From: "Chauhan, Shekhar" <shekhar.chauhan@intel.com>
To: Matt Atwood <matthew.s.atwood@intel.com>,
<intel-xe@lists.freedesktop.org>
Cc: Gustavo Sousa <gustavo.sousa@intel.com>
Subject: Re: [PATCH 3/6] drm/xe/xe3: Add initial set of workarounds
Date: Tue, 8 Oct 2024 19:16:54 +0530 [thread overview]
Message-ID: <d8202760-57d3-4cd3-8e9a-deee3d4c5b3f@intel.com> (raw)
In-Reply-To: <20241008013509.61233-4-matthew.s.atwood@intel.com>
On 10/8/2024 7:05, Matt Atwood wrote:
> From: Gustavo Sousa <gustavo.sousa@intel.com>
>
> Implement the initial set of workarounds for Xe3 IPs.
>
> v2: correct steppings for 14021490052, 22019338487
>
> Signed-off-by: Gustavo Sousa <gustavo.sousa@intel.com>
> Signed-off-by: Matt Atwood <matthew.s.atwood@intel.com>
> ---
> drivers/gpu/drm/xe/regs/xe_engine_regs.h | 1 +
> drivers/gpu/drm/xe/regs/xe_gt_regs.h | 3 ++
> drivers/gpu/drm/xe/xe_wa.c | 52 ++++++++++++++++++++++--
> drivers/gpu/drm/xe/xe_wa_oob.rules | 1 +
> 4 files changed, 54 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/gpu/drm/xe/regs/xe_engine_regs.h b/drivers/gpu/drm/xe/regs/xe_engine_regs.h
> index 81b71903675e..7c78496e6213 100644
> --- a/drivers/gpu/drm/xe/regs/xe_engine_regs.h
> +++ b/drivers/gpu/drm/xe/regs/xe_engine_regs.h
> @@ -186,6 +186,7 @@
>
> #define VDBOX_CGCTL3F10(base) XE_REG((base) + 0x3f10)
> #define IECPUNIT_CLKGATE_DIS REG_BIT(22)
> +#define RAMDFTUNIT_CLKGATE_DIS REG_BIT(9)
>
> #define VDBOX_CGCTL3F18(base) XE_REG((base) + 0x3f18)
> #define ALNUNIT_CLKGATE_DIS REG_BIT(13)
> diff --git a/drivers/gpu/drm/xe/regs/xe_gt_regs.h b/drivers/gpu/drm/xe/regs/xe_gt_regs.h
> index fb80042cbe0d..ac5a9ae94eb1 100644
> --- a/drivers/gpu/drm/xe/regs/xe_gt_regs.h
> +++ b/drivers/gpu/drm/xe/regs/xe_gt_regs.h
> @@ -286,6 +286,9 @@
> #define GAMTLBVEBOX0_CLKGATE_DIS REG_BIT(16)
> #define LTCDD_CLKGATE_DIS REG_BIT(10)
>
> +#define UNSLCGCTL9454 XE_REG(0x9454)
> +#define LSCFE_CLKGATE_DIS REG_BIT(4)
> +
> #define XEHP_SLICE_UNIT_LEVEL_CLKGATE XE_REG_MCR(0x94d4)
> #define L3_CR2X_CLKGATE_DIS REG_BIT(17)
> #define L3_CLKGATE_DIS REG_BIT(16)
> diff --git a/drivers/gpu/drm/xe/xe_wa.c b/drivers/gpu/drm/xe/xe_wa.c
> index 94ea76b098ed..a47f43166651 100644
> --- a/drivers/gpu/drm/xe/xe_wa.c
> +++ b/drivers/gpu/drm/xe/xe_wa.c
> @@ -252,6 +252,34 @@ static const struct xe_rtp_entry_sr gt_was[] = {
> XE_RTP_ENTRY_FLAG(FOREACH_ENGINE),
> },
>
> + /* Xe3_LPG */
> +
> + { XE_RTP_NAME("14021871409"),
> + XE_RTP_RULES(GRAPHICS_VERSION(3000), GRAPHICS_STEP(A0, B0)),
> + XE_RTP_ACTIONS(SET(UNSLCGCTL9454, LSCFE_CLKGATE_DIS))
> + },
> +
> + /* Xe3_LPM */
> +
> + { XE_RTP_NAME("16021867713"),
