user/sven/linux.git/kernel/trace/trace_eprobe.c, branch v6.2.9

Merge tag 'trace-probes-v6.1' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace

2022-11-20T23:31:20Z

Pull tracing/probes fixes from Steven Rostedt: - Fix possible NULL pointer dereference on trace_event_file in kprobe_event_gen_test_exit() - Fix NULL pointer dereference for trace_array in kprobe_event_gen_test_exit() - Fix memory leak of filter string for eprobes - Fix a possible memory leak in rethook_alloc() - Skip clearing aggrprobe's post_handler in kprobe-on-ftrace case which can cause a possible use-after-free - Fix warning in eprobe filter creation - Fix eprobe filter creation as it picked the wrong event for the fields * tag 'trace-probes-v6.1' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace: tracing/eprobe: Fix eprobe filter to make a filter correctly tracing/eprobe: Fix warning in filter creation kprobes: Skip clearing aggrprobe's post_handler in kprobe-on-ftrace case rethook: fix a potential memleak in rethook_alloc() tracing/eprobe: Fix memory leak of filter string tracing: kprobe: Fix potential null-ptr-deref on trace_array in kprobe_event_gen_test_exit() tracing: kprobe: Fix potential null-ptr-deref on trace_event_file in kprobe_event_gen_test_exit()

tracing: Fix race where eprobes can be called before the event

2022-11-20T19:05:50Z

The flag that tells the event to call its triggers after reading the event is set for eprobes after the eprobe is enabled. This leads to a race where the eprobe may be triggered at the beginning of the event where the record information is NULL. The eprobe then dereferences the NULL record causing a NULL kernel pointer bug. Test for a NULL record to keep this from happening. Link: https://lore.kernel.org/linux-trace-kernel/20221116192552.1066630-1-rafaelmendsr@gmail.com/ Link: https://lore.kernel.org/linux-trace-kernel/20221117214249.2addbe10@gandalf.local.home Cc: Linux Trace Kernel Cc: Tzvetomir Stoyanov Cc: Tom Zanussi Cc: stable@vger.kernel.org Fixes: 7491e2c442781 ("tracing: Add a probe that attaches to trace events") Acked-by: Masami Hiramatsu (Google) Reported-by: Rafael Mendonca Signed-off-by: Steven Rostedt (Google)

tracing/eprobe: Fix eprobe filter to make a filter correctly

2022-11-18T01:15:34Z

Since the eprobe filter was defined based on the eprobe's trace event itself, it doesn't work correctly. Use the original trace event of the eprobe when making the filter so that the filter works correctly. Without this fix: # echo 'e syscalls/sys_enter_openat \ flags_rename=$flags:u32 if flags < 1000' >> dynamic_events # echo 1 > events/eprobes/sys_enter_openat/enable [ 114.551550] event trace: Could not enable event sys_enter_openat -bash: echo: write error: Invalid argument With this fix: # echo 'e syscalls/sys_enter_openat \ flags_rename=$flags:u32 if flags < 1000' >> dynamic_events # echo 1 > events/eprobes/sys_enter_openat/enable # tail trace cat-241 [000] ...1. 266.498449: sys_enter_openat: (syscalls.sys_enter_openat) flags_rename=0 cat-242 [000] ...1. 266.977640: sys_enter_openat: (syscalls.sys_enter_openat) flags_rename=0 Link: https://lore.kernel.org/all/166823166395.1385292.8931770640212414483.stgit@devnote3/ Fixes: 752be5c5c910 ("tracing/eprobe: Add eprobe filter support") Reported-by: Rafael Mendonca Tested-by: Rafael Mendonca Signed-off-by: Masami Hiramatsu (Google)

