mesa-panvk-bifrost: iter13 — implement VK_EXT_transform_feedback for Bifrost

iter12 hit a wall: Brave's ANGLE-Vulkan path requires GLES3, which
requires VK_EXT_transform_feedback, which PanVk-Bifrost did not
implement. This iter implements that extension, unlocking the full
ANGLE-Vulkan-on-Bifrost stack.

The implementation follows Panfrost-Gallium's well-validated XFB lowering
(nir_io_add_intrinsic_xfb_info + pan_nir_lower_xfb) wired into the PanVk
shader pipeline after nir_lower_io. Adds 4 XFB buffer address sysvals
plus per-draw num_vertices to the graphics sysval struct. Buffer state
is tracked on the cmd buffer; per-draw sysval upload populates either
the bound buffer's GPU address or PAN_SHADER_OOB_ADDRESS (memory-sink)
so XFB-capable pipelines used outside Begin/End survive without GPU
fault — the Panfrost-Gallium idiom from gallium/drivers/panfrost/
pan_cmdstream.c:1350.

Verified on PineTab2 (Mali-G52 r1 MC1, RK3566):
- /tmp/panvk-iter13/probe_xfb: 3 vertices captured byte-exact
- /tmp/panvk-iter13/probe_xfb_nodraw: XFB pipeline used without Bind/
  Begin/End survives — DEVICE_LOST regression closed
- Brave 148 with --use-angle=vulkan: WebGL 2.0 (OpenGL ES 3.0) creates
  cleanly, renderer reports
  "ANGLE (ARM, Vulkan 1.2.335 (Mali-G52 r1 MC1), panvk)"
- chrome://gpu graphics feature status: Canvas/Compositing/OpenGL/
  Rasterization/WebGL/WebGL2/WebGPU/Video Decode all hardware accelerated

Phase docs:
- ~/src/panvk-bifrost/phase4_iter13_close.md  (build green)
- ~/src/panvk-bifrost/phase5_iter13_close.md  (review fixes applied)
- ~/src/panvk-bifrost/phase6_iter13_close.md  (Brave integration green)

pkgver bumped 26.0.6.r2 -> 26.0.6.r3; iter13 patch applied via
unified-diff (the 328-line change scope is past sed-of-individual-
lines territory). Sanity checks in prepare() verify the patch landed.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

