From: Rong Tao <rongtao@cestc.cn>
Date: Thu, 7 Sep 2023 01:59:14 +0000 (+0800)
Subject: selftests/bpf: trace_helpers.c: Add a global ksyms initialization mutex
X-Git-Tag: v6.7-rc1~160^2~316^2~12^2
X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=a28b1ba25934f24b9aabfc5cb86247150b7bb67d;p=thirdparty%2Fkernel%2Flinux.git

selftests/bpf: trace_helpers.c: Add a global ksyms initialization mutex

As Jirka said [0], we just need to make sure that global ksyms
initialization won't race.

[0] https://lore.kernel.org/lkml/ZPCbAs3ItjRd8XVh@krava/

Signed-off-by: Rong Tao <rongtao@cestc.cn>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/tencent_5D0A837E219E2CFDCB0495DAD7D5D1204407@qq.com
---

diff --git a/tools/testing/selftests/bpf/trace_helpers.c b/tools/testing/selftests/bpf/trace_helpers.c
index dc4efaf538ae9..4faa898ff7fc4 100644
--- a/tools/testing/selftests/bpf/trace_helpers.c
+++ b/tools/testing/selftests/bpf/trace_helpers.c
@@ -7,6 +7,7 @@
 #include <errno.h>
 #include <fcntl.h>
 #include <poll.h>
+#include <pthread.h>
 #include <unistd.h>
 #include <linux/perf_event.h>
 #include <sys/mman.h>
@@ -26,6 +27,7 @@ struct ksyms {
 };
 
 static struct ksyms *ksyms;
+static pthread_mutex_t ksyms_mutex = PTHREAD_MUTEX_INITIALIZER;
 
 static int ksyms__add_symbol(struct ksyms *ksyms, const char *name,
 			     unsigned long addr)
@@ -109,8 +111,10 @@ error:
 
 int load_kallsyms(void)
 {
+	pthread_mutex_lock(&ksyms_mutex);
 	if (!ksyms)
 		ksyms = load_kallsyms_local();
+	pthread_mutex_unlock(&ksyms_mutex);
 	return ksyms ? 0 : 1;
 }