Fix possible memory leak on TLS handshake failure - #796
Merged
Conversation
On an unrelated PR a memory leak was reported by CI on a TLS test. ``` ==11838== VALGRIND-ERROR-BEGIN ==11838== 71,890 (112 direct, 71,778 indirect) bytes in 1 blocks are definitely lost in loss record 178 of 178 ==11838== at 0x483DD99: calloc (in /usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so) ==11838== by 0x15B06E: tls_new (tls.c:240) ==11838== by 0x16210B: tls_server_conn (tls_server.c:47) ==11838== by 0x16248E: tls_accept_fds (tls_server.c:141) ==11838== by 0x1380EB: sbuf_tls_accept (sbuf.c:1127) ==11838== by 0x11E15A: handle_client_startup (client.c:729) ==11838== by 0x11EF17: client_proto (client.c:1041) ==11838== by 0x136069: sbuf_call_proto (sbuf.c:390) ==11838== by 0x136B58: sbuf_process_pending (sbuf.c:583) ==11838== by 0x13722C: sbuf_main_loop (sbuf.c:763) ==11838== by 0x134FAB: sbuf_accept (sbuf.c:137) ==11838== by 0x12D220: accept_client (objects.c:1540) ==11838== ==11838== VALGRIND-ERROR-END ``` Source: https://cirrus-ci.com/task/5909005202096128?logs=test#L91 When running the same test multiple times locally on my own machine this reproduced ~100 runs. After adding some logging the reason turned out to be that we were not always handling errors from `handle_tls_handshake`. This PR fixes that by handling handshake errors in all places that we call `handle_tls_handshake`. I confirmed this fixed the issue by running the originally failing test 1000 times with valgrind and all runs passed now.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
On an unrelated PR a memory leak was reported by CI on a TLS test.
Source: https://cirrus-ci.com/task/5909005202096128?logs=test#L91
When running the same test multiple times locally on my own machine this
reproduced ~100 runs. After adding some logging the reason turned out to
be that we were not always handling errors from
handle_tls_handshake.This PR fixes that by handling handshake errors in all places that we
call
handle_tls_handshake. I confirmed this fixed the issue by runningthe originally failing test 1000 times with valgrind and all runs passed
now.