Repository navigation
Conversation
…ities pass (Closes #7394) Ledger: dqn.t27 and string_dualities.t27 move blocked -> pass, and the new conformance spec builtin_abs_max_min_pi.t27 is a pass row. Corpus on the lab: pass 824 -> 827, blocked 538 -> 535, jit_interp_mismatch 0. Part of #6063 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
gHashTag
added
the
owner-approved-foreign
|
Foreign-line budget (owner rule #7371, merged in #7399): this PR is over it, so it is now a draft with auto-merge off. Master's
The Part of #6063. |
|
Superseded by #7614. It lowers the same builtins ( |
#7394) (#7614) * t27b: @abs, @max, @min and std.math.pi / e lower from a t27 plan (Closes #7394) What t27b decides about these calls is specs/tri/t27b/builtin_plan.t27: which operands each call takes, where the result type comes from, the select chain each call lowers to (rows over numbered slots), the operator of each row, the names a refusal reports, and Zig's digits of std.math.pi and std.math.e. `t27c gen-rust` gives gen/rust/tri/t27b/builtin_plan.rs, mounted in cli/t27b/src/lower.rs the way a64.rs and macho.rs mount theirs. The Rust glue classifies the lowered operands, evaluates each one once and turns the rows into IR selects. A bare `abs(x)` now goes through the same plan; its IR is unchanged. Hand Rust added: lower.rs 40, lower_float.rs 37 (34 deleted), 77 in all, under the #7371 cap of 40 per file and 80 per PR. Generated: 261 lines. Spec: 395 lines, 12 tests, 0 vacuous on `t27c test-report`; 10 of 10 mutants killed. The conformance spec is #7411's, 12 of 12 on the reference. Closes #7394 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * t27b ledger: six specs move to pass, the two new specs are listed (Closes #7394) `tri t27b ratchet` of a private t27b lab run of 0ab8f32 (the deployed lab.py lab_run), compared with the lab's own run of its base 95182e9: the only verdict moves are blocked -> pass for dqn, string_dualities, residual_connection, huber_loss, mse_loss and rmsprop, and both new specs pass. Those eight rows come from `ratchet --bless` of the branch run; every other row is master's. pass 846 -> 854, not_pass 67 -> 61, max_not_pass 67 -> 61; 1065 rows, no duplicate path. Closes #7394 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * t27b: builtin_plan.t27 tests compare strings with std.mem.eql (Closes #7394) The spec's own `same_text` helper had the body of `same` in specs/tri/utils/template.t27, a new duplicate group for tools/dupe_scan.py. The tests now use `std.mem.eql(u8, ...)`, which gen-zig prints as written and t27b lowers, and the helper is gone; the generated Rust loses it too (261 -> 247 lines). Resealed with t27c from master 95182e9 on the t27c lab: 12/12 pass, 0 vacuous, all hashes MATCH; 10 of 10 mutants still killed; on the t27b lab `t27b test --check` gives 12 passed and 103 runtime asserts, so the ledger row stays `pass`. Closes #7394 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * t27b: no conformance test pins the sign of @max / @min of -0.0 and +0.0 (Closes #7394) The t27b-native CI corpus (ubuntu-24.04-arm) had the reference fail amp_equal_operands_give_the_first, so the ratchet called the new conformance spec STALE. Zig 0.16 leaves that sign to the target. Measured on the t27b lab, `zig test` natively and for aarch64-linux-musl under qemu-aarch64: x86_64: max(-0,+0) = -0, max(+0,-0) = +0, min(-0,+0) = -0, min(+0,-0) = +0 aarch64: max gives +0 both ways, min gives -0 both ways The conformance test now asserts that equal operands give an equal value, which both targets do. builtin_plan.t27 says in its header that the first operand is t27b's choice (the x86_64 answer, which the t27b lab's reference gives); its run_plan test of that choice is plain compare and select and passes on both. Both specs: `t27c test-report` 12/12, 0 vacuous; `zig test` for aarch64 under qemu: all 12 pass. Resealed with t27c from master 706788b (bootstrap/src/compiler.rs unchanged since 95182e9); builtin_plan's generated Rust is byte-identical. Closes #7394 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> --------- Co-authored-by: Claude <claude@anthropic.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
1 participant
What
t27b lowers four Zig builtins that gen-zig prints exactly as written:
@abs(x)of an f64 / f32. It reuses the bareabs(x)lowering. An integer operand is still refused, because Zig's@absof aniNis auN.@max(a, b)/@min(a, b):-0.0vs+0.0) give the first, matching Zig 0.16 Debug.std.math.piandstd.math.e: Zig's comptime_float digits, held as binary128 and rounded once where a typed float is needed.The conformance spec came first:
specs/tri/t27b/conformance/builtin_abs_max_min_pi.t27passes 12/12 undert27c test-report, with 0 vacuous tests. Its names are unique in the corpus (amp_/AMP_/T27bConformanceBuiltinAbsMaxMinPi).Ledger
ExprCall(@max)ExprFieldAccessCounts: pass 533 -> 536, not_pass 25 -> 23, max_not_pass 25 -> 23. The file has 821 entries and no duplicate paths.
Corpus (Railway lab t27b-lab, aarch64 under qemu,
t27b corpus specs --jobs 16 --timeout-ms 60000)Per-file diff: three specs move to pass (dqn, string_dualities, the conformance spec). Six others change only their blocker list:
@abs/@minnow lower, and their next blocker is reported.@minof an int and a literal is honestly refused (type mismatch), because Zig narrows that result type.Foreign code
This PR edits
cli/t27b/src/lower.rsandcli/t27b/src/lower_float.rs. They are listed at the top oftools/policy/foreign-exceptions.txt, and the PR carriesowner-approved-foreignunder the owner's standing rule.Closes #7394
Part of #6063
🤖 Generated with Claude Code