[PPC64LE] poly_tomont: trim dead callee-saved spills (frame 320->224) [full CI] by mkannwischer · Pull Request #1799 · pq-code-package/mlkem-native

mkannwischer · 2026-07-15T01:34:45Z

Running full CI + benchmarks on #1773

The prologue saved and restored v20-v30, but the body only clobbers v23, v24, and v27-v30. v20, v21, v22, v25, and v26 were spilled and reloaded but never used: five dead stxvx/lxvx pairs and 80 bytes of frame. This saves only the registers the body touches and trims the frame from 320 to 224 bytes. The Montgomery multiply is unchanged. The mlkem/ copy was regenerated with scripts/simpasm. Closes #1722. Signed-off-by: Scott Boudreaux <scottbphone12@gmail.com>

oqs-bot

Mac Mini (M1, 2020) benchmarks

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`12318` cycles	`12317` cycles	`1.00`
`ML-KEM-512 encaps`	`14761` cycles	`14761` cycles	`1`
`ML-KEM-512 decaps`	`19316` cycles	`19315` cycles	`1.00`
`ML-KEM-768 keypair`	`21263` cycles	`21263` cycles	`1`
`ML-KEM-768 encaps`	`23509` cycles	`23508` cycles	`1.00`
`ML-KEM-768 decaps`	`30056` cycles	`30054` cycles	`1.00`
`ML-KEM-1024 keypair`	`30314` cycles	`30315` cycles	`1.00`
`ML-KEM-1024 encaps`	`34084` cycles	`34080` cycles	`1.00`
`ML-KEM-1024 decaps`	`43695` cycles	`43695` cycles	`1`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

ppc64le (POWER10) benchmarks

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`38267` cycles	`37977` cycles	`1.01`
`ML-KEM-512 encaps`	`43846` cycles	`43448` cycles	`1.01`
`ML-KEM-512 decaps`	`54160` cycles	`53423` cycles	`1.01`
`ML-KEM-768 keypair`	`67792` cycles	`67467` cycles	`1.00`
`ML-KEM-768 encaps`	`76830` cycles	`76412` cycles	`1.01`
`ML-KEM-768 decaps`	`91045` cycles	`90618` cycles	`1.00`
`ML-KEM-1024 keypair`	`108949` cycles	`108562` cycles	`1.00`
`ML-KEM-1024 encaps`	`119891` cycles	`118622` cycles	`1.01`
`ML-KEM-1024 decaps`	`137980` cycles	`137281` cycles	`1.01`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Arm Cortex-A76 (Raspberry Pi 5) benchmarks

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`28231` cycles	`28242` cycles	`1.00`
`ML-KEM-512 encaps`	`34124` cycles	`34125` cycles	`1.00`
`ML-KEM-512 decaps`	`44476` cycles	`44473` cycles	`1.00`
`ML-KEM-768 keypair`	`47608` cycles	`47611` cycles	`1.00`
`ML-KEM-768 encaps`	`53860` cycles	`53862` cycles	`1.00`
`ML-KEM-768 decaps`	`68525` cycles	`68521` cycles	`1.00`
`ML-KEM-1024 keypair`	`70194` cycles	`70247` cycles	`1.00`
`ML-KEM-1024 encaps`	`78680` cycles	`78704` cycles	`1.00`
`ML-KEM-1024 decaps`	`98310` cycles	`98273` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 4th gen (c7i)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`11900` cycles	`11784` cycles	`1.01`
`ML-KEM-512 encaps`	`13179` cycles	`13079` cycles	`1.01`
`ML-KEM-512 decaps`	`17044` cycles	`17091` cycles	`1.00`
`ML-KEM-768 keypair`	`19386` cycles	`19364` cycles	`1.00`
`ML-KEM-768 encaps`	`20656` cycles	`20596` cycles	`1.00`
`ML-KEM-768 decaps`	`26322` cycles	`26281` cycles	`1.00`
`ML-KEM-1024 keypair`	`28203` cycles	`28109` cycles	`1.00`
`ML-KEM-1024 encaps`	`30419` cycles	`30378` cycles	`1.00`
`ML-KEM-1024 decaps`	`37508` cycles	`37448` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 4th gen (c7a)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`12832` cycles	`12801` cycles	`1.00`
`ML-KEM-512 encaps`	`14225` cycles	`14237` cycles	`1.00`
`ML-KEM-512 decaps`	`18996` cycles	`18975` cycles	`1.00`
`ML-KEM-768 keypair`	`21586` cycles	`21566` cycles	`1.00`
`ML-KEM-768 encaps`	`22746` cycles	`22732` cycles	`1.00`
`ML-KEM-768 decaps`	`29730` cycles	`29717` cycles	`1.00`
`ML-KEM-1024 keypair`	`30609` cycles	`30514` cycles	`1.00`
`ML-KEM-1024 encaps`	`32599` cycles	`32678` cycles	`1.00`
`ML-KEM-1024 decaps`	`41885` cycles	`41973` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 3rd gen (c6a)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`14454` cycles	`14443` cycles	`1.00`
`ML-KEM-512 encaps`	`15947` cycles	`15927` cycles	`1.00`
`ML-KEM-512 decaps`	`21345` cycles	`21327` cycles	`1.00`
`ML-KEM-768 keypair`	`23796` cycles	`23806` cycles	`1.00`
`ML-KEM-768 encaps`	`25259` cycles	`25250` cycles	`1.00`
`ML-KEM-768 decaps`	`33015` cycles	`33030` cycles	`1.00`
`ML-KEM-1024 keypair`	`33335` cycles	`33344` cycles	`1.00`
`ML-KEM-1024 encaps`	`35858` cycles	`35877` cycles	`1.00`
`ML-KEM-1024 decaps`	`46182` cycles	`46192` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton5

