Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Chaitanya Kumar Borah <chaitanya.kumar.borah@intel.com>
To: intel-xe@lists.freedesktop.org, intel-gfx@lists.freedesktop.org
Cc: ville.syrjala@linux.intel.com, uma.shankar@intel.com,
	chaitanya.kumar.borah@intel.com, animesh.manna@intel.com
Subject: [PATCH 04/11] drm/i915/dsb: Implement intel_dsb_gosub()
Date: Tue,  8 Apr 2025 16:30:15 +0530	[thread overview]
Message-ID: <20250408110022.1907802-5-chaitanya.kumar.borah@intel.com> (raw)
In-Reply-To: <20250408110022.1907802-1-chaitanya.kumar.borah@intel.com>

From: Ville Syrjälä <ville.syrjala@linux.intel.com>

Add support for the new GOSUB DSB instruction (available on ptl+),
which instructs the DSB to jump to a different buffer, executie
the commands there, and then return execution to the next
instruction in the original buffer.

There are a few alignment related workarounds that need to
be dealt with when emitting GOSUB instruction.

v2: Right shift head and tail pointer passed to gosub command (chaitanya)

Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
---
 drivers/gpu/drm/i915/display/intel_dsb.c | 52 ++++++++++++++++++++++++
 drivers/gpu/drm/i915/display/intel_dsb.h |  2 +
 2 files changed, 54 insertions(+)

diff --git a/drivers/gpu/drm/i915/display/intel_dsb.c b/drivers/gpu/drm/i915/display/intel_dsb.c
index 0de15e3a9a56..2cda6fc7857b 100644
--- a/drivers/gpu/drm/i915/display/intel_dsb.c
+++ b/drivers/gpu/drm/i915/display/intel_dsb.c
@@ -93,6 +93,9 @@ struct intel_dsb {
 /* see DSB_REG_VALUE_MASK */
 #define DSB_OPCODE_POLL			0xA
 /* see DSB_REG_VALUE_MASK */
+#define DSB_OPCODE_GOSUB		0xC /* ptl+ */
+#define   DSB_GOSUB_HEAD_SHIFT		26
+#define   DSB_GOSUB_TAIL_SHIFT		0
 
 static bool pre_commit_is_vrr_active(struct intel_atomic_state *state,
 				     struct intel_crtc *crtc)
@@ -533,6 +536,55 @@ static void intel_dsb_align_tail(struct intel_dsb *dsb)
 	dsb->free_pos = aligned_tail / 4;
 }
 
+static void intel_dsb_gosub_align(struct intel_dsb *dsb)
+{
+	u32 aligned_tail, tail;
+
+	intel_dsb_ins_align(dsb);
+
+	tail = dsb->free_pos * 4;
+	aligned_tail = ALIGN(tail, CACHELINE_BYTES);
+
+	/*
+	 * "The GOSUB instruction cannot be placed in
+	 *  cacheline QW slot 6 or 7 (numbered 0-7)"
+	 */
+	if (aligned_tail - tail <= 2 * 8)
+		intel_dsb_buffer_memset(&dsb->dsb_buf, dsb->free_pos, 0,
+					aligned_tail - tail);
+
+	dsb->free_pos = aligned_tail / 4;
+}
+
+void intel_dsb_gosub(struct intel_dsb *dsb,
+		     struct intel_dsb *sub_dsb)
+{
+	struct intel_crtc *crtc = dsb->crtc;
+	struct intel_display *display = to_intel_display(crtc->base.dev);
+	u64 head_tail;
+
+	if (drm_WARN_ON(display->drm, dsb->id != sub_dsb->id))
+		return;
+
+	if (!assert_dsb_tail_is_aligned(sub_dsb))
+		return;
+
+	intel_dsb_gosub_align(dsb);
+
+	head_tail = ((u64)(intel_dsb_head(sub_dsb) >> 6) << DSB_GOSUB_HEAD_SHIFT) |
+		((u64)(intel_dsb_tail(sub_dsb) >> 6) << DSB_GOSUB_TAIL_SHIFT);
+
+	intel_dsb_emit(dsb, lower_32_bits(head_tail),
+		       (DSB_OPCODE_GOSUB << DSB_OPCODE_SHIFT) |
+		       upper_32_bits(head_tail));
+
+	/*
+	 * "NOTE: the instructions within the cacheline
+	 *  FOLLOWING the GOSUB instruction must be NOPs."
+	 */
+	intel_dsb_align_tail(dsb);
+}
+
 void intel_dsb_finish(struct intel_dsb *dsb)
 {
 	struct intel_crtc *crtc = dsb->crtc;
diff --git a/drivers/gpu/drm/i915/display/intel_dsb.h b/drivers/gpu/drm/i915/display/intel_dsb.h
index e843c52bf97c..8b2cf0a7b7e6 100644
--- a/drivers/gpu/drm/i915/display/intel_dsb.h
+++ b/drivers/gpu/drm/i915/display/intel_dsb.h
@@ -57,6 +57,8 @@ void intel_dsb_vblank_evade(struct intel_atomic_state *state,
 void intel_dsb_poll(struct intel_dsb *dsb,
 		    i915_reg_t reg, u32 mask, u32 val,
 		    int wait_us, int count);
+void intel_dsb_gosub(struct intel_dsb *dsb,
+		     struct intel_dsb *sub_dsb);
 void intel_dsb_chain(struct intel_atomic_state *state,
 		     struct intel_dsb *dsb,
 		     struct intel_dsb *chained_dsb,
-- 
2.25.1


  parent reply	other threads:[~2025-04-08 11:15 UTC|newest]

