[Tehshortbus]

What

The synchronous upload in Texture.cpp printed StorageTexture upload submit and StorageTexture upload done unconditionally, around the blocking submit of every storage image upload. Both now use APS5_LOG_CHARS_OUT_DEBUG, as the matching writeback lines do since #12244, so they are compiled out unless APS5_DEBUG_LOG is defined. +2/−2, no behaviour change.

Tested

CachyOS Linux (kernel 7.3.0-4-cachyos), GCC 16.2.1, Release, NVIDIA GeForce RTX 4070 SUPER (615.78.08).

• Focused tests: main f99d7fece3 and this change on it (6eec740e13), with ANYPS5_REQUIRE_VULKAN=1 ANYPS5_NO_SHADER_CACHE=1 VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation ctest --test-dir build-tests -V -R '^(agc_driver_prepared_shaders_indirect|agc_driver_recorder|agc_driver_array_layer_target_cpu|agc_driver_depth_plane_partial_store_buffer_copy|agc_driver_image_table_cache|agc_recorder_cube|agc_storage_split_writeback|agc_driver_bc_block_copy|agc_driver_vertex_target_range_copied|agc_slice_dirty|agc_driver_staged_clear_sampled_read_draw_pipeline|agc_driver_depth_surface_from_storage)$'. These are the twelve tests that print the lines in a full agc_ run on main (1,110 of 1,110 pass there). After rebasing onto 365ad48546 (head 93568c8644; Texture.cpp is unchanged on main in between), the same twelve pass again with no VUID- line and no StorageTexture upload line.

  	main	this change
  tests	12/12 pass	12/12 pass
  VUID- lines	0	0
  was not found but was requested by env var VK_INSTANCE_LAYERS	0	0
  StorageTexture upload lines	286	0

• PPSA26344 (02.002.000), main f99d7fece3 and this change, one run each from launch, with anyps5-entitlements.ini listing the add-on label the title needs. Both runs stop about 116 s in on AGC: Ghost of Yōtei (PPSA26344) stops on a draw that blends unclamped normalized color into two color targets #12796, which this does not touch:

  	StorageTexture upload lines	stdout lines
  main	856	978 (88 % of them)
  this change	0	122

  GTA V (PPSA04264) takes no synchronous storage upload, so its logs have none of these lines on either build.

• python3 tools/check_conventions.py --base origin/main and git diff --check pass.

Overlap and merge order: I merged each pull request tools/pr_overlap.py lists with this PR, and separately with main, and compared the conflicts in Texture.cpp, the only file this PR changes.

• Affected (3), all older, so this lands after them (Depends on below). Whichever of them merges, I rebase this onto it and switch the two lines to APS5_LOG_CHARS_OUT_DEBUG where they end up:
  • feat(agc): back extra guest mip levels with storage images #7085 replaces the vkCmdPipelineBarrier call on the line before upload submit with imageBarrier;
  • perf(agc): expand single-key DCC storage uploads on the GPU from the host import #8484 moves both lines, unchanged, into an if (batch) block;
  • refactor(agc): create, bind and free buffers through the device function table #11920 resolves the same vkCmdPipelineBarrier call through the device function table.
• Not affected (32): merging with this PR gives the same Texture.cpp result as merging with main, either the same conflicts as with main (feat(agc): retain and refresh sampled depth array storage #1368, feat(agc): add experimental depth clear and copy recognition #1395, perf(libSceAgcDriver): keep storage surfaces tracked around unwatching host imports #4448, perf(agc): reuse cached draw stages when only buffer base addresses moved #5609, perf(agc): an untracked comparison memo holds until memory it covers is forgotten #6187, perf(agc): compile uncached draws ahead of queue 0 #6188, perf(agc): fast revalidation over untracked memory proves a surface by its stamp until the next ordering point #6192, perf(agc): import storage surfaces that cross the import window through a joined import #6197, perf(agc): a DCC key proof over untracked memory holds until the next ordering point #6209, fix(agc): keep texture images inside their extent's mip chain #6280, fix(agc): fault image table entries the device or guest memory cannot hold #6481, fix(agc): keep guest stores over freed storage-image memory #6960, fix(agc): sample multisampled targets in read-only layout #6971, feat(agc): render layered draws into a depth view of several slices #7256, feat(agc): trace ranges that leave the write watch #7604, fix(agc): build a partially committed sampled 3D view from its committed slices #10064) or none in this file (feat: run PS4 games on the AGC driver — Limbo (CUSA01664) in game (integration branch) #6977, perf(agc): answer Refresh's registry and host-import lookups from an index and a cache #7514, feat(agc): look up table images at run time through a device image table on the GPU heap (APS5_GPU_PATH) #7642, fix(agc): preserve CPU writes after resident linear target draws #8246, feat(libSceGnmDriver): build PS4 command buffers and submit them #9432, perf(agc): GPU descriptor heap on VK_EXT_descriptor_heap (APS5_GPU_HEAP) #9508, fix(agc): decode PS4 texture bases, linear-aligned targets and thin-tiled 3D textures #9891, feat(agc): draw PS4 multisampled color targets as host-resident targets #9892, feat(agc): write RELEASE_MEM DATA_SEL 4 for PS4 guests as an 800 MHz count #9893, fix(agc): keep PS4 depth planes coherent with buffer reads of their memory #9895, feat(agc): serve lane-keyed, pointer and runtime image-table columns #10193, fix: define decoded AGC state and bound size arithmetic #10249, feat(agc): suballocate sampled and storage image memory #10472, feat(agc): honour the DEPTH row pitch of linear 2D textures as RDNA2 does #11280, fix(agc): do not wait for a storage image's own stamped writes before flushing it #12070, feat(agc): apply a T#'s MIN_LOD clamp in the shader without VK_EXT_image_view_min_lod #12662).

