Consolidate GPU runtimes and extend JPEG and HTJ2K validation - #80
Draft
jcwal1516 wants to merge 3 commits into
Draft
Consolidate GPU runtimes and extend JPEG and HTJ2K validation#80jcwal1516 wants to merge 3 commits into
jcwal1516 wants to merge 3 commits into
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Metal runtime initialization is split into lazy decode, encode, profiling and buffer-operation groups with separate shader inputs. A new
j2k-mpsgraph-supportpackage owns graph submission lifetime shared with JXR. Compact native encoding reuses checked memory accounting, and obsolete timing fields are deprecated.This review also includes pending JPEG ICC/transfer-syntax work, bounded HTJ2K candidate rate control, Metal IDWT and resident-encoding changes, benchmark instrumentation, release/API inventory updates, and removal of obsolete execution plans.
Resident lossy HT execution is wired into the tile path, and Classic batching respects the private-allocation cap before submission. The previously failing regressions now pass. Lint fixes retain strict gates, and the core/JPEG/Metal/MPSGraph API inventory changes have explicit reviewed fingerprints and migration rationale; snapshots were regenerated with the canonical tools.
Validation on Apple Silicon macOS:
J2K_REQUIRE_METAL_RUNTIME=1 RUST_TEST_THREADS=1 cargo test -p j2k-metal --tests --locked: 519 passed, 26 ignored.cargo xtask metal-compile,cargo xtask clippy, andcargo xtask clippy-strict: passed.cargo xtask semver --write-report: passed, including all pinned compatibility checks.cargo xtask stable-api --write,cargo xtask release-integrity, andcargo xtask repo-lint: passed (101 repository-policy tests).Exact-head GPU validation passed for CUDA and Metal at commit
8bba69d6f820d8c547c8d5b79f1cd68a904313d6: https://github.com/frames-sg/j2k/actions/runs/33944361504. The unsafe benchmark audit and all four command-orchestration integration tests also pass. All hosted PR checks, including the exact-SHA GPU evidence policy and aggregate required gate, passed after the evidence rerun. Publishing the new shared graph package remains a separate registry-release action.