From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Deucher, Alexander" Subject: Re: [PATCH] drm/amdgpu: no job timeout setting on compute queues Date: Fri, 16 Mar 2018 17:17:14 +0000 Message-ID: References: <1521175952-21758-1-git-send-email-evan.quan@amd.com> , Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============1618828545==" Return-path: In-Reply-To: Content-Language: en-US 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: =?iso-8859-1?Q?Michel_D=E4nzer?= , "Quan, Evan" Cc: "amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org" --===============1618828545== Content-Language: en-US Content-Type: multipart/alternative; boundary="_000_CY4PR12MB16535036642155291C39F030F7D70CY4PR12MB1653namp_" --_000_CY4PR12MB16535036642155291C39F030F7D70CY4PR12MB1653namp_ Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable How about something like this: /* Disable timeout when GPU reset is disabled to avoid confusing timeout me= ssages in the kernel log */ If (amdgpu_gpu_recovery =3D=3D 0 || amdgpu_gpu_recovery =3D=3D -1) amdgpu_lockup_timeout =3D MAX_SCHEDULE_TIMEOUT; Alex ________________________________ From: Michel D=E4nzer Sent: Friday, March 16, 2018 12:25:03 PM To: Deucher, Alexander; Quan, Evan Cc: amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org Subject: Re: [PATCH] drm/amdgpu: no job timeout setting on compute queues On 2018-03-16 05:14 PM, Deucher, Alexander wrote: > Since GPU reset is not enabled yet anyway, a timeout will just print a > message, can we just change amdgpu_lockup_timeout to > MAX_SCHEDULE_TIMEOUT until we enable GPU reset? Wouldn't that be rather surprising for somebody who sets amdgpu.gpu_recovery=3D1 ? -- Earthling Michel D=E4nzer | http://www.amd.com Libre software enthusiast | Mesa and X developer --_000_CY4PR12MB16535036642155291C39F030F7D70CY4PR12MB1653namp_ Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

How about something like this:

/* Disable timeout when GPU reset is disabled t= o avoid confusing timeout messages in the kernel log */

If (amdgpu_gpu_recovery =3D=3D 0 || amdgpu_gpu_recovery =3D=3D -1)

    amdgpu_lockup_timeout =3D MAX_SCHEDULE_TIMEOUT;

Ale= x

From: Michel D=E4nzer <m= ichel-otUistvHUpPR7s880joybQ@public.gmane.org>
Sent: Friday, March 16, 2018 12:25:03 PM
To: Deucher, Alexander; Quan, Evan
Cc: amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
Subject: Re: [PATCH] drm/amdgpu: no job timeout setting on compute q= ueues
 
On 2018-03-16 05:14 PM, Deucher, Alexander wrote:<= br> > Since GPU reset is not enabled yet anyway, a timeout will just print a=
> message, can we just change amdgpu_lockup_timeout to
> MAX_SCHEDULE_TIMEOUT until we enable GPU reset?

Wouldn't that be rather surprising for somebody who sets
amdgpu.gpu_recovery=3D1 ?


--
Earthling Michel D=E4nzer        &n= bsp;      |      &nb= sp;        http://www.amd.com
Libre software enthusiast        &n= bsp;    |        &nb= sp;    Mesa and X developer
--_000_CY4PR12MB16535036642155291C39F030F7D70CY4PR12MB1653namp_-- --===============1618828545== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: inline X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18KYW1kLWdmeCBt YWlsaW5nIGxpc3QKYW1kLWdmeEBsaXN0cy5mcmVlZGVza3RvcC5vcmcKaHR0cHM6Ly9saXN0cy5m cmVlZGVza3RvcC5vcmcvbWFpbG1hbi9saXN0aW5mby9hbWQtZ2Z4Cg== --===============1618828545==--