From: sashiko-bot@kernel.org
To: "Maarten Lankhorst" <dev@lankhorst.se>
Cc: dri-devel@lists.freedesktop.org
Subject: Re: [PATCH v8 14/27] drm/i915/display: Make set_pipeconf use the fw variants
Date: Mon, 15 Jun 2026 20:44:03 +0000 [thread overview]
Message-ID: <20260615204403.954891F00A3F@smtp.kernel.org> (raw)
In-Reply-To: <20260615201846.307297-15-dev@lankhorst.se>
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Incomplete migration to _fw register access variants in VRR handling causes 'sleeping in atomic context' BUG on PREEMPT_RT.
--
commit 3e8bb0c4e5f9821d1b47fa3563431c605e96b128
Author: Maarten Lankhorst <dev@lankhorst.se>
drm/i915/display: Make set_pipeconf use the fw variants
The calls are used inside the critical section when updating
the gamma mode, and thus should use the _fw variants to prevent
locks.
> diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
> index 2e3ade44f28ee..03b15765843a0 100644
> --- a/drivers/gpu/drm/i915/display/intel_display.c
> +++ b/drivers/gpu/drm/i915/display/intel_display.c
[ ... ]
> @@ -3042,8 +3042,9 @@ void i9xx_set_pipeconf(const struct intel_crtc_state *crtc_state)
>
> val |= TRANSCONF_FRAME_START_DELAY(crtc_state->framestart_delay - 1);
>
> - intel_de_write(display, TRANSCONF(display, cpu_transcoder), val);
> - intel_de_posting_read(display, TRANSCONF(display, cpu_transcoder));
> + intel_de_write_fw(display, TRANSCONF(display, cpu_transcoder), val);
> + /* posting read */
> + intel_de_read_fw(display, TRANSCONF(display, cpu_transcoder));
> }
[Severity: High]
This is a pre-existing issue, but since the goal is to prevent sleeping
in atomic context on PREEMPT_RT during the vblank evasion section, are
there other register accesses in this path that need the fw variants?
Several VRR functions still use the sleepable non-fw register access
variants inside the atomic section where local IRQs are disabled:
1. intel_vrr_set_vrr_timings() and intel_vrr_enable_dc_balancing() use
intel_de_write() and intel_de_read() and are called from
intel_vrr_enable(), which executes via commit_pipe_post_planes().
2. intel_vrr_set_transcoder_timings() uses intel_de_write() and executes
via commit_pipe_pre_planes() -> intel_pipe_fastset().
For example, in drivers/gpu/drm/i915/display/intel_vrr.c:
intel_vrr_set_vrr_timings() {
...
intel_de_write(display, TRANS_VRR_VMIN(display, cpu_transcoder),
intel_vrr_hw_vmin(crtc_state) - 1);
...
}
Could these non-fw accesses trigger the same PREEMPT_RT splat?
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260615201846.307297-1-dev@lankhorst.se?part=14
next prev parent reply other threads:[~2026-06-15 20:44 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-15 20:18 [PATCH v8 00/27] drm/i915/display: All patches to make PREEMPT_RT work on i915 + xe Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 01/27] drm/vblank_work: Add methods to schedule vblank_work in 2 stages Maarten Lankhorst
2026-06-15 20:30 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 02/27] drm/vblank: Add a 2-stage version of drm_crtc_arm_vblank_event Maarten Lankhorst
2026-06-15 20:31 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 03/27] drm/intel/display: Make intel_crtc_arm_vblank_event static Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 04/27] drm/intel/display: Convert vblank event handling to 2-stage arming Maarten Lankhorst
2026-06-15 20:35 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 05/27] drm/i915/display: Move vblank put until after critical section Maarten Lankhorst
2026-06-15 20:25 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 06/27] drm/i915/display: Remove locking from intel_vblank_evade " Maarten Lankhorst
2026-06-15 20:39 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 07/27] drm/i915/display: Handle vlv dsi workaround in scanline_in_safe_range too Maarten Lankhorst
2026-06-15 20:29 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 08/27] drm/i915: Use preempt_disable/enable_rt() where recommended Maarten Lankhorst
2026-06-15 20:32 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 09/27] drm/i915/display: Make get_vblank_counter use intel_de_read_fw() Maarten Lankhorst
2026-06-15 20:39 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 10/27] drm/i915/display: Do not take uncore lock in i915_get_vblank_counter Maarten Lankhorst
2026-06-15 20:35 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 11/27] drm/i915/display: Make icl_dsi_frame_update use _fw too Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 12/27] drm/i915/display: Use intel_de_read/write_fw in colorops Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 13/27] drm/i915/display: Use intel_de_write_fw in intel_pipe_fastset Maarten Lankhorst
2026-06-15 20:46 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 14/27] drm/i915/display: Make set_pipeconf use the fw variants Maarten Lankhorst
2026-06-15 20:44 ` sashiko-bot [this message]
2026-06-15 20:18 ` [PATCH v8 15/27] drm/i915/gt: Use spin_lock_irq() instead of local_irq_disable() + spin_lock() Maarten Lankhorst
2026-06-15 20:39 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 16/27] drm/i915: Drop the irqs_disabled() check Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 17/27] drm/i915/guc: Consider also RCU depth in busy loop Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 18/27] drm/i915/gt: Fix selftests on PREEMPT_RT Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 19/27] drm/i915/gt: Set stop_timeout() correctly on PREEMPT-RT Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 20/27] drm/i915/display: Remove uncore lock from vlv_atomic_update_fifo Maarten Lankhorst
2026-06-15 20:36 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 21/27] drm/i915: Use sleeping selftests for igt_atomic on PREEMPT_RT Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 22/27] Revert "drm/i915: Depend on !PREEMPT_RT." Maarten Lankhorst
2026-06-15 20:18 ` [PATCH v8 23/27] PREEMPT_RT injection Maarten Lankhorst
2026-06-15 20:39 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 24/27] FOR-CI: bump MAX_STACK_TRACE_ENTRIES Maarten Lankhorst
2026-06-15 20:35 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 25/27] drm/i915/gt: Add a spinlock to prevent starvation of irq_work Maarten Lankhorst
2026-06-15 20:38 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 26/27] drm/xe/display: Always use system memory on PREEMPT_RT for DPT Maarten Lankhorst
2026-06-15 20:39 ` sashiko-bot
2026-06-15 20:18 ` [PATCH v8 27/27] drm/xe/display: Prefer not to allocate a framebuffers in stolen memory Maarten Lankhorst
2026-06-15 20:41 ` sashiko-bot
2026-06-15 20:33 ` ✗ CI.checkpatch: warning for drm/i915/display: All patches to make PREEMPT_RT work on i915 + xe. (rev16) Patchwork
2026-06-15 20:34 ` ✓ CI.KUnit: success " Patchwork
2026-06-15 20:50 ` ✗ CI.checksparse: warning " Patchwork
2026-06-15 21:27 ` ✗ Xe.CI.BAT: failure " Patchwork
2026-06-15 21:48 ` ✗ i915.CI.BAT: " Patchwork
2026-06-15 23:22 ` ✗ Xe.CI.FULL: " Patchwork
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260615204403.954891F00A3F@smtp.kernel.org \
--to=sashiko-bot@kernel.org \
--cc=dev@lankhorst.se \
--cc=dri-devel@lists.freedesktop.org \
--cc=sashiko-reviews@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.