* [PATCH] dma-buf: avoid scheduling on fence status query
@ 2017-04-26 1:36 Andres Rodriguez
2017-04-26 2:50 ` Andres Rodriguez
0 siblings, 1 reply; 8+ messages in thread
From: Andres Rodriguez @ 2017-04-26 1:36 UTC (permalink / raw)
To: dri-devel; +Cc: andresx7
When a timeout of zero is specified, the caller is only interested in
the fence status.
In the current implementation, dma_fence_default_wait will always call
schedule_timeout() at least once for an unsignaled fence. This adds a
significant overhead to a fence status query.
Avoid this overhead by returning early if a zero timeout is specified.
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
---
This heavily affects the performance of the Source2 engine running on
radv.
This patch improves dota2(radv) perf on a i7-6700k+RX480 system from
72fps->81fps.
drivers/dma-buf/dma-fence.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c
index 0918d3f..348e9e2 100644
--- a/drivers/dma-buf/dma-fence.c
+++ b/drivers/dma-buf/dma-fence.c
@@ -380,6 +380,9 @@ dma_fence_default_wait(struct dma_fence *fence, bool intr, signed long timeout)
if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
return ret;
+ if (!timeout)
+ return 0;
+
spin_lock_irqsave(fence->lock, flags);
if (intr && signal_pending(current)) {
--
2.9.3
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH] dma-buf: avoid scheduling on fence status query
2017-04-26 1:36 [PATCH] dma-buf: avoid scheduling on fence status query Andres Rodriguez
@ 2017-04-26 2:50 ` Andres Rodriguez
2017-04-26 7:20 ` Christian König
0 siblings, 1 reply; 8+ messages in thread
From: Andres Rodriguez @ 2017-04-26 2:50 UTC (permalink / raw)
To: dri-devel; +Cc: deathsimple, linaro-mm-sig, sumit.semwal, linux-media
CC a few extra lists I missed.
Regards,
Andres
On 2017-04-25 09:36 PM, Andres Rodriguez wrote:
> When a timeout of zero is specified, the caller is only interested in
> the fence status.
>
> In the current implementation, dma_fence_default_wait will always call
> schedule_timeout() at least once for an unsignaled fence. This adds a
> significant overhead to a fence status query.
>
> Avoid this overhead by returning early if a zero timeout is specified.
>
> Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
> ---
>
> This heavily affects the performance of the Source2 engine running on
> radv.
>
> This patch improves dota2(radv) perf on a i7-6700k+RX480 system from
> 72fps->81fps.
>
> drivers/dma-buf/dma-fence.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c
> index 0918d3f..348e9e2 100644
> --- a/drivers/dma-buf/dma-fence.c
> +++ b/drivers/dma-buf/dma-fence.c
> @@ -380,6 +380,9 @@ dma_fence_default_wait(struct dma_fence *fence, bool intr, signed long timeout)
> if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
> return ret;
>
> + if (!timeout)
> + return 0;
> +
> spin_lock_irqsave(fence->lock, flags);
>
> if (intr && signal_pending(current)) {
>
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] dma-buf: avoid scheduling on fence status query
2017-04-26 2:50 ` Andres Rodriguez
@ 2017-04-26 7:20 ` Christian König
2017-04-26 9:59 ` Dave Airlie
0 siblings, 1 reply; 8+ messages in thread
From: Christian König @ 2017-04-26 7:20 UTC (permalink / raw)
To: Andres Rodriguez, dri-devel; +Cc: linaro-mm-sig, sumit.semwal, linux-media
NAK, I'm wondering how often I have to reject that change. We should
probably add a comment here.
Even with a zero timeout we still need to enable signaling, otherwise
some fence will never signal if userspace just polls on them.
If a caller is only interested in the fence status without enabling the
signaling it should call dma_fence_is_signaled() instead.
Regards,
Christian.
Am 26.04.2017 um 04:50 schrieb Andres Rodriguez:
> CC a few extra lists I missed.
>
> Regards,
> Andres
>
> On 2017-04-25 09:36 PM, Andres Rodriguez wrote:
>> When a timeout of zero is specified, the caller is only interested in
>> the fence status.
>>
>> In the current implementation, dma_fence_default_wait will always call
>> schedule_timeout() at least once for an unsignaled fence. This adds a
>> significant overhead to a fence status query.
>>
>> Avoid this overhead by returning early if a zero timeout is specified.
>>
>> Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
>> ---
>>
>> This heavily affects the performance of the Source2 engine running on
>> radv.
>>
>> This patch improves dota2(radv) perf on a i7-6700k+RX480 system from
>> 72fps->81fps.
>>
>> drivers/dma-buf/dma-fence.c | 3 +++
>> 1 file changed, 3 insertions(+)
>>
>> diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c
>> index 0918d3f..348e9e2 100644
>> --- a/drivers/dma-buf/dma-fence.c
>> +++ b/drivers/dma-buf/dma-fence.c
>> @@ -380,6 +380,9 @@ dma_fence_default_wait(struct dma_fence *fence,
>> bool intr, signed long timeout)
>> if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
>> return ret;
>>
>> + if (!timeout)
>> + return 0;
>> +
>> spin_lock_irqsave(fence->lock, flags);
>>
>> if (intr && signal_pending(current)) {
>>
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] dma-buf: avoid scheduling on fence status query
2017-04-26 7:20 ` Christian König
@ 2017-04-26 9:59 ` Dave Airlie
0 siblings, 0 replies; 8+ messages in thread
From: Dave Airlie @ 2017-04-26 9:59 UTC (permalink / raw)
To: Christian König
Cc: Andres Rodriguez, dri-devel, linaro-mm-sig, Linux Media Mailing List
On 26 April 2017 at 17:20, Christian König <deathsimple@vodafone.de> wrote:
> NAK, I'm wondering how often I have to reject that change. We should
> probably add a comment here.
