From d5812cb30d528d1dd17a19c53ba82a44f24dccba Mon Sep 17 00:00:00 2001
From: Boris Brezillon <boris.brezillon@collabora.com>
Date: Wed, 7 Feb 2024 21:48:57 +0100
Subject: [PATCH] panfrost: Pad compute jobs with zeros on v4

Apparently, Midgard GPUs don't like when the last 2 words of
compute/vertex jobs contain garbage. Extend the compute job definition
to include a padding section thus aligning the job on a 64-byte boundary,
and add the according pan_section_pack() calls where we have a
compute job filled.

Fixes: b76420be1f1f ("panfrost: Split command stream descriptor definitions per-gen")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/10558
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Acked-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Tested-by: Anton Bambura <jenneron@postmarketos.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27515>
(cherry picked from commit 5b1b76e9cdd8b44538aa327c7705381fa68b891b)
---
 .pick_status.json                            |  2 +-
 src/gallium/drivers/panfrost/pan_cmdstream.c | 10 ++++++++++
 src/panfrost/lib/genxml/v4.xml               |  4 ++++
 3 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/.pick_status.json b/.pick_status.json
index ade93212d53..71ca7c6b05c 100644
--- a/.pick_status.json
+++ b/.pick_status.json
@@ -2634,7 +2634,7 @@
         "description": "panfrost: Pad compute jobs with zeros on v4",
         "nominated": true,
         "nomination_type": 1,
-        "resolution": 0,
+        "resolution": 1,
         "main_sha": null,
         "because_sha": "b76420be1f1f3a0ceb7808131b9c8aa8e970a415",
         "notes": null
diff --git a/src/gallium/drivers/panfrost/pan_cmdstream.c b/src/gallium/drivers/panfrost/pan_cmdstream.c
index a728f4a5e68..bec2617f1fa 100644
--- a/src/gallium/drivers/panfrost/pan_cmdstream.c
+++ b/src/gallium/drivers/panfrost/pan_cmdstream.c
@@ -2801,6 +2801,11 @@ panfrost_draw_emit_vertex(struct panfrost_batch *batch,
    section = pan_section_ptr(job, COMPUTE_JOB, DRAW);
    panfrost_draw_emit_vertex_section(batch, vs_vary, varyings, attribs,
                                      attrib_bufs, section);
+
+#if PAN_ARCH == 4
+   pan_section_pack(job, COMPUTE_JOB, COMPUTE_PADDING, cfg)
+      ;
+#endif
 }
 #endif
 
@@ -3822,6 +3827,11 @@ panfrost_launch_grid_on_batch(struct pipe_context *pipe,
       cfg.textures = batch->textures[PIPE_SHADER_COMPUTE];
       cfg.samplers = batch->samplers[PIPE_SHADER_COMPUTE];
    }
+
+#if PAN_ARCH == 4
+   pan_section_pack(t.cpu, COMPUTE_JOB, COMPUTE_PADDING, cfg)
+      ;
+#endif
 #else
    struct panfrost_compiled_shader *cs = ctx->prog[PIPE_SHADER_COMPUTE];
 
diff --git a/src/panfrost/lib/genxml/v4.xml b/src/panfrost/lib/genxml/v4.xml
index 63b7f7f57ac..a804cd2acfc 100644
--- a/src/panfrost/lib/genxml/v4.xml
+++ b/src/panfrost/lib/genxml/v4.xml
@@ -992,12 +992,16 @@
     <field name="Job Task Split" size="4" start="0:26" type="uint"/>
   </struct>
 
+  <struct name="Compute Padding" size="2">
+  </struct>
+
   <!-- Compute job also covers vertex and geometry operations -->
   <aggregate name="Compute Job" align="64">
     <section name="Header" offset="0" type="Job Header"/>
     <section name="Invocation" offset="32" type="Invocation"/>
     <section name="Parameters" offset="40" type="Compute Job Parameters"/>
     <section name="Draw" offset="64" type="Draw"/>
+    <section name="Compute padding" offset="184" type="Compute Padding"/>
   </aggregate>
 
   <struct name="Primitive Size">