Nicolas Sarlin
312ce494bf
chore(zk): add 1 * 64 benches with production CRS
2025-12-17 15:06:37 +01:00
Nicolas Sarlin
5f2e7e31f1
chore(zk): align wasm bench and integer bench
2025-12-17 15:06:37 +01:00
Thomas Montaigu
d394af7f4d
chore: bump dyn-stack to 0.13
...
Notable changes:
- StackReq methods no longer returns Result<StackReq, SizeOverflow>
instead, StackReq contains the invalid state.
Now, its when we create a PodBuffer that we can check/catch if the
size req is invalid by catching errors when calling
`PodBuffer::try_new`. Its also possible to manually check that
`stack_req != StackReq::OVERFLOW`
- GlobalaPodBuffer is now PodBuffer
2025-12-15 10:02:17 +01:00
Andrei Stoian
78d1ce18c1
feat(gpu): support keyswitch 64/32
2025-12-12 22:01:49 +01:00
Arthur Meyre
14d49f0891
chore: add possibility to manually populate tfhe FFT plan cache
2025-12-12 19:06:47 +01:00
Nicolas Sarlin
e544dfc08e
fix(integer): handle large string size in DataKind::num_blocks
2025-12-12 13:10:14 +01:00
Nicolas Sarlin
5891a4d78a
fix(hl): handles empty compact lists
2025-12-12 13:10:14 +01:00
Nicolas Sarlin
f17cd9bd37
fix(integer): handle message mod 0 in num_blocks
2025-12-12 13:10:14 +01:00
Nicolas Sarlin
1479315725
fix(integer): handles num_blocks_per_integer is 0 in ct list upgrade
2025-12-12 13:10:14 +01:00
Nicolas Sarlin
b5e5058759
fix(core): handle lwe dim of 0 when computing ct list size
2025-12-12 13:10:14 +01:00
Nicolas Sarlin
d98033c71d
fix(integer): check overflows when computing expected list size
2025-12-12 13:10:14 +01:00
Nicolas Sarlin
c7b869c956
fix(core): use saturating_* to convert from lwe dim and size
2025-12-12 13:10:14 +01:00
Nicolas Sarlin
50b76817c9
fix(shortint): use saturating_sub to get degree from message modulus
2025-12-12 13:10:14 +01:00
Thomas Montaigu
aa49d141c7
feat: add MetaParameterFinder
2025-12-12 11:20:03 +01:00
Thomas Montaigu
be1de6ef2b
chore: impl Versionize for MetaParameters
2025-12-12 11:20:03 +01:00
Guillermo Oyarzun
11579bd3d0
feat(gpu): create noise and pfail tests for pbs + ks + ms
2025-12-12 09:41:11 +01:00
Arthur Meyre
403cabb70a
chore: bump CSPRNG to 0.8.0 since MSRV was bumped
2025-12-11 13:12:36 +01:00
Arthur Meyre
63b46c3b99
chore: bump tfhe-versionable to 0.7 since the MSRV was changed
2025-12-11 13:12:36 +01:00
Arthur Meyre
a5aa3c366f
chore: bump tfhe-hpu-backend
2025-12-11 13:12:36 +01:00
Arthur Meyre
5818b08f2c
chore: bump tfhe-cuda-backend to 0.13.0
2025-12-11 13:12:36 +01:00
Thomas Montaigu
4f14343668
chore: impl Versionize for NormalizedHammingWeightBound
2025-12-10 10:41:37 +01:00
Arthur Meyre
4da1e19d5c
chore: fix typo in js wasm API doc
2025-12-09 16:33:37 +01:00
Nicolas Sarlin
4d6aeebd0c
chore: detect degree in compact list conformance
2025-12-04 13:59:03 +01:00
Arthur Meyre
f8ee0b3649
chore: add comment on the value causing issues in 64 bits decomp tests
2025-12-04 09:43:32 +01:00
Arthur Meyre
37c872786e
chore: add ref
2025-12-04 09:43:32 +01:00
Arthur Meyre
85a281efc5
chore: update FFT128 params to match production params in core
...
- allows to be even more confident about fft128 tests
2025-12-04 09:43:32 +01:00
Arthur Meyre
a9d2137424
chore: add test and fix for specialized decomposition code
...
