bpf: Fix bpf_put_raw_tracepoint()'s use of __module_address()
authorAndrii Nakryiko <andrii@kernel.org>
Thu, 3 Dec 2020 20:46:21 +0000 (12:46 -0800)
committerAlexei Starovoitov <ast@kernel.org>
Fri, 4 Dec 2020 01:38:20 +0000 (17:38 -0800)
__module_address() needs to be called with preemption disabled or with
module_mutex taken. preempt_disable() is enough for read-only uses, which is
what this fix does. Also, module_put() does internal check for NULL, so drop
it as well.

Fixes: a38d1107f937 ("bpf: support raw tracepoints in modules")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20201203204634.1325171-2-andrii@kernel.org
kernel/trace/bpf_trace.c

index d255bc9..23a390a 100644 (file)
@@ -2060,10 +2060,12 @@ struct bpf_raw_event_map *bpf_get_raw_tracepoint(const char *name)
 
 void bpf_put_raw_tracepoint(struct bpf_raw_event_map *btp)
 {
-       struct module *mod = __module_address((unsigned long)btp);
+       struct module *mod;
 
-       if (mod)
-               module_put(mod);
+       preempt_disable();
+       mod = __module_address((unsigned long)btp);
+       module_put(mod);
+       preempt_enable();
 }
 
 static __always_inline