All of lore.kernel.org
 help / color / mirror / Atom feed
* error messages from running uvd on polaris10
@ 2018-01-29 18:45 Tom St Denis
       [not found] ` <1fe2acde-4518-5bdd-0584-1c3e19aaa114-5C7GfCeVMHo@public.gmane.org>
  0 siblings, 1 reply; 4+ messages in thread
From: Tom St Denis @ 2018-01-29 18:45 UTC (permalink / raw)
  To: amd-gfx mailing list

Hi all,

Getting these:

[  537.515830] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.516057] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.516267] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.516485] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.516823] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.517080] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.517323] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.517605] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.517848] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.518090] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.518330] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.518581] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.518807] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.519034] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.519260] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!
[  537.519486] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
writing more dwords to the ring than expected!


from running

DRI_PRIME=1 mpv --hwdec=auto *mkv

on my CZ+Polaris10 setup (with drm-next up to date).  I don't get the 
messages though if I run without DRI_PRIME=1 and they both have the same 
UVD6 right?

Is it because one is in physical vs VM mode?

Cheers,
Tom
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: error messages from running uvd on polaris10
       [not found] ` <1fe2acde-4518-5bdd-0584-1c3e19aaa114-5C7GfCeVMHo@public.gmane.org>
@ 2018-01-29 18:50   ` Christian König
       [not found]     ` <db8ebbbb-c678-d977-102c-ab3e782fda35-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
  0 siblings, 1 reply; 4+ messages in thread
From: Christian König @ 2018-01-29 18:50 UTC (permalink / raw)
  To: Tom St Denis, amd-gfx mailing list

Hi Tom,

well that is an interesting one. Looks like we use more dw on the UVD 
ring when prime is enabled.

Going to take a look tomorrow,
Christian

Am 29.01.2018 um 19:45 schrieb Tom St Denis:
> Hi all,
>
> Getting these:
>
> [  537.515830] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.516057] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.516267] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.516485] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.516823] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.517080] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.517323] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.517605] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.517848] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.518090] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.518330] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.518581] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.518807] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.519034] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.519260] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
> [  537.519486] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
> writing more dwords to the ring than expected!
>
>
> from running
>
> DRI_PRIME=1 mpv --hwdec=auto *mkv
>
> on my CZ+Polaris10 setup (with drm-next up to date).  I don't get the 
> messages though if I run without DRI_PRIME=1 and they both have the 
> same UVD6 right?
>
> Is it because one is in physical vs VM mode?
>
> Cheers,
> Tom
> _______________________________________________
> amd-gfx mailing list
> amd-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: error messages from running uvd on polaris10
       [not found]     ` <db8ebbbb-c678-d977-102c-ab3e782fda35-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
@ 2018-01-29 18:53       ` Tom St Denis
       [not found]         ` <d0149fb1-c220-cbfa-9935-c95a5b5cfd2a-5C7GfCeVMHo@public.gmane.org>
  0 siblings, 1 reply; 4+ messages in thread
From: Tom St Denis @ 2018-01-29 18:53 UTC (permalink / raw)
  To: christian.koenig-5C7GfCeVMHo, amd-gfx mailing list

On 29/01/18 01:50 PM, Christian König wrote:
> Hi Tom,
> 
> well that is an interesting one. Looks like we use more dw on the UVD 
> ring when prime is enabled.
> 
> Going to take a look tomorrow,
> Christian

Hi Christian,

In this case prime=1 means polaris10 where VM is used right?

prime=0 means CZ (on my setup) which is physical I think.