arch/mesa-panvk-bifrost/0003-panvk-bifrost-vk-ext-transform-feedback.patch

@@ -0,0 +1,328 @@
+--- a/src/panfrost/vulkan/panvk_shader.h	2026-04-29 22:19:00.000000000 +0200
++++ b/src/panfrost/vulkan/panvk_shader.h	2026-05-20 18:52:53.312698258 +0200
+@@ -150,6 +150,10 @@
+    struct {
+ #if PAN_ARCH < 9
+       int32_t raw_vertex_offset;
++      uint32_t num_vertices;       /* iter13: XFB needs per-draw vertex count */
++      /* aligned_u64 attribute below inserts the 4-byte alignment gap
++       * after num_vertices automatically — no explicit pad needed. */
++      aligned_u64 xfb_address[4];  /* iter13: 4 transform feedback buffer base addresses */
+ #endif
+       int32_t first_vertex;
+       int32_t base_instance;
+--- a/src/panfrost/vulkan/panvk_vX_physical_device.c	2026-05-20 19:09:29.711145446 +0200
++++ b/src/panfrost/vulkan/panvk_vX_physical_device.c	2026-05-20 18:52:54.832720445 +0200
+@@ -169,6 +169,7 @@
+       .EXT_provoking_vertex = true,
+       .EXT_queue_family_foreign = true,
+       .EXT_robustness2 = true,
++      .EXT_transform_feedback = PAN_ARCH < 9,   /* iter13: JM-class only for now */
+       .EXT_sampler_filter_minmax = PAN_ARCH >= 10,
+       .EXT_scalar_block_layout = true,
+       .EXT_separate_stencil_usage = true,
+@@ -495,6 +496,10 @@
+       .robustImageAccess2 = false,
+       .nullDescriptor = true,
+ 
++      /* VK_EXT_transform_feedback (iter13) */
++      .transformFeedback = PAN_ARCH < 9,
++      .geometryStreams = false,
++
+       /* VK_KHR_shader_clock */
+       .shaderSubgroupClock = device->kmod.dev->props.gpu_can_query_timestamp,
+       .shaderDeviceClock = device->kmod.dev->props.timestamp_device_coherent,
+@@ -1020,6 +1025,18 @@
+       .robustStorageBufferAccessSizeAlignment = 1,
+       .robustUniformBufferAccessSizeAlignment = 1,
+ 
++      /* VK_EXT_transform_feedback (iter13) */
++      .maxTransformFeedbackStreams = 1,
++      .maxTransformFeedbackBuffers = 4,
++      .maxTransformFeedbackBufferSize = UINT32_MAX,
++      .maxTransformFeedbackStreamDataSize = 512,
++      .maxTransformFeedbackBufferDataSize = 512,
++      .maxTransformFeedbackBufferDataStride = 2048,
++      .transformFeedbackQueries = false,
++      .transformFeedbackStreamsLinesTriangles = false,
++      .transformFeedbackRasterizationStreamSelect = false,
++      .transformFeedbackDraw = false,
++
+       /* VK_EXT_shader_object */
+       /* We do not currently support VK_EXT_shader_object but this is used
+        * internally by vk_shader
+--- a/src/panfrost/vulkan/panvk_vX_shader.c	2026-04-29 22:19:00.000000000 +0200
++++ b/src/panfrost/vulkan/panvk_vX_shader.c	2026-05-20 18:52:56.556745611 +0200
+@@ -21,6 +21,7 @@
+ #include "panvk_physical_device.h"
+ #include "panvk_sampler.h"
+ #include "panvk_shader.h"
++#include "pan_nir.h"   /* iter13: pan_nir_lower_xfb */
+ 
+ #include "spirv/nir_spirv.h"
+ #include "util/memstream.h"
+@@ -100,6 +101,20 @@
+    case nir_intrinsic_load_raw_vertex_offset_pan:
+       val = load_sysval(b, graphics, bit_size, vs.raw_vertex_offset);
+       break;
++   case nir_intrinsic_load_num_vertices:    /* iter13: XFB index calc */
++      val = load_sysval(b, graphics, bit_size, vs.num_vertices);
++      break;
++   case nir_intrinsic_load_xfb_address: {   /* iter13: XFB buffer N base address */
++      unsigned idx = nir_intrinsic_base(intr);
++      switch (idx) {
++      case 0: val = load_sysval(b, graphics, bit_size, vs.xfb_address[0]); break;
++      case 1: val = load_sysval(b, graphics, bit_size, vs.xfb_address[1]); break;
++      case 2: val = load_sysval(b, graphics, bit_size, vs.xfb_address[2]); break;
++      case 3: val = load_sysval(b, graphics, bit_size, vs.xfb_address[3]); break;
++      default: return false;
++      }
++      break;
++   }
+    case nir_intrinsic_load_layer_id:
+       assert(b->shader->info.stage == MESA_SHADER_FRAGMENT);
+       val = load_sysval(b, graphics, bit_size, layer_id);
+@@ -457,6 +472,7 @@
+             core_max_id);
+ 
+    pan_preprocess_nir(nir, pdev->kmod.dev->props.gpu_id);
++
+ }
+ 
+ static void
+@@ -870,6 +886,18 @@
+             nir_var_shader_in | nir_var_shader_out, UINT32_MAX);
+    NIR_PASS(_, nir, nir_lower_io, nir_var_shader_in | nir_var_shader_out,
+             glsl_type_size, nir_lower_io_use_interpolated_input_intrinsics);
++
++#if PAN_ARCH < 9
++   /* iter13: VK_EXT_transform_feedback — runs AFTER nir_lower_io so that
++    * shader outputs are now store_output intrinsics that pan_nir_lower_xfb
++    * can rewrite to nir_store_global+nir_load_xfb_address. */
++   if (nir->info.stage == MESA_SHADER_VERTEX &&
++       nir->info.has_transform_feedback_varyings) {
++      NIR_PASS(_, nir, nir_opt_constant_folding);
++      NIR_PASS(_, nir, nir_io_add_intrinsic_xfb_info);
++      NIR_PASS(_, nir, pan_nir_lower_xfb);
++   }
++#endif
+ }
+ 
+ static VkResult
+@@ -1288,6 +1316,9 @@
+       .view_mask = (state && state->rp) ? state->rp->view_mask : 0,
+       .robust2_modes = robust2_modes,
+       .robust_descriptors = dev->vk.enabled_features.nullDescriptor,
++      /* iter13: XFB shaders must disable IDVS (matches Panfrost-Gallium). */
++      .no_idvs = (info->stage == MESA_SHADER_VERTEX) &&
++                 info->nir->info.has_transform_feedback_varyings,
+    };
+ 
+    switch (info->stage) {
+--- a/src/panfrost/vulkan/panvk_cmd_draw.h	2026-04-29 22:19:00.000000000 +0200
++++ b/src/panfrost/vulkan/panvk_cmd_draw.h	2026-05-20 18:52:57.748763011 +0200
+@@ -135,6 +135,19 @@
+    struct panvk_graphics_sysvals sysvals;
+ 
+ #if PAN_ARCH < 9
++   /* iter13: VK_EXT_transform_feedback state (JM-class only for now). */
++   struct {
++      bool active;
++      uint32_t buffer_count;
++      struct {
++         uint64_t addr;
++         uint64_t offset;
++         uint64_t size;
++      } buffers[4];
++   } xfb;
++#endif
++
++#if PAN_ARCH < 9
+    struct panvk_shader_link link;
+ #endif
+ 
+--- a/src/panfrost/vulkan/panvk_vX_cmd_draw.c	2026-04-29 22:19:00.000000000 +0200
++++ b/src/panfrost/vulkan/panvk_vX_cmd_draw.c	2026-05-20 19:10:23.031919662 +0200
+@@ -10,6 +10,7 @@
+ #include "panvk_entrypoints.h"
+ 
+ #include "pan_desc.h"
++#include "pan_compiler.h"   /* PAN_SHADER_OOB_ADDRESS */
+ #include "pan_util.h"
+ 
+ static void
+@@ -722,6 +723,35 @@
+    set_gfx_sysval(cmdbuf, dirty_sysvals, vs.raw_vertex_offset,
+                   info->vertex.raw_offset);
+    set_gfx_sysval(cmdbuf, dirty_sysvals, layer_id, info->layer_id);
++
++   /* iter13: VK_EXT_transform_feedback sysvals — always set (per draw),
++    * reflect bound XFB state. set_gfx_sysval is a no-op if value unchanged. */
++   set_gfx_sysval(cmdbuf, dirty_sysvals, vs.num_vertices, info->vertex.count);
++   {
++      const struct panvk_cmd_graphics_state *_gfx = &cmdbuf->state.gfx;
++      /* iter13: default each XFB buffer address to PAN_SHADER_OOB_ADDRESS
++       * (= 1<<63). This is the Panfrost-Gallium memory-sink idiom — the
