From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Koenig, Christian" Subject: Re: Fixing SDMA TO after GPU reset Date: Tue, 11 Sep 2018 13:57:04 +0000 Message-ID: References: <059118f3-2729-12a1-7c8d-e306f69369aa@amd.com> , Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============1885738265==" Return-path: In-Reply-To: Content-Language: de-DE List-Id: Discussion list for AMD gfx List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: amd-gfx-bounces-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org Sender: "amd-gfx" To: "Grodzovsky, Andrey" Cc: "Deucher, Alexander" , "amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org" --===============1885738265== Content-Language: de-DE Content-Type: multipart/alternative; boundary="_000_e72109580a4d44158dbfdd61c73564cbemailandroidcom_" --_000_e72109580a4d44158dbfdd61c73564cbemailandroidcom_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Yes, that is actually desired. We should not flush the system VM just because we do some clearing command = submission. Christian. Am 11.09.2018 15:54 schrieb "Grodzovsky, Andrey" : By current code logic job->vm_pd_addr is never going to be set unless the j= ob is created for user SC or for buffer copy in amdgpu_copy_buffer So in any other case we are going to skip VM flush. But amdgpu_vm_flush wan= ts a flush to happen in case GPU reset just happend (amdgpu_vmid_had_gpu_re= set is true) so we will be skipping that VM flush (as in my case with amdgpu_driver_open= _kms->amdgpu_vm_init->amdgpu_vm_clear_bo right after GPU reset occured) Is it safe ? Andrey On 09/11/2018 07:46 AM, Christian K=F6nig wrote: It would probably be better to initialize job->vm_pd_addr with AMDGPU_BO_IN= VALID_OFFSET. And then just drop the vm flush alltogether when the vm_pd_addr isn't set t= o something sane. Thanks, Christian. Am 11.09.2018 um 00:52 schrieb Andrey Grodzovsky: Attached patch fixes SDMA TO after GPU reset, it's a regression caused by c= bd5285 drm/amdgpu: move setting the GART addr into TTM. But to me it looks safer just to revert the original patch all together sin= ce we never can predict for sure if VM flush will take place and so it's sa= fer to just always assign job->vm_pd_addr. Andrey _______________________________________________ amd-gfx mailing list amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org https://lists.freedesktop.org/mailman/listinfo/amd-gfx --_000_e72109580a4d44158dbfdd61c73564cbemailandroidcom_ Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Yes, that is actually desired.

We should not flush the system VM just because we do some= clearing command submission.

Christian.

Am 11.09.2018 15:54 schrieb "Grodzovsky, An= drey" <Andrey.Grodzovsky-5C7GfCeVMHo@public.gmane.org>:
By current code logic job->vm_pd_addr is never going to be set unle= ss the job is created for user SC or for buffer copy in amdgpu_copy_buffer<= br> So in any other case we are going to skip VM flush. But amdgpu_vm_flush wan= ts a flush to happen in case GPU reset just happend (amdgpu_vmid_had_gpu_re= set is true)
so we will be skipping that VM flush (as in my case with amdgpu_driver_open= _kms->amdgpu_vm_init->amdgpu_vm_clear_bo right after GPU reset occure= d)
Is it safe ?

Andrey

On 09/11/2018 07:46 AM, Christian K=F6nig wr= ote:
It would probably be better to initialize jo= b->vm_pd_addr with AMDGPU_BO_INVALID_OFFSET.

And then just drop the vm flush alltogether when the vm_pd_addr isn't set t= o something sane.

Thanks,
Christian.

Am 11.09.2018 um 00:52 schrieb Andrey Grodzovsky:
Attached patch fixes SDMA TO after GPU reset, it'= s a regression caused by cbd5285 drm/amdgpu: move setting the GART addr int= o TTM.

But to me it looks safer just to revert the original patch all together sin= ce we never can predict for sure if VM flush will take place and so it's sa= fer to just always assign job->vm_pd_addr.

Andrey 



_______________________________________________
amd-gfx mailing list
amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
https://lists.freedesktop.org/mailman/listinfo/amd-=
gfx


--_000_e72109580a4d44158dbfdd61c73564cbemailandroidcom_-- --===============1885738265== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: inline X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18KYW1kLWdmeCBt YWlsaW5nIGxpc3QKYW1kLWdmeEBsaXN0cy5mcmVlZGVza3RvcC5vcmcKaHR0cHM6Ly9saXN0cy5m cmVlZGVza3RvcC5vcmcvbWFpbG1hbi9saXN0aW5mby9hbWQtZ2Z4Cg== --===============1885738265==--