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`13478` cycles	`13452` cycles	`1.00`
`ML-KEM-512 encaps`	`15781` cycles	`15781` cycles	`1`
`ML-KEM-512 decaps`	`21311` cycles	`21312` cycles	`1.00`
`ML-KEM-768 keypair`	`22907` cycles	`22907` cycles	`1`
`ML-KEM-768 encaps`	`25228` cycles	`25225` cycles	`1.00`
`ML-KEM-768 decaps`	`33166` cycles	`33173` cycles	`1.00`
`ML-KEM-1024 keypair`	`33291` cycles	`33232` cycles	`1.00`
`ML-KEM-1024 encaps`	`36952` cycles	`37044` cycles	`1.00`
`ML-KEM-1024 decaps`	`47102` cycles	`47104` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 3rd gen (c6i)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`17904` cycles	`17868` cycles	`1.00`
`ML-KEM-512 encaps`	`20089` cycles	`20057` cycles	`1.00`
`ML-KEM-512 decaps`	`26603` cycles	`26606` cycles	`1.00`
`ML-KEM-768 keypair`	`30018` cycles	`29950` cycles	`1.00`
`ML-KEM-768 encaps`	`33031` cycles	`31996` cycles	`1.03`
`ML-KEM-768 decaps`	`41440` cycles	`41491` cycles	`1.00`
`ML-KEM-1024 keypair`	`43053` cycles	`42078` cycles	`1.02`
`ML-KEM-1024 encaps`	`45071` cycles	`45040` cycles	`1.00`
`ML-KEM-1024 decaps`	`58042` cycles	`57971` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Intel Xeon 3rd gen (c6i)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-768 encaps`	`33031` cycles	`31996` cycles	`1.03`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 4th gen (c7i) (no-opt)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`27637` cycles	`27523` cycles	`1.00`
`ML-KEM-512 encaps`	`35576` cycles	`35607` cycles	`1.00`
`ML-KEM-512 decaps`	`44399` cycles	`44308` cycles	`1.00`
`ML-KEM-768 keypair`	`44163` cycles	`44159` cycles	`1.00`
`ML-KEM-768 encaps`	`54984` cycles	`54927` cycles	`1.00`
`ML-KEM-768 decaps`	`68472` cycles	`68284` cycles	`1.00`
`ML-KEM-1024 keypair`	`67856` cycles	`67850` cycles	`1.00`
`ML-KEM-1024 encaps`	`79098` cycles	`78888` cycles	`1.00`
`ML-KEM-1024 decaps`	`95998` cycles	`96098` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton3

