perf trace: Validate syscall id before growing syscall table
authorArnaldo Carvalho de Melo <acme@redhat.com>
Sat, 20 Oct 2012 15:39:03 +0000 (12:39 -0300)
committerArnaldo Carvalho de Melo <acme@redhat.com>
Mon, 22 Oct 2012 01:07:36 +0000 (23:07 -0200)
In some cases the ID for a syscall read thru the raw_syscalls tracepoint
is bogus, still needs to be investigated why, but to make the tool more
robust first try to resolve the ID to a name via libaudit and if it
fails, don't grow the table.

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-0lsokw3xor7c4ijo45u6bauh@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
tools/perf/builtin-trace.c

index dec8ced..83c6515 100644 (file)
@@ -56,6 +56,10 @@ static int trace__read_syscall_info(struct trace *trace, int id)
 {
        char tp_name[128];
        struct syscall *sc;
+       const char *name = audit_syscall_to_name(id, trace->audit_machine);
+
+       if (name == NULL)
+               return -1;
 
        if (id > trace->syscalls.max) {
                struct syscall *nsyscalls = realloc(trace->syscalls.table, (id + 1) * sizeof(*sc));
@@ -75,11 +79,8 @@ static int trace__read_syscall_info(struct trace *trace, int id)
        }
 
        sc = trace->syscalls.table + id;
-       sc->name = audit_syscall_to_name(id, trace->audit_machine);
-       if (sc->name == NULL)
-               return -1;
-
-       sc->fmt = syscall_fmt__find(sc->name);
+       sc->name = name;
+       sc->fmt  = syscall_fmt__find(sc->name);
 
        snprintf(tp_name, sizeof(tp_name), "sys_enter_%s", sc->name);
        sc->tp_format = event_format__new("syscalls", tp_name);