Lean Kernel Arena / perf/ite-taken-branch-bad

Test "perf/ite-taken-branch-bad"

Expected: ✋ reject · Size: 4.3 MB · Lines: 88.8 k · lean4export: 3.1.0 · Lean: 4.36.0 · Timeout: 1.83 min · 📄 Declaration · 🔗 Source

ite-taken-branch with the branch the condition does not take:

(if false = true then piA else tA) = piA

must be rejected: the left side is the permutation table tA, the right side its inverse. A checker must decide the condition rather than assume the branch that matches the other side; refuting tA = piA may evaluate piA on the way.

Checker Result ⏱️ 🧠
nano-lean ✋ 273 ms (÷25) 324.0 MB (÷5.2)
gruszka ✋ 624 ms (÷11) 470.4 MB (÷3.6)
sokonanoda ✋ 1.06 s (÷6.4) 1.2 GB (-27%)
rpylean ✋ 27.7 s (×4.1) 612.4 MB (÷2.8)
pestka ✋ 850 ms (÷8.0) 194.7 MB (÷8.7)
kefir ✋ 2.11 s (÷3.2) 381.2 MB (÷4.4)
palladium ✋ 2.17 s (÷3.1) 881.1 MB (-48%)
con-ron ✋ 4.07 s (-40%) 612.8 MB (÷2.8)
kefir-wasm ✋ 3.34 s (÷2.0) 680.8 MB (÷2.5)
con-agua ✋ 5.49 s (-19%) 720.5 MB (÷2.3)
nanobruijn ✋ 4.61 s (-32%) 652.5 MB (÷2.6)
nanoda ✋ 4.77 s (-30%) 677.9 MB (÷2.5)
con-leche ✋ 5.14 s (-24%) 1.1 GB (-31%)
pestka-wasm ✋ 2.82 s (÷2.4) 794.7 MB (÷2.1)
tenet-jvm ✋ 7.86 s (+16%) 1.9 GB (+18%)
official-nightly ✋ 4.81 s (-29%) 1.7 GB (+2%)
eink0rn ✋ 6.40 s (-6%) 2.0 GB (+19%)
official ✋ 6.77 s 1.7 GB
ind-models ✋ 6.80 s (0%) 1.7 GB (+1%)
tenet ✋ 21.8 s (×3.2) 4.3 GB (×2.6)
tenet-hs ✋ 35.0 s (×5.2) 1.1 GB (-35%)
kiota ✋ 14.5 s (×2.1) 1.8 GB (+7%)
ocaml4lean ✋ 12.5 s (+84%) 1.6 GB (-4%)
logos-kernel ✋ 873 ms (÷7.8) 1.0 GB (-39%)
evmlean 🚫 305 ms 105.8 MB
mini 🚫 59.4 ms 65.8 MB
overfull ⌛ — —
lean4cobol ✋ 1.70 s (÷4.0) 21.6 MB (÷78)
lean4lean ✋ 292 ms (÷23) 91.7 MB (÷18)
canonical-min ✋ 653 ms (÷10) 1.0 GB (-38%)
nanoclo ✋ 867 ms (÷7.8) 567.6 MB (÷3.0)
vow-lean-kernel 🚫 18.8 s 101.2 MB
lazylean ✋ 2.71 s (÷2.5) 369.7 MB (÷4.6)
metalean-slow 🚫 301 ms 67.0 MB
metalean-fast ✋ 11.6 s (+71%) 1.6 GB (-2%)
nanoclo-fortran ✋ 3.44 s (-49%) 1.5 GB (-10%)
mathgraph ✋ 2.50 s (÷2.7) 3.1 GB (+87%)