Repository navigation
Conversation
…set sources (J.2) Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
Inventories for the vendored trees, and a table of pins for files that are named but not held. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
…a64 (J.2) ACL2 x86isa, the Sail x86 models, Sail Arm and Arm's machine-readable specification, each with SOURCE.md and licence; the Intel SDM and Arm's Instructions.json pinned by SHA-256. The READMEs record that the Sail x86 model is a translation of the ACL2 one and the Sail Arm model derives from Arm's ASL. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
…s and benchmark proposal Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
…pin the Arm ARM (J.2) The second transcriptions come from the vendors' manuals (Intel SDM, Arm ARM) written by a different session; the Sail x86 translation is only a check on our reading of ACL2. shina-lab/sail-x86 is removed (single author, unreviewed, floating point external). The Arm Architecture Reference Manual issue M.d is pinned by URL, size and SHA-256, not vendored. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
atiernan
had a problem deploying
to
full-verification
November 16, 2026 15:49 — with
GitHub Actions
Error
The J.3 transcription found the floating-point instructions defined with Russinoff's RTL library (books/rtl/rel11/lib), and ucomisd in machine/instructions/fp/logical.lisp, neither of which J.2 vendored. Both come from the same acl2/acl2 commit, 33b05575. The library is a new suite, acl2-rtl (0.3 MB, BSD-3-Clause), since its paths are relative to books/rtl/rel11, not books/projects/x86isa. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
atiernan
had a problem deploying
to
full-verification
November 16, 2026 22:56 — with
GitHub Actions
Error
…h forbids file access Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…c8' into worktree-agent-a5b3f92b0ee132842
atiernan
had a problem deploying
to
full-verification
November 16, 2026 23:14 — with
GitHub Actions
Error
Miri (Stacked Borrows) reports undefined behaviour in iced-x86 1.22.0's decoder (decoder/handlers/legacy.rs:2006), which failed CI's Miri job. yaxpeax-x86 2.2.1 (0BSD) passes the same tests under Miri. It is used without features: its std would turn on yaxpeax-arch's, which yaxpeax-arm 0.5.0 does not build with. iced-x86 and lazy_static leave the allowlist; yaxpeax-x86 joins it, and its other dependencies are already listed. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
atiernan
had a problem deploying
to
full-verification
November 17, 2026 07:56 — with
GitHub Actions
Error
atiernan
changed the base branch from
claude/js-jit-compiler-next-21ddc8
to
claude/js-vm-spec
November 17, 2026 13:11
atiernan
had a problem deploying
to
full-verification
November 17, 2026 14:00 — with
GitHub Actions
Error
atiernan
had a problem deploying
to
full-verification
November 17, 2026 21:06 — with
GitHub Actions
Error
atiernan
temporarily deployed
to
full-verification
November 18, 2026 13:59 — with
GitHub Actions
Inactive
atiernan
requested a deployment
to
full-verification
November 20, 2026 19:51 — with
GitHub Actions
Queued
This branch is being deployed
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.
JIT plan item J.2, test data: the instruction-set sources that J.3 (
clazel-x64) and J.4 (clazel-a64) are transcribed from, plus test-only reference disassemblers. There is no implementation yet.Stacked on #176. Merge that PR first, or retarget this one to
mainafterwards.New crates
clazel-x64andclazel-a64are empty: Primitive, tier 0, target tier 2.Sources
acl2-x86isa33b05575,books/projects/x86isasubsetacl2-rtl33b05575, Russinoff's RTL library (books/rtl/rel11:lib/and the package), which the x86isa SSE books callsail-x86-from-acl20c6c023dintel-sdmsail-arm5416ef6b,arm-v9.4-asubsetarm-mrsInstructions.json, releaseFAT-2026-09, by SHA-256arm-armEach suite has a
SOURCE.mdand licence, and an xtask inventory table (path, size, SHA-256) inxtask/src/testdata/isa.rs.Decision recorded in
docs/phases/jit.mdThe maintainer made this decision on 2026-11-16, because the formal sources turned out not to be independent:
What the plan now says:
shina-lab/sail-x86was considered and rejected: single author, unreviewed, floating point left external.Test dependencies
Every new package has a
testrow indocs/DEPENDENCIES.md, and the allowlist is now 232 packages.yaxpeax-x86 =2.2.1, used without features. It replacediced-x861.22.0, whose decoder Miri reports as undefined behaviour under Stacked Borrows.yaxpeax-arm =0.5.0andyaxpeax-arch =0.3.3, withbitvec,funty,radium,tapandwyz.For a human to check
arm-mrshash was taken from the kernel.googlesource.com mirror. It should be compared with Arm's own tarball.Verification
cargo xtask testdata: TEST DATA UP TO DATE.cargo xtask check: STATIC CHECKS PASSED.cargo test -p xtask -p clazel-x64 -p clazel-a64passes, and clippy is clean.cargo xtask verify --changedwas not run locally, because the new crates changeCargo.lock. CI's full verification is the check, and its proof jobs need approval.🤖 Generated with Claude Code