test regression - qemu:block-slow+slow / io-raw-055

Posted by Pierrick Bouvier 2 weeks, 4 days ago

Hi,

this test was recently broken by 34a889 (migration: Drop 
migration_is_idle()).

Reproduce with:
meson test -C build -t 1 --setup slow  --num-processes 1 
--print-errorlogs io-raw-055 --verbose

1/1 qemu:block-slow+slow / io-raw-055        RUNNING
 >>> ASAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1 
UBSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1 
MALLOC_PERTURB_=29 MESON_TEST_ITERATION=1 
PYTHON=/home/user/.work/qemu/build/pyvenv/bin/python3 G_TEST_SLOW=1 
SPEED=slow 
MSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1 
/home/user/.work/qemu/build/pyvenv/bin/python3 
/home/user/.work/qemu/build/../tests/qemu-iotests/check -tap -raw 055 
--source-dir /home/user/.work/qemu/tests/qemu-iotests --build-dir 
/home/user/.work/qemu/build/tests/qemu-iotests
▶ 1/1 raw 055                                FAIL
1/1 qemu:block-slow+slow / io-raw-055        ERROR           11.06s 
exit status 1
――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――― 
✀ 
―――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――
stderr:
--- /home/user/.work/qemu/tests/qemu-iotests/055.out
+++ /home/user/.work/qemu/build/scratch/raw-file-055/055.out.bad
@@ -1,5 +1,215 @@
-........................................
+......ERROR:qemu.qmp.qmp_client.qemu-1856388:Failed to receive 
Greeting: EOFError
+ERROR:qemu.qmp.qmp_client.qemu-1856388:Failed to establish session: 
EOFError

---

Regards,
Pierrick

Re: test regression - qemu:block-slow+slow / io-raw-055

Posted by Peter Xu 2 weeks, 4 days ago

On Mon, Nov 04, 2024 at 07:41:44PM -0800, Pierrick Bouvier wrote:
> Hi,
> 
> this test was recently broken by 34a889 (migration: Drop
> migration_is_idle()).
> 
> Reproduce with:
> meson test -C build -t 1 --setup slow  --num-processes 1 --print-errorlogs
> io-raw-055 --verbose
> 
> 1/1 qemu:block-slow+slow / io-raw-055        RUNNING
> >>> ASAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1 UBSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1
> MALLOC_PERTURB_=29 MESON_TEST_ITERATION=1
> PYTHON=/home/user/.work/qemu/build/pyvenv/bin/python3 G_TEST_SLOW=1
> SPEED=slow MSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1
> /home/user/.work/qemu/build/pyvenv/bin/python3
> /home/user/.work/qemu/build/../tests/qemu-iotests/check -tap -raw 055
> --source-dir /home/user/.work/qemu/tests/qemu-iotests --build-dir
> /home/user/.work/qemu/build/tests/qemu-iotests
> ▶ 1/1 raw 055                                FAIL
> 1/1 qemu:block-slow+slow / io-raw-055        ERROR           11.06s exit
> status 1
> ―――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――
> ✀ ―――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――
> stderr:
> --- /home/user/.work/qemu/tests/qemu-iotests/055.out
> +++ /home/user/.work/qemu/build/scratch/raw-file-055/055.out.bad
> @@ -1,5 +1,215 @@
> -........................................
> +......ERROR:qemu.qmp.qmp_client.qemu-1856388:Failed to receive Greeting:
> EOFError
> +ERROR:qemu.qmp.qmp_client.qemu-1856388:Failed to establish session:
> EOFError

Thanks for the report.

I don't know how this was triggered, looks like current_migration wasn't
set at all for some reason..

Could you try below change to see whether it can fix there too (it worked
here)?

===8<===
From d33ad33854d36c04260150ee817b984f48da46c6 Mon Sep 17 00:00:00 2001
From: Peter Xu <peterx@redhat.com>
Date: Tue, 5 Nov 2024 11:29:07 -0500
Subject: [PATCH] migration: Check current_migration in migration_is_running()

Report shows that commit 34a8892dec broke iotest 055:

https://lore.kernel.org/r/b8806360-a2b6-4608-83a3-db67e264c733@linaro.org

When replacing migration_is_idle() with "!migration_is_running()", it was
overlooked that the idle helper also checks for current_migration being
available first.

The check would be there if the whole series was applied, but since the
last patches in the previous series rely on some other patches to land
first, we need to recover the behavior of migration_is_idle() first before
that whole set will be merged.

I left migration_is_active / migration_is_device alone, as I don't think
it's possible for them to hit his case (current_migration not initialized).
Also they're prone to removal soon from VFIO side.

