verify memory usage: Unpack hints h on the fly #897

mkannwischer · 2026-01-19T13:41:19Z

This commit splits up the signature unpacking into two parts: unpack_sig_c_z and unpack_sig_h.
This allows delaying unpacking of the hints h until later during the verification which in turn allows re-using the buffer used for the matrix. This cuts memory consumption by L KiB.

Hoisted out from [TEST] Reduce crypto_sign_verify_internal stack usage #751

This commit splits up the signature unpacking into two parts: unpack_sig_c_z and unpack_sig_h. This allows delaying unpacking of the hints h until later during the verification which in turn allows re-using the buffer used for the matrix. This cuts memory consumption by L KiB. - Hoisted out from #751 Signed-off-by: Matthias J. Kannwischer <matthias@kannwischer.eu>

github-actions

Mac Mini (M1, 2020) benchmarks (opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`46204` cycles	`46204` cycles	`1`
`ML-DSA-44 sign`	`131270` cycles	`131278` cycles	`1.00`
`ML-DSA-44 verify`	`47776` cycles	`47765` cycles	`1.00`
`ML-DSA-65 keypair`	`81015` cycles	`81031` cycles	`1.00`
`ML-DSA-65 sign`	`215698` cycles	`215800` cycles	`1.00`
`ML-DSA-65 verify`	`80361` cycles	`80058` cycles	`1.00`
`ML-DSA-87 keypair`	`132160` cycles	`132172` cycles	`1.00`
`ML-DSA-87 sign`	`276871` cycles	`276892` cycles	`1.00`
`ML-DSA-87 verify`	`130562` cycles	`130410` cycles	`1.00`

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

github-actions

Mac Mini (M1, 2020) benchmarks (no-opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`114193` cycles	`114156` cycles	`1.00`
`ML-DSA-44 sign`	`418150` cycles	`418052` cycles	`1.00`
`ML-DSA-44 verify`	`122661` cycles	`122255` cycles	`1.00`
`ML-DSA-65 keypair`	`195474` cycles	`195490` cycles	`1.00`
`ML-DSA-65 sign`	`682605` cycles	`682460` cycles	`1.00`
`ML-DSA-65 verify`	`198053` cycles	`197755` cycles	`1.00`
`ML-DSA-87 keypair`	`322541` cycles	`322642` cycles	`1.00`
`ML-DSA-87 sign`	`864349` cycles	`864696` cycles	`1.00`
`ML-DSA-87 verify`	`328718` cycles	`328644` cycles	`1.00`

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

github-actions

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`113417` cycles	`113348` cycles	`1.00`
`ML-DSA-44 sign`	`356123` cycles	`355932` cycles	`1.00`
`ML-DSA-44 verify`	`117980` cycles	`118018` cycles	`1.00`
`ML-DSA-65 keypair`	`196577` cycles	`196546` cycles	`1.00`
`ML-DSA-65 sign`	`589169` cycles	`589059` cycles	`1.00`
`ML-DSA-65 verify`	`194566` cycles	`194746` cycles	`1.00`
`ML-DSA-87 keypair`	`322422` cycles	`322709` cycles	`1.00`
`ML-DSA-87 sign`	`752735` cycles	`753072` cycles	`1.00`
`ML-DSA-87 verify`	`319942` cycles	`320128` 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: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`35141` cycles	`34372` cycles	`1.02`
`ML-DSA-44 sign`	`120136` cycles	`120027` cycles	`1.00`
`ML-DSA-44 verify`	`38187` cycles	`38103` cycles	`1.00`
`ML-DSA-65 keypair`	`60745` cycles	`60659` cycles	`1.00`
`ML-DSA-65 sign`	`200543` cycles	`200397` cycles	`1.00`
`ML-DSA-65 verify`	`62781` cycles	`62812` cycles	`1.00`
`ML-DSA-87 keypair`	`94010` cycles	`93992` cycles	`1.00`
`ML-DSA-87 sign`	`239553` cycles	`238255` cycles	`1.01`
`ML-DSA-87 verify`	`96178` cycles	`94920` cycles	`1.01`

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

oqs-bot

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

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`93587` cycles	`94201` cycles	`0.99`
`ML-DSA-44 sign`	`333236` cycles	`334218` cycles	`1.00`
`ML-DSA-44 verify`	`99980` cycles	`99867` cycles	`1.00`
`ML-DSA-65 keypair`	`159709` cycles	`160034` cycles	`1.00`
`ML-DSA-65 sign`	`545589` cycles	`545365` cycles	`1.00`
`ML-DSA-65 verify`	`161523` cycles	`160650` cycles	`1.01`
`ML-DSA-87 keypair`	`267333` cycles	`268199` cycles	`1.00`
`ML-DSA-87 sign`	`706937` cycles	`707903` cycles	`1.00`
`ML-DSA-87 verify`	`270917` cycles	`270461` cycles	`1.00`

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

github-actions

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (no-opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`212633` cycles	`212419` cycles	`1.00`
`ML-DSA-44 sign`	`759965` cycles	`759723` cycles	`1.00`
`ML-DSA-44 verify`	`228830` cycles	`228844` cycles	`1.00`
`ML-DSA-65 keypair`	`380315` cycles	`380545` cycles	`1.00`
`ML-DSA-65 sign`	`1253538` cycles	`1252640` cycles	`1.00`
`ML-DSA-65 verify`	`371994` cycles	`371680` cycles	`1.00`
`ML-DSA-87 keypair`	`604406` cycles	`604445` cycles	`1.00`
`ML-DSA-87 sign`	`1597737` cycles	`1594215` cycles	`1.00`
`ML-DSA-87 verify`	`617054` cycles	`618697` 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: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`68963` cycles	`68940` cycles	`1.00`
`ML-DSA-44 sign`	`187390` cycles	`187846` cycles	`1.00`
`ML-DSA-44 verify`	`69100` cycles	`69282` cycles	`1.00`
`ML-DSA-65 keypair`	`119833` cycles	`119446` cycles	`1.00`
`ML-DSA-65 sign`	`300090` cycles	`301247` cycles	`1.00`
`ML-DSA-65 verify`	`115173` cycles	`115447` cycles	`1.00`
`ML-DSA-87 keypair`	`203799` cycles	`202992` cycles	`1.00`
`ML-DSA-87 sign`	`394125` cycles	`393314` cycles	`1.00`
`ML-DSA-87 verify`	`194984` cycles	`195680` 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: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`56510` cycles	`57044` cycles	`0.99`
`ML-DSA-44 sign`	`181510` cycles	`181848` cycles	`1.00`
`ML-DSA-44 verify`	`61226` cycles	`61082` cycles	`1.00`
`ML-DSA-65 keypair`	`98821` cycles	`98510` cycles	`1.00`
`ML-DSA-65 sign`	`298523` cycles	`299264` cycles	`1.00`
`ML-DSA-65 verify`	`100377` cycles	`100372` cycles	`1.00`
`ML-DSA-87 keypair`	`152801` cycles	`152400` cycles	`1.00`
`ML-DSA-87 sign`	`354475` cycles	`355097` cycles	`1.00`
`ML-DSA-87 verify`	`153732` cycles	`153647` 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: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`41579` cycles	`41768` cycles	`1.00`
`ML-DSA-44 sign`	`134840` cycles	`134643` cycles	`1.00`
`ML-DSA-44 verify`	`44687` cycles	`44801` cycles	`1.00`
`ML-DSA-65 keypair`	`72607` cycles	`73157` cycles	`0.99`
`ML-DSA-65 sign`	`215866` cycles	`217074` cycles	`0.99`
`ML-DSA-65 verify`	`75115` cycles	`73831` cycles	`1.02`
`ML-DSA-87 keypair`	`107561` cycles	`108344` cycles	`0.99`
`ML-DSA-87 sign`	`251470` cycles	`252157` cycles	`1.00`
`ML-DSA-87 verify`	`109366` cycles	`110026` cycles	`0.99`

