* [PATCH] drm/amdgpu: treat negative lockup timeout as 'infinite timeout'
@ 2019-05-05 14:23 Evan Quan
[not found] ` <20190505142353.30369-1-evan.quan-5C7GfCeVMHo@public.gmane.org>
0 siblings, 1 reply; 3+ messages in thread
From: Evan Quan @ 2019-05-05 14:23 UTC (permalink / raw)
To: amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW
Cc: Evan Quan, christian.koenig-5C7GfCeVMHo
Negative lockup timeout is valid and will be treated as
'infinite timeout'.
Change-Id: I0d8387956a9c744073c0281ef2e1a547d4f16dec
Signed-off-by: Evan Quan <evan.quan@amd.com>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 14 ++++++++++----
1 file changed, 10 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 5b03e17e6e06..4d6dff6855f8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -233,13 +233,14 @@ module_param_named(msi, amdgpu_msi, int, 0444);
* Set GPU scheduler timeout value in ms.
*
* The format can be [Non-Compute] or [GFX,Compute,SDMA,Video]. That is there can be one or
- * multiple values specified. 0 and negative values are invalidated. They will be adjusted
- * to default timeout.
+ * multiple values specified.
* - With one value specified, the setting will apply to all non-compute jobs.
* - With multiple values specified, the first one will be for GFX. The second one is for Compute.
* And the third and fourth ones are for SDMA and Video.
* By default(with no lockup_timeout settings), the timeout for all non-compute(GFX, SDMA and Video)
* jobs is 10000. And there is no timeout enforced on compute jobs.
+ * Value 0 is invalidated, will be adjusted to default timeout settings.
+ * Negative values mean 'infinite timeout' (MAX_JIFFY_OFFSET).
*/
MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms (default: 10000 for non-compute jobs and no timeout for compute jobs), "
"format is [Non-Compute] or [GFX,Compute,SDMA,Video]");
@@ -1248,11 +1249,16 @@ int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev)
if (ret)
return ret;
- /* Invalidate 0 and negative values */
- if (timeout <= 0) {
+ /*
+ * Value 0 will be adjusted to default timeout settings.
+ * Negative values mean 'infinite timeout' (MAX_JIFFY_OFFSET).
+ */
+ if (!timeout) {
index++;
continue;
}
+ if (timeout < 0)
+ timeout = MAX_JIFFY_OFFSET;
switch (index++) {
case 0:
--
2.21.0
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] drm/amdgpu: treat negative lockup timeout as 'infinite timeout'
[not found] ` <20190505142353.30369-1-evan.quan-5C7GfCeVMHo@public.gmane.org>
@ 2019-05-06 11:26 ` Christian König
[not found] ` <03378dca-c0bb-443e-818a-3a01b69dcccb-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
0 siblings, 1 reply; 3+ messages in thread
From: Christian König @ 2019-05-06 11:26 UTC (permalink / raw)
To: Evan Quan, amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW
Cc: christian.koenig-5C7GfCeVMHo
Am 05.05.19 um 16:23 schrieb Evan Quan:
> Negative lockup timeout is valid and will be treated as
> 'infinite timeout'.
>
> Change-Id: I0d8387956a9c744073c0281ef2e1a547d4f16dec
> Signed-off-by: Evan Quan <evan.quan@amd.com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 14 ++++++++++----
> 1 file changed, 10 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> index 5b03e17e6e06..4d6dff6855f8 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> @@ -233,13 +233,14 @@ module_param_named(msi, amdgpu_msi, int, 0444);
> * Set GPU scheduler timeout value in ms.
> *
> * The format can be [Non-Compute] or [GFX,Compute,SDMA,Video]. That is there can be one or
> - * multiple values specified. 0 and negative values are invalidated. They will be adjusted
> - * to default timeout.
> + * multiple values specified.
> * - With one value specified, the setting will apply to all non-compute jobs.
> * - With multiple values specified, the first one will be for GFX. The second one is for Compute.
> * And the third and fourth ones are for SDMA and Video.
> * By default(with no lockup_timeout settings), the timeout for all non-compute(GFX, SDMA and Video)
> * jobs is 10000. And there is no timeout enforced on compute jobs.
> + * Value 0 is invalidated, will be adjusted to default timeout settings.
> + * Negative values mean 'infinite timeout' (MAX_JIFFY_OFFSET).
> */
> MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms (default: 10000 for non-compute jobs and no timeout for compute jobs), "
> "format is [Non-Compute] or [GFX,Compute,SDMA,Video]");
> @@ -1248,11 +1249,16 @@ int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev)
> if (ret)
> return ret;
>
> - /* Invalidate 0 and negative values */
> - if (timeout <= 0) {
> + /*
> + * Value 0 will be adjusted to default timeout settings.
