Commit 0df50e0b authored Dec 12, 2018 by Arnaldo Carvalho de Melo

perf trace: Avoid using raw_syscalls in duplicity with eBPF augmentation

So when we do something like:

   # perf trace -e open*,augmented_raw_syscalls.o

We need to set trace->trace_syscalls because there is logic that use
that when mixing strace-like output with other events, such as scheduler
tracepoints, but with that set we ended up having multiple
raw_syscalls:sys_{enter,exit} setup, which garbled the output, so
check if trace->augmented_raw_syscalls is set and avoid the two extra
tracepoints.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: https://lkml.kernel.org/n/tip-kjmnbrlgu0c38co1ye8egbsb@git.kernel.org


Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

parent 246fbe03

tools/perf/builtin-trace.c

+10 −6

Original line number	Diff line number	Diff line
		@@ -2721,11 +2721,13 @@ static int trace__run(struct trace trace, int argc, const char *argv)

		trace->live = true;

		if (!trace->raw_augmented_syscalls) {
		if (trace->trace_syscalls && trace__add_syscall_newtp(trace))
		goto out_error_raw_syscalls;

		if (trace->trace_syscalls)
		trace->vfs_getname = perf_evlist__add_vfs_getname(evlist);
		}

		if ((trace->trace_pgfaults & TRACE_PFMAJ)) {
		pgfault_maj = perf_evsel__new_pgfault(PERF_COUNT_SW_PAGE_FAULTS_MAJ);
		@@ -2825,9 +2827,11 @@ static int trace__run(struct trace trace, int argc, const char *argv)
		if (err < 0)
		goto out_errno;

		if (trace->syscalls.events.sys_exit) {
		pr_debug("event qualifier tracepoint filter: %s\n",
		trace->syscalls.events.sys_exit->filter);
		}
		}

		err = perf_evlist__apply_filters(evlist, &evsel);
		if (err < 0)

Admin message