Project

General

Profile

Bug #5466

raring tarball builds mysteriously fail make check

Added by Dan Mick almost 11 years ago. Updated over 10 years ago.

Status:
Can't reproduce
Priority:
High
Assignee:
-
Category:
-
Target version:
-
% Done:

0%

Source:
other
Tags:
Backport:
Regression:
No
Severity:
3 - minor
Reviewed:
Affected Versions:
ceph-qa-suite:
Pull request ID:
Crash signature (v1):
Crash signature (v2):

Description

I've noticed this intermittently: make check on raring tarball builds fails like this:

    [==========] Running 4 tests from 1 test case.
    [----------] Global test environment set-up.
    [----------] 4 tests from FlatIndex
    [ RUN ] FlatIndex.FlatIndex
    [ OK ] FlatIndex.FlatIndex (0 ms)
    [ RUN ] FlatIndex.collection

    [WARNING] ./src/gtest-death-test.cc:741:: Death tests use fork(), which is unsafe particularly in a threaded context. For this test, Google Test couldn't detect the number of threads.
    make[4]: *** [check-TESTS] Terminated
    make[2]: *** [check-recursive] Terminated
        make: *** [check-recursive] Terminated 
    make[1]: *** [check] Terminated
    make[3]: *** [check-am] Terminated
    + exit 5

>>> Result code: 40
FAIL
‘out/log’ -> ‘out/fail/d1d902846d8baa6f0d6b49959aaee0ec9d8435fe’
Done: d1d902846d8baa6f0d6b49959aaee0ec9d8435fe

I don't know what's terminating the run, or why; it seems to Greg and me that it's something
external to the make run. ??

I looked into the constant warnings about death tests: "Death tests use fork(), which is unsafe particularly..."
The trigger of the message is bogus; it tries to count running threads, or rather doesn't try:

size_t GetThreadCount() {
  // There's no portable way to detect the number of threads, so we just
  // return 0 to indicate that we cannot detect it.
  return 0;
}

so that's probably a red herring.

History

#1 Updated by Sage Weil over 10 years ago

  • Status changed from New to Can't reproduce

this hasn't happened since this has been migrated to lxc

Also available in: Atom PDF