> + XE_RTP_RULES(MEDIA_VERSION(3000),
> + ENGINE_CLASS(VIDEO_DECODE)),
> + XE_RTP_ACTIONS(SET(VDBOX_CGCTL3F1C(0), MFXPIPE_CLKGATE_DIS)),
> + XE_RTP_ENTRY_FLAG(FOREACH_ENGINE),
> + },
> + { XE_RTP_NAME("16021865536"),
> + XE_RTP_RULES(MEDIA_VERSION(3000),
> + ENGINE_CLASS(VIDEO_DECODE)),
> + XE_RTP_ACTIONS(SET(VDBOX_CGCTL3F10(0), IECPUNIT_CLKGATE_DIS)),
> + XE_RTP_ENTRY_FLAG(FOREACH_ENGINE),
> + },
> + { XE_RTP_NAME("14021486841"),
> + XE_RTP_RULES(MEDIA_VERSION(3000), MEDIA_STEP(A0, B0),
> + ENGINE_CLASS(VIDEO_DECODE)),
> + XE_RTP_ACTIONS(SET(VDBOX_CGCTL3F10(0), RAMDFTUNIT_CLKGATE_DIS)),
> + XE_RTP_ENTRY_FLAG(FOREACH_ENGINE),
> + },
> +
> {}
> };
>
> @@ -568,6 +596,13 @@ static const struct xe_rtp_entry_sr engine_was[] = {
> XE_RTP_ACTION_FLAG(ENGINE_BASE)))
> },
>
> + /* Xe3_LPG */
> +
> + { XE_RTP_NAME("14021402888"),
> + XE_RTP_RULES(GRAPHICS_VERSION_RANGE(3000, 3001), FUNC(xe_rtp_match_first_render_or_compute)),
> + XE_RTP_ACTIONS(SET(HALF_SLICE_CHICKEN7, CLEAR_OPTIMIZATION_DISABLE))
> + },
> +
> {}
> };
>
> @@ -656,8 +691,7 @@ static const struct xe_rtp_entry_sr lrc_was[] = {
> /* Xe2_LPG */
>
> { XE_RTP_NAME("16020518922"),
> - XE_RTP_RULES(GRAPHICS_VERSION(2004), GRAPHICS_STEP(A0, B0),
> - ENGINE_CLASS(RENDER)),
> + XE_RTP_RULES(GRAPHICS_VERSION(2004), ENGINE_CLASS(RENDER)),
I think you got confused between 16020518922 and 14021490052 (right
below). 16020518922 applies to steppings A0 to B0 and 14021490052
applies to all the steppings. Having said that, I believe we shouldn't
be touching workarounds targeting GRAPHICS_VERSION(2004) at all, since
the patch stands for Xe3. The original code was correct, maybe some
confusion resulted in this, if you could revert these two w/a changes,
the patch looks good to go.
> XE_RTP_ACTIONS(SET(FF_MODE,
> DIS_TE_AUTOSTRIP |
> DIS_MESH_PARTIAL_AUTOSTRIP |
> @@ -703,7 +737,7 @@ static const struct xe_rtp_entry_sr lrc_was[] = {
> XE_RTP_ACTIONS(SET(WM_CHICKEN3, HIZ_PLANE_COMPRESSION_DIS))
> },
> { XE_RTP_NAME("14021490052"),
> - XE_RTP_RULES(GRAPHICS_VERSION(2004), ENGINE_CLASS(RENDER)),
> + XE_RTP_RULES(GRAPHICS_VERSION(2004), ENGINE_CLASS(RENDER), GRAPHICS_STEP(A0, B0)),
> XE_RTP_ACTIONS(SET(FF_MODE,
> DIS_MESH_PARTIAL_AUTOSTRIP |
> DIS_MESH_AUTOSTRIP),
> @@ -739,6 +773,18 @@ static const struct xe_rtp_entry_sr lrc_was[] = {
> XE_RTP_ACTIONS(SET(CHICKEN_RASTER_1, DIS_CLIP_NEGATIVE_BOUNDING_BOX))
> },
>
> + /* Xe3_LPG */
> + { XE_RTP_NAME("14021490052"),
> + XE_RTP_RULES(GRAPHICS_VERSION(3000), GRAPHICS_STEP(A0, B0),
> + ENGINE_CLASS(RENDER)),
> + XE_RTP_ACTIONS(SET(FF_MODE,
> + DIS_MESH_PARTIAL_AUTOSTRIP |
> + DIS_MESH_AUTOSTRIP),
> + SET(VFLSKPD,