Thread overview: 37+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-04-08 11:00 [PATCH 00/11] drm/xe/display: Program double buffered LUT registers Chaitanya Kumar Borah
2025-04-08 11:00 ` [PATCH 01/11] drm/i915/dsb: Extract intel_dsb_ins_align() Chaitanya Kumar Borah
2025-04-21 12:40   ` Manna, Animesh
2025-04-08 11:00 ` [PATCH 02/11] drm/i915/dsb: Extract assert_dsb_tail_is_aligned() Chaitanya Kumar Borah
2025-04-21 12:43   ` Manna, Animesh
2025-04-08 11:00 ` [PATCH 03/11] drm/i915/dsb: Extract intel_dsb_{head,tail}() Chaitanya Kumar Borah
2025-04-21 12:44   ` Manna, Animesh
2025-04-08 11:00 ` Chaitanya Kumar Borah [this message]
2025-04-16  7:28   ` [PATCH 04/11] drm/i915/dsb: Implement intel_dsb_gosub() Manna, Animesh
2025-04-08 11:00 ` [PATCH v2 05/11] drm/i915/dsb: add intel_dsb_gosub_finish() Chaitanya Kumar Borah
2025-05-14  9:19   ` Shankar, Uma
2025-04-08 11:00 ` [PATCH v2 06/11] drm/i915/dsb: Add support for GOSUB interrupt Chaitanya Kumar Borah
2025-05-14  9:46   ` Shankar, Uma
2025-04-08 11:00 ` [PATCH 07/11] drm/i915: s/dsb_color_vblank/dsb_color Chaitanya Kumar Borah
2025-04-08 11:00 ` [PATCH v2 08/11] drm/i915: use GOSUB to program doubled buffered LUT registers Chaitanya Kumar Borah
2025-05-14 11:40   ` Shankar, Uma
2025-04-08 11:00 ` [PATCH v2 09/11] drm/i915: Program DB LUT registers before vblank Chaitanya Kumar Borah
2025-05-14 11:44   ` Shankar, Uma
2025-04-08 11:00 ` [PATCH 10/11] drm/i915/color: Do not pre-load LUTs with DB registers Chaitanya Kumar Borah
2025-05-14 11:46   ` Shankar, Uma
2025-04-08 11:00 ` [PATCH 11/11] drm/i915: Disable updating of LUT values during vblank Chaitanya Kumar Borah
2025-05-14 11:50   ` Shankar, Uma
2025-04-08 11:21 ` ✓ CI.Patch_applied: success for drm/xe/display: Program double buffered LUT registers (rev6) Patchwork
2025-04-08 11:21 ` ✓ CI.checkpatch: " Patchwork
2025-04-08 11:22 ` ✓ CI.KUnit: " Patchwork
2025-04-08 11:26 ` ✗ CI.Build: failure " Patchwork
2025-04-08 21:37 ` ✓ CI.Patch_applied: success for drm/xe/display: Program double buffered LUT registers (rev7) Patchwork
2025-04-08 21:37 ` ✓ CI.checkpatch: " Patchwork
2025-04-08 21:38 ` ✓ CI.KUnit: " Patchwork
2025-04-08 22:05 ` ✓ CI.Build: " Patchwork
2025-04-08 22:08 ` ✓ CI.Hooks: " Patchwork
2025-04-08 22:10 ` ✗ CI.checksparse: warning " Patchwork
2025-04-08 22:55 ` ✓ Xe.CI.BAT: success " Patchwork
2025-04-08 23:58 ` ✗ Xe.CI.Full: failure " Patchwork
2025-04-09  6:27 ` Patchwork
  -- strict thread matches above, loose matches on Subject: below --
2025-04-07 14:23 [PATCH 00/11] drm/xe/display: Program double buffered LUT registers Chaitanya Kumar Borah
2025-04-07 14:23 ` [PATCH 04/11] drm/i915/dsb: Implement intel_dsb_gosub() Chaitanya Kumar Borah
2025-05-14  9:18   ` Shankar, Uma

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=20250408110022.1907802-5-chaitanya.kumar.borah@intel.com \
    --to=chaitanya.kumar.borah@intel.com \
    --cc=animesh.manna@intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=uma.shankar@intel.com \
    --cc=ville.syrjala@linux.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