Proposal: API Extension #715

mkannwischer · 2025-01-29T00:29:31Z

This PR rebases an old proposal for extending the API to allow operating on validated+expanded keys (splitting out the serialization and deserialization into separate functions which contain the input validation).
Benefit of this is that in case you can keep the keys expanded, you get much better performance. This is particularly useful for an ephemeral use-case (e.g., TLS), where the secret key never has to leave memory.

The benchmarks below show why we should consider doing this: Decapsulation gets up to 3x faster if you can cache the expanded secret key from key generation. (Encapsulation gets even up to 5x faster if you can cache, but I don't think this is useful for any major use case of ML-KEM).

This is far from being ready, but I'd like to kick off a discussion on how the API should look like.
One point I do not like about this anymore is that it changes the signature of the previous API - It would be cleaner to just add new functions and implement the old 3-function API on top of the new API. Downside of that is that we will have a lot of functions.

Once we have consensus on how to do this, there is still some work to do here: I have not addressed C90 complaince, constant-time tests, kat tests, and CBMC proofs. Also currently the user (through mlkem_native.h) requires to know the details of the key structs - this should be changed to be a generic uint8_t array of the right size.

oqs-bot

Arm Cortex-A76 (Raspberry Pi 5) benchmarks

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`28989` cycles	`29044` cycles	`1.00`
`ML-KEM-512 encaps`	`18345` cycles	`34919` cycles	`0.53`
`ML-KEM-512 decaps`	`29401` cycles	`45430` cycles	`0.65`
`ML-KEM-768 keypair`	`49525` cycles	`49260` cycles	`1.01`
`ML-KEM-768 encaps`	`22398` cycles	`55570` cycles	`0.40`
`ML-KEM-768 decaps`	`38088` cycles	`70407` cycles	`0.54`
`ML-KEM-1024 keypair`	`72470` cycles	`71976` cycles	`1.01`
`ML-KEM-1024 encaps`	`27886` cycles	`80729` cycles	`0.35`
`ML-KEM-1024 decaps`	`48885` cycles	`100536` cycles	`0.49`

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

oqs-bot

Intel Xeon 4th gen (c7i)

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`9548` cycles	`9322` cycles	`1.02`
`ML-KEM-512 encaps`	`4648` cycles	`10791` cycles	`0.43`
`ML-KEM-512 decaps`	`8705` cycles	`14879` cycles	`0.59`
`ML-KEM-768 keypair`	`16204` cycles	`16110` cycles	`1.01`
`ML-KEM-768 encaps`	`5153` cycles	`17328` cycles	`0.30`
`ML-KEM-768 decaps`	`11148` cycles	`22971` cycles	`0.49`
`ML-KEM-1024 keypair`	`21778` cycles	`21807` cycles	`1.00`
`ML-KEM-1024 encaps`	`6529` cycles	`23551` cycles	`0.28`
`ML-KEM-1024 decaps`	`14544` cycles	`31213` cycles	`0.47`

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

oqs-bot

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

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`29127` cycles	`28793` cycles	`1.01`
`ML-KEM-512 encaps`	`21816` cycles	`34819` cycles	`0.63`
`ML-KEM-512 decaps`	`32954` cycles	`44962` cycles	`0.73`
`ML-KEM-768 keypair`	`48722` cycles	`48491` cycles	`1.00`
`ML-KEM-768 encaps`	`28989` cycles	`58558` cycles	`0.50`
`ML-KEM-768 decaps`	`44767` cycles	`72121` cycles	`0.62`
`ML-KEM-1024 keypair`	`72995` cycles	`71744` cycles	`1.02`
`ML-KEM-1024 encaps`	`37349` cycles	`86145` cycles	`0.43`
`ML-KEM-1024 decaps`	`57599` cycles	`103558` cycles	`0.56`

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

oqs-bot

AMD EPYC 3rd gen (c6a)

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`16990` cycles	`16963` cycles	`1.00`
`ML-KEM-512 encaps`	`7511` cycles	`18638` cycles	`0.40`
`ML-KEM-512 decaps`	`13017` cycles	`23952` cycles	`0.54`
`ML-KEM-768 keypair`	`28794` cycles	`28897` cycles	`1.00`
`ML-KEM-768 encaps`	`7667` cycles	`29934` cycles	`0.26`
`ML-KEM-768 decaps`	`15695` cycles	`37633` cycles	`0.42`
`ML-KEM-1024 keypair`	`42024` cycles	`42144` cycles	`1.00`
`ML-KEM-1024 encaps`	`9464` cycles	`44254` cycles	`0.21`
`ML-KEM-1024 decaps`	`19905` cycles	`54692` cycles	`0.36`

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

oqs-bot

Intel Xeon 3rd gen (c6i)

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`15875` cycles	`15779` cycles	`1.01`
`ML-KEM-512 encaps`	`7482` cycles	`17843` cycles	`0.42`
`ML-KEM-512 decaps`	`14268` cycles	`24358` cycles	`0.59`
`ML-KEM-768 keypair`	`27408` cycles	`27053` cycles	`1.01`
`ML-KEM-768 encaps`	`8093` cycles	`28661` cycles	`0.28`
`ML-KEM-768 decaps`	`17951` cycles	`38119` cycles	`0.47`
`ML-KEM-1024 keypair`	`36840` cycles	`36575` cycles	`1.01`
`ML-KEM-1024 encaps`	`10331` cycles	`39488` cycles	`0.26`
`ML-KEM-1024 decaps`	`23356` cycles	`51951` cycles	`0.45`

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

oqs-bot

AMD EPYC 4th gen (c7a)

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`11470` cycles	`11577` cycles	`0.99`
`ML-KEM-512 encaps`	`5326` cycles	`13033` cycles	`0.41`
`ML-KEM-512 decaps`	`10347` cycles	`17949` cycles	`0.58`
`ML-KEM-768 keypair`	`20014` cycles	`19635` cycles	`1.02`
`ML-KEM-768 encaps`	`5405` cycles	`20577` cycles	`0.26`
`ML-KEM-768 decaps`	`12744` cycles	`27635` cycles	`0.46`
`ML-KEM-1024 keypair`	`26754` cycles	`26223` cycles	`1.02`
`ML-KEM-1024 encaps`	`7000` cycles	`28249` cycles	`0.25`
`ML-KEM-1024 decaps`	`16757` cycles	`37534` cycles	`0.45`

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

oqs-bot

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

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`39672` cycles	`39696` cycles	`1.00`
`ML-KEM-512 encaps`	`30099` cycles	`48393` cycles	`0.62`
`ML-KEM-512 decaps`	`47036` cycles	`63017` cycles	`0.75`
`ML-KEM-768 keypair`	`65600` cycles	`65547` cycles	`1.00`
`ML-KEM-768 encaps`	`41304` cycles	`77185` cycles	`0.54`
`ML-KEM-768 decaps`	`64125` cycles	`96276` cycles	`0.67`
`ML-KEM-1024 keypair`	`97560` cycles	`98056` cycles	`0.99`
`ML-KEM-1024 encaps`	`53345` cycles	`112583` cycles	`0.47`
`ML-KEM-1024 decaps`	`82066` cycles	`137084` cycles	`0.60`

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

oqs-bot

Graviton4

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`18174` cycles	`18124` cycles	`1.00`
`ML-KEM-512 encaps`	`10438` cycles	`21280` cycles	`0.49`
`ML-KEM-512 decaps`	`17391` cycles	`27915` cycles	`0.62`
`ML-KEM-768 keypair`	`30748` cycles	`30584` cycles	`1.01`
`ML-KEM-768 encaps`	`12009` cycles	`33621` cycles	`0.36`
`ML-KEM-768 decaps`	`21958` cycles	`43181` cycles	`0.51`
`ML-KEM-1024 keypair`	`44751` cycles	`44190` cycles	`1.01`
`ML-KEM-1024 encaps`	`15924` cycles	`49652` cycles	`0.32`
`ML-KEM-1024 decaps`	`29486` cycles	`62638` cycles	`0.47`