> + DIS_PARTIAL_AUTOSTRIP |
> + DIS_AUTOSTRIP))
> + },
> +
> {}
> };
>
> diff --git a/drivers/gpu/drm/xe/xe_wa_oob.rules b/drivers/gpu/drm/xe/xe_wa_oob.rules
> index 920ca5060146..51d7208d7681 100644
> --- a/drivers/gpu/drm/xe/xe_wa_oob.rules
> +++ b/drivers/gpu/drm/xe/xe_wa_oob.rules
> @@ -33,6 +33,7 @@
> GRAPHICS_VERSION(2004)
> 22019338487 MEDIA_VERSION(2000)
> GRAPHICS_VERSION(2001)
> + MEDIA_VERSION(3000), GRAPHICS_STEP(A0, B0)
> 22019338487_display PLATFORM(LUNARLAKE)
> 16023588340 GRAPHICS_VERSION(2001)
> 14019789679 GRAPHICS_VERSION(1255)
--
-shekhar
next prev parent reply other threads:[~2024-10-08 13:47 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-08 1:35 [PATCH 0/6 v3] Add Xe3 and Panther Lake support Matt Atwood
2024-10-08 1:35 ` [PATCH 1/6] drm/xe/xe3: Xe3 uses the same PAT settings as Xe2 Matt Atwood
2024-10-08 1:35 ` [PATCH 2/6] drm/xe/xe3: Define Xe3 feature flags Matt Atwood
2024-10-08 2:23 ` Chauhan, Shekhar
2024-10-08 16:15 ` Matt Roper
2024-10-08 17:49 ` Chauhan, Shekhar
2024-10-08 18:06 ` Matt Roper
2024-10-08 18:28 ` Chauhan, Shekhar
2024-10-08 1:35 ` [PATCH 3/6] drm/xe/xe3: Add initial set of workarounds Matt Atwood
2024-10-08 13:46 ` Chauhan, Shekhar [this message]
2024-10-08 16:17 ` Matt Roper
2024-10-08 1:35 ` [PATCH 4/6] drm/xe/ptl: PTL re-uses Xe2 MOCS table Matt Atwood
2024-10-08 13:19 ` Chauhan, Shekhar
2024-10-08 1:35 ` [PATCH 5/6] drm/xe/ptl: Add PTL platform definition Matt Atwood
2024-10-08 13:25 ` Chauhan, Shekhar
2024-10-08 1:35 ` [PATCH 6/6] drm/xe/xe3lpm: Add new "instance0" steering table Matt Atwood
2024-10-08 1:40 ` ✓ CI.Patch_applied: success for Add Xe3 and Panther Lake support (rev3) Patchwork
2024-10-08 1:40 ` ✗ CI.checkpatch: warning " Patchwork
2024-10-08 1:41 ` ✓ CI.KUnit: success " Patchwork
2024-10-08 1:53 ` ✓ CI.Build: " Patchwork
2024-10-08 1:56 ` ✓ CI.Hooks: " Patchwork
2024-10-08 1:58 ` ✓ CI.checksparse: " Patchwork
2024-10-08 2:23 ` ✓ CI.BAT: " Patchwork
2024-10-08 10:01 ` ✗ CI.FULL: failure " Patchwork
2024-10-08 16:26 ` Matt Roper
2024-10-08 16:08 ` [PATCH 0/6 v3] Add Xe3 and Panther Lake support Matt Roper
-- strict thread matches above, loose matches on Subject: below --
2024-10-04 22:05 [PATCH 0/6] " Matt Atwood
2024-10-04 22:05 ` [PATCH 3/6] drm/xe/xe3: Add initial set of workarounds Matt Atwood
2024-10-04 22:40 ` Matt Roper
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=d8202760-57d3-4cd3-8e9a-deee3d4c5b3f@intel.com \
--to=shekhar.chauhan@intel.com \
--cc=gustavo.sousa@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=matthew.s.atwood@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