From: Varun Gupta <varun.gupta@intel.com>
To: intel-xe@lists.freedesktop.org
Cc: matthew.brost@intel.com, matthew.d.roper@intel.com,
himal.prasad.ghimiray@intel.com, priyanka.dandamudi@intel.com
Subject: [PATCH v3 0/2] Add prefetch fault support for Xe3p
Date: Wed, 28 Jan 2026 14:41:51 +0530 [thread overview]
Message-ID: <20260128091153.25543-1-varun.gupta@intel.com> (raw)
Xe3p hardware prefetches memory ranges and notifies software via an
additional bit (bit 11) in the page fault descriptor that the fault
was caused by prefetch.
Patch 1 adds a dedicated stats counter to track invalid prefetch
pagefaults separately from regular faults.
Patch 2 implements the prefetch support by extracting the prefetch bit
from the fault descriptor and echoing it in the response (bit 6) only
when handling fails. This prevents CAT errors for prefetch faults.
Stats are incremented for prefetch errors without verbose logging to
avoid log spam. The prefetch bit value is included in error messages
to distinguish prefetch faults from regular faults.
Based on original patches by Brian Welty and Priyanka Dandamudi.
v3:
- Split stats counter into separate patch (Matt Brost)
- Drop xe_guc_pagefault_types.h rename patch (Matt Brost)
- Move prefetch check to caller of xe_pagefault_print (Matt Brost)
- Rename XE3P_PFD_PREFETCH/PFR_PREFETCH to remove XE3P_ prefix,
add comments about Xe3+ platform support (Matt Brost)
- Rename PREFETCH_PAGEFAULT_COUNT to INVALID_PREFETCH_PAGEFAULT_COUNT
for clarity (Matt Brost)
- Include prefetch bit value in error message (Matt Brost)
v2:
- Initial implementation combining all changes
Varun Gupta (2):
drm/xe: Add counter for invalid prefetch pagefaults
drm/xe: Add prefetch fault support for Xe3p
drivers/gpu/drm/xe/xe_gt_stats.c | 1 +
drivers/gpu/drm/xe/xe_gt_stats_types.h | 1 +
drivers/gpu/drm/xe/xe_guc_fwif.h | 5 +++--
drivers/gpu/drm/xe/xe_guc_pagefault.c | 2 ++
drivers/gpu/drm/xe/xe_pagefault.c | 16 +++++++++++++---
drivers/gpu/drm/xe/xe_pagefault_types.h | 8 +++++++-
6 files changed, 27 insertions(+), 6 deletions(-)
--
2.43.0
next reply other threads:[~2026-01-28 9:12 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-01-28 9:11 Varun Gupta [this message]
2026-01-28 9:11 ` [PATCH v3 1/2] drm/xe: Add counter for invalid prefetch pagefaults Varun Gupta
2026-01-28 9:11 ` [PATCH v3 2/2] drm/xe: Add prefetch fault support for Xe3p Varun Gupta
2026-01-28 16:43 ` ✗ CI.checkpatch: warning for " Patchwork
2026-01-28 16:44 ` ✓ CI.KUnit: success " Patchwork
2026-01-28 17:18 ` ✓ Xe.CI.BAT: " Patchwork
2026-02-02 3:48 ` [PATCH v3 0/2] " Gupta, Varun
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=20260128091153.25543-1-varun.gupta@intel.com \
--to=varun.gupta@intel.com \
--cc=himal.prasad.ghimiray@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=matthew.brost@intel.com \
--cc=matthew.d.roper@intel.com \
--cc=priyanka.dandamudi@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