Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Lucas De Marchi <lucas.demarchi@intel.com>
To: intel-xe@lists.freedesktop.org
Cc: Lucas De Marchi <lucas.demarchi@intel.com>,
	Shekhar Chauhan <shekhar.chauhan@intel.com>,
	Balasubramani Vivekanandan <balasubramani.vivekanandan@intel.com>,
	Matt Roper <matthew.d.roper@intel.com>,
	Tejas Upadhyay <tejas.upadhyay@intel.com>,
	Gustavo Sousa <gustavo.sousa@intel.com>
Subject: [PATCH v3 08/24] drm/xe/xe3p: Determine service copy availability from fuse
Date: Thu, 16 Oct 2025 19:26:27 -0700	[thread overview]
Message-ID: <20251016-xe3p-v3-8-3dd173a3097a@intel.com> (raw)
In-Reply-To: <20251016-xe3p-v3-0-3dd173a3097a@intel.com>

From: Matt Roper <matthew.d.roper@intel.com>

Xe3p introduces a dedicated SERVICE_COPY_ENABLE fuse register to reflect
the availability of the service copy engines (BCS1-BCS8).

Bspec: 74624
Signed-off-by: Matt Roper <matthew.d.roper@intel.com>
Reviewed-by: Gustavo Sousa <gustavo.sousa@intel.com>
Signed-off-by: Lucas De Marchi <lucas.demarchi@intel.com>
---
 drivers/gpu/drm/xe/regs/xe_gt_regs.h |  3 +++
 drivers/gpu/drm/xe/xe_hw_engine.c    | 43 ++++++++++++++++++++++++++++--------
 2 files changed, 37 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/xe/regs/xe_gt_regs.h b/drivers/gpu/drm/xe/regs/xe_gt_regs.h
index 21c15441c4537..228de47c0f3f5 100644
--- a/drivers/gpu/drm/xe/regs/xe_gt_regs.h
+++ b/drivers/gpu/drm/xe/regs/xe_gt_regs.h
@@ -239,6 +239,9 @@
 #define XE2_GT_GEOMETRY_DSS_1			XE_REG(0x9150)
 #define XE2_GT_GEOMETRY_DSS_2			XE_REG(0x9154)
 
+#define SERVICE_COPY_ENABLE			XE_REG(0x9170)
+#define   FUSE_SERVICE_COPY_ENABLE_MASK		REG_GENMASK(7, 0)
+
 #define GDRST					XE_REG(0x941c)
 #define   GRDOM_GUC				REG_BIT(3)
 #define   GRDOM_FULL				REG_BIT(0)
diff --git a/drivers/gpu/drm/xe/xe_hw_engine.c b/drivers/gpu/drm/xe/xe_hw_engine.c
index b08a6d42c8ffc..073ecd263e543 100644
--- a/drivers/gpu/drm/xe/xe_hw_engine.c
+++ b/drivers/gpu/drm/xe/xe_hw_engine.c
@@ -718,27 +718,52 @@ static void read_media_fuses(struct xe_gt *gt)
 	}
 }
 
+static u32 infer_svccopy_from_meml3(struct xe_gt *gt)
+{
+	u32 meml3 = REG_FIELD_GET(MEML3_EN_MASK,
+				  xe_mmio_read32(&gt->mmio, MIRROR_FUSE3));
+	u32 svccopy_mask = 0;
+
+	/*
+	 * Each of the four meml3 bits determines the fusing of two service
+	 * copy engines.
+	 */
+	for (int i = 0; i < 4; i++)
+		svccopy_mask |= (meml3 & BIT(i)) ? 0b11 << 2 * i : 0;
+
+	return svccopy_mask;
+}
+
+static u32 read_svccopy_fuses(struct xe_gt *gt)
+{
+	return REG_FIELD_GET(FUSE_SERVICE_COPY_ENABLE_MASK,
+			     xe_mmio_read32(&gt->mmio, SERVICE_COPY_ENABLE));
+}
+
 static void read_copy_fuses(struct xe_gt *gt)
 {
 	struct xe_device *xe = gt_to_xe(gt);
 	u32 bcs_mask;
 
-	if (GRAPHICS_VERx100(xe) < 1260 || GRAPHICS_VERx100(xe) >= 1270)
-		return;
-
 	xe_force_wake_assert_held(gt_to_fw(gt), XE_FW_GT);
 
-	bcs_mask = xe_mmio_read32(&gt->mmio, MIRROR_FUSE3);
-	bcs_mask = REG_FIELD_GET(MEML3_EN_MASK, bcs_mask);
+	if (GRAPHICS_VER(xe) >= 35)
+		bcs_mask = read_svccopy_fuses(gt);
+	else if (GRAPHICS_VERx100(xe) == 1260)
+		bcs_mask = infer_svccopy_from_meml3(gt);
+	else
+		return;
 