tracing/eprobe: Fix warning in filter creation

2022-11-18T01:15:34Z

The filter pointer (filterp) passed to create_filter() function must be a pointer that references a NULL pointer, otherwise, we get a warning when adding a filter option to the event probe: root@localhost:/sys/kernel/tracing# echo 'e:egroup/stat_runtime_4core sched/sched_stat_runtime \ runtime=$runtime:u32 if cpu < 4' >> dynamic_events [ 5034.340439] ------------[ cut here ]------------ [ 5034.341258] WARNING: CPU: 0 PID: 223 at kernel/trace/trace_events_filter.c:1939 create_filter+0x1db/0x250 [...] stripped [ 5034.345518] RIP: 0010:create_filter+0x1db/0x250 [...] stripped [ 5034.351604] Call Trace: [ 5034.351803] [ 5034.351959] ? process_preds+0x1b40/0x1b40 [ 5034.352241] ? rcu_read_lock_bh_held+0xd0/0xd0 [ 5034.352604] ? kasan_set_track+0x29/0x40 [ 5034.352904] ? kasan_save_alloc_info+0x1f/0x30 [ 5034.353264] create_event_filter+0x38/0x50 [ 5034.353573] __trace_eprobe_create+0x16f4/0x1d20 [ 5034.353964] ? eprobe_dyn_event_release+0x360/0x360 [ 5034.354363] ? mark_held_locks+0xa6/0xf0 [ 5034.354684] ? _raw_spin_unlock_irqrestore+0x35/0x60 [ 5034.355105] ? trace_hardirqs_on+0x41/0x120 [ 5034.355417] ? _raw_spin_unlock_irqrestore+0x35/0x60 [ 5034.355751] ? __create_object+0x5b7/0xcf0 [ 5034.356027] ? lock_is_held_type+0xaf/0x120 [ 5034.356362] ? rcu_read_lock_bh_held+0xb0/0xd0 [ 5034.356716] ? rcu_read_lock_bh_held+0xd0/0xd0 [ 5034.357084] ? kasan_set_track+0x29/0x40 [ 5034.357411] ? kasan_save_alloc_info+0x1f/0x30 [ 5034.357715] ? __kasan_kmalloc+0xb8/0xc0 [ 5034.357985] ? write_comp_data+0x2f/0x90 [ 5034.358302] ? __sanitizer_cov_trace_pc+0x25/0x60 [ 5034.358691] ? argv_split+0x381/0x460 [ 5034.358949] ? write_comp_data+0x2f/0x90 [ 5034.359240] ? eprobe_dyn_event_release+0x360/0x360 [ 5034.359620] trace_probe_create+0xf6/0x110 [ 5034.359940] ? trace_probe_match_command_args+0x240/0x240 [ 5034.360376] eprobe_dyn_event_create+0x21/0x30 [ 5034.360709] create_dyn_event+0xf3/0x1a0 [ 5034.360983] trace_parse_run_command+0x1a9/0x2e0 [ 5034.361297] ? dyn_event_release+0x500/0x500 [ 5034.361591] dyn_event_write+0x39/0x50 [ 5034.361851] vfs_write+0x311/0xe50 [ 5034.362091] ? dyn_event_seq_next+0x40/0x40 [ 5034.362376] ? kernel_write+0x5b0/0x5b0 [ 5034.362637] ? write_comp_data+0x2f/0x90 [ 5034.362937] ? __sanitizer_cov_trace_pc+0x25/0x60 [ 5034.363258] ? ftrace_syscall_enter+0x544/0x840 [ 5034.363563] ? write_comp_data+0x2f/0x90 [ 5034.363837] ? __sanitizer_cov_trace_pc+0x25/0x60 [ 5034.364156] ? write_comp_data+0x2f/0x90 [ 5034.364468] ? write_comp_data+0x2f/0x90 [ 5034.364770] ksys_write+0x158/0x2a0 [ 5034.365022] ? __ia32_sys_read+0xc0/0xc0 [ 5034.365344] __x64_sys_write+0x7c/0xc0 [ 5034.365669] ? syscall_enter_from_user_mode+0x53/0x70 [ 5034.366084] do_syscall_64+0x60/0x90 [ 5034.366356] entry_SYSCALL_64_after_hwframe+0x63/0xcd [ 5034.366767] RIP: 0033:0x7ff0b43938f3 [...] stripped [ 5034.371892] [ 5034.374720] ---[ end trace 0000000000000000 ]--- Link: https://lore.kernel.org/all/20221108202148.1020111-1-rafaelmendsr@gmail.com/ Fixes: 752be5c5c910 ("tracing/eprobe: Add eprobe filter support") Signed-off-by: Rafael Mendonca Acked-by: Masami Hiramatsu (Google) Signed-off-by: Masami Hiramatsu (Google)

