Peter Xu <pet...@redhat.com> wrote: > On Fri, Jun 09, 2023 at 12:49:03AM +0200, Juan Quintela wrote: >> Signed-off-by: Juan Quintela <quint...@redhat.com> >> --- >> tests/qtest/migration-test.c | 10 ++++------ >> 1 file changed, 4 insertions(+), 6 deletions(-) >> >> diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c >> index c5e0c69c6b..73b2f01427 100644 >> --- a/tests/qtest/migration-test.c >> +++ b/tests/qtest/migration-test.c >> @@ -602,7 +602,7 @@ static int test_migrate_start(QTestState **from, >> QTestState **to, >> g_autofree gchar *arch_target = NULL; >> g_autofree gchar *cmd_source = NULL; >> g_autofree gchar *cmd_target = NULL; >> - const gchar *ignore_stderr; >> + const gchar *ignore_stderr = NULL; >> g_autofree char *bootpath = NULL; >> g_autofree char *shmem_opts = NULL; >> g_autofree char *shmem_path = NULL; >> @@ -672,10 +672,7 @@ static int test_migrate_start(QTestState **from, >> QTestState **to, >> * IO redirection does not work, so don't bother adding IO >> redirection >> * to the command line. >> */ >> - ignore_stderr = ""; > > Personally maybe I won't bother with such a patch as a whole..
I am trying that all the options are handled the same. The whole idea here is that I have to split guest definition and guest creation. See GuestState later on the series. If all the options are handled the same, it is much easier than if I have to do "magic" for one vs others. One thing that I wonder is that migration-test can't be the only test that needs to redirect stderr, but that is a completely different issue. > but if we do > want to, we could also remove "#else" here and move comment above it. Done.