Cc: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Cc: Fabiano Rosas <farosas@suse.de>
Fixes: 34a8892dec ("migration: Drop migration_is_idle()")
Signed-off-by: Peter Xu <peterx@redhat.com>
---
 migration/migration.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/migration/migration.c b/migration/migration.c
index aedf7f0751..8c5bd0a75c 100644
--- a/migration/migration.c
+++ b/migration/migration.c
@@ -1117,6 +1117,10 @@ bool migration_is_running(void)
 {
     MigrationState *s = current_migration;
 
+    if (!s) {
+        return false;
+    }
+
     switch (s->state) {
     case MIGRATION_STATUS_ACTIVE:
     case MIGRATION_STATUS_POSTCOPY_ACTIVE:
-- 
2.45.0


-- 
Peter Xu

Re: test regression - qemu:block-slow+slow / io-raw-055

Posted by Pierrick Bouvier 2 weeks, 3 days ago

On 11/5/24 08:41, Peter Xu wrote:
> On Mon, Nov 04, 2024 at 07:41:44PM -0800, Pierrick Bouvier wrote:
>> Hi,
>>
>> this test was recently broken by 34a889 (migration: Drop
>> migration_is_idle()).
>>
>> Reproduce with:
>> meson test -C build -t 1 --setup slow  --num-processes 1 --print-errorlogs
>> io-raw-055 --verbose
>>
>> 1/1 qemu:block-slow+slow / io-raw-055        RUNNING
>>>>> ASAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1 UBSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1
>> MALLOC_PERTURB_=29 MESON_TEST_ITERATION=1
>> PYTHON=/home/user/.work/qemu/build/pyvenv/bin/python3 G_TEST_SLOW=1
>> SPEED=slow MSAN_OPTIONS=halt_on_error=1:abort_on_error=1:print_summary=1:print_stacktrace=1
>> /home/user/.work/qemu/build/pyvenv/bin/python3
>> /home/user/.work/qemu/build/../tests/qemu-iotests/check -tap -raw 055
>> --source-dir /home/user/.work/qemu/tests/qemu-iotests --build-dir
>> /home/user/.work/qemu/build/tests/qemu-iotests
>> ▶ 1/1 raw 055                                FAIL
>> 1/1 qemu:block-slow+slow / io-raw-055        ERROR           11.06s exit
>> status 1
>> ―――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――
>> ✀ ―――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――――
>> stderr:
>> --- /home/user/.work/qemu/tests/qemu-iotests/055.out
>> +++ /home/user/.work/qemu/build/scratch/raw-file-055/055.out.bad
>> @@ -1,5 +1,215 @@
>> -........................................
>> +......ERROR:qemu.qmp.qmp_client.qemu-1856388:Failed to receive Greeting:
>> EOFError
>> +ERROR:qemu.qmp.qmp_client.qemu-1856388:Failed to establish session:
>> EOFError
> 
> Thanks for the report.
> 
> I don't know how this was triggered, looks like current_migration wasn't
> set at all for some reason..
> 
> Could you try below change to see whether it can fix there too (it worked
> here)?
> 
> ===8<===
>  From d33ad33854d36c04260150ee817b984f48da46c6 Mon Sep 17 00:00:00 2001
> From: Peter Xu <peterx@redhat.com>
> Date: Tue, 5 Nov 2024 11:29:07 -0500
> Subject: [PATCH] migration: Check current_migration in migration_is_running()
> 
> Report shows that commit 34a8892dec broke iotest 055:
> 
> https://lore.kernel.org/r/b8806360-a2b6-4608-83a3-db67e264c733@linaro.org
> 
> When replacing migration_is_idle() with "!migration_is_running()", it was
> overlooked that the idle helper also checks for current_migration being
> available first.
> 
> The check would be there if the whole series was applied, but since the
> last patches in the previous series rely on some other patches to land
> first, we need to recover the behavior of migration_is_idle() first before
> that whole set will be merged.
> 
> I left migration_is_active / migration_is_device alone, as I don't think
> it's possible for them to hit his case (current_migration not initialized).
> Also they're prone to removal soon from VFIO side.
> 
> Cc: Pierrick Bouvier <pierrick.bouvier@linaro.org>
> Cc: Fabiano Rosas <farosas@suse.de>
> Fixes: 34a8892dec ("migration: Drop migration_is_idle()")
> Signed-off-by: Peter Xu <peterx@redhat.com>
> ---
>   migration/migration.c | 4 ++++
>   1 file changed, 4 insertions(+)
> 
> diff --git a/migration/migration.c b/migration/migration.c
> index aedf7f0751..8c5bd0a75c 100644
> --- a/migration/migration.c
> +++ b/migration/migration.c
> @@ -1117,6 +1117,10 @@ bool migration_is_running(void)
>   {
>       MigrationState *s = current_migration;
>   
> +    if (!s) {
> +        return false;
> +    }
> +
>       switch (s->state) {
>       case MIGRATION_STATUS_ACTIVE:
>       case MIGRATION_STATUS_POSTCOPY_ACTIVE:

Hi Peter,
I confirm this fix the test mentioned. You can post this patch.

Regards,
Pierrick