tracing/eprobe: Fix memory leak of filter string

2022-11-18T01:15:34Z

The filter string doesn't get freed when a dynamic event is deleted. If a filter is set, then memory is leaked: root@localhost:/sys/kernel/tracing# echo 'e:egroup/stat_runtime_4core \ sched/sched_stat_runtime runtime=$runtime:u32 if cpu < 4' >> dynamic_events root@localhost:/sys/kernel/tracing# echo "-:egroup/stat_runtime_4core" >> dynamic_events root@localhost:/sys/kernel/tracing# echo scan > /sys/kernel/debug/kmemleak [ 224.416373] kmemleak: 1 new suspected memory leaks (see /sys/kernel/debug/kmemleak) root@localhost:/sys/kernel/tracing# cat /sys/kernel/debug/kmemleak unreferenced object 0xffff88810156f1b8 (size 8): comm "bash", pid 224, jiffies 4294935612 (age 55.800s) hex dump (first 8 bytes): 63 70 75 20 3c 20 34 00 cpu < 4. backtrace: [<000000009f880725>] __kmem_cache_alloc_node+0x18e/0x720 [<0000000042492946>] __kmalloc+0x57/0x240 [<0000000034ea7995>] __trace_eprobe_create+0x1214/0x1d30 [<00000000d70ef730>] trace_probe_create+0xf6/0x110 [<00000000915c7b16>] eprobe_dyn_event_create+0x21/0x30 [<000000000d894386>] create_dyn_event+0xf3/0x1a0 [<00000000e9af57d5>] trace_parse_run_command+0x1a9/0x2e0 [<0000000080777f18>] dyn_event_write+0x39/0x50 [<0000000089f0ec73>] vfs_write+0x311/0xe50 [<000000003da1bdda>] ksys_write+0x158/0x2a0 [<00000000bb1e616e>] __x64_sys_write+0x7c/0xc0 [<00000000e8aef1f7>] do_syscall_64+0x60/0x90 [<00000000fe7fe8ba>] entry_SYSCALL_64_after_hwframe+0x63/0xcd Additionally, in __trace_eprobe_create() function, if an error occurs after the call to trace_eprobe_parse_filter(), which allocates the filter string, then memory is also leaked. That can be reproduced by creating the same event probe twice: root@localhost:/sys/kernel/tracing# echo 'e:egroup/stat_runtime_4core \ sched/sched_stat_runtime runtime=$runtime:u32 if cpu < 4' >> dynamic_events root@localhost:/sys/kernel/tracing# echo 'e:egroup/stat_runtime_4core \ sched/sched_stat_runtime runtime=$runtime:u32 if cpu < 4' >> dynamic_events -bash: echo: write error: File exists root@localhost:/sys/kernel/tracing# echo scan > /sys/kernel/debug/kmemleak [ 207.871584] kmemleak: 1 new suspected memory leaks (see /sys/kernel/debug/kmemleak) root@localhost:/sys/kernel/tracing# cat /sys/kernel/debug/kmemleak unreferenced object 0xffff8881020d17a8 (size 8): comm "bash", pid 223, jiffies 4294938308 (age 31.000s) hex dump (first 8 bytes): 63 70 75 20 3c 20 34 00 cpu < 4. backtrace: [<000000000e4f5f31>] __kmem_cache_alloc_node+0x18e/0x720 [<0000000024f0534b>] __kmalloc+0x57/0x240 [<000000002930a28e>] __trace_eprobe_create+0x1214/0x1d30 [<0000000028387903>] trace_probe_create+0xf6/0x110 [<00000000a80d6a9f>] eprobe_dyn_event_create+0x21/0x30 [<000000007168698c>] create_dyn_event+0xf3/0x1a0 [<00000000f036bf6a>] trace_parse_run_command+0x1a9/0x2e0 [<00000000014bde8b>] dyn_event_write+0x39/0x50 [<0000000078a097f7>] vfs_write+0x311/0xe50 [<00000000996cb208>] ksys_write+0x158/0x2a0 [<00000000a3c2acb0>] __x64_sys_write+0x7c/0xc0 [<0000000006b5d698>] do_syscall_64+0x60/0x90 [<00000000780e8ecf>] entry_SYSCALL_64_after_hwframe+0x63/0xcd Fix both issues by releasing the filter string in trace_event_probe_cleanup(). Link: https://lore.kernel.org/all/20221108235738.1021467-1-rafaelmendsr@gmail.com/ Fixes: 752be5c5c910 ("tracing/eprobe: Add eprobe filter support") Signed-off-by: Rafael Mendonca Acked-by: Masami Hiramatsu (Google) Signed-off-by: Masami Hiramatsu (Google)

