linux-kernel - Re: [PATCH v3 3/3] drm/msm/dpu: capture snapshot on the first commit

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: <3dd62c7e-e419-9ea2-b383-4c1c0959fecf@quicinc.com>
Date: Sun, 25 Feb 2024 17:35:56 -0800
From: Abhinav Kumar <quic_abhinavk@...cinc.com>
To: Dmitry Baryshkov <dmitry.baryshkov@...aro.org>,
        Rob Clark
	<robdclark@...il.com>, Sean Paul <sean@...rly.run>,
        Marijn Suijten
	<marijn.suijten@...ainline.org>,
        David Airlie <airlied@...il.com>, Daniel
 Vetter <daniel@...ll.ch>
CC: Steev Klimaszewski <steev@...i.org>, <linux-arm-msm@...r.kernel.org>,
        <dri-devel@...ts.freedesktop.org>, <freedreno@...ts.freedesktop.org>,
        <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3 3/3] drm/msm/dpu: capture snapshot on the first
 commit_done timeout



On 2/25/2024 11:57 AM, Abhinav Kumar wrote:
> 
> 
> On 2/25/2024 6:12 AM, Dmitry Baryshkov wrote:
>> In order to debug commit_done timeouts, capture the devcoredump state
>> when the first timeout occurs after the encoder has been enabled.
>>
>> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@...aro.org>
>> ---
>>   drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c | 9 +++++++++
>>   1 file changed, 9 insertions(+)
>>
> 
> This looks fine now. Once we discuss patch 2, I can ack this.
> 

Not entirely onboard with patch 2, but lets see how that code evolves

Reviewed-by: Abhinav Kumar <quic_abhinavk@...cinc.com>

>> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c 
>> b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
>> index 30f349c8a1e5..3cae07bf0b9b 100644
>> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
>> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
>> @@ -126,6 +126,8 @@ enum dpu_enc_rc_states {
>>    * @base:        drm_encoder base class for registration with DRM
>>    * @enc_spinlock:    Virtual-Encoder-Wide Spin Lock for IRQ purposes
>>    * @enabled:        True if the encoder is active, protected by 
>> enc_lock
>> + * @commit_done_timedout: True if there has been a timeout on commit 
>> after
>> + *            enabling the encoder.
>>    * @num_phys_encs:    Actual number of physical encoders contained.
>>    * @phys_encs:        Container of physical encoders managed.
>>    * @cur_master:        Pointer to the current master in this mode. 
>> Optimization
>> @@ -172,6 +174,7 @@ struct dpu_encoder_virt {
>>       spinlock_t enc_spinlock;
>>       bool enabled;
>> +    bool commit_done_timedout;
>>       unsigned int num_phys_encs;
>>       struct dpu_encoder_phys *phys_encs[MAX_PHYS_ENCODERS_PER_VIRTUAL];
>> @@ -1226,6 +1229,8 @@ static void 
>> dpu_encoder_virt_atomic_enable(struct drm_encoder *drm_enc,
>>       else if (disp_info->intf_type == INTF_DSI)
>>           dpu_enc->wide_bus_en = 
>> msm_dsi_wide_bus_enabled(priv->dsi[index]);
>> +    dpu_enc->commit_done_timedout = false;
>> +
>>       mutex_lock(&dpu_enc->enc_lock);
>>       cur_mode = &dpu_enc->base.crtc->state->adjusted_mode;
>> @@ -2436,6 +2441,10 @@ int dpu_encoder_wait_for_commit_done(struct 
>> drm_encoder *drm_enc)
>>               DPU_ATRACE_BEGIN("wait_for_commit_done");
>>               ret = phys->ops.wait_for_commit_done(phys);
>>               DPU_ATRACE_END("wait_for_commit_done");
>> +            if (ret == -ETIMEDOUT && !dpu_enc->commit_done_timedout) {
>> +                dpu_enc->commit_done_timedout = true;
>> +                msm_disp_snapshot_state(drm_enc->dev);
>> +            }
>>               if (ret)
>>                   return ret;
>>           }
>>