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

oqs-bot

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

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`134907` cycles	`134815` cycles	`1.00`
`ML-DSA-44 sign`	`523677` cycles	`524851` cycles	`1.00`
`ML-DSA-44 verify`	`147315` cycles	`147535` cycles	`1.00`
`ML-DSA-65 keypair`	`226868` cycles	`226816` cycles	`1.00`
`ML-DSA-65 sign`	`860135` cycles	`861519` cycles	`1.00`
`ML-DSA-65 verify`	`234724` cycles	`234798` cycles	`1.00`
`ML-DSA-87 keypair`	`371386` cycles	`372163` cycles	`1.00`
`ML-DSA-87 sign`	`1083150` cycles	`1081624` cycles	`1.00`
`ML-DSA-87 verify`	`383992` cycles	`383383` 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: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`157793` cycles	`158774` cycles	`0.99`
`ML-DSA-44 sign`	`549648` cycles	`553689` cycles	`0.99`
`ML-DSA-44 verify`	`169942` cycles	`170266` cycles	`1.00`
`ML-DSA-65 keypair`	`268891` cycles	`268386` cycles	`1.00`
`ML-DSA-65 sign`	`907860` cycles	`904101` cycles	`1.00`
`ML-DSA-65 verify`	`275702` cycles	`274733` cycles	`1.00`
`ML-DSA-87 keypair`	`448617` cycles	`448848` cycles	`1.00`
`ML-DSA-87 sign`	`1158163` cycles	`1158629` cycles	`1.00`
`ML-DSA-87 verify`	`460215` cycles	`459042` cycles	`1.00`

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

oqs-bot

Graviton4

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`68464` cycles	`68285` cycles	`1.00`
`ML-DSA-44 sign`	`201880` cycles	`201946` cycles	`1.00`
`ML-DSA-44 verify`	`70902` cycles	`70642` cycles	`1.00`
`ML-DSA-65 keypair`	`121022` cycles	`121073` cycles	`1.00`
`ML-DSA-65 sign`	`330661` cycles	`331230` cycles	`1.00`
`ML-DSA-65 verify`	`117884` cycles	`117970` cycles	`1.00`
`ML-DSA-87 keypair`	`198310` cycles	`198361` cycles	`1.00`
`ML-DSA-87 sign`	`426820` cycles	`426811` cycles	`1.00`
`ML-DSA-87 verify`	`194029` cycles	`194224` 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: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`120072` cycles	`121340` cycles	`0.99`
`ML-DSA-44 sign`	`447197` cycles	`450417` cycles	`0.99`
`ML-DSA-44 verify`	`130084` cycles	`131705` cycles	`0.99`
`ML-DSA-65 keypair`	`204835` cycles	`206768` cycles	`0.99`
`ML-DSA-65 sign`	`729102` cycles	`730995` cycles	`1.00`
`ML-DSA-65 verify`	`210498` cycles	`211843` cycles	`0.99`
`ML-DSA-87 keypair`	`337641` cycles	`340681` cycles	`0.99`
`ML-DSA-87 sign`	`923064` cycles	`929434` cycles	`0.99`
`ML-DSA-87 verify`	`346540` cycles	`347157` cycles	`1.00`

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

oqs-bot

Graviton3

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`72250` cycles	`72206` cycles	`1.00`
`ML-DSA-44 sign`	`212030` cycles	`212085` cycles	`1.00`
`ML-DSA-44 verify`	`75754` cycles	`75732` cycles	`1.00`
`ML-DSA-65 keypair`	`127381` cycles	`127427` cycles	`1.00`
`ML-DSA-65 sign`	`350265` cycles	`350098` cycles	`1.00`
`ML-DSA-65 verify`	`125532` cycles	`125335` cycles	`1.00`
`ML-DSA-87 keypair`	`205895` cycles	`208160` cycles	`0.99`
`ML-DSA-87 sign`	`444237` cycles	`448888` cycles	`0.99`
`ML-DSA-87 verify`	`205727` cycles	`205075` cycles	`1.00`

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

oqs-bot

Graviton4 (no-opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`128554` cycles	`128290` cycles	`1.00`
`ML-DSA-44 sign`	`447128` cycles	`447683` cycles	`1.00`
`ML-DSA-44 verify`	`138110` cycles	`138350` cycles	`1.00`
`ML-DSA-65 keypair`	`220383` cycles	`220301` cycles	`1.00`
`ML-DSA-65 sign`	`726363` cycles	`727535` cycles	`1.00`
`ML-DSA-65 verify`	`222829` cycles	`223194` cycles	`1.00`
`ML-DSA-87 keypair`	`365105` cycles	`365100` cycles	`1.00`
`ML-DSA-87 sign`	`926679` cycles	`926074` cycles	`1.00`
`ML-DSA-87 verify`	`372574` cycles	`372724` cycles	`1.00`

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

oqs-bot

Graviton2

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`114074` cycles	`113737` cycles	`1.00`
`ML-DSA-44 sign`	`356790` cycles	`356305` cycles	`1.00`
`ML-DSA-44 verify`	`118382` cycles	`118106` cycles	`1.00`
`ML-DSA-65 keypair`	`196886` cycles	`196661` cycles	`1.00`
`ML-DSA-65 sign`	`590040` cycles	`589237` cycles	`1.00`
`ML-DSA-65 verify`	`194865` cycles	`194763` cycles	`1.00`
`ML-DSA-87 keypair`	`323008` cycles	`323538` cycles	`1.00`
`ML-DSA-87 sign`	`753229` cycles	`754524` cycles	`1.00`
`ML-DSA-87 verify`	`320370` cycles	`320437` cycles	`1.00`

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

oqs-bot

Graviton3 (no-opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`138678` cycles	`138511` cycles	`1.00`
`ML-DSA-44 sign`	`483594` cycles	`484146` cycles	`1.00`
`ML-DSA-44 verify`	`148450` cycles	`148699` cycles	`1.00`
`ML-DSA-65 keypair`	`242267` cycles	`242282` cycles	`1.00`
`ML-DSA-65 sign`	`792332` cycles	`792625` cycles	`1.00`
`ML-DSA-65 verify`	`240741` cycles	`241147` cycles	`1.00`
`ML-DSA-87 keypair`	`395864` cycles	`396266` cycles	`1.00`
`ML-DSA-87 sign`	`1012467` cycles	`1012387` cycles	`1.00`
`ML-DSA-87 verify`	`402710` cycles	`402523` cycles	`1.00`

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

oqs-bot

Graviton2 (no-opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`212963` cycles	`212757` cycles	`1.00`
`ML-DSA-44 sign`	`760596` cycles	`760583` cycles	`1.00`
`ML-DSA-44 verify`	`229256` cycles	`229095` cycles	`1.00`
`ML-DSA-65 keypair`	`380644` cycles	`380969` cycles	`1.00`
`ML-DSA-65 sign`	`1252949` cycles	`1254036` cycles	`1.00`
`ML-DSA-65 verify`	`372381` cycles	`371934` cycles	`1.00`
`ML-DSA-87 keypair`	`605930` cycles	`604475` cycles	`1.00`
`ML-DSA-87 sign`	`1596320` cycles	`1595118` cycles	`1.00`
`ML-DSA-87 verify`	`618008` cycles	`618516` cycles	`1.00`

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

github-actions

SpacemiT K1 8 (Banana Pi F3) benchmarks (no-opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`829217` cycles	`827059` cycles	`1.00`
`ML-DSA-44 sign`	`3235321` cycles	`3231963` cycles	`1.00`
`ML-DSA-44 verify`	`922407` cycles	`922300` cycles	`1.00`
`ML-DSA-65 keypair`	`1408325` cycles	`1412898` cycles	`1.00`
`ML-DSA-65 sign`	`5337316` cycles	`5354291` cycles	`1.00`
`ML-DSA-65 verify`	`1478483` cycles	`1480666` cycles	`1.00`
`ML-DSA-87 keypair`	`2309529` cycles	`2312958` cycles	`1.00`
`ML-DSA-87 sign`	`6650066` cycles	`6668645` cycles	`1.00`
`ML-DSA-87 verify`	`2408322` cycles	`2409942` cycles	`1.00`

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

github-actions

Arm Cortex-A72 (Raspberry Pi 4) benchmarks (opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`229171` cycles	`228876` cycles	`1.00`
`ML-DSA-44 sign`	`629704` cycles	`614814` cycles	`1.02`
`ML-DSA-44 verify`	`234317` cycles	`217070` cycles	`1.08`
`ML-DSA-65 keypair`	`386832` cycles	`394718` cycles	`0.98`
`ML-DSA-65 sign`	`1018531` cycles	`991561` cycles	`1.03`
`ML-DSA-65 verify`	`370168` cycles	`376152` cycles	`0.98`
`ML-DSA-87 keypair`	`658975` cycles	`646097` cycles	`1.02`
`ML-DSA-87 sign`	`1406876` cycles	`1332158` cycles	`1.06`
`ML-DSA-87 verify`	`647532` cycles	`625886` cycles	`1.03`

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

github-actions

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'Arm Cortex-A72 (Raspberry Pi 4) benchmarks (opt)'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.03.

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 verify`	`234317` cycles	`217070` cycles	`1.08`
`ML-DSA-87 sign`	`1406876` cycles	`1332158` cycles	`1.06`
`ML-DSA-87 verify`	`647532` cycles	`625886` cycles	`1.03`

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

github-actions

Arm Cortex-A55 (Snapdragon 888) benchmarks (opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`272086` cycles	`273072` cycles	`1.00`
`ML-DSA-44 sign`	`799452` cycles	`805254` cycles	`0.99`
`ML-DSA-44 verify`	`272361` cycles	`273857` cycles	`0.99`
`ML-DSA-65 keypair`	`468476` cycles	`468169` cycles	`1.00`
`ML-DSA-65 sign`	`1317665` cycles	`1319237` cycles	`1.00`
`ML-DSA-65 verify`	`451223` cycles	`449707` cycles	`1.00`
`ML-DSA-87 keypair`	`805715` cycles	`804538` cycles	`1.00`
`ML-DSA-87 sign`	`1808353` cycles	`1809292` cycles	`1.00`
`ML-DSA-87 verify`	`778420` cycles	`779109` cycles	`1.00`

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

github-actions

Arm Cortex-A72 (Raspberry Pi 4) benchmarks (no-opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`309867` cycles	`312735` cycles	`0.99`
`ML-DSA-44 sign`	`1154040` cycles	`1199505` cycles	`0.96`
`ML-DSA-44 verify`	`333978` cycles	`345578` cycles	`0.97`
`ML-DSA-65 keypair`	`565534` cycles	`582435` cycles	`0.97`
`ML-DSA-65 sign`	`1932069` cycles	`1976366` cycles	`0.98`
`ML-DSA-65 verify`	`526835` cycles	`554587` cycles	`0.95`
`ML-DSA-87 keypair`	`856162` cycles	`873951` cycles	`0.98`
`ML-DSA-87 sign`	`2425164` cycles	`2499834` cycles	`0.97`
`ML-DSA-87 verify`	`882491` cycles	`890351` cycles	`0.99`

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

github-actions

Arm Cortex-A55 (Snapdragon 888) benchmarks (no-opt)

Details

Benchmark suite	Current: `f6503be`	Previous: `8933db9`	Ratio
`ML-DSA-44 keypair`	`462570` cycles	`461655` cycles	`1.00`
`ML-DSA-44 sign`	`2138045` cycles	`2136241` cycles	`1.00`
`ML-DSA-44 verify`	`547668` cycles	`547092` cycles	`1.00`
`ML-DSA-65 keypair`	`774877` cycles	`774826` cycles	`1.00`
`ML-DSA-65 sign`	`3499244` cycles	`3500411` cycles	`1.00`
`ML-DSA-65 verify`	`849141` cycles	`850053` cycles	`1.00`
`ML-DSA-87 keypair`	`1251533` cycles	`1249873` cycles	`1.00`
`ML-DSA-87 sign`	`4339029` cycles	`4335460` cycles	`1.00`
`ML-DSA-87 verify`	`1364449` cycles	`1366205` cycles	`1.00`

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

oqs-bot · 2026-01-19T14:27:40Z

CBMC Results (ML-DSA-44)

⚠️ Attention Required

Proof	Status	Current	Previous	Change
`sign_verify_internal`	❌	-	162s	-

Full Results (173 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	1731s	1942s	-10.9%
`mld_attempt_signature_generation`	✅	220s	207s	+6%
`polyvecl_pointwise_acc_montgomery_c`	✅	206s	217s	-5%
`poly_pointwise_montgomery_c`	✅	129s	141s	-9%
`rej_uniform_native`	✅	125s	130s	-4%
`mld_ct_memcmp`	✅	78s	85s	-8%
`mld_invntt_layer`	✅	71s	74s	-4%
`keccak_squeezeblocks_x4`	✅	42s	44s	-5%
`mld_ntt_layer`	✅	40s	45s	-11%
`sign_signature_internal`	✅	32s	35s	-9%
`rej_uniform`	✅	19s	23s	-17%
`polyvec_matrix_expand`	✅	18s	19s	-5%
`rej_uniform_c`	✅	18s	18s	+0%
`fqmul`	✅	17s	21s	-19%
`poly_uniform_eta_4x`	✅	17s	16s	+6%
`polymat_permute_bitrev_to_custom`	✅	17s	15s	+13%
`poly_add`	✅	16s	14s	+14%
`poly_chknorm_c`	✅	16s	19s	-16%
`poly_uniform_4x`	✅	14s	14s	+0%
`polyt0_unpack`	✅	14s	16s	-12%
`keccakf1600x4_permute_native`	✅	13s	13s	+0%
`keccak_absorb_once_x4`	✅	12s	15s	-20%
`mld_ntt_butterfly_block`	✅	12s	12s	+0%
`polyeta_unpack`	✅	11s	12s	-8%
`polyz_unpack_c`	✅	10s	13s	-23%
`keccakf1600_permute`	✅	9s	6s	+50%
`keccakf1600_permute_native`	✅	9s	8s	+12%
`mld_compute_t0_t1_tr_from_sk_components`	✅	9s	15s	-40%
`poly_invntt_tomont_c`	✅	9s	10s	-10%
`polyvec_matrix_expand_serial`	✅	8s	7s	+14%
`polyvec_matrix_pointwise_montgomery`	✅	8s	8s	+0%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	7s	9s	-22%
`polyveck_chknorm`	✅	7s	4s	+75%
`polyveck_decompose`	✅	7s	8s	-12%
`polyvecl_uniform_gamma1`	✅	7s	4s	+75%
`mld_check_pct`	✅	6s	6s	+0%
`mld_sample_s1_s2_serial`	✅	6s	3s	+100%
`poly_challenge`	✅	6s	5s	+20%
`poly_decompose`	✅	6s	4s	+50%
`polyveck_ntt`	✅	6s	4s	+50%
`polyveck_pack_t0`	✅	6s	3s	+100%
`polyveck_power2round`	✅	6s	7s	-14%
`polyveck_shiftl`	✅	6s	4s	+50%
`polyveck_sub`	✅	6s	6s	+0%
`polyvecl_ntt`	✅	6s	5s	+20%
`sign`	✅	6s	6s	+0%
`sign_keypair`	✅	6s	5s	+20%
`sign_open`	✅	6s	4s	+50%
`mld_compute_pack_z`	✅	5s	4s	+25%
`mld_sample_s1_s2`	✅	5s	5s	+0%
`pack_sk`	✅	5s	5s	+0%
`poly_uniform`	✅	5s	5s	+0%
`poly_uniform_gamma1_4x`	✅	5s	4s	+25%
`polyveck_add`	✅	5s	8s	-38%
`polyveck_caddq`	✅	5s	4s	+25%
`polyveck_invntt_tomont`	✅	5s	5s	+0%
`polyveck_pointwise_poly_montgomery`	✅	5s	4s	+25%
`polyveck_reduce`	✅	5s	4s	+25%
`polyveck_use_hint`	✅	5s	4s	+25%
`polyvecl_chknorm`	✅	5s	4s	+25%
`polyvecl_unpack_z`	✅	5s	5s	+0%
`rej_eta_native`	✅	5s	3s	+67%
`shake256x4_squeezeblocks`	✅	5s	3s	+67%
`sign_keypair_internal`	✅	5s	6s	-17%
`sign_signature`	✅	5s	4s	+25%
`sign_signature_extmu`	✅	5s	3s	+67%
`keccak_absorb`	✅	4s	6s	-33%
`keccak_finalize`	✅	4s	4s	+0%
`keccakf1600_xor_bytes`	✅	4s	4s	+0%
`make_hint`	✅	4s	4s	+0%
`mld_ct_get_optblocker_i64`	✅	4s	3s	+33%
`mld_ct_get_optblocker_u32`	✅	4s	3s	+33%
`mld_h`	✅	4s	4s	+0%
`montgomery_reduce`	✅	4s	2s	+100%
`ntt_native_x86_64`	✅	4s	3s	+33%
`pack_pk`	✅	4s	3s	+33%
`pack_sig_c_h`	✅	4s	2s	+100%
`pack_sig_z`	✅	4s	1s	+300%
`poly_decompose_c`	✅	4s	2s	+100%
`poly_decompose_native`	✅	4s	3s	+33%
`poly_invntt_tomont`	✅	4s	5s	-20%
`poly_ntt_c`	✅	4s	4s	+0%
`poly_pointwise_montgomery`	✅	4s	3s	+33%
`poly_uniform_eta`	✅	4s	3s	+33%
`poly_use_hint_c`	✅	4s	4s	+0%
`poly_use_hint_native`	✅	4s	7s	-43%
`polyt0_pack`	✅	4s	2s	+100%
`polyt1_unpack`	✅	4s	2s	+100%
`polyveck_make_hint`	✅	4s	4s	+0%
`polyveck_unpack_t0`	✅	4s	3s	+33%
`polyvecl_permute_bitrev_to_custom`	✅	4s	3s	+33%
`polyvecl_pointwise_acc_montgomery`	✅	4s	3s	+33%
`polyvecl_unpack_eta`	✅	4s	3s	+33%
`polyz_pack`	✅	4s	2s	+100%
`reduce32`	✅	4s	2s	+100%
`rej_eta_c`	✅	4s	6s	-33%
`shake128_absorb`	✅	4s	3s	+33%
`shake128_release`	✅	4s	3s	+33%
`sign_pk_from_sk`	✅	4s	4s	+0%
`sign_signature_pre_hash_shake256`	✅	4s	4s	+0%
`sign_verify_extmu`	✅	4s	4s	+0%
`sign_verify_pre_hash_internal`	✅	4s	3s	+33%
`sign_verify_pre_hash_shake256`	✅	4s	3s	+33%
`unpack_sig_h`	✅	4s	-	new
`fqscale`	✅	3s	3s	+0%
`keccak_squeeze`	✅	3s	5s	-40%
`keccakf1600x4_extract_bytes`	✅	3s	1s	+200%
`keccakf1600x4_permute`	✅	3s	4s	-25%
`mld_ct_cmask_nonzero_u8`	✅	3s	4s	-25%
`mld_ct_get_optblocker_u8`	✅	3s	2s	+50%
`mld_ct_sel_int32`	✅	3s	2s	+50%
`mld_keccakf1600_extract_bytes`	✅	3s	3s	+0%
`mld_prepare_domain_separation_prefix`	✅	3s	3s	+0%
`mld_value_barrier_i64`	✅	3s	4s	-25%
`poly_caddq_native`	✅	3s	2s	+50%
`poly_chknorm`	✅	3s	3s	+0%
`poly_invntt_tomont_native`	✅	3s	3s	+0%
`poly_make_hint`	✅	3s	3s	+0%
`poly_ntt`	✅	3s	5s	-40%
`poly_ntt_native`	✅	3s	5s	-40%
`poly_power2round`	✅	3s	3s	+0%
`poly_shiftl`	✅	3s	4s	-25%
`poly_sub`	✅	3s	2s	+50%
`poly_uniform_gamma1`	✅	3s	2s	+50%
`polyeta_pack`	✅	3s	2s	+50%
`polyt1_pack`	✅	3s	4s	-25%
`polyveck_pack_eta`	✅	3s	3s	+0%
`polyvecl_pack_eta`	✅	3s	2s	+50%
`polyvecl_pointwise_acc_montgomery_native`	✅	3s	4s	-25%
`polyz_unpack`	✅	3s	2s	+50%
`polyz_unpack_native`	✅	3s	2s	+50%
`shake128_init`	✅	3s	6s	-50%
`shake256_init`	✅	3s	3s	+0%
`shake256x4_absorb_once`	✅	3s	2s	+50%
`sign_signature_pre_hash_internal`	✅	3s	3s	+0%
`sys_check_capability`	✅	3s	1s	+200%
`unpack_sk`	✅	3s	4s	-25%
`use_hint`	✅	3s	3s	+0%
`decompose`	✅	2s	3s	-33%
`keccak_init`	✅	2s	1s	+100%
`keccakf1600_extract_bytes (big endian)`	✅	2s	3s	-33%
`keccakf1600x4_xor_bytes`	✅	2s	4s	-50%
`mld_ct_abs_i32`	✅	2s	2s	+0%
`mld_ct_cmask_neg_i32`	✅	2s	2s	+0%
`mld_ct_cmask_nonzero_u32`	✅	2s	3s	-33%
`mld_value_barrier_u32`	✅	2s	2s	+0%
`poly_caddq`	✅	2s	3s	-33%
`poly_caddq_c`	✅	2s	3s	-33%
`poly_pointwise_montgomery_native`	✅	2s	3s	-33%
`poly_reduce`	✅	2s	4s	-50%
`poly_use_hint`	✅	2s	3s	-33%
`polyveck_pack_w1`	✅	2s	3s	-33%
`polyveck_unpack_eta`	✅	2s	3s	-33%
`polyvecl_uniform_gamma1_serial`	✅	2s	3s	-33%
`polyw1_pack`	✅	2s	2s	+0%
`power2round`	✅	2s	2s	+0%
`rej_eta`	✅	2s	4s	-50%
`shake128_finalize`	✅	2s	3s	-33%
`shake128_squeeze`	✅	2s	2s	+0%
`shake256_absorb`	✅	2s	2s	+0%
`shake256_finalize`	✅	2s	2s	+0%
`shake256_release`	✅	2s	2s	+0%
`sign_verify`	✅	2s	5s	-60%
`unpack_pk`	✅	2s	4s	-50%
`unpack_sig_c_z`	✅	2s	-	new
`sign_verify_internal`	❌	-	162s	-
`caddq`	✅	1s	3s	-67%
`keccakf1600_xor_bytes (big endian)`	✅	1s	6s	-83%
`mld_value_barrier_u8`	✅	1s	2s	-50%
`poly_chknorm_native`	✅	1s	3s	-67%
`shake128x4_absorb_once`	✅	1s	5s	-80%
`shake128x4_squeezeblocks`	✅	1s	4s	-75%
`shake256`	✅	1s	5s	-80%
`shake256_squeeze`	✅	1s	3s	-67%

oqs-bot · 2026-01-19T14:27:53Z

CBMC Results (ML-DSA-87)

⚠️ Attention Required

Proof	Status	Current	Previous	Change
`sign_verify_internal`	❌	-	140s	-

Full Results (173 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	2103s	2119s	-0.8%
`mld_attempt_signature_generation`	✅	225s	193s	+17%
`polyvecl_pointwise_acc_montgomery_c`	✅	187s	158s	+18%
`polyvec_matrix_expand`	✅	148s	134s	+10%
`poly_pointwise_montgomery_c`	✅	145s	132s	+10%
`rej_uniform_native`	✅	132s	126s	+5%
`polyvec_matrix_expand_serial`	✅	122s	116s	+5%
`mld_ct_memcmp`	✅	85s	77s	+10%
`mld_invntt_layer`	✅	63s	57s	+11%
`keccak_squeezeblocks_x4`	✅	45s	43s	+5%
`mld_ntt_layer`	✅	45s	44s	+2%
`sign_signature_internal`	✅	45s	48s	-6%
`polymat_permute_bitrev_to_custom`	✅	25s	26s	-4%
`rej_uniform`	✅	22s	21s	+5%
`fqmul`	✅	21s	18s	+17%
`mld_compute_t0_t1_tr_from_sk_components`	✅	19s	21s	-10%
`poly_uniform_eta_4x`	✅	18s	17s	+6%
`rej_uniform_c`	✅	18s	17s	+6%
`poly_chknorm_c`	✅	17s	18s	-6%
`poly_uniform_4x`	✅	16s	16s	+0%
`polyeta_unpack`	✅	15s	15s	+0%
`mld_ntt_butterfly_block`	✅	14s	11s	+27%
`poly_add`	✅	14s	10s	+40%
`polyveck_add`	✅	14s	13s	+8%
`keccak_absorb_once_x4`	✅	13s	15s	-13%
`keccakf1600x4_permute_native`	✅	13s	13s	+0%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	13s	12s	+8%
`polyt0_unpack`	✅	13s	12s	+8%
`poly_invntt_tomont_c`	✅	10s	8s	+25%
`polyvecl_ntt`	✅	10s	8s	+25%
`sign`	✅	10s	6s	+67%
`mld_check_pct`	✅	9s	7s	+29%
`polyvec_matrix_pointwise_montgomery`	✅	9s	9s	+0%
`polyveck_decompose`	✅	9s	9s	+0%
`polyveck_use_hint`	✅	9s	9s	+0%
`keccakf1600_permute`	✅	8s	7s	+14%
`keccakf1600_permute_native`	✅	8s	7s	+14%
`poly_decompose_c`	✅	8s	10s	-20%
`polyveck_ntt`	✅	8s	6s	+33%
`polyveck_pointwise_poly_montgomery`	✅	8s	8s	+0%
`polyveck_power2round`	✅	8s	7s	+14%
`polyveck_shiftl`	✅	8s	5s	+60%
`sign_pk_from_sk`	✅	8s	9s	-11%
`unpack_sig_h`	✅	8s	-	new
`mld_sample_s1_s2_serial`	✅	7s	6s	+17%
`polyveck_caddq`	✅	7s	6s	+17%
`polyveck_invntt_tomont`	✅	7s	9s	-22%
`polyveck_reduce`	✅	7s	9s	-22%
`polyveck_sub`	✅	7s	9s	-22%
`polyvecl_uniform_gamma1`	✅	7s	5s	+40%
`keccak_absorb`	✅	6s	4s	+50%
`mld_compute_pack_z`	✅	6s	8s	-25%
`mld_sample_s1_s2`	✅	6s	6s	+0%
`poly_invntt_tomont`	✅	6s	2s	+200%
`poly_use_hint`	✅	6s	4s	+50%
`poly_use_hint_native`	✅	6s	3s	+100%
`polyveck_chknorm`	✅	6s	5s	+20%
`rej_eta_c`	✅	6s	5s	+20%
`sign_keypair_internal`	✅	6s	3s	+100%
`unpack_sk`	✅	6s	7s	-14%
`keccakf1600_xor_bytes`	✅	5s	2s	+150%
`mld_prepare_domain_separation_prefix`	✅	5s	3s	+67%
`montgomery_reduce`	✅	5s	4s	+25%
`poly_challenge`	✅	5s	3s	+67%
`poly_shiftl`	✅	5s	5s	+0%
`poly_uniform`	✅	5s	4s	+25%
`polyveck_unpack_t0`	✅	5s	5s	+0%
`polyvecl_chknorm`	✅	5s	7s	-29%
`polyvecl_pointwise_acc_montgomery_native`	✅	5s	6s	-17%
`polyw1_pack`	✅	5s	4s	+25%
`rej_eta_native`	✅	5s	3s	+67%
`sign_keypair`	✅	5s	4s	+25%
`sign_signature`	✅	5s	2s	+150%
`sign_signature_extmu`	✅	5s	4s	+25%
`sign_signature_pre_hash_internal`	✅	5s	4s	+25%
`decompose`	✅	4s	4s	+0%
`keccak_finalize`	✅	4s	1s	+300%
`keccak_init`	✅	4s	3s	+33%
`keccakf1600_xor_bytes (big endian)`	✅	4s	3s	+33%
`mld_value_barrier_u8`	✅	4s	2s	+100%
`poly_caddq_c`	✅	4s	4s	+0%
`poly_decompose_native`	✅	4s	3s	+33%
`poly_pointwise_montgomery`	✅	4s	5s	-20%
`poly_uniform_eta`	✅	4s	3s	+33%
`poly_uniform_gamma1_4x`	✅	4s	4s	+0%
`polyt0_pack`	✅	4s	4s	+0%
`polyveck_make_hint`	✅	4s	3s	+33%
`polyvecl_pack_eta`	✅	4s	4s	+0%
`polyvecl_unpack_z`	✅	4s	4s	+0%
`polyz_unpack_c`	✅	4s	6s	-33%
`power2round`	✅	4s	4s	+0%
`shake128_release`	✅	4s	3s	+33%
`shake256x4_squeezeblocks`	✅	4s	2s	+100%
`sign_signature_pre_hash_shake256`	✅	4s	6s	-33%
`sign_verify`	✅	4s	4s	+0%
`sign_verify_pre_hash_shake256`	✅	4s	4s	+0%
`unpack_sig_c_z`	✅	4s	-	new
`use_hint`	✅	4s	1s	+300%
`keccak_squeeze`	✅	3s	4s	-25%
`keccakf1600x4_xor_bytes`	✅	3s	1s	+200%
`mld_ct_get_optblocker_i64`	✅	3s	2s	+50%
`mld_ct_get_optblocker_u32`	✅	3s	2s	+50%
`mld_value_barrier_u32`	✅	3s	3s	+0%
`ntt_native_x86_64`	✅	3s	5s	-40%
`pack_pk`	✅	3s	3s	+0%
`pack_sig_c_h`	✅	3s	4s	-25%
`pack_sk`	✅	3s	2s	+50%
`poly_caddq`	✅	3s	5s	-40%
`poly_caddq_native`	✅	3s	4s	-25%
`poly_chknorm_native`	✅	3s	2s	+50%
`poly_invntt_tomont_native`	✅	3s	2s	+50%
`poly_ntt_c`	✅	3s	2s	+50%
`poly_pointwise_montgomery_native`	✅	3s	8s	-62%
`poly_power2round`	✅	3s	4s	-25%
`poly_reduce`	✅	3s	7s	-57%
`poly_use_hint_c`	✅	3s	3s	+0%
`polyeta_pack`	✅	3s	3s	+0%
`polyt1_unpack`	✅	3s	3s	+0%
`polyveck_pack_eta`	✅	3s	3s	+0%
`polyveck_pack_t0`	✅	3s	4s	-25%
`polyveck_pack_w1`	✅	3s	3s	+0%
`polyvecl_permute_bitrev_to_custom`	✅	3s	4s	-25%
`polyvecl_pointwise_acc_montgomery`	✅	3s	1s	+200%
`polyvecl_unpack_eta`	✅	3s	4s	-25%
`polyz_pack`	✅	3s	6s	-50%
`polyz_unpack`	✅	3s	3s	+0%
`polyz_unpack_native`	✅	3s	3s	+0%
`reduce32`	✅	3s	2s	+50%
`shake128_absorb`	✅	3s	2s	+50%
`shake128_finalize`	✅	3s	4s	-25%
`shake128_init`	✅	3s	3s	+0%
`shake128x4_absorb_once`	✅	3s	3s	+0%
`shake128x4_squeezeblocks`	✅	3s	3s	+0%
`shake256`	✅	3s	2s	+50%
`shake256_finalize`	✅	3s	2s	+50%
`shake256x4_absorb_once`	✅	3s	5s	-40%
`sign_open`	✅	3s	6s	-50%
`sign_verify_extmu`	✅	3s	5s	-40%
`sign_verify_pre_hash_internal`	✅	3s	4s	-25%
`unpack_pk`	✅	3s	1s	+200%
`caddq`	✅	2s	2s	+0%
`fqscale`	✅	2s	3s	-33%
`keccakf1600_extract_bytes (big endian)`	✅	2s	3s	-33%
`keccakf1600x4_extract_bytes`	✅	2s	2s	+0%
`keccakf1600x4_permute`	✅	2s	4s	-50%
`make_hint`	✅	2s	4s	-50%
`mld_ct_abs_i32`	✅	2s	3s	-33%
`mld_ct_cmask_nonzero_u8`	✅	2s	6s	-67%
`mld_ct_get_optblocker_u8`	✅	2s	3s	-33%
`mld_ct_sel_int32`	✅	2s	1s	+100%
`mld_h`	✅	2s	4s	-50%
`mld_keccakf1600_extract_bytes`	✅	2s	3s	-33%
`mld_value_barrier_i64`	✅	2s	4s	-50%
`pack_sig_z`	✅	2s	3s	-33%
`poly_chknorm`	✅	2s	3s	-33%
`poly_decompose`	✅	2s	4s	-50%
`poly_make_hint`	✅	2s	3s	-33%
`poly_ntt`	✅	2s	3s	-33%
`poly_ntt_native`	✅	2s	3s	-33%
`poly_sub`	✅	2s	2s	+0%
`poly_uniform_gamma1`	✅	2s	4s	-50%
`polyt1_pack`	✅	2s	3s	-33%
`polyveck_unpack_eta`	✅	2s	3s	-33%
`polyvecl_uniform_gamma1_serial`	✅	2s	4s	-50%
`rej_eta`	✅	2s	3s	-33%
`shake128_squeeze`	✅	2s	2s	+0%
`shake256_absorb`	✅	2s	2s	+0%
`shake256_init`	✅	2s	2s	+0%
`shake256_release`	✅	2s	5s	-60%
`sys_check_capability`	✅	2s	3s	-33%
`sign_verify_internal`	❌	-	140s	-
`mld_ct_cmask_neg_i32`	✅	1s	2s	-50%
`mld_ct_cmask_nonzero_u32`	✅	1s	3s	-67%
`shake256_squeeze`	✅	1s	2s	-50%

oqs-bot · 2026-01-19T14:27:59Z

CBMC Results (ML-DSA-65)

⚠️ Attention Required

Proof	Status	Current	Previous	Change
`sign_verify_internal`	❌	-	97s	-

Full Results (173 proofs)

Proof	Status	Current	Previous	Change
`TOTAL`	✅	2174s	2163s	+0.5%
`mld_attempt_signature_generation`	✅	272s	240s	+13%
`polyvecl_pointwise_acc_montgomery_c`	✅	248s	227s	+9%
`poly_pointwise_montgomery_c`	✅	146s	137s	+7%
`polyvec_matrix_expand`	✅	138s	131s	+5%
`rej_uniform_native`	✅	136s	130s	+5%
`mld_ct_memcmp`	✅	85s	86s	-1%
`mld_invntt_layer`	✅	66s	63s	+5%
`polyvec_matrix_expand_serial`	✅	58s	59s	-2%
`mld_ntt_layer`	✅	47s	45s	+4%
`keccak_squeezeblocks_x4`	✅	46s	45s	+2%
`sign_signature_internal`	✅	45s	48s	-6%
`polyveck_ntt`	✅	24s	25s	-4%
`fqmul`	✅	22s	20s	+10%
`polymat_permute_bitrev_to_custom`	✅	21s	20s	+5%
`rej_uniform`	✅	21s	21s	+0%
`rej_uniform_c`	✅	21s	19s	+11%
`poly_uniform_eta_4x`	✅	17s	17s	+0%
`polyveck_decompose`	✅	17s	16s	+6%
`mld_compute_t0_t1_tr_from_sk_components`	✅	16s	17s	-6%
`polyt0_unpack`	✅	15s	11s	+36%
`poly_uniform_4x`	✅	14s	16s	-12%
`keccak_absorb_once_x4`	✅	13s	13s	+0%
`keccakf1600x4_permute_native`	✅	13s	13s	+0%
`mld_ntt_butterfly_block`	✅	13s	14s	-7%
`poly_chknorm_c`	✅	12s	13s	-8%
`polyvec_matrix_pointwise_montgomery`	✅	12s	14s	-14%
`polyveck_add`	✅	11s	11s	+0%
`polyveck_power2round`	✅	11s	11s	+0%
`polyveck_pointwise_poly_montgomery`	✅	10s	7s	+43%
`sign`	✅	10s	11s	-9%
`mld_polyvecl_permute_bitrev_to_custom_native`	✅	9s	9s	+0%
`poly_add`	✅	9s	9s	+0%
`polyveck_caddq`	✅	9s	5s	+80%
`polyveck_sub`	✅	9s	10s	-10%
`poly_invntt_tomont_c`	✅	8s	10s	-20%
`polyeta_unpack`	✅	8s	7s	+14%
`polyveck_invntt_tomont`	✅	8s	10s	-20%
`polyveck_use_hint`	✅	8s	7s	+14%
`polyz_unpack_c`	✅	8s	5s	+60%
`sign_open`	✅	8s	5s	+60%
`sign_pk_from_sk`	✅	8s	8s	+0%
`keccak_absorb`	✅	7s	7s	+0%
`keccakf1600_permute`	✅	7s	7s	+0%
`keccakf1600_permute_native`	✅	7s	8s	-12%
`poly_decompose_c`	✅	7s	10s	-30%
`poly_pointwise_montgomery_native`	✅	7s	2s	+250%
`poly_use_hint_c`	✅	7s	5s	+40%
`polyvecl_ntt`	✅	7s	7s	+0%
`sign_keypair_internal`	✅	7s	5s	+40%
`mld_check_pct`	✅	6s	5s	+20%
`mld_compute_pack_z`	✅	6s	6s	+0%
`poly_caddq_c`	✅	6s	5s	+20%
`poly_caddq_native`	✅	6s	7s	-14%
`poly_challenge`	✅	6s	5s	+20%
`poly_make_hint`	✅	6s	5s	+20%
`poly_power2round`	✅	6s	4s	+50%
`poly_use_hint_native`	✅	6s	3s	+100%
`polyeta_pack`	✅	6s	3s	+100%
`polyveck_reduce`	✅	6s	6s	+0%
`polyveck_shiftl`	✅	6s	10s	-40%
`sign_signature_extmu`	✅	6s	5s	+20%
`sign_verify`	✅	6s	4s	+50%
`sign_verify_pre_hash_shake256`	✅	6s	4s	+50%
`decompose`	✅	5s	2s	+150%
`mld_h`	✅	5s	5s	+0%
`mld_keccakf1600_extract_bytes`	✅	5s	2s	+150%
`mld_sample_s1_s2`	✅	5s	5s	+0%
`mld_sample_s1_s2_serial`	✅	5s	6s	-17%
`ntt_native_x86_64`	✅	5s	3s	+67%
`polyt0_pack`	✅	5s	3s	+67%
`polyt1_unpack`	✅	5s	2s	+150%
`polyveck_chknorm`	✅	5s	5s	+0%
`polyveck_unpack_eta`	✅	5s	2s	+150%
`polyvecl_pointwise_acc_montgomery`	✅	5s	2s	+150%
`polyvecl_unpack_eta`	✅	5s	4s	+25%
`rej_eta_c`	✅	5s	6s	-17%
`shake256_release`	✅	5s	3s	+67%
`unpack_sig_h`	✅	5s	-	new
`unpack_sk`	✅	5s	4s	+25%
`caddq`	✅	4s	4s	+0%
`keccakf1600_extract_bytes (big endian)`	✅	4s	3s	+33%
