From: Matthew Auld <matthew.auld@intel.com>
To: intel-xe@lists.freedesktop.org
Subject: [PATCH v2 0/5] VRAM health check + CCS fix
Date: Wed, 2 Sep 2026 11:39:32 +0100 [thread overview]
Message-ID: <20260902103931.826063-7-matthew.auld@intel.com> (raw)
Add a basic VRAM health check routine, which does some basic sanity
checking on the last VRAM page. This could be extended in the future.
This seems to reproduce the same CCS issue as Linus, on my b570.
This also picks up the fix from Linus. I went with leaving it as-is,
but that means we need to fixup the identity mapping to support 4K
pages for the tail part of usable VRAM.
It doesn't look like anything in our CI actually has the CCS alignment
issue present, so merging the fix from Linus without the identity map
changes likely shouldn't trigger any asserts, if that does end up
coming in via another tree, AFAICT.
v2:
- Give more details around what the VRAM health check actually does
and how it catches the CCS bug (Matt B).
- Tweak the CCS base + alignment print (Matt B).
- Drop the fix from Linus; this is now already in drm-tip.
- Add some hardening for VRAM CPU mapping also.
--
2.55.0
next reply other threads:[~2026-09-02 10:39 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-02 10:39 Matthew Auld [this message]
2026-09-02 10:39 ` [PATCH v2 1/5] drm/xe/migrate: support 4K PTEs for identity map Matthew Auld
2026-09-02 10:39 ` [PATCH v2 2/5] drm/xe/vram: report FLAT_CCS base misalignment Matthew Auld
2026-09-02 10:39 ` [PATCH v2 3/5] drm/xe/vram: revamp CPU VRAM mapping Matthew Auld
2026-09-02 10:49 ` sashiko-bot
2026-09-02 10:56 ` Matthew Auld
2026-09-02 10:39 ` [PATCH v2 4/5] drm/xe: add force option for global invalidation Matthew Auld
2026-09-02 10:39 ` [PATCH v2 5/5] drm/xe/vram: add early VRAM health check Matthew Auld
2026-09-02 10:47 ` ✓ CI.KUnit: success for VRAM health check + CCS fix (rev2) Patchwork
2026-09-02 11:35 ` ✗ Xe.CI.BAT: failure " 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=20260902103931.826063-7-matthew.auld@intel.com \
--to=matthew.auld@intel.com \
--cc=intel-xe@lists.freedesktop.org \
/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;
as well as URLs for NNTP newsgroup(s).