Tom

> 
> Am 29.01.2018 um 19:45 schrieb Tom St Denis:
>> Hi all,
>>
>> Getting these:
>>
>> [  537.515830] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.516057] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.516267] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.516485] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.516823] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.517080] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.517323] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.517605] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.517848] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.518090] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.518330] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.518581] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.518807] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.519034] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.519260] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>> [  537.519486] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>> writing more dwords to the ring than expected!
>>
>>
>> from running
>>
>> DRI_PRIME=1 mpv --hwdec=auto *mkv
>>
>> on my CZ+Polaris10 setup (with drm-next up to date).  I don't get the 
>> messages though if I run without DRI_PRIME=1 and they both have the 
>> same UVD6 right?
>>
>> Is it because one is in physical vs VM mode?
>>
>> Cheers,
>> Tom
>> _______________________________________________
>> amd-gfx mailing list
>> amd-gfx@lists.freedesktop.org
>> https://lists.freedesktop.org/mailman/listinfo/amd-gfx
> 

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: error messages from running uvd on polaris10
       [not found]         ` <d0149fb1-c220-cbfa-9935-c95a5b5cfd2a-5C7GfCeVMHo@public.gmane.org>
@ 2018-01-30 14:16           ` Tom St Denis
  0 siblings, 0 replies; 4+ messages in thread
From: Tom St Denis @ 2018-01-30 14:16 UTC (permalink / raw)
  To: christian.koenig-5C7GfCeVMHo, amd-gfx mailing list

On 29/01/18 01:53 PM, Tom St Denis wrote:
> On 29/01/18 01:50 PM, Christian König wrote:
>> Hi Tom,
>>
>> well that is an interesting one. Looks like we use more dw on the UVD 
>> ring when prime is enabled.
>>
>> Going to take a look tomorrow,
>> Christian
> 
> Hi Christian,
> 
> In this case prime=1 means polaris10 where VM is used right?
> 
> prime=0 means CZ (on my setup) which is physical I think.

Hi Christian,

Reverting 75965041675b1427a8d4cffc13dd070de9414619 eliminates the error 
messages.  I don't claim that that's the "fix" but it's certainly the 
commit that introduces the error :-)


Cheers,
Tom

> 
> Tom
> 
>>
>> Am 29.01.2018 um 19:45 schrieb Tom St Denis:
>>> Hi all,
>>>
>>> Getting these:
>>>
>>> [  537.515830] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* 
>>> amdgpu: writing more dwords to the ring than expected!
>>> [  537.516057] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* 
>>> amdgpu: writing more dwords to the ring than expected!
>>> [  537.516267] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* 
>>> amdgpu: writing more dwords to the ring than expected!
>>> [  537.516485] [drm:uvd_v6_0_ring_emit_fence [amdgpu]] *ERROR* 
>>> amdgpu: writing more dwords to the ring than expected!
>>> [  537.516823] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.517080] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.517323] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.517605] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.517848] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.518090] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.518330] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.518581] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.518807] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.519034] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.519260] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>> [  537.519486] [drm:amdgpu_ring_insert_nop [amdgpu]] *ERROR* amdgpu: 
>>> writing more dwords to the ring than expected!
>>>
>>>
>>> from running
>>>
>>> DRI_PRIME=1 mpv --hwdec=auto *mkv
>>>
>>> on my CZ+Polaris10 setup (with drm-next up to date).  I don't get the 
>>> messages though if I run without DRI_PRIME=1 and they both have the 
>>> same UVD6 right?
>>>
>>> Is it because one is in physical vs VM mode?
>>>
>>> Cheers,
>>> Tom
>>> _______________________________________________
>>> amd-gfx mailing list
>>> amd-gfx@lists.freedesktop.org
>>> https://lists.freedesktop.org/mailman/listinfo/amd-gfx
>>
> 
> _______________________________________________
> amd-gfx mailing list
> amd-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/amd-gfx

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2018-01-30 14:16 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-01-29 18:45 error messages from running uvd on polaris10 Tom St Denis
     [not found] ` <1fe2acde-4518-5bdd-0584-1c3e19aaa114-5C7GfCeVMHo@public.gmane.org>
2018-01-29 18:50   ` Christian König
     [not found]     ` <db8ebbbb-c678-d977-102c-ab3e782fda35-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2018-01-29 18:53       ` Tom St Denis
     [not found]         ` <d0149fb1-c220-cbfa-9935-c95a5b5cfd2a-5C7GfCeVMHo@public.gmane.org>
2018-01-30 14:16           ` Tom St Denis

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.