ppc64 with TCG seems to no longer be failing this test. Let's try to
enable it. s390x is still hanging about 1 in 10 runs.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 tests/qtest/migration-test.c | 16 +++-------------
 1 file changed, 3 insertions(+), 13 deletions(-)

diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
index c13535c37d..b8617cc843 100644
--- a/tests/qtest/migration-test.c
+++ b/tests/qtest/migration-test.c
@@ -3454,19 +3454,9 @@ int main(int argc, char **argv)
 #endif
 
     /*
-     * On ppc64, the test only works with kvm-hv, but not with kvm-pr and TCG
-     * is touchy due to race conditions on dirty bits (especially on PPC for
-     * some reason)
-     */
-    if (g_str_equal(arch, "ppc64") &&
-        (!has_kvm || access("/sys/module/kvm_hv", F_OK))) {
-        g_test_message("Skipping tests: kvm_hv not available");
-        goto test_add_done;
-    }
-
-    /*
-     * Similar to ppc64, s390x seems to be touchy with TCG, so disable it
-     * there until the problems are resolved
+     * On s390x, the test seems to be touchy with TCG, perhaps due to race
+     * conditions on dirty bits, so disable it there until the problems are
+     * resolved.
      */
     if (g_str_equal(arch, "s390x") && !has_kvm) {
         g_test_message("Skipping tests: s390x host with KVM is required");
-- 
2.43.0

On 25/05/2024 05.13, Nicholas Piggin wrote:
> ppc64 with TCG seems to no longer be failing this test. Let's try to
> enable it. s390x is still hanging about 1 in 10 runs.
> 
> Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
> ---
>   tests/qtest/migration-test.c | 16 +++-------------
>   1 file changed, 3 insertions(+), 13 deletions(-)
> 
> diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
> index c13535c37d..b8617cc843 100644
> --- a/tests/qtest/migration-test.c
> +++ b/tests/qtest/migration-test.c
> @@ -3454,19 +3454,9 @@ int main(int argc, char **argv)
>   #endif
>   
>       /*
> -     * On ppc64, the test only works with kvm-hv, but not with kvm-pr and TCG
> -     * is touchy due to race conditions on dirty bits (especially on PPC for
> -     * some reason)
> -     */
> -    if (g_str_equal(arch, "ppc64") &&
> -        (!has_kvm || access("/sys/module/kvm_hv", F_OK))) {
> -        g_test_message("Skipping tests: kvm_hv not available");
> -        goto test_add_done;
> -    }
> -
> -    /*
> -     * Similar to ppc64, s390x seems to be touchy with TCG, so disable it
> -     * there until the problems are resolved
> +     * On s390x, the test seems to be touchy with TCG, perhaps due to race
> +     * conditions on dirty bits, so disable it there until the problems are
> +     * resolved.
>        */
>       if (g_str_equal(arch, "s390x") && !has_kvm) {
>           g_test_message("Skipping tests: s390x host with KVM is required");

Since you've identified the problem on s390x, you could maybe adjust the 
comment in case you respin ... OTOH, it will get removed anyway once we 
merge the s390x fix, so no need to respin just because of this.

Reviewed-by: Thomas Huth <thuth@redhat.com>

On Mon May 27, 2024 at 5:35 PM AEST, Thomas Huth wrote:
> On 25/05/2024 05.13, Nicholas Piggin wrote:
> > ppc64 with TCG seems to no longer be failing this test. Let's try to
> > enable it. s390x is still hanging about 1 in 10 runs.
> > 
> > Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
> > ---
> >   tests/qtest/migration-test.c | 16 +++-------------
> >   1 file changed, 3 insertions(+), 13 deletions(-)
> > 
> > diff --git a/tests/qtest/migration-test.c b/tests/qtest/migration-test.c
> > index c13535c37d..b8617cc843 100644
> > --- a/tests/qtest/migration-test.c
> > +++ b/tests/qtest/migration-test.c
> > @@ -3454,19 +3454,9 @@ int main(int argc, char **argv)
> >   #endif
> >   
> >       /*
> > -     * On ppc64, the test only works with kvm-hv, but not with kvm-pr and TCG
> > -     * is touchy due to race conditions on dirty bits (especially on PPC for
> > -     * some reason)
> > -     */
> > -    if (g_str_equal(arch, "ppc64") &&
> > -        (!has_kvm || access("/sys/module/kvm_hv", F_OK))) {
> > -        g_test_message("Skipping tests: kvm_hv not available");
> > -        goto test_add_done;
> > -    }
> > -
> > -    /*
> > -     * Similar to ppc64, s390x seems to be touchy with TCG, so disable it
> > -     * there until the problems are resolved
> > +     * On s390x, the test seems to be touchy with TCG, perhaps due to race
> > +     * conditions on dirty bits, so disable it there until the problems are
> > +     * resolved.
> >        */
> >       if (g_str_equal(arch, "s390x") && !has_kvm) {
> >           g_test_message("Skipping tests: s390x host with KVM is required");
>
> Since you've identified the problem on s390x, you could maybe adjust the 
> comment in case you respin ... OTOH, it will get removed anyway once we 
> merge the s390x fix, so no need to respin just because of this.
>
> Reviewed-by: Thomas Huth <thuth@redhat.com>

Thanks. I will respin to at least fix the subjet "Enable on ppc64 TCG"
so I can do the other minor things too.

Thanks,
Nick