Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Maarten Lankhorst" <dev@lankhorst.se>, intel-xe@lists.freedesktop.org
Subject: Re: [PATCH v4 4/5] HACK: Always fallback
Date: Thu, 05 Nov 2026 16:40:54 +0000	[thread overview]
Message-ID: <sashiko-outbox-199515@kernel.org> (raw)
In-Reply-To: <20261105161444.593259-11-dev@lankhorst.se>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] drm/xe/display: debug return false in xe_display_bo_fbdev_prefer_stolen()
--

commit 6c3a580cc3d1b156a2c5d921989c5c7a936adda3
Author: Maarten Lankhorst <dev@lankhorst.se>

HACK: Always fallback

This commit adds an unconditional return false to
xe_display_bo_fbdev_prefer_stolen() in the xe display driver.

> diff --git a/drivers/gpu/drm/xe/display/xe_display_bo.c b/drivers/gpu/drm/xe/display/xe_display_bo.c
> index 9385026a80b97..d78d893210c4f 100644
> --- a/drivers/gpu/drm/xe/display/xe_display_bo.c
> +++ b/drivers/gpu/drm/xe/display/xe_display_bo.c
> @@ -142,6 +142,8 @@ bool xe_display_bo_fbdev_prefer_stolen(struct xe_device *xe, unsigned int size)
>  	if (XE_DEVICE_WA(xe, 22019338487_display))
>  		return false;
>  
> +	return false;

[Severity: Medium]
As reported by Sashiko on v2 [1], is this temporary debugging commit
intended to be dropped before merging?

Adding an unconditional return false in xe_display_bo_fbdev_prefer_stolen()
in drivers/gpu/drm/xe/display/xe_display_bo.c makes the subsequent stolen
memory size check unreachable dead code and renders all earlier checks and
the size parameter unused:

drivers/gpu/drm/xe/display/xe_display_bo.c:xe_display_bo_fbdev_prefer_stolen() {
    ...
	if (XE_DEVICE_WA(xe, 22019338487_display))
		return false;

	return false;

	/*
	 * If the FB is too big, just don't use it since fbdev is not very
	 * important and we should probably use that space with FBC or other
	 * features.
	 */
	return stolen->size >= (size * 2) >> PAGE_SHIFT;
}

During display initialization on integrated platforms with
CONFIG_FRAMEBUFFER_CONSOLE and CONFIG_DRM_FBDEV_EMULATION enabled,
initial_plane_bo() in drivers/gpu/drm/xe/display/xe_initial_plane.c calls
xe_display_bo_fbdev_prefer_stolen():

drivers/gpu/drm/xe/display/xe_initial_plane.c:initial_plane_bo() {
    ...
		if (IS_ENABLED(CONFIG_FRAMEBUFFER_CONSOLE) &&
		    IS_ENABLED(CONFIG_DRM_FBDEV_EMULATION) &&
		    !xe_display_bo_fbdev_prefer_stolen(xe, plane_config->size)) {
			drm_info(&xe->drm, "Initial FB size exceeds half of stolen, discarding\n");
			return NULL;
		}
    ...
}

Because xe_display_bo_fbdev_prefer_stolen() now unconditionally returns
false, initial_plane_bo() always discards the initial BIOS framebuffer with
a misleading log message, and even with the subsequent takeover helper
xe_display_bo_fbdev_bios_fb_takeover() it will always fall back to
allocating a new system memory BO and blit-copying the BIOS framebuffer
instead of keeping it in stolen memory when eligible.

[1] https://lore.kernel.org/r/20260721142610.93804-11-dev@lankhorst.se

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261105161444.593259-7-dev@lankhorst.se?part=4

  reply	other threads:[~2026-11-05 16:41 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-05 16:14 [PATCH v4 0/5] drm/xe/display: Transparant fallback from stolen to sysmem Maarten Lankhorst
2026-11-05 16:14 ` [PATCH v4 1/5] drm/xe/display: Avoid using stolen memory for framebuffer when media gt exists Maarten Lankhorst
2026-11-05 16:14 ` [PATCH v4 2/5] drm/xe/migrate: Support copying between sysmem and stolen Maarten Lankhorst
2026-11-05 16:32   ` sashiko-bot
2026-11-05 16:14 ` [PATCH v4 3/5] drm/xe: Raise gt frequency slightly earlier Maarten Lankhorst
2026-11-05 16:14 ` [PATCH v4 4/5] HACK: Always fallback Maarten Lankhorst
2026-11-05 16:40   ` sashiko-bot [this message]
2026-11-05 16:14 ` [PATCH v4 5/5] drm/i915: Introduce intel_bo_fbdev_bios_fb_takeover() Maarten Lankhorst
2026-11-05 17:04   ` sashiko-bot
2026-11-05 18:12 ` ✗ CI.checkpatch: warning for drm/xe/display: Transparant fallback from stolen to sysmem. (rev4) Patchwork
2026-11-05 18:14 ` ✓ CI.KUnit: success " Patchwork
2026-11-05 18:48 ` ✗ Xe.CI.BAT: failure " Patchwork
2026-11-06 19:21 ` ✗ 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=sashiko-outbox-199515@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=dev@lankhorst.se \
    --cc=intel-xe@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox