summaryrefslogtreecommitdiffstats
path: root/tools/perf
diff options
context:
space:
mode:
authorStanislav Fomichev <sdf@google.com>2019-02-04 16:20:55 -0800
committerAlexei Starovoitov <ast@kernel.org>2019-02-04 17:45:31 -0800
commita8a1f7d09cfc7e18874786c7634c9e71384fcd4e (patch)
tree80f7023de9935668f7299c2eda0ba951eecc9e85 /tools/perf
parent1728b11110f1f86db97a2f3d18cb89552637e391 (diff)
downloadlinux-a8a1f7d09cfc7e18874786c7634c9e71384fcd4e.tar.bz2
libbpf: fix libbpf_print
With the recent print rework we now have the following problem: pr_{warning,info,debug} expand to __pr which calls libbpf_print. libbpf_print does va_start and calls __libbpf_pr with va_list argument. In __base_pr we again do va_start. Because the next argument is a va_list, we don't get correct pointer to the argument (and print noting in my case, I don't know why it doesn't crash tbh). Fix this by changing libbpf_print_fn_t signature to accept va_list and remove unneeded calls to va_start in the existing users. Alternatively, this can we solved by exporting __libbpf_pr and changing __pr macro to (and killing libbpf_print): { if (__libbpf_pr) __libbpf_pr(level, "libbpf: " fmt, ##__VA_ARGS__) } Signed-off-by: Stanislav Fomichev <sdf@google.com> Acked-by: Yonghong Song <yhs@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Diffstat (limited to 'tools/perf')
-rw-r--r--tools/perf/util/bpf-loader.c10
1 files changed, 2 insertions, 8 deletions
diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c
index 38afdbe6a9e0..037d8ff6a634 100644
--- a/tools/perf/util/bpf-loader.c
+++ b/tools/perf/util/bpf-loader.c
@@ -25,15 +25,9 @@
#include "c++/clang-c.h"
static int libbpf_perf_print(enum libbpf_print_level level __attribute__((unused)),
- const char *fmt, ...)
+ const char *fmt, va_list args)
{
- va_list args;
- int ret;
-
- va_start(args, fmt);
- ret = veprintf(1, verbose, pr_fmt(fmt), args);
- va_end(args);
- return ret;
+ return veprintf(1, verbose, pr_fmt(fmt), args);
}
struct bpf_prog_priv {