linux-kernel - Re: [PATCH 3/6] drm/msm/dpu: don't use active in atomic

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

Message-ID: <b4f1d7c2-c9eb-4b9a-b8b7-f335910601d8@quicinc.com>
Date: Wed, 8 Jan 2025 17:19:40 -0800
From: Abhinav Kumar <quic_abhinavk@...cinc.com>
To: Dmitry Baryshkov <dmitry.baryshkov@...aro.org>,
        Maarten Lankhorst
	<maarten.lankhorst@...ux.intel.com>,
        Maxime Ripard <mripard@...nel.org>,
        Thomas Zimmermann <tzimmermann@...e.de>,
        David Airlie <airlied@...il.com>, Simona Vetter <simona@...ll.ch>,
        Rob Clark <robdclark@...il.com>, Sean Paul
	<sean@...rly.run>,
        Marijn Suijten <marijn.suijten@...ainline.org>,
        Chandan
 Uddaraju <chandanu@...eaurora.org>,
        Jeykumar Sankaran
	<jsanka@...eaurora.org>,
        Jordan Crouse <jordan@...micpenguin.net>,
        Sravanthi
 Kollukuduru <skolluku@...eaurora.org>
CC: <dri-devel@...ts.freedesktop.org>, <linux-kernel@...r.kernel.org>,
        Archit
 Taneja <architt@...eaurora.org>,
        Rajesh Yadav <ryadav@...eaurora.org>, <linux-arm-msm@...r.kernel.org>,
        <freedreno@...ts.freedesktop.org>,
        Simona
 Vetter <simona.vetter@...ll.ch>
Subject: Re: [PATCH 3/6] drm/msm/dpu: don't use active in atomic_check()



On 12/21/2024 9:00 PM, Dmitry Baryshkov wrote:
> The driver isn't supposed to consult crtc_state->active/active_check for
> resource allocation. Instead all resources should be allocated if
> crtc_state->enabled is set. Stop consulting active / active_changed in
> order to determine whether the hardware resources should be
> (re)allocated.
> 
> Fixes: 25fdd5933e4c ("drm/msm: Add SDM845 DPU support")
> Reported-by: Simona Vetter <simona.vetter@...ll.ch>
> Closes: https://lore.kernel.org/dri-devel/ZtW_S0j5AEr4g0QW@phenom.ffwll.local/
> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@...aro.org>
> ---
>   drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c    | 4 ----
>   drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c | 3 +--
>   2 files changed, 1 insertion(+), 6 deletions(-)
> 
> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
> index 7191b1a6d41b3a96f956d199398f12b2923e8c82..65e33eba61726929b740831c95330756b2817e28 100644
> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c
> @@ -1264,10 +1264,6 @@ static int dpu_crtc_atomic_check(struct drm_crtc *crtc,
>   
>   	DRM_DEBUG_ATOMIC("%s: check\n", dpu_crtc->name);
>   
> -	/* force a full mode set if active state changed */
> -	if (crtc_state->active_changed)
> -		crtc_state->mode_changed = true;
> -
>   	if (cstate->num_mixers) {
>   		rc = _dpu_crtc_check_and_setup_lm_bounds(crtc, crtc_state);
>   		if (rc)
> diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
> index 83de7564e2c1fe14fcf8c4f82335cafc937e1b99..d1ccdca6044353f110bf5b507788efe368f223a3 100644
> --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
> +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c
> @@ -793,12 +793,11 @@ static int dpu_encoder_virt_atomic_check(
>   		crtc_state->mode_changed = true;
>   	/*
>   	 * Release and Allocate resources on every modeset
> -	 * Dont allocate when active is false.
>   	 */
>   	if (drm_atomic_crtc_needs_modeset(crtc_state)) {
>   		dpu_rm_release(global_state, drm_enc);
>   
> -		if (!crtc_state->active_changed || crtc_state->enable)

I think this was leftover code.

What happened was, we used to have dpu_rm_reserve() both in 
dpu_encoder's atomic_check and mode_set. Hence this is checking 
!active_changed because that case was expected to get handled in 
mode_set to avoid duplicate dpu_rm_reserve() calls. Code has progressed 
since then to drop the dpu_rm_reserve() from mode_set and only use 
atomic_check.

So the correct fixes tag for this should be:

Fixes: de3916c70a24 ("drm/msm/dpu: Track resources in global state")

With that addressed, this is

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

> +		if (crtc_state->enable)
>   			ret = dpu_rm_reserve(&dpu_kms->rm, global_state,
>   					drm_enc, crtc_state, topology);
>   		if (!ret)
>