From: Maarten Lankhorst <dev@lankhorst.se>
To: Junxiao Chang <junxiao.chang@intel.com>,
jani.nikula@linux.intel.com, rodrigo.vivi@intel.com,
matthew.brost@intel.com, thomas.hellstrom@linux.intel.com,
airlied@gmail.com, simona@ffwll.ch, bigeasy@linutronix.de,
clrkwllms@kernel.org, rostedt@goodmis.org,
intel-gfx@lists.freedesktop.org, intel-xe@lists.freedesktop.org,
dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org,
linux-rt-devel@lists.linux.dev
Subject: Re: [PATCH 1/2] drm/xe/display: Skip initial plane allocation from stolen memory on RT
Date: Tue, 3 Nov 2026 11:45:09 +0100 [thread overview]
Message-ID: <65fae402-e053-4a4f-ba55-6b5b51690eba@lankhorst.se> (raw)
In-Reply-To: <20261104032732.30681-2-junxiao.chang@intel.com>
Hey,
Is a single performance hit during boot allowed?
If so, this is a better solution, as it helps on !PREEMPT_RT as well:
https://patchwork.freedesktop.org/series/170759/
The first patch in that series is a better solution than unconditionally disabling
stolen memory, as I from what I understand the problem only exists when media gt is enabled.
If not, we should probably unconditionally disable stolen framebuffer all the time.
But with the series, we preserve the initial framebuffer contents and ensure a smooth
handover, even without stolen.
The warning in 2/2 is probably fine.
Kind regards,
~Maarten Lankhorst
On 11/4/26 04:27, Junxiao Chang wrote:
> CPU access to stolen memory ranges can introduce latency spikes that
> compromise real-time (RT) guarantees on PREEMPT_RT kernels.
>
> On PREEMPT_RT configurations, bypass retrieving the initial plane's
> framebuffer from stolen memory. This forces the driver to fall back to
> standard allocations, preventing RT latency degradation at the expense
> of seamless boot display takeover.
>
> Signed-off-by: Junxiao Chang <junxiao.chang@intel.com>
> ---
> drivers/gpu/drm/xe/display/xe_initial_plane.c | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/drivers/gpu/drm/xe/display/xe_initial_plane.c b/drivers/gpu/drm/xe/display/xe_initial_plane.c
> index 0f86b73036d03..fd0e2962412dd 100644
> --- a/drivers/gpu/drm/xe/display/xe_initial_plane.c
> +++ b/drivers/gpu/drm/xe/display/xe_initial_plane.c
> @@ -87,6 +87,12 @@ initial_plane_bo(struct xe_device *xe,
> struct ttm_resource_manager *stolen;
> u64 pte;
>
> + if (IS_ENABLED(CONFIG_PREEMPT_RT)) {
> + drm_dbg_kms(&xe->drm,
> + "Skipping FB allocation from stolen mem due to PREEMPT_RT\n");
> + return NULL;
> + }
> +
> stolen = ttm_manager_type(&xe->ttm, XE_PL_STOLEN);
> if (!stolen) {
> drm_dbg_kms(&xe->drm, "No stolen for initial FB\n");
next prev parent reply other threads:[~2026-11-03 10:44 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-04 3:27 [PATCH 0/2] drm/xe: drm/xe: Improve real-time performance by limiting stolen memory usage Junxiao Chang
2026-11-04 3:27 ` [PATCH 1/2] drm/xe/display: Skip initial plane allocation from stolen memory on RT Junxiao Chang
2026-11-03 10:45 ` Maarten Lankhorst [this message]
2026-11-04 3:27 ` [PATCH 2/2] drm/xe/gsc: warn about GSC stolen memory usage on RT kernels Junxiao Chang
2026-11-03 11:22 ` Maarten Lankhorst
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=65fae402-e053-4a4f-ba55-6b5b51690eba@lankhorst.se \
--to=dev@lankhorst.se \
--cc=airlied@gmail.com \
--cc=bigeasy@linutronix.de \
--cc=clrkwllms@kernel.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=intel-gfx@lists.freedesktop.org \
--cc=intel-xe@lists.freedesktop.org \
--cc=jani.nikula@linux.intel.com \
--cc=junxiao.chang@intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rt-devel@lists.linux.dev \
--cc=matthew.brost@intel.com \
--cc=rodrigo.vivi@intel.com \
--cc=rostedt@goodmis.org \
--cc=simona@ffwll.ch \
--cc=thomas.hellstrom@linux.intel.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®