linux-kernel - Re: [PATCH] drm: msm: a6x: Disable interrupts before recovery

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20200302164324.GA708@jcrouse1-lnx.qualcomm.com>
Date:   Mon, 2 Mar 2020 09:43:24 -0700
From:   Jordan Crouse <jcrouse@...eaurora.org>
To:     Sharat Masetty <smasetty@...eaurora.org>
Cc:     freedreno@...ts.freedesktop.org, dri-devel@...edesktop.org,
        linux-arm-msm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] drm: msm: a6x: Disable interrupts before recovery

On Mon, Mar 02, 2020 at 02:17:16PM +0530, Sharat Masetty wrote:
> This patch disables interrupts in the GPU RBBM hang detect fault handler
> before going to recovery.

Okay, but why?  What were you seeing?  A little bit of extra description would
make it easier to understand this change in the future.

> 
> Signed-off-by: Sharat Masetty <smasetty@...eaurora.org>
> ---
>  drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
> index dc8ec2c..4dd0f62 100644
> --- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
> +++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
> @@ -676,6 +676,9 @@ static void a6xx_fault_detect_irq(struct msm_gpu *gpu)
>  		gpu_read64(gpu, REG_A6XX_CP_IB2_BASE, REG_A6XX_CP_IB2_BASE_HI),
>  		gpu_read(gpu, REG_A6XX_CP_IB2_REM_SIZE));
>  
> +	/* Disable interrupts before going for a recovery*/
> +	gpu_write(gpu, REG_A6XX_RBBM_INT_0_MASK, 0);
> +

And this is turning off all the interrupts, but the commit log only mentions the
hang detect. In my experience, the hang detect usually only fires once until
after reset, but if there are other interrupts that are bothering you then it
makes sense to disable them but, again, this is good information for the commit
log and or a code comment.

Jordan

>  	/* Turn off the hangcheck timer to keep it from bothering us */
>  	del_timer(&gpu->hangcheck_timer);
>  
> -- 
> 1.9.1
> 
-- 
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project