Re: [Intel-gfx] [igt-dev] [PATCH i-g-t v2 2/2] tests/gem_eio: Add reset and unwedge stress testing

[Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>]

On 03/04/2018 19:34, Antonio Argenziano wrote:
> 
> 
> On 03/04/18 11:24, Antonio Argenziano wrote:
>>
>>
>> On 03/04/18 04:36, Tvrtko Ursulin wrote:
>>> From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>>
>>> Reset and unwedge stress testing is supposed to trigger wedging or 
>>> resets
>>> at incovenient times and then re-use the context so either the 
>>> context or
>>> driver tracking might get confused and break.
>>>
>>> v2:
>>>   * Renamed for more sensible naming.
>>>   * Added some comments to explain what the test is doing. (Chris 
>>> Wilson)
>>>
>>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>>> ---
>>>   tests/gem_eio.c | 74 
>>> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>>   1 file changed, 74 insertions(+)
>>>
>>> diff --git a/tests/gem_eio.c b/tests/gem_eio.c
>>> index b7c5047f0816..9599e73db736 100644
>>> --- a/tests/gem_eio.c
>>> +++ b/tests/gem_eio.c
>>> @@ -591,6 +591,74 @@ static void test_inflight_internal(int fd, 
>>> unsigned int wait)
>>>       close(fd);
>>>   }
>>> +/*
>>> + * Verify that we can submit and execute work after unwedging the GPU.
>>> + */
>>> +static void test_reset_stress(int fd, unsigned int flags)
>>> +{
>>> +    uint32_t ctx0 = gem_context_create(fd);
>>> +
>>> +    igt_until_timeout(5) {
>>> +        struct drm_i915_gem_execbuffer2 execbuf = { };
>>> +        struct drm_i915_gem_exec_object2 obj = { };
>>> +        uint32_t bbe = MI_BATCH_BUFFER_END;
>>> +        igt_spin_t *hang;
>>> +        unsigned int i;
>>> +        uint32_t ctx;
>>> +
>>> +        gem_quiescent_gpu(fd);
>>> +
>>> +        igt_require(i915_reset_control(flags & TEST_WEDGE ?
>>> +                           false : true));
>>> +
>>> +        ctx = context_create_safe(fd);
>>> +
>>> +        /*
>>> +         * Start executing a spin batch with some queued batches
>>> +         * against a different context after it.
>>> +         */
>>
>> Aren't all batches queued on ctx0? Or is this a reference to the check 
>> on ctx you have later in the test.

Yes, a mistake in comment text.

>>> +        hang = spin_sync(fd, ctx0, 0);
> 
> I think you meant to send this^ on ctx.

Why do you think so? Did you find a different or better way to trigger 
the bug this test is trying to hit?

Regards,

Tvrtko

> Antonio.
> 
>>> +
>>> +        obj.handle = gem_create(fd, 4096);
>>> +        gem_write(fd, obj.handle, 0, &bbe, sizeof(bbe));
>>> +
>>> +        execbuf.buffers_ptr = to_user_pointer(&obj);
>>> +        execbuf.buffer_count = 1;
>>> +        execbuf.rsvd1 = ctx0;
>>> +
>>> +        for (i = 0; i < 10; i++)
>>> +            gem_execbuf(fd, &execbuf);
>>> +
>>> +        /* Wedge after a small delay. */
>>> +        igt_assert_eq(__check_wait(fd, obj.handle, 100e3), 0);
>>> +
>>> +        /* Unwedge by forcing a reset. */
>>> +        igt_assert(i915_reset_control(true));
>>> +        trigger_reset(fd);
>>> +
>>> +        gem_quiescent_gpu(fd);
>>> +
>>> +        /*
>>> +         * Verify that we are able to submit work after unwedging from
>>> +         * both contexts.
>>> +         */
>>> +        execbuf.rsvd1 = ctx;
>>> +        for (i = 0; i < 5; i++)
>>> +            gem_execbuf(fd, &execbuf);
>>> +
>>> +        execbuf.rsvd1 = ctx0;
>>> +        for (i = 0; i < 5; i++)
>>> +            gem_execbuf(fd, &execbuf);
>>> +
>>> +        gem_sync(fd, obj.handle);
>>> +        igt_spin_batch_free(fd, hang);
>>> +        gem_context_destroy(fd, ctx);
>>> +        gem_close(fd, obj.handle);
>>> +    }
>>> +
>>> +    gem_context_destroy(fd, ctx0);
>>> +}
>>> +
>>>   static int fd = -1;
>>>   static void
>>> @@ -635,6 +703,12 @@ igt_main
>>>       igt_subtest("in-flight-suspend")
>>>           test_inflight_suspend(fd);
>>> +    igt_subtest("reset-stress")
>>> +        test_reset_stress(fd, 0);
>>> +
>>> +    igt_subtest("unwedge-stress")
>>> +        test_reset_stress(fd, TEST_WEDGE);
>>> +
>>>       igt_subtest_group {
>>>           const struct {
>>>               unsigned int wait;
>>>
>> _______________________________________________
>> igt-dev mailing list
>> igt-dev@lists.freedesktop.org
>> https://lists.freedesktop.org/mailman/listinfo/igt-dev
> _______________________________________________
> igt-dev mailing list
> igt-dev@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/igt-dev
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx