WIP: perf(vk): a draw's write masks from its program's word, not its stages walked for each target #2049
No reviewers
Labels
No labels
area/accessibility
area/audio
area/input
area/install
area/menu
area/renderer
area/saves
area/tooling
area/upscaling
beanstalk: approved
beanstalk: failed
beanstalk: needs approval
beanstalk: running
beanstalk: tested
ci: approved
good first issue
gpu/amd
gpu/igpu
gpu/intel
gpu/nvidia
help wanted
kind
bug
kind
chore
kind
docs
kind
feature
kind
performance
kind
question
needs measurements
platform/linux
platform/macos
platform/steam-deck
platform/windows
preview
priority
critical
priority
high
priority
low
regression
risk
high
risk
low
risk
medium
status
blocked
status
confirmed
status
duplicate
status
invalid
status
needs-info
status
wontfix
ci: approved
No milestone
No project
No assignees
2 participants
Notifications
Due date
No due date set.
Dependencies
No dependencies set
Reference
MagicBeanDev/Longshot!2049
Loading…
Reference in a new issue
No description provided.
Delete branch "siantm/Longshot:perf/vk-write-mask-word"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Summary
target_write_mask(gpu/vk/vk_draw.cpp) walked the program's stages for its fragment stage at every target of every draw that worked out its dynamic state, and at every target indraw_hazards.program_wm_sigalready keeps all that function reads of a program as one word (6 bits a target), read bydyn_pack;write_mask_ofmakes each mask from that word and the channel mask, anddynamic_statereads the word once fordyn_packand the masks. The masks are the same; the result expected is less work a draw.Depends on #2048: it needs only
RX_VK_BIND_CHANGED's switch from #2047, and goes after #2048 so that the next pull request, which needs both, has one base. Their commits show in this diff until they merge.Switch:
RX_VK_BIND_CHANGED(kill, on).=0: the stages walked for each target, main's code.From bd3ea307 on
exp/x86-asm-sync.Read first:
write_mask_ofand its use indynamic_stateanddraw_hazardsingpu/vk/vk_draw.cpp.Testing
Tested: Windows 11, i7-9700, RTX 4060, the builder's llvm-mingw clang 23; head
3562fe6bon #2048's head3432e729(on main @75317e1f). Every check ran throughtests/run_all.py --strict --exactin build jobs of the PC's job queue, in a hidden console (logt-dc3-vk-write-mask-word):tests/test_write_mask_word.py(new:write_mask_ofof the word istarget_write_maskfor every target, stored or not, components and channel masks; the switch indynamic_stateanddraw_hazards);tests/test_same_dyn.py(updated: reads the word asdynamic_statenow takes it): pass.--changed 3432e729; its own are among them): 157 passed, 21 failed. Failed:tests/test_sampler_admission.py,tests/test_fsr_uniform_ownership.py,tests/test_pipeline_play_cache.py,tests/test_draw_stream_visibility.py,tests/test_device_lost_exit.py,tests/test_dev_packs.py,tests/test_dev_ab.py,tests/test_draw_binds.py,tests/test_crash_dialog.py,tests/test_frame_overlap.py,tests/test_linux_pressure.py,tests/test_mesh_conv.py,tests/test_prepare_cache_save.py,tests/test_present_queue_default.py,tests/test_recorder_own_idle.py,tests/test_target_reuse_memory.py,tests/test_upscaler_rows.py,tests/test_vk_occlusion_queries.py: each fails on main @75317e1ftoo (the same checks on main in the same queue batch, logt-dc3-base).tests/test_fallback_buffer_cleanup.py,tests/test_mesh_args_cleanup.pyandtests/test_pipeline_cache_save.pyalso failed in that run, each stopped at its 60-second limit, and passed when run again on this head (logt-dc4-vk-write-mask-word).It has NOT run in the game on this PC yet.