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`18652` cycles	`18665` cycles	`1.00`
`ML-KEM-512 encaps`	`21831` cycles	`21857` cycles	`1.00`
`ML-KEM-512 decaps`	`28832` cycles	`28837` cycles	`1.00`
`ML-KEM-768 keypair`	`31535` cycles	`31576` cycles	`1.00`
`ML-KEM-768 encaps`	`34755` cycles	`34667` cycles	`1.00`
`ML-KEM-768 decaps`	`44805` cycles	`44803` cycles	`1.00`
`ML-KEM-1024 keypair`	`46202` cycles	`46143` cycles	`1.00`
`ML-KEM-1024 encaps`	`51426` cycles	`51481` cycles	`1.00`
`ML-KEM-1024 decaps`	`64959` cycles	`64971` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 4th gen (c7a) (no-opt)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`37180` cycles	`36791` cycles	`1.01`
`ML-KEM-512 encaps`	`42866` cycles	`42839` cycles	`1.00`
`ML-KEM-512 decaps`	`55627` cycles	`55590` cycles	`1.00`
`ML-KEM-768 keypair`	`58040` cycles	`58027` cycles	`1.00`
`ML-KEM-768 encaps`	`66931` cycles	`66923` cycles	`1.00`
`ML-KEM-768 decaps`	`83611` cycles	`83737` cycles	`1.00`
`ML-KEM-1024 keypair`	`88618` cycles	`88594` cycles	`1.00`
`ML-KEM-1024 encaps`	`99182` cycles	`99087` cycles	`1.00`
`ML-KEM-1024 decaps`	`120546` cycles	`120547` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton4

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`17701` cycles	`17686` cycles	`1.00`
`ML-KEM-512 encaps`	`20597` cycles	`20594` cycles	`1.00`
`ML-KEM-512 decaps`	`27023` cycles	`27005` cycles	`1.00`
`ML-KEM-768 keypair`	`29897` cycles	`29896` cycles	`1.00`
`ML-KEM-768 encaps`	`32632` cycles	`32630` cycles	`1.00`
`ML-KEM-768 decaps`	`41899` cycles	`41919` cycles	`1.00`
`ML-KEM-1024 keypair`	`43741` cycles	`43717` cycles	`1.00`
`ML-KEM-1024 encaps`	`48636` cycles	`48686` cycles	`1.00`
`ML-KEM-1024 decaps`	`61382` cycles	`61381` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton5 (no-opt)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`27495` cycles	`27472` cycles	`1.00`
`ML-KEM-512 encaps`	`31694` cycles	`31683` cycles	`1.00`
`ML-KEM-512 decaps`	`40492` cycles	`40493` cycles	`1.00`
`ML-KEM-768 keypair`	`44186` cycles	`44191` cycles	`1.00`
`ML-KEM-768 encaps`	`50712` cycles	`50710` cycles	`1.00`
`ML-KEM-768 decaps`	`62166` cycles	`62158` cycles	`1.00`
`ML-KEM-1024 keypair`	`68106` cycles	`67977` cycles	`1.00`
`ML-KEM-1024 encaps`	`75946` cycles	`76071` cycles	`1.00`
`ML-KEM-1024 decaps`	`90899` cycles	`90823` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

AMD EPYC 3rd gen (c6a) (no-opt)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`40253` cycles	`40230` cycles	`1.00`
`ML-KEM-512 encaps`	`48410` cycles	`48379` cycles	`1.00`
`ML-KEM-512 decaps`	`62049` cycles	`62191` cycles	`1.00`
`ML-KEM-768 keypair`	`62616` cycles	`62555` cycles	`1.00`
`ML-KEM-768 encaps`	`74850` cycles	`74853` cycles	`1.00`
`ML-KEM-768 decaps`	`92457` cycles	`92436` cycles	`1.00`
`ML-KEM-1024 keypair`	`94994` cycles	`95010` cycles	`1.00`
`ML-KEM-1024 encaps`	`110006` cycles	`109881` cycles	`1.00`
`ML-KEM-1024 decaps`	`132463` cycles	`132297` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Intel Xeon 3rd gen (c6i) (no-opt)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`46767` cycles	`46764` cycles	`1.00`
`ML-KEM-512 encaps`	`55608` cycles	`55568` cycles	`1.00`
`ML-KEM-512 decaps`	`71169` cycles	`71228` cycles	`1.00`
`ML-KEM-768 keypair`	`74400` cycles	`74430` cycles	`1.00`
`ML-KEM-768 encaps`	`85935` cycles	`85934` cycles	`1.00`
`ML-KEM-768 decaps`	`107079` cycles	`107113` cycles	`1.00`
`ML-KEM-1024 keypair`	`111483` cycles	`111586` cycles	`1.00`
`ML-KEM-1024 encaps`	`126167` cycles	`126184` cycles	`1.00`
`ML-KEM-1024 decaps`	`152299` cycles	`152270` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton2

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`28213` cycles	`28237` cycles	`1.00`
`ML-KEM-512 encaps`	`34078` cycles	`34121` cycles	`1.00`
`ML-KEM-512 decaps`	`44501` cycles	`44477` cycles	`1.00`
`ML-KEM-768 keypair`	`47627` cycles	`47614` cycles	`1.00`
`ML-KEM-768 encaps`	`53925` cycles	`53872` cycles	`1.00`
`ML-KEM-768 decaps`	`68633` cycles	`68558` cycles	`1.00`
`ML-KEM-1024 keypair`	`70171` cycles	`70259` cycles	`1.00`
`ML-KEM-1024 encaps`	`78584` cycles	`78638` cycles	`1.00`
`ML-KEM-1024 decaps`	`98350` cycles	`98289` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton3 (no-opt)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`38903` cycles	`38899` cycles	`1.00`
`ML-KEM-512 encaps`	`44395` cycles	`44383` cycles	`1.00`
`ML-KEM-512 decaps`	`56201` cycles	`56203` cycles	`1.00`
`ML-KEM-768 keypair`	`62481` cycles	`62495` cycles	`1.00`
`ML-KEM-768 encaps`	`70438` cycles	`70438` cycles	`1`
`ML-KEM-768 decaps`	`86367` cycles	`86368` cycles	`1.00`
`ML-KEM-1024 keypair`	`95778` cycles	`95867` cycles	`1.00`
`ML-KEM-1024 encaps`	`106020` cycles	`106090` cycles	`1.00`
`ML-KEM-1024 decaps`	`126082` cycles	`125739` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton4 (no-opt)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`35296` cycles	`35280` cycles	`1.00`
`ML-KEM-512 encaps`	`40039` cycles	`40025` cycles	`1.00`
`ML-KEM-512 decaps`	`50625` cycles	`50626` cycles	`1.00`
`ML-KEM-768 keypair`	`56805` cycles	`56804` cycles	`1.00`
`ML-KEM-768 encaps`	`64056` cycles	`64045` cycles	`1.00`
`ML-KEM-768 decaps`	`78487` cycles	`78496` cycles	`1.00`
`ML-KEM-1024 keypair`	`87310` cycles	`87379` cycles	`1.00`
`ML-KEM-1024 encaps`	`96602` cycles	`96676` cycles	`1.00`
`ML-KEM-1024 decaps`	`114927` cycles	`114706` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

oqs-bot

Graviton2 (no-opt)

Details

Benchmark suite	Current: `2a7a6c0`	Previous: `cfd1912`	Ratio
`ML-KEM-512 keypair`	`58704` cycles	`58755` cycles	`1.00`
`ML-KEM-512 encaps`	`68452` cycles	`68467` cycles	`1.00`
`ML-KEM-512 decaps`	`87317` cycles	`87387` cycles	`1.00`
`ML-KEM-768 keypair`	`94946` cycles	`94524` cycles	`1.00`
`ML-KEM-768 encaps`	`109088` cycles	`108935` cycles	`1.00`
`ML-KEM-768 decaps`	`134055` cycles	`134030` cycles	`1.00`
`ML-KEM-1024 keypair`	`151095` cycles	`150855` cycles	`1.00`
`ML-KEM-1024 encaps`	`165174` cycles	`165563` cycles	`1.00`
`ML-KEM-1024 decaps`	`199104` cycles	`198753` cycles	`1.00`

This comment was automatically generated by workflow using github-action-benchmark.

mkannwischer requested a review from a team as a code owner July 15, 2026 01:34

mkannwischer added the benchmark this PR should be benchmarked in CI label Jul 15, 2026

mkannwischer mentioned this pull request Jul 15, 2026

[PPC64LE] poly_tomont: trim dead callee-saved spills (frame 320->224) #1773

Open

oqs-bot reviewed Jul 15, 2026

View reviewed changes

mkannwischer closed this Jul 15, 2026

Uh oh!

Conversation

mkannwischer commented Jul 15, 2026

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Mac Mini (M1, 2020) benchmarks

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

ppc64le (POWER10) benchmarks

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A76 (Raspberry Pi 5) benchmarks

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 4th gen (c7a)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton5

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 3rd gen (c6i)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

⚠️ Performance Alert ⚠️

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i) (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 4th gen (c7a) (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton5 (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a) (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 3rd gen (c6i) (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton2

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3 (no-opt)

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4 (no-opt)

Uh oh!