From: "Summers, Stuart" <stuart.summers@intel.com>
To: "intel-xe@lists.freedesktop.org" <intel-xe@lists.freedesktop.org>,
"Brost, Matthew" <matthew.brost@intel.com>
Subject: Re: [PATCH v4 00/12] Context based TLB invalidations
Date: Fri, 16 Jan 2026 00:25:56 +0000 [thread overview]
Message-ID: <5f976f6fa03d6e508849593a936281394735af23.camel@intel.com> (raw)
In-Reply-To: <20260113025232.3504648-1-matthew.brost@intel.com>
Ok at least today, this definitely is going to need to be included with
[1]. Without that, I'm seeing DMAR faults and various other issues. But
with [1], I haven't seen any issues yet around the
registration/deregistration, so I think the ref count model here is
sound (as discussed inline of course as well).
So:
Tested-by: Stuart Summers <stuart.summers@intel.com>
The only concern I have at this point is around the ring invalidations.
Maybe we can discuss more in that review, but I'm a little hessitant to
put a full R-B on this series unless we also include that other one or
fixes that you had suggested (which needs more debug still).
Thanks,
Stuart
[1]: https://patchwork.freedesktop.org/series/160057/
On Mon, 2026-01-12 at 18:52 -0800, Matthew Brost wrote:
> Add support for context based TLB invalidations.
>
> Matt
>
> Matthew Brost (12):
> drm/xe: Add normalize_invalidation_range
> drm/xe: Make usm.asid_to_vm allocation use GFP_NOWAIT
> drm/xe: Add has_ctx_tlb_inval to device info
> drm/xe: Add xe_device_asid_to_vm helper
> drm/xe: Add vm to exec queues association
> drm/xe: Taint TLB invalidation seqno lock with GFP_KERNEL
> drm/xe: Rename send_tlb_inval_ppgtt to send_tlb_inval_asid_ppgtt
> drm/xe: Add send_tlb_inval_ppgtt helper
> drm/xe: Add xe_tlb_inval_idle helper
> drm/xe: Add exec queue active vfunc
> drm/xe: Add context-based invalidation to GuC TLB invalidation
> backend
> drm/xe: Enable context TLB invalidations for CI
>
> drivers/gpu/drm/xe/xe_device.c | 25 +++
> drivers/gpu/drm/xe/xe_device.h | 11 +-
> drivers/gpu/drm/xe/xe_device_types.h | 9 +
> drivers/gpu/drm/xe/xe_exec_queue.c | 7 +-
> drivers/gpu/drm/xe/xe_exec_queue_types.h | 5 +
> drivers/gpu/drm/xe/xe_execlist.c | 7 +
> drivers/gpu/drm/xe/xe_guc_submit.c | 6 +
> drivers/gpu/drm/xe/xe_guc_tlb_inval.c | 238 ++++++++++++++++++---
> --
> drivers/gpu/drm/xe/xe_pci.c | 2 +
> drivers/gpu/drm/xe/xe_pci_types.h | 1 +
> drivers/gpu/drm/xe/xe_tlb_inval.c | 33 ++++
> drivers/gpu/drm/xe/xe_tlb_inval.h | 2 +
> drivers/gpu/drm/xe/xe_vm.c | 64 +++++-
> drivers/gpu/drm/xe/xe_vm.h | 3 +
> drivers/gpu/drm/xe/xe_vm_types.h | 16 ++
> 15 files changed, 375 insertions(+), 54 deletions(-)
>
next prev parent reply other threads:[~2026-01-16 0:26 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-01-13 2:52 [PATCH v4 00/12] Context based TLB invalidations Matthew Brost
2026-01-13 2:52 ` [PATCH v4 01/12] drm/xe: Add normalize_invalidation_range Matthew Brost
2026-01-13 2:52 ` [PATCH v4 02/12] drm/xe: Make usm.asid_to_vm allocation use GFP_NOWAIT Matthew Brost
2026-01-13 2:52 ` [PATCH v4 03/12] drm/xe: Add has_ctx_tlb_inval to device info Matthew Brost
2026-01-13 21:24 ` Summers, Stuart
2026-01-13 2:52 ` [PATCH v4 04/12] drm/xe: Add xe_device_asid_to_vm helper Matthew Brost
2026-01-13 2:52 ` [PATCH v4 05/12] drm/xe: Add vm to exec queues association Matthew Brost
2026-01-13 21:26 ` Summers, Stuart
2026-01-13 2:52 ` [PATCH v4 06/12] drm/xe: Taint TLB invalidation seqno lock with GFP_KERNEL Matthew Brost
2026-01-13 2:52 ` [PATCH v4 07/12] drm/xe: Rename send_tlb_inval_ppgtt to send_tlb_inval_asid_ppgtt Matthew Brost
2026-01-13 2:52 ` [PATCH v4 08/12] drm/xe: Add send_tlb_inval_ppgtt helper Matthew Brost
2026-01-13 2:52 ` [PATCH v4 09/12] drm/xe: Add xe_tlb_inval_idle helper Matthew Brost
2026-01-13 2:52 ` [PATCH v4 10/12] drm/xe: Add exec queue active vfunc Matthew Brost
2026-01-13 21:32 ` Summers, Stuart
2026-01-13 22:16 ` Matthew Brost
2026-01-13 22:21 ` Summers, Stuart
2026-01-14 21:17 ` Summers, Stuart
2026-01-13 2:52 ` [PATCH v4 11/12] drm/xe: Add context-based invalidation to GuC TLB invalidation backend Matthew Brost
2026-01-13 21:23 ` Summers, Stuart
2026-01-13 22:08 ` Matthew Brost
2026-01-13 23:25 ` Summers, Stuart
2026-01-14 0:25 ` Matthew Brost
2026-01-13 2:52 ` [PATCH v4 12/12] drm/xe: Enable context TLB invalidations for CI Matthew Brost
2026-01-13 3:00 ` ✓ CI.KUnit: success for Context based TLB invalidations (rev4) Patchwork
2026-01-13 3:41 ` ✓ Xe.CI.BAT: " Patchwork
2026-01-13 9:45 ` ✓ Xe.CI.Full: " Patchwork
2026-01-16 0:25 ` Summers, Stuart [this message]
2026-01-16 1:52 ` [PATCH v4 00/12] Context based TLB invalidations Matthew Brost
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=5f976f6fa03d6e508849593a936281394735af23.camel@intel.com \
--to=stuart.summers@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=matthew.brost@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