++       * Bifrost MMU silently discards stores to this address, so a pipeline
++       * with XFB outputs used in a non-XFB draw (or in an XFB draw with
++       * fewer bound buffers than the shader declares) is safe instead of
++       * faulting. See gallium/drivers/panfrost/pan_cmdstream.c PAN_SYSVAL_XFB. */
++      uint64_t _xa0 = PAN_SHADER_OOB_ADDRESS, _xa1 = PAN_SHADER_OOB_ADDRESS,
++               _xa2 = PAN_SHADER_OOB_ADDRESS, _xa3 = PAN_SHADER_OOB_ADDRESS;
++      if (_gfx->xfb.active) {
++         if (_gfx->xfb.buffer_count > 0 && _gfx->xfb.buffers[0].addr)
++            _xa0 = _gfx->xfb.buffers[0].addr + _gfx->xfb.buffers[0].offset;
++         if (_gfx->xfb.buffer_count > 1 && _gfx->xfb.buffers[1].addr)
++            _xa1 = _gfx->xfb.buffers[1].addr + _gfx->xfb.buffers[1].offset;
++         if (_gfx->xfb.buffer_count > 2 && _gfx->xfb.buffers[2].addr)
++            _xa2 = _gfx->xfb.buffers[2].addr + _gfx->xfb.buffers[2].offset;
++         if (_gfx->xfb.buffer_count > 3 && _gfx->xfb.buffers[3].addr)
++            _xa3 = _gfx->xfb.buffers[3].addr + _gfx->xfb.buffers[3].offset;
++      }
++      set_gfx_sysval(cmdbuf, dirty_sysvals, vs.xfb_address[0], _xa0);
++      set_gfx_sysval(cmdbuf, dirty_sysvals, vs.xfb_address[1], _xa1);
++      set_gfx_sysval(cmdbuf, dirty_sysvals, vs.xfb_address[2], _xa2);
++      set_gfx_sysval(cmdbuf, dirty_sysvals, vs.xfb_address[3], _xa3);
++   }
+ #endif
+ 
+    if (dyn_gfx_state_dirty(cmdbuf, CB_BLEND_CONSTANTS)) {
+--- a/src/panfrost/vulkan/meson.build	2026-04-29 22:19:00.000000000 +0200
++++ b/src/panfrost/vulkan/meson.build	2026-05-20 18:53:04.484861338 +0200
+@@ -73,6 +73,7 @@
+ jm_inc_dir = ['jm']
+ jm_files = [
+   'jm/panvk_vX_bind_queue.c',
++  'jm/panvk_vX_cmd_xfb.c',   # iter13
+   'jm/panvk_vX_cmd_buffer.c',
+   'jm/panvk_vX_cmd_dispatch.c',
+   'jm/panvk_vX_cmd_draw.c',
+--- a/src/panfrost/vulkan/jm/panvk_vX_cmd_buffer.c	2026-04-29 22:19:00.000000000 +0200
++++ b/src/panfrost/vulkan/jm/panvk_vX_cmd_buffer.c	2026-05-20 19:10:26.163965149 +0200
+@@ -473,5 +473,12 @@
+ 
+    vk_command_buffer_begin(&cmdbuf->vk, pBeginInfo);
+ 
++#if PAN_ARCH < 9
++   /* iter13: clear XFB state on Begin so a reused command buffer does not
++    * inherit stale xfb.buffer_count / xfb.active / xfb.buffers[] from a
++    * prior recording. */
++   memset(&cmdbuf->state.gfx.xfb, 0, sizeof(cmdbuf->state.gfx.xfb));
++#endif
++
+    return VK_SUCCESS;
+ }
+--- a/src/panfrost/vulkan/jm/panvk_vX_cmd_xfb.c	2026-05-18 12:50:53.067999996 +0200
++++ b/src/panfrost/vulkan/jm/panvk_vX_cmd_xfb.c	2026-05-20 19:10:27.175979847 +0200
+@@ -0,0 +1,111 @@
++/*
++ * Copyright © 2026 mfritsche / claude-noether
++ * SPDX-License-Identifier: MIT
++ *
++ * iter13: VK_EXT_transform_feedback command handlers for the JM
++ * architecture path (Bifrost v6/v7 + Valhall-JM v9).
++ *
++ * The runtime contract:
++ *   - vkCmdBindTransformFeedbackBuffersEXT: stash (gpu_addr, offset, size)
++ *     for each slot into cmdbuf->state.gfx.xfb.buffers[].
++ *   - vkCmdBeginTransformFeedbackEXT: set cmdbuf->state.gfx.xfb.active = true.
++ *     Mark sysvals dirty so the next draw re-emits vs.xfb_address[].
++ *   - vkCmdEndTransformFeedbackEXT: set active = false.
++ *
++ * Counter buffers (firstCounterBuffer/counterBufferCount/pCounterBuffers/
++ * pCounterBufferOffsets) are accepted by API but ignored — v1 doesn't
++ * support pause/resume. transformFeedbackDraw is advertised as false.