Not checked: a build with APS5_DEBUG_LOG defined (the macro's debug branch is unchanged and already used elsewhere), and Windows or other GPUs.

Checklist

• Based on current main; no other open PR implements the same functions
• One topic per PR; follow-ups go in a new PR
• No comments in code except technical debt
• Unimplemented paths throw (NotImplemented_nid_no_patch); silent stubs are listed under Silent stubs in their library's technical debt file
• No notes, investigation .md files or images added to the repository (attach them to this PR)
• New third-party code is a submodule built from source
• Generic behaviour, not specific to one title
• Depends on: feat(agc): back extra guest mip levels with storage images #7085, perf(agc): expand single-key DCC storage uploads on the GPU from the host import #8484, refactor(agc): create, bind and free buffers through the device function table #11920 (merge order only: this rebases onto each as it lands)
• Fixes: none
• Unblocks: none
• GPUs tested: NVIDIA GeForce RTX 4070 SUPER, driver 615.78.08, CachyOS Linux (kernel 7.3.0-4-cachyos)
• AI-assisted: yes (Claude Code)

🤖 Generated with Claude Code

[github-actions]

Some comparisons could not be completed because a pull request conflicts with the base branch. Review these shared files manually before concluding that there is no overlap.

Pull request	Shared files not compared
PR 1368	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 1395	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 5609	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 6187	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 6188	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 6192	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 6197	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 6209	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 6481	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 7256	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 7514	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 7604	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 7642	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 8246	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 9508	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 10193	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 10249	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 11444	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 13486	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 13796	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 13863	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 13920	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 13928	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 13929	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 14943	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp
PR 15529	core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp

[Tehshortbus]

Checked the overlap report by hand. This needs no Depends on:

• refactor(agc): create, bind and free buffers through the device function table #11920 (Texture.cpp L2963-2964): it changes the vkCmdPipelineBarrier call on the line before upload submit to resolve it through the device function table, and leaves the two log lines alone. The changes are independent; whichever merges later rebases that line.
• The 34 it could not compare: I searched each one's diff for StorageTexture upload submit and StorageTexture upload done. Only perf(agc): expand single-key DCC storage uploads on the GPU from the host import #8484 touches them: it moves both lines, unchanged, into an if (batch) block. If perf(agc): expand single-key DCC storage uploads on the GPU from the host import #8484 merges first, this rebases to switch them to APS5_LOG_CHARS_OUT_DEBUG inside that block. None of the others edits either line.

AI-assisted: yes (Claude Code)

[Tehshortbus]

A correction and a complete answer to the overlap report, since my first reply only searched the diffs for the two lines and missed #7085.

I merged each of the 35 listed pull requests with this PR, and separately with main 365ad48546, and compared the conflicts in core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp, the only file this PR changes.