`keccakf1600_xor_bytes`	✅	4s	3s	+33%
`pack_sig_z`	✅	4s	5s	-20%
`pack_sk`	✅	4s	4s	+0%
`poly_decompose`	✅	4s	2s	+100%
`poly_invntt_tomont`	✅	4s	4s	+0%
`poly_invntt_tomont_native`	✅	4s	2s	+100%
`poly_ntt`	✅	4s	3s	+33%
`poly_ntt_native`	✅	4s	5s	-20%
`poly_sub`	✅	4s	2s	+100%
`poly_uniform`	✅	4s	6s	-33%
`poly_uniform_eta`	✅	4s	4s	+0%
`polyveck_make_hint`	✅	4s	6s	-33%
`polyveck_pack_t0`	✅	4s	2s	+100%
`polyvecl_chknorm`	✅	4s	5s	-20%
`polyvecl_pack_eta`	✅	4s	3s	+33%
`polyvecl_pointwise_acc_montgomery_native`	✅	4s	3s	+33%
`polyvecl_uniform_gamma1_serial`	✅	4s	2s	+100%
`polyw1_pack`	✅	4s	5s	-20%
`polyz_unpack_native`	✅	4s	4s	+0%
`rej_eta_native`	✅	4s	6s	-33%
`shake128_init`	✅	4s	2s	+100%
`shake128x4_squeezeblocks`	✅	4s	4s	+0%
`shake256_absorb`	✅	4s	3s	+33%
`shake256x4_squeezeblocks`	✅	4s	3s	+33%
`sign_signature_pre_hash_shake256`	✅	4s	5s	-20%
`sys_check_capability`	✅	4s	3s	+33%
`fqscale`	✅	3s	3s	+0%
`keccak_finalize`	✅	3s	1s	+200%
`keccak_init`	✅	3s	2s	+50%
`keccak_squeeze`	✅	3s	3s	+0%
`keccakf1600x4_extract_bytes`	✅	3s	5s	-40%
`keccakf1600x4_permute`	✅	3s	2s	+50%
`keccakf1600x4_xor_bytes`	✅	3s	3s	+0%
`make_hint`	✅	3s	1s	+200%
`mld_ct_cmask_neg_i32`	✅	3s	2s	+50%
`mld_ct_cmask_nonzero_u32`	✅	3s	2s	+50%
`mld_ct_cmask_nonzero_u8`	✅	3s	3s	+0%
`mld_ct_get_optblocker_i64`	✅	3s	4s	-25%
`mld_ct_get_optblocker_u32`	✅	3s	1s	+200%
`mld_value_barrier_i64`	✅	3s	2s	+50%
`mld_value_barrier_u32`	✅	3s	2s	+50%
`pack_pk`	✅	3s	2s	+50%
`poly_caddq`	✅	3s	3s	+0%
`poly_chknorm_native`	✅	3s	5s	-40%
`poly_decompose_native`	✅	3s	7s	-57%
`poly_ntt_c`	✅	3s	4s	-25%
`poly_pointwise_montgomery`	✅	3s	2s	+50%
`poly_reduce`	✅	3s	4s	-25%
`poly_shiftl`	✅	3s	2s	+50%
`poly_uniform_gamma1`	✅	3s	2s	+50%
`poly_uniform_gamma1_4x`	✅	3s	3s	+0%
`poly_use_hint`	✅	3s	3s	+0%
`polyveck_pack_w1`	✅	3s	5s	-40%
`polyveck_unpack_t0`	✅	3s	2s	+50%
`polyvecl_uniform_gamma1`	✅	3s	4s	-25%
`power2round`	✅	3s	3s	+0%
`reduce32`	✅	3s	4s	-25%
`rej_eta`	✅	3s	3s	+0%
`shake128_finalize`	✅	3s	3s	+0%
`shake128_squeeze`	✅	3s	4s	-25%
`shake256_finalize`	✅	3s	2s	+50%
`shake256_squeeze`	✅	3s	3s	+0%
`shake256x4_absorb_once`	✅	3s	2s	+50%
`sign_signature`	✅	3s	6s	-50%
`sign_signature_pre_hash_internal`	✅	3s	5s	-40%
`sign_verify_pre_hash_internal`	✅	3s	5s	-40%
`unpack_pk`	✅	3s	2s	+50%
`use_hint`	✅	3s	1s	+200%
`keccakf1600_xor_bytes (big endian)`	✅	2s	3s	-33%
`mld_ct_abs_i32`	✅	2s	2s	+0%
`mld_ct_sel_int32`	✅	2s	2s	+0%
`mld_prepare_domain_separation_prefix`	✅	2s	2s	+0%
`mld_value_barrier_u8`	✅	2s	3s	-33%
`montgomery_reduce`	✅	2s	2s	+0%
`pack_sig_c_h`	✅	2s	3s	-33%
`poly_chknorm`	✅	2s	3s	-33%
`polyt1_pack`	✅	2s	3s	-33%
`polyveck_pack_eta`	✅	2s	3s	-33%
`polyvecl_permute_bitrev_to_custom`	✅	2s	3s	-33%
`polyvecl_unpack_z`	✅	2s	3s	-33%
`polyz_pack`	✅	2s	2s	+0%
`polyz_unpack`	✅	2s	5s	-60%
`shake128_absorb`	✅	2s	3s	-33%
`shake128_release`	✅	2s	5s	-60%
`shake128x4_absorb_once`	✅	2s	3s	-33%
`shake256_init`	✅	2s	2s	+0%
`sign_keypair`	✅	2s	3s	-33%
`sign_verify_extmu`	✅	2s	3s	-33%
`unpack_sig_c_z`	✅	2s	-	new
`sign_verify_internal`	❌	-	97s	-
`mld_ct_get_optblocker_u8`	✅	1s	3s	-67%
`shake256`	✅	1s	3s	-67%