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

oqs-bot

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

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`47319` cycles	`47172` cycles	`1.00`
`ML-KEM-512 encaps`	`33047` cycles	`56388` cycles	`0.59`
`ML-KEM-512 decaps`	`51798` cycles	`72631` cycles	`0.71`
`ML-KEM-768 keypair`	`78567` cycles	`77979` cycles	`1.01`
`ML-KEM-768 encaps`	`45414` cycles	`89785` cycles	`0.51`
`ML-KEM-768 decaps`	`70661` cycles	`111101` cycles	`0.64`
`ML-KEM-1024 keypair`	`114683` cycles	`115046` cycles	`1.00`
`ML-KEM-1024 encaps`	`59010` cycles	`130352` cycles	`0.45`
`ML-KEM-1024 decaps`	`91673` cycles	`157863` cycles	`0.58`

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

oqs-bot

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

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`37564` cycles	`37785` cycles	`0.99`
`ML-KEM-512 encaps`	`27686` cycles	`44424` cycles	`0.62`
`ML-KEM-512 decaps`	`43785` cycles	`58614` cycles	`0.75`
`ML-KEM-768 keypair`	`61384` cycles	`61433` cycles	`1.00`
`ML-KEM-768 encaps`	`37479` cycles	`70047` cycles	`0.54`
`ML-KEM-768 decaps`	`59200` cycles	`89231` cycles	`0.66`
`ML-KEM-1024 keypair`	`88513` cycles	`88826` cycles	`1.00`
`ML-KEM-1024 encaps`	`46691` cycles	`101176` cycles	`0.46`
`ML-KEM-1024 decaps`	`73050` cycles	`123493` cycles	`0.59`

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

oqs-bot

Graviton2

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`29015` cycles	`29045` cycles	`1.00`
`ML-KEM-512 encaps`	`18346` cycles	`34950` cycles	`0.52`
`ML-KEM-512 decaps`	`29404` cycles	`45444` cycles	`0.65`
`ML-KEM-768 keypair`	`49526` cycles	`49244` cycles	`1.01`
`ML-KEM-768 encaps`	`22398` cycles	`55555` cycles	`0.40`
`ML-KEM-768 decaps`	`38095` cycles	`70408` cycles	`0.54`
`ML-KEM-1024 keypair`	`72489` cycles	`71971` cycles	`1.01`
`ML-KEM-1024 encaps`	`27885` cycles	`80753` cycles	`0.35`
`ML-KEM-1024 decaps`	`48884` cycles	`100437` cycles	`0.49`

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

oqs-bot

Graviton4 (no-opt)

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`37901` cycles	`37939` cycles	`1.00`
`ML-KEM-512 encaps`	`24585` cycles	`43313` cycles	`0.57`
`ML-KEM-512 decaps`	`38593` cycles	`55545` cycles	`0.69`
`ML-KEM-768 keypair`	`63366` cycles	`63144` cycles	`1.00`
`ML-KEM-768 encaps`	`33427` cycles	`70605` cycles	`0.47`
`ML-KEM-768 decaps`	`52681` cycles	`86945` cycles	`0.61`
`ML-KEM-1024 keypair`	`95216` cycles	`94525` cycles	`1.01`
`ML-KEM-1024 encaps`	`44220` cycles	`105293` cycles	`0.42`
`ML-KEM-1024 decaps`	`69331` cycles	`126582` cycles	`0.55`

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

oqs-bot

Graviton2 (no-opt)

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`60752` cycles	`61376` cycles	`0.99`
`ML-KEM-512 encaps`	`38938` cycles	`69997` cycles	`0.56`
`ML-KEM-512 decaps`	`60893` cycles	`88962` cycles	`0.68`
`ML-KEM-768 keypair`	`102315` cycles	`102056` cycles	`1.00`
`ML-KEM-768 encaps`	`52908` cycles	`114152` cycles	`0.46`
`ML-KEM-768 decaps`	`82771` cycles	`139501` cycles	`0.59`
`ML-KEM-1024 keypair`	`155041` cycles	`154590` cycles	`1.00`
`ML-KEM-1024 encaps`	`67320` cycles	`170334` cycles	`0.40`
`ML-KEM-1024 decaps`	`106294` cycles	`202511` cycles	`0.52`

