All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ramalingam C <ramalingam.c@intel.com>
To: intel-gfx <intel-gfx@lists.freedesktop.org>,
	dri-devel <dri-devel@lists.freedesktop.org>
Cc: Hellstrom Thomas <thomas.hellstrom@intel.com>,
	Matthew Auld <matthew.auld@intel.com>,
	Chris Wilson <chris@chris-wilson.co.uk>
Subject: [PATCH v3 1/6] drm/i915/gt: Use XY_FASR_COLOR_BLT to clear obj on graphics ver 12+
Date: Mon,  7 Mar 2022 19:10:33 +0530	[thread overview]
Message-ID: <20220307134038.30525-2-ramalingam.c@intel.com> (raw)
In-Reply-To: <20220307134038.30525-1-ramalingam.c@intel.com>

XY_FAST_COLOR_BLT cmd is faster than the older XY_COLOR_BLT. Hence for
clearing (Zero out) the pages of the newly allocated object, faster cmd
is used.

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/gt/intel_gpu_commands.h |  5 ++
 drivers/gpu/drm/i915/gt/intel_migrate.c      | 51 +++++++++++++++++---
 2 files changed, 49 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
index d112ffd56418..925e55b6a94f 100644
--- a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
+++ b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
@@ -205,6 +205,11 @@
 
 #define COLOR_BLT_CMD			(2 << 29 | 0x40 << 22 | (5 - 2))
 #define XY_COLOR_BLT_CMD		(2 << 29 | 0x50 << 22)
+#define XY_FAST_COLOR_BLT_CMD		(2 << 29 | 0x44 << 22)
+#define   XY_FAST_COLOR_BLT_DEPTH_32	(2 << 19)
+#define   XY_FAST_COLOR_BLT_DW		16
+#define   XY_FAST_COLOR_BLT_MOCS_MASK	GENMASK(27, 21)
+#define   XY_FAST_COLOR_BLT_MEM_TYPE_SHIFT 31
 #define SRC_COPY_BLT_CMD		(2 << 29 | 0x43 << 22)
 #define GEN9_XY_FAST_COPY_BLT_CMD	(2 << 29 | 0x42 << 22)
 #define XY_SRC_COPY_BLT_CMD		(2 << 29 | 0x53 << 22)
diff --git a/drivers/gpu/drm/i915/gt/intel_migrate.c b/drivers/gpu/drm/i915/gt/intel_migrate.c
index 20444d6ceb3c..cb68f7bf6b28 100644
--- a/drivers/gpu/drm/i915/gt/intel_migrate.c
+++ b/drivers/gpu/drm/i915/gt/intel_migrate.c
@@ -16,6 +16,8 @@ struct insert_pte_data {
 };
 
 #define CHUNK_SZ SZ_8M /* ~1ms at 8GiB/s preemption delay */
+#define GET_CCS_BYTES(i915, size)	(HAS_FLAT_CCS(i915) ? \
+					 DIV_ROUND_UP(size, NUM_BYTES_PER_CCS_BYTE) : 0)
 
 static bool engine_supports_migration(struct intel_engine_cs *engine)
 {
@@ -614,20 +616,56 @@ intel_context_migrate_copy(struct intel_context *ce,
 	return err;
 }
 
-static int emit_clear(struct i915_request *rq, u64 offset, int size, u32 value)
+static int emit_clear(struct i915_request *rq, u64 offset, int size,
+		      u32 value, bool is_lmem)
 {
-	const int ver = GRAPHICS_VER(rq->engine->i915);
-	u32 *cs;
+	struct drm_i915_private *i915 = rq->engine->i915;
+	int mocs = rq->engine->gt->mocs.uc_index << 1;
+	const int ver = GRAPHICS_VER(i915);
+	u32 *cs, mem_type = 0;
+	int ring_sz;
 
 	GEM_BUG_ON(size >> PAGE_SHIFT > S16_MAX);
 
 	offset += (u64)rq->engine->instance << 32;
 
-	cs = intel_ring_begin(rq, ver >= 8 ? 8 : 6);
+	if (ver >= 12)
+		ring_sz = 16;
+	else if (ver >= 8)
+		ring_sz = 8;
+	else
+		ring_sz = 6;
+
+	if (!is_lmem)
+		mem_type = 1 << XY_FAST_COLOR_BLT_MEM_TYPE_SHIFT;
+
+	cs = intel_ring_begin(rq, ring_sz);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
 
-	if (ver >= 8) {
+	if (ver >= 12) {
+		*cs++ = XY_FAST_COLOR_BLT_CMD | XY_FAST_COLOR_BLT_DEPTH_32 |
+			(XY_FAST_COLOR_BLT_DW - 2);
+		*cs++ = FIELD_PREP(XY_FAST_COLOR_BLT_MOCS_MASK, mocs) |
+			(PAGE_SIZE - 1);
+		*cs++ = 0;
+		*cs++ = size >> PAGE_SHIFT << 16 | PAGE_SIZE / 4;
+		*cs++ = lower_32_bits(offset);
+		*cs++ = upper_32_bits(offset);
+		*cs++ = mem_type;
+		/* BG7 */
+		*cs++ = value;
+		*cs++ = 0;
+		*cs++ = 0;
+		*cs++ = 0;
+		/* BG11 */
+		*cs++ = 0;
+		*cs++ = 0;
+		/* BG13 */
+		*cs++ = 0;
+		*cs++ = 0;
+		*cs++ = 0;
+	} else if (ver >= 8) {
 		*cs++ = XY_COLOR_BLT_CMD | BLT_WRITE_RGBA | (7 - 2);
 		*cs++ = BLT_DEPTH_32 | BLT_ROP_COLOR_COPY | PAGE_SIZE;
 		*cs++ = 0;
@@ -645,7 +683,6 @@ static int emit_clear(struct i915_request *rq, u64 offset, int size, u32 value)
 		*cs++ = lower_32_bits(offset);
 		*cs++ = value;
 	}
-
 	intel_ring_advance(rq, cs);
 	return 0;
 }
@@ -711,7 +748,7 @@ intel_context_migrate_clear(struct intel_context *ce,
 		if (err)
 			goto out_rq;
 
-		err = emit_clear(rq, offset, len, value);
+		err = emit_clear(rq, offset, len, value, is_lmem);
 
 		/* Arbitration is re-enabled between requests. */
 out_rq:
-- 
2.20.1


WARNING: multiple messages have this Message-ID (diff)
From: Ramalingam C <ramalingam.c@intel.com>
To: intel-gfx <intel-gfx@lists.freedesktop.org>,
	dri-devel <dri-devel@lists.freedesktop.org>
Cc: Hellstrom Thomas <thomas.hellstrom@intel.com>,
	Matthew Auld <matthew.auld@intel.com>,
	Chris Wilson <chris@chris-wilson.co.uk>
Subject: [Intel-gfx] [PATCH v3 1/6] drm/i915/gt: Use XY_FASR_COLOR_BLT to clear obj on graphics ver 12+
Date: Mon,  7 Mar 2022 19:10:33 +0530	[thread overview]
Message-ID: <20220307134038.30525-2-ramalingam.c@intel.com> (raw)
In-Reply-To: <20220307134038.30525-1-ramalingam.c@intel.com>

XY_FAST_COLOR_BLT cmd is faster than the older XY_COLOR_BLT. Hence for
clearing (Zero out) the pages of the newly allocated object, faster cmd
is used.

Signed-off-by: Ramalingam C <ramalingam.c@intel.com>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
---
 drivers/gpu/drm/i915/gt/intel_gpu_commands.h |  5 ++
 drivers/gpu/drm/i915/gt/intel_migrate.c      | 51 +++++++++++++++++---
 2 files changed, 49 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
index d112ffd56418..925e55b6a94f 100644
--- a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
+++ b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
@@ -205,6 +205,11 @@
 
 #define COLOR_BLT_CMD			(2 << 29 | 0x40 << 22 | (5 - 2))
 #define XY_COLOR_BLT_CMD		(2 << 29 | 0x50 << 22)
+#define XY_FAST_COLOR_BLT_CMD		(2 << 29 | 0x44 << 22)
+#define   XY_FAST_COLOR_BLT_DEPTH_32	(2 << 19)
+#define   XY_FAST_COLOR_BLT_DW		16
+#define   XY_FAST_COLOR_BLT_MOCS_MASK	GENMASK(27, 21)
+#define   XY_FAST_COLOR_BLT_MEM_TYPE_SHIFT 31
 #define SRC_COPY_BLT_CMD		(2 << 29 | 0x43 << 22)
 #define GEN9_XY_FAST_COPY_BLT_CMD	(2 << 29 | 0x42 << 22)
 #define XY_SRC_COPY_BLT_CMD		(2 << 29 | 0x53 << 22)
diff --git a/drivers/gpu/drm/i915/gt/intel_migrate.c b/drivers/gpu/drm/i915/gt/intel_migrate.c
index 20444d6ceb3c..cb68f7bf6b28 100644
--- a/drivers/gpu/drm/i915/gt/intel_migrate.c
+++ b/drivers/gpu/drm/i915/gt/intel_migrate.c
@@ -16,6 +16,8 @@ struct insert_pte_data {
 };
 
 #define CHUNK_SZ SZ_8M /* ~1ms at 8GiB/s preemption delay */
+#define GET_CCS_BYTES(i915, size)	(HAS_FLAT_CCS(i915) ? \
+					 DIV_ROUND_UP(size, NUM_BYTES_PER_CCS_BYTE) : 0)
 
 static bool engine_supports_migration(struct intel_engine_cs *engine)
 {
@@ -614,20 +616,56 @@ intel_context_migrate_copy(struct intel_context *ce,
 	return err;
 }
 
-static int emit_clear(struct i915_request *rq, u64 offset, int size, u32 value)
+static int emit_clear(struct i915_request *rq, u64 offset, int size,
+		      u32 value, bool is_lmem)
 {
-	const int ver = GRAPHICS_VER(rq->engine->i915);
-	u32 *cs;
+	struct drm_i915_private *i915 = rq->engine->i915;
+	int mocs = rq->engine->gt->mocs.uc_index << 1;
+	const int ver = GRAPHICS_VER(i915);
+	u32 *cs, mem_type = 0;
+	int ring_sz;
 
 	GEM_BUG_ON(size >> PAGE_SHIFT > S16_MAX);
 
 	offset += (u64)rq->engine->instance << 32;
 
-	cs = intel_ring_begin(rq, ver >= 8 ? 8 : 6);
+	if (ver >= 12)
+		ring_sz = 16;
+	else if (ver >= 8)
+		ring_sz = 8;
+	else
+		ring_sz = 6;
+
+	if (!is_lmem)
+		mem_type = 1 << XY_FAST_COLOR_BLT_MEM_TYPE_SHIFT;
+
+	cs = intel_ring_begin(rq, ring_sz);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
 