++ *
++ * Per-draw integration: jm/panvk_vX_cmd_draw.c reads cmdbuf->state.gfx.xfb
++ * and populates vs.xfb_address[i] for shader use. The pan_nir_lower_xfb
++ * pass in panvk_vX_shader.c emits nir_load_xfb_address(i) which lowers
++ * (via panvk_vX_shader.c sysval handler) to a load from the per-draw
++ * sysval push area.
++ */
++
++#include "vk_log.h"
++#include "util/log.h"
++
++#include "panvk_cmd_buffer.h"
++#include "panvk_cmd_draw.h"
++#include "panvk_buffer.h"
++#include "panvk_entrypoints.h"
++
++VKAPI_ATTR void VKAPI_CALL
++panvk_per_arch(CmdBindTransformFeedbackBuffersEXT)(
++   VkCommandBuffer commandBuffer,
++   uint32_t firstBinding,
++   uint32_t bindingCount,
++   const VkBuffer *pBuffers,
++   const VkDeviceSize *pOffsets,
++   const VkDeviceSize *pSizes)
++{
++   VK_FROM_HANDLE(panvk_cmd_buffer, cmdbuf, commandBuffer);
++   struct panvk_cmd_graphics_state *gfx = &cmdbuf->state.gfx;
++
++   for (uint32_t i = 0; i < bindingCount; i++) {
++      uint32_t slot = firstBinding + i;
++      if (slot >= 4)
++         continue;
++
++      VK_FROM_HANDLE(panvk_buffer, buf, pBuffers[i]);
++      gfx->xfb.buffers[slot].addr = panvk_buffer_gpu_ptr(buf, 0);
++      gfx->xfb.buffers[slot].offset = pOffsets[i];
++      gfx->xfb.buffers[slot].size =
++         (pSizes != NULL && pSizes[i] != VK_WHOLE_SIZE)
++            ? pSizes[i]
++            : (buf->vk.size - pOffsets[i]);
++   }
++
++   if (firstBinding + bindingCount > gfx->xfb.buffer_count)
++      gfx->xfb.buffer_count = firstBinding + bindingCount;
++}
++
++VKAPI_ATTR void VKAPI_CALL
++panvk_per_arch(CmdBeginTransformFeedbackEXT)(
++   VkCommandBuffer commandBuffer,
++   uint32_t firstCounterBuffer,
++   uint32_t counterBufferCount,
++   const VkBuffer *pCounterBuffers,
++   const VkDeviceSize *pCounterBufferOffsets)
++{
++   VK_FROM_HANDLE(panvk_cmd_buffer, cmdbuf, commandBuffer);
++   struct panvk_cmd_graphics_state *gfx = &cmdbuf->state.gfx;
++
++   /* Counter buffers ignored in v1 — see VkPhysicalDeviceTransformFeedback
++    * PropertiesEXT.transformFeedbackDraw = false in panvk_vX_physical_device.c.
++    * App is spec-compliant if it does not pass counter buffers (which our
++    * features advertisement allows), but warn loudly if it does so we do not
++    * silently produce wrong capture state. */
++   (void)firstCounterBuffer;
++   (void)pCounterBufferOffsets;
++   if (counterBufferCount > 0 && pCounterBuffers != NULL) {
++      mesa_logw("panvk: CmdBeginTransformFeedbackEXT: counter buffers not "
++                "implemented (transformFeedbackDraw=false); XFB resume will "
++                "restart at buffer offset 0");
++   }
++
++   gfx->xfb.active = true;
++   /* Per-draw set_gfx_sysval picks up the change automatically — no
++    * explicit dirty marking required (set_gfx_sysval uses memcmp +
++    * BITSET to detect state diffs and re-emit sysvals). */
++}
++
++VKAPI_ATTR void VKAPI_CALL
++panvk_per_arch(CmdEndTransformFeedbackEXT)(
++   VkCommandBuffer commandBuffer,
++   uint32_t firstCounterBuffer,
++   uint32_t counterBufferCount,
++   const VkBuffer *pCounterBuffers,
++   const VkDeviceSize *pCounterBufferOffsets)
++{
++   VK_FROM_HANDLE(panvk_cmd_buffer, cmdbuf, commandBuffer);
++   struct panvk_cmd_graphics_state *gfx = &cmdbuf->state.gfx;
++
++   (void)firstCounterBuffer;
++   (void)counterBufferCount;
++   (void)pCounterBuffers;
++   (void)pCounterBufferOffsets;
++
++   gfx->xfb.active = false;
++}