Affected by this PR (3). All three are older, so this lands after them, recorded in Depends on. Whichever of them merges, I rebase this onto it and switch the two lines to APS5_LOG_CHARS_OUT_DEBUG where they end up:

• feat(agc): back extra guest mip levels with storage images #7085 (Amoenus): replaces the vkCmdPipelineBarrier call on the line before upload submit with imageBarrier. The merge with this PR gives one conflict more than with main, around those lines.
• perf(agc): expand single-key DCC storage uploads on the GPU from the host import #8484 (moong15): moves both lines, unchanged, into an if (batch) block. Its conflict at that spot contains them.
• refactor(agc): create, bind and free buffers through the device function table #11920 (milescoviello): resolves the same vkCmdPipelineBarrier call through the device function table. It merges cleanly with main and gets one conflict with this PR, on the adjacent lines.

Not affected (32). Merging with this PR gives the same Texture.cpp result as merging with main:

• the same conflicts as with main, none at these lines: feat(agc): retain and refresh sampled depth array storage #1368, feat(agc): add experimental depth clear and copy recognition #1395, perf(libSceAgcDriver): keep storage surfaces tracked around unwatching host imports #4448, perf(agc): reuse cached draw stages when only buffer base addresses moved #5609, perf(agc): an untracked comparison memo holds until memory it covers is forgotten #6187, perf(agc): compile uncached draws ahead of queue 0 #6188, perf(agc): fast revalidation over untracked memory proves a surface by its stamp until the next ordering point #6192, perf(agc): import storage surfaces that cross the import window through a joined import #6197, perf(agc): a DCC key proof over untracked memory holds until the next ordering point #6209, fix(agc): keep texture images inside their extent's mip chain #6280, fix(agc): fault image table entries the device or guest memory cannot hold #6481, fix(agc): keep guest stores over freed storage-image memory #6960, fix(agc): sample multisampled targets in read-only layout #6971, feat(agc): render layered draws into a depth view of several slices #7256, feat(agc): trace ranges that leave the write watch #7604, fix(agc): build a partially committed sampled 3D view from its committed slices #10064;
• no conflict in Texture.cpp (their conflicts with main are in other files): feat: run PS4 games on the AGC driver — Limbo (CUSA01664) in game (integration branch) #6977, perf(agc): answer Refresh's registry and host-import lookups from an index and a cache #7514, feat(agc): look up table images at run time through a device image table on the GPU heap (APS5_GPU_PATH) #7642, fix(agc): preserve CPU writes after resident linear target draws #8246, feat(libSceGnmDriver): build PS4 command buffers and submit them #9432, perf(agc): GPU descriptor heap on VK_EXT_descriptor_heap (APS5_GPU_HEAP) #9508, fix(agc): decode PS4 texture bases, linear-aligned targets and thin-tiled 3D textures #9891, feat(agc): draw PS4 multisampled color targets as host-resident targets #9892, feat(agc): write RELEASE_MEM DATA_SEL 4 for PS4 guests as an 800 MHz count #9893, fix(agc): keep PS4 depth planes coherent with buffer reads of their memory #9895, feat(agc): serve lane-keyed, pointer and runtime image-table columns #10193, fix: define decoded AGC state and bound size arithmetic #10249, feat(agc): suballocate sampled and storage image memory #10472, feat(agc): honour the DEPTH row pitch of linear 2D textures as RDNA2 does #11280, fix(agc): do not wait for a storage image's own stamped writes before flushing it #12070, feat(agc): apply a T#'s MIN_LOD clamp in the shader without VK_EXT_image_view_min_lod #12662.

AI-assisted: yes (Claude Code)

[TechLevelZero]

✅ NVIDIA GPU test: passed

PR head 93568c8 merged onto main 88e804b · tested 2026-11-02 11:53 UTC · 21 min

GPU	NVIDIA RTX 2000 Ada Generation, driver 595.99.02, Vulkan 1.4.329
Host	Ubuntu 26.04.1 LTS, kernel 7.0.0-38-generic, container ubuntu:24.04
Logs	build and test logs for this run
Build	✅ passed (Release, -DANYPS5_ENABLE_SPIRV_TOOLS=ON)
Tests	✅ 2,497 passed · 4 failed (0 new, 3 also fail on main, 1 unstable on main) · 45 skipped (ANYPS5_REQUIRE_VULKAN=1)
Test setup	CPU-only tests without a GPU; GPU tests (and any CPU-only test that failed or skipped without one) once on the GPU with VK_LAYER_KHRONOS_validation and ANYPS5_NO_SHADER_CACHE=1; main ran the same way
Test selection	2,546 of 3,234 tests run: the ones this change can reach, chosen by tools/select_tests.py from main's build graph
Vulkan validation	✅ no new validation IDs vs main (970 tests run on the GPU)
Slow tests (>30 s, GPU tests timed with validation layers)	⚠️ agc_driver_pipeline_journal 120.1 s, ⚠️ agc_driver_graphics_maps_fallback 30.7 s