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

oqs-bot

Graviton3

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`19004` cycles	`18956` cycles	`1.00`
`ML-KEM-512 encaps`	`10896` cycles	`22521` cycles	`0.48`
`ML-KEM-512 decaps`	`18360` cycles	`29631` cycles	`0.62`
`ML-KEM-768 keypair`	`32475` cycles	`32343` cycles	`1.00`
`ML-KEM-768 encaps`	`12794` cycles	`35917` cycles	`0.36`
`ML-KEM-768 decaps`	`23503` cycles	`46302` cycles	`0.51`
`ML-KEM-1024 keypair`	`47062` cycles	`46668` cycles	`1.01`
`ML-KEM-1024 encaps`	`16623` cycles	`52528` cycles	`0.32`
`ML-KEM-1024 decaps`	`31165` cycles	`66312` cycles	`0.47`

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

oqs-bot

Graviton3 (no-opt)

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`39326` cycles	`39346` cycles	`1.00`
`ML-KEM-512 encaps`	`24749` cycles	`45274` cycles	`0.55`
`ML-KEM-512 decaps`	`38758` cycles	`57142` cycles	`0.68`
`ML-KEM-768 keypair`	`65940` cycles	`65866` cycles	`1.00`
`ML-KEM-768 encaps`	`33405` cycles	`73883` cycles	`0.45`
`ML-KEM-768 decaps`	`52545` cycles	`89852` cycles	`0.58`
`ML-KEM-1024 keypair`	`99448` cycles	`99001` cycles	`1.00`
`ML-KEM-1024 encaps`	`43458` cycles	`109983` cycles	`0.40`
`ML-KEM-1024 decaps`	`68410` cycles	`130860` cycles	`0.52`

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

oqs-bot

Bananapi bpi-f3 benchmarks

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`329727` cycles	`331334` cycles	`1.00`
`ML-KEM-512 encaps`	`308683` cycles	`439896` cycles	`0.70`
`ML-KEM-512 decaps`	`472974` cycles	`588436` cycles	`0.80`
`ML-KEM-768 keypair`	`546924` cycles	`548599` cycles	`1.00`
`ML-KEM-768 encaps`	`433011` cycles	`688259` cycles	`0.63`
`ML-KEM-768 decaps`	`648468` cycles	`880050` cycles	`0.74`
`ML-KEM-1024 keypair`	`812647` cycles	`814706` cycles	`1.00`
`ML-KEM-1024 encaps`	`564190` cycles	`988517` cycles	`0.57`
`ML-KEM-1024 decaps`	`830107` cycles	`1222676` cycles	`0.68`

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

oqs-bot

Arm Cortex-A55 (Snapdragon 888) benchmarks

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`58513` cycles	`58238` cycles	`1.00`
`ML-KEM-512 encaps`	`31247` cycles	`65650` cycles	`0.48`
`ML-KEM-512 decaps`	`51216` cycles	`84458` cycles	`0.61`
`ML-KEM-768 keypair`	`100330` cycles	`98763` cycles	`1.02`
`ML-KEM-768 encaps`	`41423` cycles	`110184` cycles	`0.38`
`ML-KEM-768 decaps`	`69816` cycles	`136303` cycles	`0.51`
`ML-KEM-1024 keypair`	`152515` cycles	`150097` cycles	`1.02`
`ML-KEM-1024 encaps`	`51824` cycles	`165912` cycles	`0.31`
`ML-KEM-1024 decaps`	`89795` cycles	`202018` cycles	`0.44`

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

oqs-bot

Arm Cortex-A72 (Raspberry Pi 4) benchmarks

Benchmark suite	Current: `4c5b57e`	Previous: `3dc9642`	Ratio
`ML-KEM-512 keypair`	`51722` cycles	`51590` cycles	`1.00`
`ML-KEM-512 encaps`	`28720` cycles	`60136` cycles	`0.48`
`ML-KEM-512 decaps`	`45557` cycles	`75769` cycles	`0.60`
`ML-KEM-768 keypair`	`88252` cycles	`88608` cycles	`1.00`
`ML-KEM-768 encaps`	`33755` cycles	`96818` cycles	`0.35`
`ML-KEM-768 decaps`	`58361` cycles	`120138` cycles	`0.49`
`ML-KEM-1024 keypair`	`132440` cycles	`131788` cycles	`1.00`
`ML-KEM-1024 encaps`	`42071` cycles	`145353` cycles	`0.29`
`ML-KEM-1024 decaps`	`75130` cycles	`175673` cycles	`0.43`