tracing: Move duplicate code of trace_kprobe/eprobe.c into header

2022-10-12T17:50:00Z

The functions: fetch_store_strlen_user() fetch_store_strlen() fetch_store_string_user() fetch_store_string() are identical in both trace_kprobe.c and trace_eprobe.c. Move them into a new header file trace_probe_kernel.h to share it. This code will later be used by the synthetic events as well. Marked for stable as a fix for a crash in synthetic events requires it. Link: https://lkml.kernel.org/r/20221012104534.467668078@goodmis.org Cc: stable@vger.kernel.org Cc: Andrew Morton Cc: Tom Zanussi Acked-by: Masami Hiramatsu (Google) Reviewed-by: Tom Zanussi Fixes: bd82631d7ccdc ("tracing: Add support for dynamic strings to synthetic events") Signed-off-by: Steven Rostedt (Google)

tracing/eprobe: Fix alloc event dir failed when event name no set

2022-09-27T18:48:26Z

The event dir will alloc failed when event name no set, using the command: "echo "e:esys/ syscalls/sys_enter_openat file=\$filename:string" >> dynamic_events" It seems that dir name="syscalls/sys_enter_openat" is not allowed in debugfs. So just use the "sys_enter_openat" as the event name. Link: https://lkml.kernel.org/r/1664028814-45923-1-git-send-email-chentao.kernel@linux.alibaba.com Cc: Ingo Molnar Cc: Tom Zanussi Cc: Linyu Yuan Cc: Tao Chen Signed-off-by: Tao Chen Signed-off-by: Steven Rostedt (Google)

tracing/eprobe: Add eprobe filter support

2022-09-26T17:01:08Z

Add the filter option to the event probe. This is useful if user wants to derive a new event based on the condition of the original event. E.g. echo 'e:egroup/stat_runtime_4core sched/sched_stat_runtime \ runtime=$runtime:u32 if cpu < 4' >> ../dynamic_events Then it can filter the events only on first 4 cores. Note that the fields used for 'if' must be the fields in the original events, not eprobe events. Link: https://lkml.kernel.org/r/165932114513.2850673.2592206685744598080.stgit@devnote2 Cc: Tzvetomir Stoyanov Cc: Ingo Molnar Signed-off-by: Masami Hiramatsu (Google) Signed-off-by: Steven Rostedt (Google)

tracing/eprobes: Have event probes be consistent with kprobes and uprobes

2022-08-21T19:56:08Z

Currently, if a symbol "@" is attempted to be used with an event probe (eprobes), it will cause a NULL pointer dereference crash. Both kprobes and uprobes can reference data other than the main registers. Such as immediate address, symbols and the current task name. Have eprobes do the same thing. For "comm", if "comm" is used and the event being attached to does not have the "comm" field, then make it the "$comm" that kprobes has. This is consistent to the way histograms and filters work. Link: https://lkml.kernel.org/r/20220820134401.136924220@goodmis.org Cc: stable@vger.kernel.org Cc: Ingo Molnar Cc: Andrew Morton Cc: Masami Hiramatsu Cc: Tzvetomir Stoyanov Cc: Tom Zanussi Fixes: 7491e2c44278 ("tracing: Add a probe that attaches to trace events") Signed-off-by: Steven Rostedt (Google)

tracing/eprobes: Fix reading of string fields

2022-08-21T19:56:08Z

Currently when an event probe (eprobe) hooks to a string field, it does not display it as a string, but instead as a number. This makes the field rather useless. Handle the different kinds of strings, dynamic, static, relational/dynamic etc. Now when a string field is used, the ":string" type can be used to display it: echo "e:sw sched/sched_switch comm=$next_comm:string" > dynamic_events Link: https://lkml.kernel.org/r/20220820134400.959640191@goodmis.org Cc: stable@vger.kernel.org Cc: Ingo Molnar Cc: Andrew Morton Cc: Tzvetomir Stoyanov Cc: Tom Zanussi Fixes: 7491e2c44278 ("tracing: Add a probe that attaches to trace events") Acked-by: Masami Hiramatsu (Google) Signed-off-by: Steven Rostedt (Google)