mkannwischer force-pushed the verify-stack branch from fa2a1b1 to f6503be Compare January 19, 2026 13:44

mkannwischer added the benchmark label Jan 19, 2026

github-actions bot reviewed Jan 19, 2026

View reviewed changes

oqs-bot reviewed Jan 19, 2026

View reviewed changes

github-actions bot reviewed Jan 19, 2026

View reviewed changes

oqs-bot reviewed Jan 19, 2026

View reviewed changes

github-actions bot reviewed Jan 19, 2026

View reviewed changes

verify memory usage: Unpack hints h on the fly #897

Are you sure you want to change the base?

verify memory usage: Unpack hints h on the fly #897

Uh oh!

Conversation

mkannwischer commented Jan 19, 2026

Uh oh!

github-actions bot left a comment

Choose a reason for hiding this comment

Mac Mini (M1, 2020) benchmarks (opt)

Uh oh!

github-actions bot left a comment

Choose a reason for hiding this comment

Mac Mini (M1, 2020) benchmarks (no-opt)

Uh oh!

github-actions bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (opt)

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

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

Uh oh!

github-actions bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A76 (Raspberry Pi 5) benchmarks (no-opt)

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

Intel Xeon 3rd gen (c6i)

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) (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

Graviton4

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

Graviton3

Uh oh!

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4 (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

Graviton2 (no-opt)

Uh oh!

github-actions bot left a comment