Re: [PATCH v3] drm/xe/ufence: Flush xe ordered_wq in case of ufence timeout

[Nirmoy Das <nirmoy.das@intel.com>]

On 10/28/2024 6:03 PM, John Harrison wrote:
> On 10/28/2024 04:49, Nirmoy Das wrote:
>> Flush xe ordered_wq in case of ufence timeout which is observed
>> on LNL and that points to recent scheduling issue with E-cores.
>>
>> This is similar to the recent fix:
>> commit e51527233804 ("drm/xe/guc/ct: Flush g2h worker in case of g2h
>> response timeout") and should be removed once there is a E-core
>> scheduling fix for LNL.
>>
>> v2: Add platform check(Himal)
>>      s/__flush_workqueue/flush_workqueue(Jani)
>> v3: Remove gfx platform check as the issue related to cpu
>>      platform(John)
>>
>> Cc: Badal Nilawar <badal.nilawar@intel.com>
>> Cc: Matthew Auld <matthew.auld@intel.com>
>> Cc: John Harrison <John.C.Harrison@Intel.com>
>> Cc: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com>
>> Cc: Lucas De Marchi <lucas.demarchi@intel.com>
>> Cc: <stable@vger.kernel.org> # v6.11+
>> Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/2754
>> Suggested-by: Matthew Brost <matthew.brost@intel.com>
>> Signed-off-by: Nirmoy Das <nirmoy.das@intel.com>
>> Reviewed-by: Matthew Brost <matthew.brost@intel.com>
>> ---
>>   drivers/gpu/drm/xe/xe_wait_user_fence.c | 11 +++++++++++
>>   1 file changed, 11 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/xe/xe_wait_user_fence.c b/drivers/gpu/drm/xe/xe_wait_user_fence.c
>> index f5deb81eba01..886c9862d89c 100644
>> --- a/drivers/gpu/drm/xe/xe_wait_user_fence.c
>> +++ b/drivers/gpu/drm/xe/xe_wait_user_fence.c
>> @@ -155,6 +155,17 @@ int xe_wait_user_fence_ioctl(struct drm_device *dev, void *data,
>>           }
>>             if (!timeout) {
>> +            /*
>> +             * This is analogous to e51527233804 ("drm/xe/guc/ct: Flush g2h worker
>> +             * in case of g2h response timeout")
>> +             *
>> +             * TODO: Drop this change once workqueue scheduling delay issue is
>> +             * fixed on LNL Hybrid CPU.
>> +             */
>> +            flush_workqueue(xe->ordered_wq);
> I thought the plan was to make this a trackable macro used by all instances of this w/a - LNL_FLUSH_WORK|WORKQUEUE()? With a single, complete description of the problem attached to the macro rather than 'this is similar to' comments scattered through the code.
>
> There was also a request to add a dmesg print if the failing condition was met after doing the flush.
>

I will resend. I misunderstood the last conversation.

> John.
>
>> +            err = do_compare(addr, args->value, args->mask, args->op);
>> +            if (err <= 0)
>> +                break;
>>               err = -ETIME;
>>               break;
>>           }
>