Re: leaks in migration-test

[Peter Xu <peterx@redhat.com>]

On Mon, Mar 09, 2026 at 09:06:45AM -0300, Fabiano Rosas wrote:
> Peter Maydell <peter.maydell@linaro.org> writes:
> 
> > I'm continuing to work through leaks reported during "make check".
> > Some of them come from the migration-test. (For these to actually
> > appear in the log you need to run with QTEST_LOG=1, otherwise the
> > migration test framework will send stderr to /dev/null and the
> > tests will fail without printing the leak report.)
> 
> I'm pretty sure something changed that caused this. In the past the
> leaks would be reported by the parent migration-test binary, so it
> wouldn't matter where QEMU sends stderr. I don't know if it's QEMU's
> build environment or something on my machine.
> 
> I'm also seeing in dmesg after migration-test aborts, which didn't use
> to be the case:
> Process 23732(migration-test) has RLIMIT_CORE set to 1
> 
> Anyway, the redirect to /dev/null is a pain for other reasons as well,
> we discussed adding an output filter step to qtest like the iotests
> have, but I couldn't find the time to work on it yet.

Just to mention another thing.. looks like asan supports this:

  export ASAN_OPTIONS="detect_leaks=1:log_path=$SOME_PATH"

Then it can dump leak reports somewhere else than stderr.

-- 
Peter Xu