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

mkannwischer · 2025-02-10T07:51:15Z

@hanno-becker - any thoughts on this?

hanno-becker · 2025-02-18T20:20:13Z

@mkannwischer What stands in the way of keeping the old API and providing the new one as an (optional) addition?

hanno-becker

This is going in a good direction I think. You may defer updating mlkem_native.h for the time being, keeping that to the core API, and using kem.h directly from the test exercising the new API.

mlkem/kem.h

mlkem/kem.c

hanno-becker · 2025-02-25T05:19:24Z

mlkem/kem.c

+
+  crypto_kem_marshal_pk(pk, &pks);
+  crypto_kem_marshal_sk(sk, &sks);
+


Referring to the PCT comment, I think we need to retain the PCT here for the time being.

No - If PCT is enabled, all public APIs need to do PCT - so we cannot move it here, otherwise crypto_kem_keypair_struct doesn't peform PCT. We can perform serialization in the PCT if needed, but that is very expensive.

The current code does not perform the PCT on the key material that is returned to the caller -- it is thus not clear if it is FIPS compliant. Please re-introduce the previous form of the PCT.

I don't know how to best introduce a PCT in the internal-format API (without duplicating the PCT in the standard API) -- but this is secondary to me at the moment, because the FIPS-compliance of that API is still questionable anyway.

Of course, one can always have mlk_indcpa_keypair_derand_internal which does not perform the PCT, and (a) call that directly from crypto_kem_keypair, with the PCT on the standard-keys afterwards, (b) call that from mlk_indcpa_keypair_derand(), with the PCT on the internal keys afterwards. This way, you also have PCTs at every top-level API, without duplication.

To be clear: I support the PCT the way you propose to do it -- but we cannot do it until we know that it's compliant. It would at the least be blocking integration into AWS-LC.

Okay - I have implemented what you said. we should be explicit somewhere that PCT is ONLY implemented when using crypto_kem_keypair, but not when using any of the other functions.
AWS-LC may want to consider changing this in the medium-term - it's much faster if you can work on the internal formats.

AWS-LC may want to consider changing this in the medium-term - it's much faster if you can work on the internal formats.

Of course it is -- but it's about FIPS compliance. When it's clear that the internal PCT is still considered compliant, I'm confident it would be changed.

hanno-becker · 2025-02-26T05:03:13Z

mlkem/kem.h

+#define crypto_kem_keypair_struct MLK_NAMESPACE_K(keypair_struct)
+#define crypto_kem_enc_derand_struct MLK_NAMESPACE_K(enc_derand_struct)
+#define crypto_kem_enc_struct MLK_NAMESPACE_K(enc_struct)
+#define crypto_kem_dec_struct MLK_NAMESPACE_K(dec_struct)


This should be MLK_ADD_LEVEL for the structs

Not sure I can follow. For the struct I am using using MLK_ADD_LEVEL,

Signed-off-by: Matthias J. Kannwischer <[email protected]>

mkannwischer added the benchmark this PR should be benchmarked in CI label Jan 29, 2025

oqs-bot reviewed Jan 29, 2025

View reviewed changes

mkannwischer changed the title ~~Proposal: New API~~ Proposal: API Extension Jan 29, 2025

oqs-bot reviewed Jan 29, 2025

View reviewed changes

mkannwischer force-pushed the newapi branch 2 times, most recently from 3ca90c7 to 1f56ed1 Compare February 25, 2025 04:26

hanno-becker reviewed Feb 25, 2025

View reviewed changes

mkannwischer force-pushed the newapi branch 5 times, most recently from 0ab73bb to c1e1a00 Compare February 25, 2025 07:26

mkannwischer force-pushed the newapi branch 10 times, most recently from c2de257 to 38e570c Compare February 26, 2025 03:46

