On 15. 1. 26 17:48, Nathan Hartman wrote:
On Thu, Jan 15, 2026 at 9:50 AM GitBox <[email protected]> wrote:
The GitHub Actions job "autoconf" on subversion.git/trunk has failed.
Run started by GitHub user asf-gitbox-commits (triggered by
asf-gitbox-commits).
Head commit for run:
2126567c6462bd1d1f246cfda5989fbd313f14ab / Evgeny Kotkov
<[email protected]>
* CHANGES: Draft initial changelog for 1.15.0.
git-svn-id:
https://svn.apache.org/repos/asf/subversion/trunk@1931332
13f79535-47bb-0310-9956-ffa450edef68
Report URL:
https://github.com/apache/subversion/actions/runs/21034688896
With regards,
GitHub Actions via GitBox
I haven't dug into this yet, but I've collected the relevant logs from
GHA here (because they are kept only for a limited time):
The test failure occurs on ubuntu-latest, target davautocheck, system
httpd. (The test appears to succeed on ARM64 as shown below.)
(snip)
At least one test FAILED, checking
/home/runner/work/subversion/subversion/tests.log
FAIL: ra-test: Unknown test failure (-11); see tests.log.
There were some XML validation errors,
checking/home/runner/work/subversion/subversion/tests.log
XML: Opening and ending tag mismatch: log line 2 and logentry, line 3,
column 12 (<string>, line 3)
Summary of test results:
2636 tests PASSED
97 tests SKIPPED
82 tests XFAILED (17 WORK-IN-PROGRESS)
1 test FAILED
Python version: 3.12.3.
SUMMARY: Some tests failed
[...]
(snip)
START: ra-test
TEST ASSERTION FAILED: child_exit_code == 0
PASS: ra-test 2: test ra_svn tunnel callback check
PASS: ra-test 1: test svn_ra_get_location_segments
PASS: ra-test 5: test ra_get_dir2
PASS: ra-test 6: commit callback failure
PASS: ra-test 7: base revision newer than youngest
PASS: ra-test 8: delete revision newer than youngest
PASS: ra-test 4: lock multiple paths
PASS: ra-test 10: check how ra layers handle errors from callbacks
PASS: ra-test 9: check how ra functions handle bad revisions
PASS: ra-test 12: verify checkout over a tunnel
PASS: ra-test 13: check how last change applies to empty commit
PASS: ra-test 14: check commit editor for a locked file
PASS: ra-test 16: test get-deleted-rev errors
PASS: ra-test 15: test get-deleted-rev no delete
FAIL: ra-test: Unknown test failure (-11); see tests.log.
END: ra-test
ELAPSED: ra-test 0:00:00.676186
(snip)
So it would appear ra-test 11 is having a bit of trouble. I hope I can
reproduce it locally...
-11 is the exit code. On Linux, signal 11 is SIGSEGV, so something is
trying to access an invalid memory address. This might be happening in
ra-test 11 (what a coincidence) or it may be that some test setup or
teardown in ra-test leaves dirty data lying around.
I've seen this very intermittently while testing on arm64 macOS, but
have never been able to reproduce it consistently.
-- Brane
