Commit graph

1933 commits

Author SHA1 Message Date
sayantn
1f779b7b40 Added runtime detection
Expanded the cache size to 93 (we will need this in near future)
Fixed detection of VAES, GFNI and VPCLMULQDQ
Could not test with `cupid` because they do not support these yet
2024-06-23 10:36:46 +02:00
Tobias Decking
2fd58a7ac7 Use generic simd for avx512 popcnt 2024-06-23 10:14:32 +02:00
Ralf Jung
90d47e9c71 set asm attributes 2024-06-21 16:59:03 +02:00
Ralf Jung
5982c0838a fix test_mm512_stream_ps test 2024-06-21 16:59:03 +02:00
Ralf Jung
5c0744a3e5 non-temporal stores: use inline assembly 2024-06-21 16:59:03 +02:00
Tobias Decking
36852a1264 Update avx2.rs 2024-06-21 14:52:17 +02:00
Ralf Jung
5e7dada0eb addcarryx: use pointers of the right type 2024-06-21 11:59:57 +02:00
sayantn
8ee3cc779a AVX512DQ: Fixes (Corrected some typos in tests, Removed intrinsics list as everything has been implemented) 2024-06-18 19:13:13 +02:00
sayantn
b21e02ad83 AVX512DQ: Fixes (Adding SSE target_feature for i586) 2024-06-18 19:13:13 +02:00
sayantn
1f4034ba50 AVX512DQ Part 7: FP-Class 2024-06-18 19:13:13 +02:00
sayantn
c0a49d908b AVX512DQ Part 6: Reduce 2024-06-18 19:13:13 +02:00
sayantn
177b75bde5 AVX512DQ Part 6: Reduce 2024-06-18 19:13:13 +02:00
sayantn
52f8b0c1a9 AVX512DQ Part 5: Range. Fixed intrinsic verification. 2024-06-18 19:13:13 +02:00
sayantn
c052982434 AVX512DQ Part 4: Mask Registers and Multiply Low 2024-06-18 19:13:13 +02:00
sayantn
54ef05ac65 AVX512DQ : Fix errors in Part 2 2024-06-18 19:13:13 +02:00
sayantn
5d2e19f5b6 AVX512DQ Part 3: Convert Intrinsics 2024-06-18 19:13:13 +02:00
sayantn
3281ecd0da AVX512DQ : Fix Instructions 2024-06-18 19:13:13 +02:00
sayantn
a0efee80a1 AVX512DQ : Fix 2024-06-18 19:13:13 +02:00
sayantn
dbb53389db AVX512DQ : Fix : Added to mod.rs 2024-06-18 19:13:13 +02:00
sayantn
011c102479 AVX512DQ : Fix : Added to mod.rs 2024-06-18 19:13:13 +02:00
sayantn
9af0ddafac AVX512DQ Part 2: Broadcast, Extract, Insert 2024-06-18 19:13:13 +02:00
sayantn
fd6d97c21e AVX512DQ Part 1: Logical Operations (and, andn, or, xor) - tests and doc 2024-06-18 19:13:13 +02:00
sayantn
d316154255 AVX512DQ Part 1: Logical Operations (and, andn, or, xor) 2024-06-18 19:13:13 +02:00
sayantn
cc2c29113f AVX512DQ Part 0: Intrinsics List 2024-06-18 19:13:13 +02:00
Tobias Decking
f0f309bc52 Replace addsub variations 2024-06-17 00:39:37 +02:00
Tobias Decking
3e95b6133a Fix SSE2 sqrt 2024-06-17 00:39:37 +02:00
Tobias Decking
628831b7e1 Fix SSE sqrt 2024-06-17 00:39:37 +02:00
Tobias Decking
4831241378 Parenthesis 2024-06-17 00:39:37 +02:00
Tobias Decking
ef989eff75 Fix compiler errors 2024-06-17 00:39:37 +02:00
Tobias Decking
746b490131 Update avx.rs 2024-06-17 00:39:37 +02:00
Tobias Decking
13da9da323 Update sse2.rs 2024-06-17 00:39:37 +02:00
Tobias Decking
2d30424922 Update sse.rs 2024-06-17 00:39:37 +02:00
Tobias Decking
b683da6b0b Update fma.rs 2024-06-17 00:39:37 +02:00
Luca Barbato
61fb419a9b Use longer associated types in the Altivec traits 2024-06-11 15:00:39 +02:00
sayantn
511bbff3b2 Update avx.rs 2024-06-11 00:18:28 +02:00
sayantn
1ccc7dbd0d Fixed _mm256_cvtsd_f64
This intrinsic should have `target_feature` AVX, (according to Intel Intrinsics Guide) but had AVX2
2024-06-11 00:18:28 +02:00
bjorn3
292c0ecffd Fix duplicated argument name in extern block
While rustc accepts it just fine due to what is arguably a bug, this
duplication does become an issue when generating functions from the llvm
intrinsic declarations while reusing the argument names.
2024-06-07 20:42:46 +02:00
Amanieu d'Antras
860145884d Ignore int3 instructions when counting instructions in tests
These are generated as padding and are not actually part of the
function.
2024-06-07 19:18:13 +02:00
Olasunkanmi Olayinka
da09b47285 feat: stabilization for stdarch_aarch64_crc32 2024-05-14 15:52:07 +02:00
Luca Barbato
adb2db7471 Add vec_insert and vec_extract 2024-05-13 19:08:20 +02:00
Daniel Paoliello
130bc86694 Remove libc dependency on Windows by using Win32 to get env vars 2024-05-07 20:50:51 +02:00
Luca Barbato
0a5325a16c Add vec_orc 2024-05-06 16:02:08 +02:00
Luca Barbato
f91e5fd6b2 Simplify vec_andc implementation 2024-05-06 16:02:08 +02:00
Luca Barbato
509d5efe2b Silence unexpected-cfgs 2024-05-06 16:02:08 +02:00
Luca Barbato
4815a9efaf Add vec_mul 2024-04-28 20:56:17 +02:00
Eduardo Sánchez Muñoz
c5f9c14154 Remove #![feature(inline_const)]
The feature has been stabilized
2024-04-28 20:33:41 +02:00
Eduardo Sánchez Muñoz
95ce20e6f2 Add #[cfg_attr(miri, ignore)] to tests of intrinsics that cannot be supported by Miri 2024-04-28 20:33:41 +02:00
Eduardo Sánchez Muñoz
ff7e6c7ac6 Implement ARM __ssat and __usat functions 2024-04-23 15:58:46 +02:00
Ben Kimock
2a21235f7a Don't depend on libc for cfg(windows) 2024-04-22 01:24:03 +02:00
Daniel Paoliello
613efc499c Enable testing for AArch64 Windows 2024-04-19 17:21:08 +02:00