hanno-becker reviewed Feb 26, 2025

View reviewed changes

mkannwischer force-pushed the newapi branch 3 times, most recently from 74ce0ac to 39391ea Compare February 27, 2025 07:57

mkannwischer added 6 commits February 27, 2025 16:39

Extended API

426258d

Signed-off-by: Matthias J. Kannwischer <[email protected]>

Add functional tests for new API

819c0ee

Signed-off-by: Matthias J. Kannwischer <[email protected]>

add monolithic_build_multilevel_native to examples CI

c60fefb

Signed-off-by: Matthias J. Kannwischer <[email protected]>

Example monolithic_build_multilevel_native: Do not build as single CU

c4e5f63

Signed-off-by: Matthias J. Kannwischer <[email protected]>

CBMC: fail on proof failure when -f is passed

a7596ab

Signed-off-by: Matthias J. Kannwischer <[email protected]>

wip: Adjust cbmc proofs

813f192

Signed-off-by: Matthias J. Kannwischer <[email protected]>

mkannwischer force-pushed the newapi branch from 39391ea to 813f192 Compare February 27, 2025 08:41

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Proposal: API Extension #715

Proposal: API Extension #715

mkannwischer commented Jan 29, 2025 •

edited

Loading

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

oqs-bot left a comment

mkannwischer commented Feb 10, 2025

hanno-becker commented Feb 18, 2025

hanno-becker left a comment

hanno-becker Feb 25, 2025

mkannwischer Feb 25, 2025

hanno-becker Feb 25, 2025 •

edited

Loading

mkannwischer Feb 26, 2025 •

edited

Loading

hanno-becker Feb 26, 2025 •

edited

Loading

hanno-becker Feb 26, 2025

mkannwischer Feb 26, 2025


		crypto_kem_marshal_pk(pk, &pks);
		crypto_kem_marshal_sk(sk, &sks);

Proposal: API Extension #715

Are you sure you want to change the base?

Proposal: API Extension #715

Conversation

mkannwischer commented Jan 29, 2025 • edited Loading

oqs-bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A76 (Raspberry Pi 5) benchmarks

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 4th gen (c7i)

oqs-bot left a comment

Choose a reason for hiding this comment

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

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 3rd gen (c6a)

oqs-bot left a comment

Choose a reason for hiding this comment

Intel Xeon 3rd gen (c6i)

oqs-bot left a comment

Choose a reason for hiding this comment

AMD EPYC 4th gen (c7a)

oqs-bot left a comment

Choose a reason for hiding this comment

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

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4

oqs-bot left a comment

Choose a reason for hiding this comment

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

oqs-bot left a comment

Choose a reason for hiding this comment

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

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton2

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton4 (no-opt)

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton2 (no-opt)

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3

oqs-bot left a comment

Choose a reason for hiding this comment

Graviton3 (no-opt)

oqs-bot left a comment

Choose a reason for hiding this comment

Bananapi bpi-f3 benchmarks

oqs-bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A55 (Snapdragon 888) benchmarks

oqs-bot left a comment

Choose a reason for hiding this comment

Arm Cortex-A72 (Raspberry Pi 4) benchmarks

mkannwischer commented Feb 10, 2025

hanno-becker commented Feb 18, 2025

hanno-becker left a comment

Choose a reason for hiding this comment

hanno-becker Feb 25, 2025

Choose a reason for hiding this comment

mkannwischer Feb 25, 2025

Choose a reason for hiding this comment

hanno-becker Feb 25, 2025 • edited Loading

Choose a reason for hiding this comment

mkannwischer Feb 26, 2025 • edited Loading

Choose a reason for hiding this comment

hanno-becker Feb 26, 2025 • edited Loading

Choose a reason for hiding this comment

hanno-becker Feb 26, 2025

Choose a reason for hiding this comment

mkannwischer Feb 26, 2025

Choose a reason for hiding this comment

mkannwischer commented Jan 29, 2025 •

edited

Loading

hanno-becker Feb 25, 2025 •

edited

Loading

mkannwischer Feb 26, 2025 •

edited

Loading

hanno-becker Feb 26, 2025 •

edited

Loading