All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Dixit, Ashutosh" <ashutosh.dixit@intel.com>
To: Matt Roper <matthew.d.roper@intel.com>
Cc: <intel-xe@lists.freedesktop.org>,
	Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Subject: Re: [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728
Date: Mon, 27 Apr 2026 15:13:31 -0700	[thread overview]
Message-ID: <87jytsjblg.wl-ashutosh.dixit@intel.com> (raw)
In-Reply-To: <20260427205241.GK2131374@mdroper-desk1.amr.corp.intel.com>

On Mon, 27 Apr 2026 13:52:41 -0700, Matt Roper wrote:
>

Hi Matt,

> On Mon, Apr 27, 2026 at 01:26:55PM -0700, Ashutosh Dixit wrote:
> > CRI Wa_14026633728 requires MERTOA buffer to be allocated in device memory,
> > not system memory (which is the default for OA buffers).
> >
> > Signed-off-by: Ashutosh Dixit <ashutosh.dixit@intel.com>
> > Reviewed-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
> > ---
> >  drivers/gpu/drm/xe/xe_oa.c         | 10 +++++++++-
> >  drivers/gpu/drm/xe/xe_wa_oob.rules |  1 +
> >  2 files changed, 10 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/gpu/drm/xe/xe_oa.c b/drivers/gpu/drm/xe/xe_oa.c
> > index 83797c9479f7c..d905302f3f151 100644
> > --- a/drivers/gpu/drm/xe/xe_oa.c
> > +++ b/drivers/gpu/drm/xe/xe_oa.c
> > @@ -250,6 +250,11 @@ static void oa_timestamp_clear(struct xe_oa_stream *stream, u32 report_offset)
> >		xe_map_wr(stream->oa->xe, map, report_offset + 4, u32, 0);
> >  }
> >
> > +static bool mert_wa_14026633728(struct xe_oa_stream *s)
> > +{
> > +	return s->oa_unit->type == DRM_XE_OA_UNIT_TYPE_MERT && XE_GT_WA(s->gt, 14026633728);
> > +}
> > +
> >  static bool xe_oa_buffer_check_unlocked(struct xe_oa_stream *stream)
> >  {
> >	u32 gtt_offset = xe_bo_ggtt_addr(stream->oa_buffer.bo);
> > @@ -893,11 +898,14 @@ static void xe_oa_stream_destroy(struct xe_oa_stream *stream)
> >
> >  static int xe_oa_alloc_oa_buffer(struct xe_oa_stream *stream, size_t size)
> >  {
> > +	u32 vram = mert_wa_14026633728(stream) ?
> > +		XE_BO_FLAG_VRAM_IF_DGFX(xe_device_get_root_tile(stream->oa->xe)) :
> > +		XE_BO_FLAG_SYSTEM;
> >	struct xe_bo *bo;
> >
> >	bo = xe_bo_create_pin_map_novm(stream->oa->xe, stream->gt->tile,
> >				       size, ttm_bo_type_kernel,
> > -				       XE_BO_FLAG_SYSTEM | XE_BO_FLAG_GGTT, false);
> > +				       vram | XE_BO_FLAG_GGTT, false);
> >	if (IS_ERR(bo))
> >		return PTR_ERR(bo);
> >
> > diff --git a/drivers/gpu/drm/xe/xe_wa_oob.rules b/drivers/gpu/drm/xe/xe_wa_oob.rules
> > index f8a185103b805..a7c1bd9bcb943 100644
> > --- a/drivers/gpu/drm/xe/xe_wa_oob.rules
> > +++ b/drivers/gpu/drm/xe/xe_wa_oob.rules
> > @@ -65,3 +65,4 @@
> >
> >  14025883347	MEDIA_VERSION_RANGE(1301, 3503)
> >		GRAPHICS_VERSION_RANGE(2004, 3005)
> > +14026633728	PLATFORM(CRESCENTISLAND)
>
> Is the underlying defect here tied to the graphics IP or the
> platform/SoC? If graphics IP, this should key off the graphics version
> instead of platform).  If platform, we should treat it as a device
> workaround rather than a GT workaround. I.e., definition in
> xe_device_wa_oob.rules and use XE_DEVICE_WA() at the check sites.

Because the WA is in MERT I'd think it would be a device WA. So I've
changed it to a device WA as suggested here in v8.

Thanks.
--
Ashutosh

  reply	other threads:[~2026-04-27 22:13 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-04-27 20:26 [PATCH v7 0/3] drm/xe/oa: Wa_14026633728 Ashutosh Dixit
2026-04-27 20:26 ` [PATCH 1/3] drm/xe/oa: Use xe_map layer Ashutosh Dixit
2026-04-27 20:26 ` [PATCH 2/3] drm/xe/oa: Use drm_gem_mmap_obj for OA buffer mmap Ashutosh Dixit
2026-04-27 20:26 ` [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728 Ashutosh Dixit
2026-04-27 20:52   ` Matt Roper
2026-04-27 22:13     ` Dixit, Ashutosh [this message]
2026-04-27 21:28 ` ✓ CI.KUnit: success for drm/xe/oa: Wa_14026633728 (rev7) Patchwork
2026-04-27 23:02 ` ✓ Xe.CI.BAT: " Patchwork
  -- strict thread matches above, loose matches on Subject: below --
2026-04-27 22:11 [PATCH v8 0/3] drm/xe/oa: Wa_14026633728 Ashutosh Dixit
2026-04-27 22:11 ` [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728 Ashutosh Dixit
2026-04-27 19:02 [PATCH v6 0/3] drm/xe/oa: Wa_14026633728 Ashutosh Dixit
2026-04-27 19:02 ` [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728 Ashutosh Dixit
2026-04-25  0:14 [PATCH v5 0/3] drm/xe/oa: Wa_14026633728 Ashutosh Dixit
2026-04-25  0:14 ` [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728 Ashutosh Dixit
2026-04-15  2:03 [PATCH v4 0/3] drm/xe/oa: Wa_14026633728 Ashutosh Dixit
2026-04-15  2:03 ` [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728 Ashutosh Dixit
2026-04-11  0:48 [PATCH v3 0/3] drm/xe/oa: Wa_14026633728 Ashutosh Dixit
2026-04-11  0:48 ` [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728 Ashutosh Dixit
2026-04-09 23:17 [PATCH v2 0/3] drm/xe/oa: Wa_14026633728 Ashutosh Dixit
2026-04-09 23:17 ` [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728 Ashutosh Dixit
2026-04-07  3:02 [PATCH 0/3] drm/xe/oa: Wa_14026633728 Ashutosh Dixit
2026-04-07  3:02 ` [PATCH 3/3] drm/xe/oa: Implement Wa_14026633728 Ashutosh Dixit
2026-04-07 23:17   ` Umesh Nerlige Ramappa

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=87jytsjblg.wl-ashutosh.dixit@intel.com \
    --to=ashutosh.dixit@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=matthew.d.roper@intel.com \
    --cc=umesh.nerlige.ramappa@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.