Still to test:
RX_VK_BIND_CHANGED=0and this PR with it unset (on), each warmed a minute in the scene first, then interleaved (base, =0, unset, unset, =0, base), two runs each, expecting the same picture all three ways (no missing colour channel or transparency) and unset the replay's time a frame the same as or lower than the base's (by hand)RX_VK_BIND_CHANGED=0and unset, warmed, in ABBA order, expecting the same picture every way and no crash or validation error (@Beanstalk linux+nvidia, linux+amd, windows+amd, windows+intel, steamdeck: Kokiri Forest a minute, its base PR alone against this PR with RX_VK_BIND_CHANGED=0 and unset, ABBA, the same picture)tests/test_write_mask_word.py,tests/test_same_dyn.pywith--cxx g++and--cxx clang++, expecting PASS (by hand)Measurements
Verdict: not measured: no micro-benchmark of this port and no run in the game on this PC yet; the in-game A/B is in Still to test.
RX_VK_BIND_CHANGEDships on (kind kill), the branch's default; it is not measured on main.Risks
A mask that differs from
target_write_mask's would write the wrong colour channels (a wrong picture) or miss a pass hazard (a flicker); the test compares them for every case. Turned off byRX_VK_BIND_CHANGED=0. The renderer is compiled again after updating.Checklist
Ready for review
main; the title is a Conventional Commit.mise run testpasses (any failure that also fails onmainis named). Not yet: only the checks in Testing ran.main, or Measurements says why there is nothing to measure. Not yet (Measurements).third_party/untouched; nothing GPL-incompatible.When it applies
gpu/vk/switches.def. Its row is in; not measured yet.tests/that reads changed source is updated, not weakened.Assisted-by:/Co-Authored-By:trailers.push_constants pushed the program's whole block (page table, task first, render area, fallback addresses) at every draw whose program reads it, though the draws of a pass push the same block one after another. It now leaves the push out when the last block this thread recorded was the same bytes, for the same stages, into the same command buffer (push_standing). vk_work.h counts every vkCmdPushConstants on the thread (push_epoch), so a push of anyone's in between makes the next draw push again. The block is forgotten where the command buffer's may be something else: at each recording's begin (binds_begin_recording), a masked clear, the temporal upscalers' stand-ins, each dispatch's end (its stand-ins call NGX / FidelityFX, which push past vk_work.h), and (new here, beside the branch's sites) in descriptors_disturbed, which every upscaler and library pass on main calls after recording its own pipelines. A direct mesh draw no longer pushes entry 0's first a second time: it is in the block just sent or left standing. The first of RX_VK_BIND_CHANGED's parts on the branch (bd3ea307, "draw-binds"); its other parts follow as their own commits. Its index-buffer part (bind at the pool buffer's start, draw from the first index) is main's RX_VK_INDEX_REBASE already and is not ported. Switch: RX_VK_BIND_CHANGED, kill, on by default (the branch's default); =0 (or false, off, no) pushes at every draw as before. Read once, at its first use. Row in gpu/vk/switches.def. Not measured on its own on the branch ("the saving is not measured yet"); not measured in play. Test: tests/test_push_changed.py runs push_standing (cut from the source) and reads the wiring: vk_work.h's count, the check before the push, the block kept after it, every place it is forgotten, the mesh first; and the switch's reader runs in fresh processes (unset, 0, off, false, 1, on and another value, twice each, set again after its first read). The three fixtures that compile draw_begin_recording stand in binds_begin_recording. Assisted-by: Claude Code@Beanstalk This changes the renderer so that every draw's dynamic state and pass hazards are tracked with one write-mask word (gpu/vk), behind RX_VK_BIND_CHANGED, a kill switch on by default; RX_VK_BIND_CHANGED=0 is main's way. It is on #2048 and #2047, whose commits are in its diff. It has not run in the game on these PCs yet. Please play Kokiri Forest and Hyrule Field, two minutes each, the PR against main on the same PC, on the PR's side with the switch unset and with RX_VK_BIND_CHANGED=0, with a picture in each compared with main's, on Linux with NVIDIA, Linux with AMD (RADV) and the Steam Deck. The same picture, no crash, device lost, validation error or FATAL line; unset, frame time p99 equal to main's or better and to the switch off.
@siantm only members of
BigBeanscan start the farm. One of them can approve this request by reacting 👍 to it or to this comment.View command line instructions
Checkout
From your project repository, check out a new branch and test the changes.