Test "perf/loop-unfolded-once-bad"
Expected: ✋ reject · Size: 4.1 MB · Lines: 84.7 k · lean4export: 3.1.0 · Lean: 4.36.0 · Timeout: 2.00 min · 📄 Declaration · 🔗 Source
loop-unfolded-once with a different test on the right:
allInK 0 N (fun i => g w i)
decide (Nat.allTR (N - 0) (fun k _ => (fun i => g w (i + 1)) (0 + k)) = true)
with N = 300, must be rejected: the right side tests the points
1 … N instead of 0 … N - 1. After the cheap steps the loop bodies
differ under a binder, which refutes nothing by itself, and checkers then
unfold the loop. N is small because official (v4.36.0) refutes in time
roughly cubic in N: about 7 s here, and more than 5 minutes for N = 1000.
| Checker | Result | ⏱️ | 🧠 | |||
|---|---|---|---|---|---|---|
| nano-lean | ✋ | 43.0 ms | (÷331) | 166.5 MB | (+61%) | |
| gruszka | ✋ | 49.9 ms | (÷286) | 93.5 MB | (-10%) | |
| sokonanoda | ✋ | 154 ms | (÷92) | 148.2 MB | (+43%) | |
| rpylean | ✋ | 77.4 ms | (÷184) | 62.4 MB | (-40%) | |
| pestka | ✋ | 128 ms | (÷111) | 74.7 MB | (-28%) | |
| kefir | ✋ | 297 ms | (÷48) | 32.1 MB | (÷3.2) | |
| palladium | ✋ | 103 ms | (÷138) | 71.6 MB | (-31%) | |
| con-ron | ✋ | 362 ms | (÷39) | 98.2 MB | (-5%) | |
| kefir-wasm | ✋ | 980 ms | (÷15) | 138.4 MB | (+34%) | |
| con-agua | ✋ | 443 ms | (÷32) | 52.1 MB | (-50%) | |
| nanobruijn | ✋ | 286 ms | (÷50) | 33.8 MB | (÷3.1) | |
| nanoda | ✋ | 234 ms | (÷61) | 14.9 MB | (÷7.0) | |
| con-leche | ✋ | 384 ms | (÷37) | 40.0 MB | (÷2.6) | |
| pestka-wasm | ✋ | 849 ms | (÷17) | 623.3 MB | (×6.0) | |
| tenet-jvm | ✋ | 2.79 s | (÷5.1) | 275.3 MB | (×2.7) | |
| official-nightly | ✋ | 15.1 s | (+6%) | 109.1 MB | (+5%) | |
| eink0rn | ✋ | 416 ms | (÷34) | 275.4 MB | (×2.7) | |
| official | ✋ | 14.2 s | 103.6 MB | |||
| ind-models | ✋ | 14.3 s | (0%) | 128.7 MB | (+24%) | |
| tenet | ✋ | 25.2 s | (+77%) | 248.7 MB | (×2.4) | |
| tenet-hs | ⌛ | — | — | |||
| kiota | ✋ | 397 ms | (÷36) | 86.3 MB | (-17%) | |
| ocaml4lean | ✋ | 3.37 s | (÷4.2) | 49.0 MB | (÷2.1) | |
| logos-kernel | ✋ | 217 ms | (÷66) | 29.4 MB | (÷3.5) | |
| evmlean | 🚫 | 306 ms | 104.0 MB | |||
| mini | 🚫 | 250 ms | 71.6 MB | |||
| overfull | ⌛ | — | — | |||
| lean4cobol | ✋ | 1.78 s | (÷8.0) | 21.3 MB | (÷4.9) | |
| lean4lean | ✋ | 279 ms | (÷51) | 92.1 MB | (-11%) | |
| canonical-min | ✋ | 641 ms | (÷22) | 1.0 GB | (×10) | |
| nanoclo | ✋ | 648 ms | (÷22) | 118.1 MB | (+14%) | |
| vow-lean-kernel | ✋ | 13.9 s | (-2%) | 41.2 MB | (÷2.5) | |
| lazylean | ✋ | 214 ms | (÷67) | 31.2 MB | (÷3.3) | |
| metalean-slow | 🚫 | 229 ms | 67.4 MB | |||
| metalean-fast | ✋ | 6.51 s | (÷2.2) | 108.7 MB | (+5%) | |
| nanoclo-fortran | ✋ | 1.12 s | (÷13) | 40.8 MB | (÷2.5) | |
| mathgraph | ✋ | 439 ms | (÷32) | 378.2 MB | (×3.7) |