Unstable on main (1, these also failed in recent main runs on this machine; not counted against this PR)

agc_driver_pipeline_journal: failed in 2 of the last 10 main runs

Failing on main too (3, not counted against this PR)

agc_driver_image_table
guest_mmap_executable
guest_socket_options_multicast

Validation IDs already present on main (4 IDs in 495 tests)

VUID-VkDeviceCreateInfo-pNext-pNext (495 tests): agc_driver, agc_driver_address_passes, agc_driver_alias_copy_label, agc_driver_alpha_to_coverage, agc_driver_array_layer_target, agc_driver_bc_block_copy, agc_driver_bda_anchor_reads, agc_driver_bda_span_reads, and 487 more
VUID-vkCmdCopyImage-srcImage-01548 (7 tests): agc_driver_depth_clear_hle, agc_driver_depth_plane_buffer_memory, agc_driver_depth_plane_partial_store, agc_driver_depth_plane_sampling, agc_driver_depth_storage_dispatch, agc_driver_depth_surface_from_storage, agc_driver_dispatch_move_depth
VUID-vkCmdCopyImage-srcImage-01551 (7 tests): agc_driver_depth_clear_hle, agc_driver_depth_plane_buffer_memory, agc_driver_depth_plane_partial_store, agc_driver_depth_plane_sampling, agc_driver_depth_storage_dispatch, agc_driver_depth_surface_from_storage, agc_driver_dispatch_move_depth
VUID-VkShaderModuleCreateInfo-pCode-08737 (1 test): agc_driver_image_sample_derivatives

Skipped (45)

agc_driver_bda_reads_shadowed_results: skipped, the device reports watched host memory written when it imports or reads it
agc_driver_bda_reads_shadowed_results_per_build: skipped, the device reports watched host memory written when it imports or reads it
agc_driver_sample_order_grid_reference: skipped, the device does not keep depth drawn at custom sample locations at those locations, even after a transition that names them: the composite of uniform renders holds the guest's depth in 0 of 3005 sloped samples, so no path can be compared with it bit for bit
agc_host_import_sparse: skipped, host imports are unavailable: /dev/udmabuf is not accessible
agc_import_settle_initial_dcc: skipped, the device has no deferred settle of host imports: imports under 1 MiB wait for one only on Windows
agc_unit_shadow_completion_store: skipped, the device reports watched host memory written when it imports or reads it
guest_http_uri_locale: 
guest_net_ctl: skipped, the host has no IPv4 address
guest_video_out_color_space: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_hdr: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_high_frame_rate: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_output_mode: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_resolution_status: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
guest_video_out_wqhd_detection: skipped, no display or Vulkan device: SDL_InitSubSystem(VIDEO | GAMECONTROLLER) failed: wayland,x11 not available
video_out_flip_aftergpu: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_backtoback: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_controls: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_cpuflip: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_cpuflip_in_order: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_dcc: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_deferred: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop_aftergpu: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_eop_main: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_out_flip_fps_limit: skipped, no display or Vulkan device: DisplayWindow: Vulkan support is either not configured in SDL or not available in current SDL video driver (offscreen) or platform
video_o

Reproduce

git fetch origin 93568c86443b91876dabcf7a0eac76d23972016e 88e804b0a7df1f1ac373a9e2f72fb378d925fdd4
git checkout 88e804b0a7df1f1ac373a9e2f72fb378d925fdd4 && git merge --no-ff 93568c86443b91876dabcf7a0eac76d23972016e
git submodule update --init --recursive
cmake -S . -B build -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTING=ON -DANYPS5_ENABLE_SPIRV_TOOLS=ON
cmake --build build && cmake --build build --target libs
VK_DRIVER_FILES=/nonexistent/icd.json ctest --test-dir build -E '^(agc|shader|recompiler|videoout|video_out|guest_video_out|vulkan)|gpu' --timeout 120 --parallel 32 --output-on-failure
ANYPS5_REQUIRE_VULKAN=1 ANYPS5_NO_SHADER_CACHE=1 VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation \
  ctest --test-dir build -R '^(agc|shader|recompiler|videoout|video_out|guest_video_out|vulkan)|gpu' --timeout 120 --parallel 4 --output-on-failure
# failures are then re-run alone: --timeout 360 --parallel 2; new validation IDs: the test alone, twice, --parallel 1