- add dedicated edge case for u128
- fix all implementations that can be used and run tests on the same
u128 edge case
2025-12-04 09:43:32 +01:00
Arthur Meyre
3dcdf9043a
chore: add arithmetic right shift for split u128 values
...
- this applies a shift that extends the sign of the value interpreted as an
i128, this will be required for the fix of the specialized decomp code for
u128
2025-12-04 09:43:32 +01:00
Arthur Meyre
dfcceefa83
chore: add (previously) failing test case for u128 decomposition
...
- the code fix was already correct, but we add another edge case because we
have specialized decomposition code for u128 that we need to update and we
need a previously failing case, this serves as reference for the fix of the
specialized implementation
2025-12-04 09:43:32 +01:00
Enzo Di Maria
6752249f7f
refactor(gpu): moving vector_comparisons's functions to the backend
2025-12-03 09:11:00 +01:00
Enzo Di Maria
cc33161b97
refactor(gpu): moving cast_to_signed to the backend
2025-12-02 12:40:47 +01:00
Nicolas Sarlin
f9c2a5d423
chore: use dedicated types for compressed modswitched conformance
2025-12-01 14:39:06 +01:00
Enzo Di Maria
3cfbaa40c3
refactor(gpu): unchecked_index_of_clear to backend
2025-11-28 14:34:53 +01:00
Andrei Stoian
e2063c8ef4
chore(gpu): bench KS latency batches
2025-11-27 17:32:44 +01:00
Enzo Di Maria
0aa0918fea
refactor(gpu): vector_find's functions to backend
2025-11-27 13:36:10 +01:00
Nicolas Sarlin
79a89c157a
chore(fft-ntt): bump to 0.10.0 and 0.7.0
2025-11-26 13:32:14 +01:00
Agnes Leroy
ef30be5086
chore(gpu): reduce long run tests number of inputs to avoid timeout on 4090
2025-11-26 13:06:14 +01:00
Nicolas Sarlin
212b925b5e
chore: use rust-toolchain.toml for default toolchain
2025-11-26 11:28:21 +01:00
Nicolas Sarlin
f8a958663b
chore(tfhe): rename nightly feature flag to avx512
2025-11-26 11:28:21 +01:00
Nicolas Sarlin
851bd01873
chore(fft): rename nightly feature flag to avx512
2025-11-26 11:28:21 +01:00
Nicolas Sarlin
8d1f6d4d06
chore(ntt): rename nightly feature flag to avx512
2025-11-26 11:28:21 +01:00
Nicolas Sarlin
8e566c5765
chore: update pulp and bytemuck
2025-11-26 11:28:21 +01:00
Thomas Montaigu
2002184da4
feat(Tag): impl From<&str>
2025-11-25 10:24:25 +01:00
Thomas Montaigu
614ca52749
feat(CompressedXofKeySet): impl key gen of ClientKey
...
This adds the implemention of the ClientKey generation
that respects the threshold specs, some points are:
* 2 random generators are used one for everything public (masks)
the other for everything private (private keys and noise).
These generators are seeded once.
* binary secret keys are generated using
`fill_slice_with_random_uniform_binary_bits` and support pmax param
2025-11-25 10:24:25 +01:00
Arthur Meyre
48cc709cf9
test(core): add exhaustive decomposition test on several levels
...
- also add a verification that recomposition works in the previously added
edge case test
2025-11-24 18:40:36 +01:00
David Testé
a75b5107ff
chore(docs): change svg benchmark table names
...
This is done to ease automated table generation through continuous integration pipeline.
2025-11-24 14:03:08 +01:00
Enzo Di Maria
32b1a7ab1d
refactor(gpu): unchecked_match_value_or to backend
2025-11-24 13:50:15 +01:00
Enzo Di Maria
1a7efbc21b
fix(gpu): constant number of blocks for outputs of match_value
2025-11-21 15:56:54 +01:00
Agnes Leroy
bc0a4fb0d6
chore(gpu): add classical pbs tests to HL API, complement missing tests, classical in long run
2025-11-21 09:15:04 +01:00
Enzo Di Maria
c6709a82c0
refactor(gpu): match_value to backend with multiple streams
2025-11-20 17:11:15 +01:00