-	if (ver >= 8) {
+	if (ver >= 12) {
+		*cs++ = XY_FAST_COLOR_BLT_CMD | XY_FAST_COLOR_BLT_DEPTH_32 |
+			(XY_FAST_COLOR_BLT_DW - 2);
+		*cs++ = FIELD_PREP(XY_FAST_COLOR_BLT_MOCS_MASK, mocs) |
+			(PAGE_SIZE - 1);
+		*cs++ = 0;
+		*cs++ = size >> PAGE_SHIFT << 16 | PAGE_SIZE / 4;
+		*cs++ = lower_32_bits(offset);
+		*cs++ = upper_32_bits(offset);
+		*cs++ = mem_type;
+		/* BG7 */
+		*cs++ = value;
+		*cs++ = 0;
+		*cs++ = 0;
+		*cs++ = 0;
+		/* BG11 */
+		*cs++ = 0;
+		*cs++ = 0;
+		/* BG13 */
+		*cs++ = 0;
+		*cs++ = 0;
+		*cs++ = 0;
+	} else if (ver >= 8) {
 		*cs++ = XY_COLOR_BLT_CMD | BLT_WRITE_RGBA | (7 - 2);
 		*cs++ = BLT_DEPTH_32 | BLT_ROP_COLOR_COPY | PAGE_SIZE;
 		*cs++ = 0;
@@ -645,7 +683,6 @@ static int emit_clear(struct i915_request *rq, u64 offset, int size, u32 value)
 		*cs++ = lower_32_bits(offset);
 		*cs++ = value;
 	}
-
 	intel_ring_advance(rq, cs);
 	return 0;
 }
@@ -711,7 +748,7 @@ intel_context_migrate_clear(struct intel_context *ce,
 		if (err)
 			goto out_rq;
 
-		err = emit_clear(rq, offset, len, value);
+		err = emit_clear(rq, offset, len, value, is_lmem);
 
 		/* Arbitration is re-enabled between requests. */
 out_rq:
-- 
2.20.1


  reply	other threads:[~2022-03-07 13:40 UTC|newest]

Thread overview: 23+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-03-07 13:40 [PATCH v3 0/6] drm/i915/ttm: Evict and restore of compressed object Ramalingam C
2022-03-07 13:40 ` [Intel-gfx] " Ramalingam C
2022-03-07 13:40 ` Ramalingam C [this message]
2022-03-07 13:40   ` [Intel-gfx] [PATCH v3 1/6] drm/i915/gt: Use XY_FASR_COLOR_BLT to clear obj on graphics ver 12+ Ramalingam C
2022-03-07 14:29   ` Hellstrom, Thomas
2022-03-07 14:29     ` [Intel-gfx] " Hellstrom, Thomas
2022-03-07 13:40 ` [PATCH v3 2/6] drm/i915/gt: Clear compress metadata for Flat-ccs objects Ramalingam C
2022-03-07 13:40   ` [Intel-gfx] " Ramalingam C
2022-03-07 14:32   ` Hellstrom, Thomas
2022-03-07 14:32     ` Hellstrom, Thomas
2022-03-07 13:40 ` [PATCH v3 3/6] drm/ttm: Add a parameter to add extra pages into ttm_tt Ramalingam C
2022-03-07 13:40   ` [Intel-gfx] " Ramalingam C
2022-03-07 13:40 ` [PATCH v3 4/6] drm/i915/gem: Add extra pages in ttm_tt for ccs data Ramalingam C
2022-03-07 13:40   ` [Intel-gfx] " Ramalingam C
2022-03-07 15:32   ` Matthew Auld
2022-03-07 15:32     ` [Intel-gfx] " Matthew Auld
2022-03-07 13:40 ` [PATCH v3 5/6] drm/i915/gt: Optimize the migration loop Ramalingam C
2022-03-07 13:40   ` [Intel-gfx] " Ramalingam C
2022-03-07 13:40 ` [PATCH v3 6/6] drm/i915/migrate: Evict and restore the flatccs capable lmem obj Ramalingam C
2022-03-07 13:40   ` [Intel-gfx] " Ramalingam C
2022-03-07 14:31 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for drm/i915/ttm: Evict and restore of compressed object Patchwork
2022-03-07 14:33 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2022-03-07 15:04 ` [Intel-gfx] ✗ Fi.CI.BAT: failure " Patchwork

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=20220307134038.30525-2-ramalingam.c@intel.com \
    --to=ramalingam.c@intel.com \
    --cc=chris@chris-wilson.co.uk \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=matthew.auld@intel.com \
    --cc=thomas.hellstrom@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.