[Bug build/25919] New: FAIL: misc/tst-atomic and FAIL: misc/tst-atomic-long non-deterministic failures

[Bug build/25919] New: FAIL: misc/tst-atomic and FAIL: misc/tst-atomic-long non-deterministic failures

[nsz at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=25919

            Bug ID: 25919
           Summary: FAIL: misc/tst-atomic and FAIL: misc/tst-atomic-long
                    non-deterministic failures
           Product: glibc
           Version: unspecified
            Status: NEW
          Severity: normal
          Priority: P2
         Component: build
          Assignee: unassigned at sourceware dot org
          Reporter: nsz at gcc dot gnu.org
                CC: carlos at redhat dot com
  Target Milestone: ---

misc/tst-atomic and misc/tst-atomic-long sometimes fails with

atomic_compare_exchange_weak_relaxed test 1 failed
atomic_compare_exchange_weak_relaxed test 2 failed

because the cas operation is not retried on failure:

  mem = 14;
  expected = 14;
  if (!atomic_compare_exchange_weak_relaxed (&mem, &expected, 25)
      || mem != 25 || expected != 14)
    {
      puts ("atomic_compare_exchange_weak_relaxed test 1 failed");
      ret = 1;
    }

on systems with ll/sc atomics the exclusive store may
fail on scheduling or other os events, success is not
guaranteed even if the memory location is otherwise
unmodified.

(i've seen this in qemu-system-aarch64 with -smp 8 and
running cross test with make check -j8 test-wrapper=...)

i think this is a glibc test bug (but i don't know the
expected semantics of atomic_compare_exchange_weak_relaxed)

-- 
You are receiving this mail because:
You are on the CC list for the bug.