-	/* BCS0 is always present; only BCS1-BCS8 may be fused off */
-	for (int i = XE_HW_ENGINE_BCS1, j = 0; i <= XE_HW_ENGINE_BCS8; ++i, ++j) {
+	/* Only BCS1-BCS8 may be fused off */
+	bcs_mask <<= XE_HW_ENGINE_BCS1;
+	for (int i = XE_HW_ENGINE_BCS1; i <= XE_HW_ENGINE_BCS8; ++i) {
 		if (!(gt->info.engine_mask & BIT(i)))
 			continue;
 
-		if (!(BIT(j / 2) & bcs_mask)) {
+		if (!(bcs_mask & BIT(i))) {
 			gt->info.engine_mask &= ~BIT(i);
-			xe_gt_info(gt, "bcs%u fused off\n", j);
+			xe_gt_info(gt, "bcs%u fused off\n",
+				   i - XE_HW_ENGINE_BCS0);
 		}
 	}
 }

-- 
2.51.0


  parent reply	other threads:[~2025-10-17  2:27 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-10-17  2:26 [PATCH v3 00/24] drm/xe: Add Xe3p support Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 01/24] drm/xe/xe3: Add support for graphics IP versions 30.04 & 30.05 Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 02/24] drm/xe/xe3p: Add support for media IP versions 35.00 & 35.03 Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 03/24] drm/xe: Drop CTC_MODE register read Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 04/24] drm/xe: Add GT_VER() to check version specific to gt type Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 05/24] drm/xe/xe3p_lpm: Skip disabling NOA on unsupported IPs Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 06/24] drm/xe/xe3p_lpm: Handle MCR steering Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 07/24] drm/xe/xe3p: Stop programming RCU_MODE's fixed slice mode setting Lucas De Marchi
2025-10-17  2:26 ` Lucas De Marchi [this message]
2025-10-17  2:26 ` [PATCH v3 09/24] drm/xe: Dump CURRENT_LRCA register Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 10/24] drm/xe/xe3p: Dump CSMQDEBUG register Lucas De Marchi
2025-10-17 15:55   ` Matt Roper
2025-10-21 16:02   ` Summers, Stuart
2025-10-17  2:26 ` [PATCH v3 11/24] drm/xe/nvl: Define NVL-S platform Lucas De Marchi
2025-10-17 13:05   ` Gustavo Sousa
2025-10-17  2:26 ` [PATCH v3 12/24] drm/xe/nvls: Define GuC firmware for NVL-S Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 13/24] drm/xe/nvls: Attach MOCS table " Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 14/24] drm/xe/xe3p_xpc: Add Xe3p_XPC IP definition Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 15/24] drm/xe/xe3p_xpc: Add L3 bank mask Lucas De Marchi
2025-10-17 17:51   ` Matt Roper
2025-10-18  3:18     ` Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 16/24] drm/xe/xe3p_xpc: Add MCR steering Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 17/24] drm/xe/irq: Rename fuse mask variables Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 18/24] drm/xe/irq: Split irq mask per engine class Lucas De Marchi
2025-10-17 16:03   ` Matt Roper
2025-10-17  2:26 ` [PATCH v3 19/24] drm/xe/irq: Rename bits used with all engines Lucas De Marchi
2025-10-17 16:05   ` Matt Roper
2025-10-17  2:26 ` [PATCH v3 20/24] drm/xe/irq: Check fuse mask for media engines Lucas De Marchi
2025-10-17 16:07   ` Matt Roper
2025-10-17  2:26 ` [PATCH v3 21/24] drm/xe/xe3p_xpc: Add support for compute walker for non-MSIx Lucas De Marchi
2025-10-17 17:04   ` Matt Roper
2025-10-17  2:26 ` [PATCH v3 22/24] drm/xe/xe3p_xpc: Skip compression tuning on platforms without flatccs Lucas De Marchi
2025-10-17  2:26 ` [PATCH v3 23/24] drm/xe/xe3p_xpc: Setup PAT table Lucas De Marchi
2025-10-17 11:05   ` Ville Syrjälä
2025-10-17 17:18     ` Matt Roper
2025-10-17 18:09       ` Ville Syrjälä
2025-10-17 20:33         ` Matt Roper
2025-10-17  2:26 ` [PATCH v3 24/24] drm/xe/xe3p: Add xe3p EU stall data format Lucas De Marchi
2025-10-17  2:35 ` ✗ CI.checkpatch: warning for drm/xe: Add Xe3p support (rev3) Patchwork
2025-10-17  2:36 ` ✓ CI.KUnit: success " Patchwork
2025-10-17  3:23 ` ✓ Xe.CI.BAT: " Patchwork
2025-10-18  1:56 ` ✗ Xe.CI.Full: failure " Patchwork
2025-10-19  2:55 ` [PATCH v3 00/24] drm/xe: Add Xe3p support Lucas De Marchi

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=20251016-xe3p-v3-8-3dd173a3097a@intel.com \
    --to=lucas.demarchi@intel.com \
    --cc=balasubramani.vivekanandan@intel.com \
    --cc=gustavo.sousa@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=matthew.d.roper@intel.com \
    --cc=shekhar.chauhan@intel.com \
    --cc=tejas.upadhyay@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