arch/mesa-panvk-bifrost/PKGBUILD

@@ -30,7 +30,7 @@
 
 pkgname=mesa-panvk-bifrost
 _mesaver=26.0.6
-pkgver=26.0.6.r2
+pkgver=26.0.6.r3
 pkgrel=1
 pkgdesc="Patched Mesa libvulkan_panfrost.so exposing Bifrost-gen Mali to Vulkan apps (panvk-bifrost campaign)"
 arch=('aarch64')
@@ -79,6 +79,7 @@ source=(
     "https://archive.mesa3d.org/mesa-${_mesaver}.tar.xz"
     "0001-panvk-expose-robustness2-nullDescriptor-bifrost.patch"
     "0002-panvk-expose-vulkan-1.1-1.2-on-bifrost.patch"
+    "0003-panvk-bifrost-vk-ext-transform-feedback.patch"
     "brave-vulkan"
     "icd.json"
 )
@@ -88,6 +89,7 @@ sha256sums=(
     'SKIP'
     'SKIP'
     'SKIP'
+    'SKIP'
 )
 
 prepare() {
@@ -107,12 +109,23 @@ prepare() {
     sed -i 's|bool has_vk1_1 = PAN_ARCH >= 10;|bool has_vk1_1 = true;|' src/panfrost/vulkan/panvk_vX_physical_device.c
     sed -i 's|bool has_vk1_2 = PAN_ARCH >= 10;|bool has_vk1_2 = true;|' src/panfrost/vulkan/panvk_vX_physical_device.c
 
+    # iter13: VK_EXT_transform_feedback implementation for Bifrost (PAN_ARCH<9).
+    # Applied as a real unified-diff patch — the change is too large for sed.
+    # Phase-doc context: ~/src/panvk-bifrost/phase{4,5,6}_iter13_close.md.
+    # Unlocks ANGLE-Vulkan → GLES3 → WebGL2 / WebGPU on Brave (chrome://gpu
+    # reports "Hardware accelerated" across the board for the affected paths).
+    patch -p1 < "${srcdir}/0003-panvk-bifrost-vk-ext-transform-feedback.patch"
+
     # Sanity-check the patches landed.
     grep -q "KHR_robustness2 = true," src/panfrost/vulkan/panvk_vX_physical_device.c
     grep -q "EXT_robustness2 = true," src/panfrost/vulkan/panvk_vX_physical_device.c
     grep -q "nullDescriptor = true," src/panfrost/vulkan/panvk_vX_physical_device.c
     grep -q "has_vk1_1 = true;" src/panfrost/vulkan/panvk_vX_physical_device.c
     grep -q "has_vk1_2 = true;" src/panfrost/vulkan/panvk_vX_physical_device.c
+    # iter13 sanity:
+    grep -q "EXT_transform_feedback = PAN_ARCH < 9," src/panfrost/vulkan/panvk_vX_physical_device.c
+    grep -q "pan_nir_lower_xfb" src/panfrost/vulkan/panvk_vX_shader.c
+    test -f src/panfrost/vulkan/jm/panvk_vX_cmd_xfb.c
 }
 
 build() {