_{Automated build and test on a contributor's NVIDIA machine; it does not replace upstream CI. A new comment is posted for each new push. Full logs.}

[Tehshortbus]

On the NVIDIA GPU test result for 93568c8: I couldn't reproduce the agc_driver_depth_clear_hle failure. Same merge as the report (main ba5b533 + 93568c8, merge c664bf62a0), RTX 4070 SUPER (615.78.08), CachyOS Linux, Release, with ANYPS5_REQUIRE_VULKAN=1 ANYPS5_NO_SHADER_CACHE=1 VK_INSTANCE_LAYERS=VK_LAYER_KHRONOS_validation:

• 3 runs through ctest -R '^agc_driver_depth_clear_hle$' -V and 20 runs of the test executable alone: 23 of 23 pass, with no VUID-vkCmdCopyBuffer-commonparent line and no crash.
• This PR only switches the two StorageTexture upload submit/done lines to APS5_LOG_CHARS_OUT_DEBUG. Their arguments are string literals, and in a build without APS5_DEBUG_LOG the macro is ((void)0). The only runtime difference is two fewer stdout writes around the synchronous upload's submit.
• In the report's rerun log (10-rerun.log), the crash comes after the test creates a second VulkanDevice (device 0x70a34016f1a0). The next vkCmdCopyBuffer uses a source and destination buffer made on the first device (0x59a29c84a990), the validation layer can't find them, and the process crashes. Here the run also creates the second device, but no copy crosses devices.

So this looks like a buffer that outlives or crosses its device in the test's driver-submit cases, and the timing on that machine decides whether it's reused. It doesn't look like an effect of this change. I haven't found the buffer yet; the order of upload completion is my best lead.

AI-assisted: yes (Claude Code)

[Tehshortbus]

The overlap report was updated today with 12 pull requests my earlier reply didn't cover. I merged each with this PR (on main d6401d794e) and separately with main, and compared core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp:

• Not affected (12): the merge with this PR gives the same Texture.cpp as the merge with main, apart from the two StorageTexture upload lines. None of them edits those lines.
  • no conflict in Texture.cpp: feat(agc): preserve initial native multisample color contents #6973, feat(agc): snapshot resident multisample color bytes #6987, feat(agc): publish completed native color snapshots #6989, feat(agc): publish native color before guest memory reads #6995, feat(agc): publish native color for read-only shader buffers #7002, perf(agc): read dropped sampled snapshots straight into their upload staging buffer #9884, fix(agc): refuse the fast revalidation of a sampled texture whose memory became a depth target #11444, feat(agc): sample 1D arrays past maxImageArrayLayers as rows of a 2D image #12246, perf(agc): keep an untracked fast proof across collect epochs on Windows while its write-watch hint has not moved #12921, perf(agc): take an aliased storage image's units on the device without a host import #13251;
  • the same single conflict as with main, away from these lines: fix(agc): save the guest bytes of results that become pending over a surface that is no longer fully write-watched #7048, fix(agc): hash an idle unit when its range leaves the write watch instead of at every upload #10026.
• Still affected (3), unchanged: feat(agc): back extra guest mip levels with storage images #7085, perf(agc): expand single-key DCC storage uploads on the GPU from the host import #8484 and refactor(agc): create, bind and free buffers through the device function table #11920 still conflict at or next to these lines, so Depends on stays as it is.

No code change.

AI-assisted: yes (Claude Code)

[BunnyFl]

Still applies to current main f6232fcfda: both lines are unconditional APS5_LOG_CHARS_OUT at core/libs/prx/libSceAgcDriver/Graphics/src/Texture.cpp:3186 and :3188. APS5_LOG_CHARS_OUT_DEBUG (core/libs/prx/libc/include/general/LogMacros.hpp:123, :126) prints only with APS5_DEBUG_LOG and is ((void)0) otherwise, so the submit/wait between them is unchanged.

Windows 11, WinLibs GCC 15.2.0: the merged Texture.cpp compiles with no warnings. I didn't run a storage-upload test; nothing outside the two log lines changes.