> + * Negative values mean 'infinite timeout' (MAX_JIFFY_OFFSET).
> + */
> + if (!timeout) {
> index++;
> continue;
> }
> + if (timeout < 0)
> + timeout = MAX_JIFFY_OFFSET;
This is superfluous and maybe even harmful, msecs_to_jiffies() should
take care of this conversion.
Maybe even convert the values directly here.
Christian.
>
> switch (index++) {
> case 0:
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply [flat|nested] 3+ messages in thread
* RE: [PATCH] drm/amdgpu: treat negative lockup timeout as 'infinite timeout'
[not found] ` <03378dca-c0bb-443e-818a-3a01b69dcccb-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
@ 2019-05-07 1:48 ` Quan, Evan
0 siblings, 0 replies; 3+ messages in thread
From: Quan, Evan @ 2019-05-07 1:48 UTC (permalink / raw)
To: Koenig, Christian,
amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
Thanks! Just sent out a V2 version with this addressed.
> -----Original Message-----
> From: Christian König <ckoenig.leichtzumerken@gmail.com>
> Sent: 2019年5月6日 19:26
> To: Quan, Evan <Evan.Quan@amd.com>; amd-gfx@lists.freedesktop.org
> Cc: Koenig, Christian <Christian.Koenig@amd.com>
> Subject: Re: [PATCH] drm/amdgpu: treat negative lockup timeout as 'infinite
> timeout'
>
> [CAUTION: External Email]
>
> Am 05.05.19 um 16:23 schrieb Evan Quan:
> > Negative lockup timeout is valid and will be treated as 'infinite
> > timeout'.
> >
> > Change-Id: I0d8387956a9c744073c0281ef2e1a547d4f16dec
> > Signed-off-by: Evan Quan <evan.quan@amd.com>
> > ---
> > drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 14 ++++++++++----
> > 1 file changed, 10 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> > b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> > index 5b03e17e6e06..4d6dff6855f8 100644
> > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
> > @@ -233,13 +233,14 @@ module_param_named(msi, amdgpu_msi, int,
> 0444);
> > * Set GPU scheduler timeout value in ms.
> > *
> > * The format can be [Non-Compute] or [GFX,Compute,SDMA,Video].
> That
> > is there can be one or
> > - * multiple values specified. 0 and negative values are invalidated.
> > They will be adjusted
> > - * to default timeout.
> > + * multiple values specified.
> > * - With one value specified, the setting will apply to all non-compute
> jobs.
> > * - With multiple values specified, the first one will be for GFX. The
> second one is for Compute.
> > * And the third and fourth ones are for SDMA and Video.
> > * By default(with no lockup_timeout settings), the timeout for all non-
> compute(GFX, SDMA and Video)
> > * jobs is 10000. And there is no timeout enforced on compute jobs.
> > + * Value 0 is invalidated, will be adjusted to default timeout settings.
> > + * Negative values mean 'infinite timeout' (MAX_JIFFY_OFFSET).
> > */
> > MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms
> (default: 10000 for non-compute jobs and no timeout for compute jobs), "
> > "format is [Non-Compute] or [GFX,Compute,SDMA,Video]");
> > @@ -1248,11 +1249,16 @@ int
> amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev)
> > if (ret)
> > return ret;
> >
> > - /* Invalidate 0 and negative values */
> > - if (timeout <= 0) {
> > + /*
> > + * Value 0 will be adjusted to default timeout settings.
> > + * Negative values mean 'infinite timeout' (MAX_JIFFY_OFFSET).
> > + */
> > + if (!timeout) {
> > index++;
> > continue;
> > }
> > + if (timeout < 0)
> > + timeout = MAX_JIFFY_OFFSET;
>
> This is superfluous and maybe even harmful, msecs_to_jiffies() should take
> care of this conversion.
>
> Maybe even convert the values directly here.
>
> Christian.
>
> >
> > switch (index++) {
> > case 0:
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2019-05-07 1:48 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-05-05 14:23 [PATCH] drm/amdgpu: treat negative lockup timeout as 'infinite timeout' Evan Quan
[not found] ` <20190505142353.30369-1-evan.quan-5C7GfCeVMHo@public.gmane.org>
2019-05-06 11:26 ` Christian König
[not found] ` <03378dca-c0bb-443e-818a-3a01b69dcccb-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2019-05-07 1:48 ` Quan, Evan
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.