace strlcpy() here with strscpy(). The negative return value is already handled by this code so no new handling is needed here. Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strlcpy [1] Link: https://github.com/KSPP/linux/issues/89 [2] Cc: Steven Rostedt Cc: Masami Hiramatsu Cc: linux-trace-kernel@vger.kernel.org Acked-by: "Masami Hiramatsu (Google)" Link: https://lore.kernel.org/r/20231130205607.work.463-kees@kernel.org Signed-off-by: Kees Cook

Merge tag 'probes-v6.6' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace

2023-09-02T18:10:50Z

Pull probes updates from Masami Hiramatsu: - kprobes: use struct_size() for variable size kretprobe_instance data structure. - eprobe: Simplify trace_eprobe list iteration. - probe events: Data structure field access support on BTF argument. - Update BTF argument support on the functions in the kernel loadable modules (only loaded modules are supported). - Move generic BTF access function (search function prototype and get function parameters) to a separated file. - Add a function to search a member of data structure in BTF. - Support accessing BTF data structure member from probe args by C-like arrow('->') and dot('.') operators. e.g. 't sched_switch next=next->pid vruntime=next->se.vruntime' - Support accessing BTF data structure member from $retval. e.g. 'f getname_flags%return +0($retval->name):string' - Add string type checking if BTF type info is available. This will reject if user specify ":string" type for non "char pointer" type. - Automatically assume the fprobe event as a function return event if $retval is used. - selftests/ftrace: Add BTF data field access test cases. - Documentation: Update fprobe event example with BTF data field. * tag 'probes-v6.6' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace: Documentation: tracing: Update fprobe event example with BTF field selftests/ftrace: Add BTF fields access testcases tracing/fprobe-event: Assume fprobe is a return event by $retval tracing/probes: Add string type check with BTF tracing/probes: Support BTF field access from $retval tracing/probes: Support BTF based data structure field access tracing/probes: Add a function to search a member of a struct/union tracing/probes: Move finding func-proto API and getting func-param API to trace_btf tracing/probes: Support BTF argument on module functions tracing/eprobe: Iterate trace_eprobe directly kernel: kprobes: Use struct_size()

tracing/probes: Support BTF argument on module functions

2023-08-23T00:39:15Z

Since the btf returned from bpf_get_btf_vmlinux() only covers functions in the vmlinux, BTF argument is not available on the functions in the modules. Use bpf_find_btf_id() instead of bpf_get_btf_vmlinux()+btf_find_name_kind() so that BTF argument can find the correct struct btf and btf_type in it. With this fix, fprobe events can use `$arg*` on module functions as below # grep nf_log_ip_packet /proc/kallsyms ffffffffa0005c00 t nf_log_ip_packet [nf_log_syslog] ffffffffa0005bf0 t __pfx_nf_log_ip_packet [nf_log_syslog] # echo 'f nf_log_ip_packet $arg*' > dynamic_events # cat dynamic_events f:fprobes/nf_log_ip_packet__entry nf_log_ip_packet net=net pf=pf hooknum=hooknum skb=skb in=in out=out loginfo=loginfo prefix=prefix To support the module's btf which is removable, the struct btf needs to be ref-counted. So this also records the btf in the traceprobe_parse_context and returns the refcount when the parse has done. Link: https://lore.kernel.org/all/169272154223.160970.3507930084247934031.stgit@devnote2/ Suggested-by: Alexei Starovoitov Signed-off-by: Masami Hiramatsu (Google) Acked-by: Steven Rostedt (Google)

bpf: Add support for bpf_get_func_ip helper for uprobe program

2023-08-07T23:42:58Z

Adding support for bpf_get_func_ip helper for uprobe program to return probed address for both uprobe and return uprobe. We discussed this in [1] and agreed that uprobe can have special use of bpf_get_func_ip helper that differs from kprobe. The kprobe bpf_get_func_ip returns: - address of the function if probe is attach on function entry for both kprobe and return kprobe - 0 if the probe is not attach on function entry The uprobe bpf_get_func_ip returns: - address of the probe for both uprobe and return uprobe The reason for this semantic change is that kernel can't really tell if the probe user space address is function entry. The uprobe program is actually kprobe type program attached as uprobe. One of the consequences of this design is that uprobes do not have its own set of helpers, but share them with kprobes. As we need different functionality for bpf_get_func_ip helper for uprobe, I'm adding the bool value to the bpf_trace_run_ctx, so the helper can detect that it's executed in uprobe context and call specific code. The is_uprobe bool is set as true in bpf_prog_run_array_sleepable, which is currently used only for executing bpf programs in uprobe. Renaming bpf_prog_run_array_sleepable to bpf_prog_run_array_uprobe to address that it's only used for uprobes and that it sets the run_ctx.is_uprobe as suggested by Yafang Shao. Suggested-by: Andrii Nakryiko Tested-by: Alan Maguire [1] https://lore.kernel.org/bpf/CAEf4BzZ=xLVkG5eurEuvLU79wAMtwho7ReR+XJAgwhFF4M-7Cg@mail.gmail.com/ Signed-off-by: Jiri Olsa Tested-by: Viktor Malik Acked-by: Yonghong Song Link: https://lore.kernel.org/r/20230807085956.2344866-2-jolsa@kernel.org Signed-off-by: Martin KaFai Lau