[PATCH] drm/amdkfd: print unmap queue status for RAS poison consumption (v3)

Zhou1, Tao Tao.Zhou1 at amd.com
Tue Mar 22 03:51:22 UTC 2022


[AMD Official Use Only]



> -----Original Message-----
> From: Paul Menzel <pmenzel at molgen.mpg.de>
> Sent: Tuesday, March 22, 2022 11:42 AM
> To: Zhou1, Tao <Tao.Zhou1 at amd.com>
> Cc: amd-gfx at lists.freedesktop.org; Zhang, Hawking
> <Hawking.Zhang at amd.com>; Kuehling, Felix <Felix.Kuehling at amd.com>; Yang,
> Stanley <Stanley.Yang at amd.com>; Chai, Thomas <YiPeng.Chai at amd.com>;
> Lazar, Lijo <Lijo.Lazar at amd.com>
> Subject: Re: [PATCH] drm/amdkfd: print unmap queue status for RAS poison
> consumption (v3)
> 
> Dear Tao,
> 
> 
> Thank you for rerolling the pacth.
> 
> Am 22.03.22 um 04:29 schrieb Tao Zhou:
> > Print the status out when it passes, and also tell user gpu reset is
> > triggered when we fallback to legacy way.
> 
> The verb *fall back* is written with a space.

[Tao] will update it before push

> 
> > v2: make the message more explicitly.
> 
> explicit

[Tao] will update it before push.

> 
> > v3: change succeeds to succeeded.
> >      replace pr_warn with dev_info.
> 
> Please give the reason, why to decrease the debug level. (I would also do it in a
> separate commit.)

[Tao] please check my reply to Lijo. It's informational message, but ras poison consumption is special event, both debug levels are OK for me.

> 
> > Signed-off-by: Tao Zhou <tao.zhou1 at amd.com>
> > ---
> >   drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c | 13 +++++++++----
> >   1 file changed, 9 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
> > b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
> > index 56902b5bb7b6..cc9ddc4b4cb8 100644
> > --- a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
> > +++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
> > @@ -105,8 +105,6 @@ static void
> event_interrupt_poison_consumption(struct kfd_dev *dev,
> >   	if (old_poison)
> >   		return;
> >
> > -	pr_warn("RAS poison consumption handling: client id %d\n", client_id);
> > -
> >   	switch (client_id) {
> >   	case SOC15_IH_CLIENTID_SE0SH:
> >   	case SOC15_IH_CLIENTID_SE1SH:
> > @@ -130,10 +128,17 @@ static void
> event_interrupt_poison_consumption(struct kfd_dev *dev,
> >   	/* resetting queue passes, do page retirement without gpu reset
> >   	 * resetting queue fails, fallback to gpu reset solution
> >   	 */
> > -	if (!ret)
> > +	if (!ret) {
> > +		dev_info(dev->adev->dev,
> > +			"RAS poison consumption, unmap queue flow
> succeeded: client id %d\n",
> > +			client_id);
> >   		amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev,
> false);
> > -	else
> > +	} else {
> > +		dev_info(dev->adev->dev,
> > +			"RAS poison consumption, fallback to gpu reset flow:
> client id
> > +%d\n",
> 
> The verb *fall back* is written with a space.

[Tao] will update it before push.

> 
> > +			client_id);
> >   		amdgpu_amdkfd_ras_poison_consumption_handler(dev->adev,
> true);
> > +	}
> >   }
> >
> >   static bool event_interrupt_isr_v9(struct kfd_dev *dev,


More information about the amd-gfx mailing list