>
> Even with a zero timeout we still need to enable signaling, otherwise some
> fence will never signal if userspace just polls on them.
>
> If a caller is only interested in the fence status without enabling the
> signaling it should call dma_fence_is_signaled() instead.
Can we not move the return 0 (with spin unlock) down after we enabling
signalling, but before
we enter the schedule_timeout(1)?
Dave.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] dma-buf: avoid scheduling on fence status query
@ 2017-04-26 9:59 ` Dave Airlie
0 siblings, 0 replies; 8+ messages in thread
From: Dave Airlie @ 2017-04-26 9:59 UTC (permalink / raw)
To: Christian König
Cc: linaro-mm-sig, Andres Rodriguez, dri-devel, Linux Media Mailing List
On 26 April 2017 at 17:20, Christian König <deathsimple@vodafone.de> wrote:
> NAK, I'm wondering how often I have to reject that change. We should
> probably add a comment here.
>
> Even with a zero timeout we still need to enable signaling, otherwise some
> fence will never signal if userspace just polls on them.
>
> If a caller is only interested in the fence status without enabling the
> signaling it should call dma_fence_is_signaled() instead.
Can we not move the return 0 (with spin unlock) down after we enabling
signalling, but before
we enter the schedule_timeout(1)?
Dave.
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] dma-buf: avoid scheduling on fence status query
2017-04-26 9:59 ` Dave Airlie
@ 2017-04-26 10:13 ` Christian König
-1 siblings, 0 replies; 8+ messages in thread
From: Christian König @ 2017-04-26 10:13 UTC (permalink / raw)
To: Dave Airlie
Cc: Andres Rodriguez, dri-devel, linaro-mm-sig, Linux Media Mailing List
Am 26.04.2017 um 11:59 schrieb Dave Airlie:
> On 26 April 2017 at 17:20, Christian König <deathsimple@vodafone.de> wrote:
>> NAK, I'm wondering how often I have to reject that change. We should
>> probably add a comment here.
>>
>> Even with a zero timeout we still need to enable signaling, otherwise some
>> fence will never signal if userspace just polls on them.
>>
>> If a caller is only interested in the fence status without enabling the
>> signaling it should call dma_fence_is_signaled() instead.
> Can we not move the return 0 (with spin unlock) down after we enabling
> signalling, but before
> we enter the schedule_timeout(1)?
Yes, that would be an option.
Christian.
>
> Dave.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] dma-buf: avoid scheduling on fence status query
@ 2017-04-26 10:13 ` Christian König
0 siblings, 0 replies; 8+ messages in thread
From: Christian König @ 2017-04-26 10:13 UTC (permalink / raw)
To: Dave Airlie
Cc: linaro-mm-sig, Andres Rodriguez, dri-devel, Linux Media Mailing List
Am 26.04.2017 um 11:59 schrieb Dave Airlie:
> On 26 April 2017 at 17:20, Christian König <deathsimple@vodafone.de> wrote:
>> NAK, I'm wondering how often I have to reject that change. We should
>> probably add a comment here.
>>
>> Even with a zero timeout we still need to enable signaling, otherwise some
>> fence will never signal if userspace just polls on them.
>>
>> If a caller is only interested in the fence status without enabling the
>> signaling it should call dma_fence_is_signaled() instead.
> Can we not move the return 0 (with spin unlock) down after we enabling
> signalling, but before
> we enter the schedule_timeout(1)?
Yes, that would be an option.
Christian.
>
> Dave.
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] dma-buf: avoid scheduling on fence status query
2017-04-26 10:13 ` Christian König
(?)
@ 2017-04-26 14:28 ` Andres Rodriguez
-1 siblings, 0 replies; 8+ messages in thread
From: Andres Rodriguez @ 2017-04-26 14:28 UTC (permalink / raw)
To: Christian König, Dave Airlie
Cc: dri-devel, linaro-mm-sig, Linux Media Mailing List
On 2017-04-26 06:13 AM, Christian König wrote:
> Am 26.04.2017 um 11:59 schrieb Dave Airlie:
>> On 26 April 2017 at 17:20, Christian König <deathsimple@vodafone.de>
>> wrote:
>>> NAK, I'm wondering how often I have to reject that change. We should
>>> probably add a comment here.
>>>
>>> Even with a zero timeout we still need to enable signaling, otherwise
>>> some
>>> fence will never signal if userspace just polls on them.
>>>
>>> If a caller is only interested in the fence status without enabling the
>>> signaling it should call dma_fence_is_signaled() instead.
>> Can we not move the return 0 (with spin unlock) down after we enabling
>> signalling, but before
>> we enter the schedule_timeout(1)?
>
> Yes, that would be an option.
>
I was actually arguing with Dave about this on IRC yesterday. Seems like
I owe him a beer now.
-Andres
> Christian.
>
>>
>> Dave.
>
>
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2017-04-26 14:28 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-04-26 1:36 [PATCH] dma-buf: avoid scheduling on fence status query Andres Rodriguez
2017-04-26 2:50 ` Andres Rodriguez
2017-04-26 7:20 ` Christian König
2017-04-26 9:59 ` Dave Airlie
2017-04-26 9:59 ` Dave Airlie
2017-04-26 10:13 ` Christian König
2017-04-26 10:13 ` Christian König
2017-04-26 14:28 ` Andres Rodriguez
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.