tracing: Fix handle leak in TracepointProvider #12652
Conversation
| @@ -39,6 +39,7 @@ void TracepointProvider::verify_config(const struct md_config_t *conf) { | |||
| void *handle = dlopen(m_library.c_str(), RTLD_NOW); | |||
| if (handle != NULL) { | |||
| m_enabled = true; | |||
| free(handle); | |||
There was a problem hiding this comment.
@xiexingguo that would unload the shared object and I'm not sure that's the intention here? You're right though that free is probably not the way to go in hindsight. @dillaman could we get your opinion here?
There was a problem hiding this comment.
Perhaps making the handle a member and calling dlclose in the destructor?
badone
force-pushed
the
wip-coverity-TracepointProvider
branch
from
fee20e2
to
6872601
Compare
|
I've changed this to be more of an RAII approach which I think is better. @dillaman still like your opinion when you are around. |
| void *handle = dlopen(m_library.c_str(), RTLD_NOW); | ||
| if (handle != NULL) { | ||
| m_handle = dlopen(m_library.c_str(), RTLD_NOW); | ||
| if (m_handle != NULL) { |
There was a problem hiding this comment.
Nit: I would now eliminate the m_enabled member variable and just use m_handle != nullptr to determine if it's enabled.
There was a problem hiding this comment.
Yes, more like @dragonylffly solution in #12673 Agreed, it's a good idea.
Silences Coverity 1397733 Signed-off-by: Brad Hubbard <bhubbard@redhat.com>
badone
force-pushed
the
wip-coverity-TracepointProvider
branch
from
6872601
to
756c201
Compare
|
Updated. Many thanks to @dragonylffly for his contribution. |
Silences Coverity 1397733
Signed-off-by: Brad Hubbard bhubbard@redhat.com