Folkert de Vries
74e2cb12d1
s390x: document the different rounding flavors
2025-07-14 17:22:38 +02:00
Sayantan Chakraborty
cd48abd478
Merge pull request #1858 from folkertdev/aarch-comparison-operators
...
aarch64: use `core::intrinsics::simd` for integer min/max
2025-07-11 23:07:13 +00:00
Marijn Schouten
a8110072ff
thumbv7neon-unknown-linux-gnueabihf clippy fixes
2025-07-11 11:13:13 +00:00
Sayantan Chakraborty
bb7a446c75
Merge pull request #1857 from folkertdev/arm-dup
...
use `splat` for the aarch64/arm dup intrinsics
2025-07-10 22:40:03 +00:00
Folkert de Vries
6d1f9552bd
use intrinsics::simd for integer max/min
2025-07-10 15:04:13 +02:00
Folkert de Vries
ef048d98ce
use splat for the aarch64/arm dup intrinsics
2025-07-10 12:08:38 +02:00
Folkert de Vries
061a941adc
remove unneeded cast
2025-07-08 10:27:34 +00:00
Folkert de Vries
92701fef20
shorten array literals
2025-07-08 10:27:34 +00:00
Folkert de Vries
323c646ec3
remote intermediate assignment
2025-07-08 10:27:34 +00:00
sayantn
658b5c8f69
Use simd_funnel_sh{l,r} and simd_round_ties_even to remove uses of LLVM intrinsics
2025-07-07 23:30:02 +00:00
Marijn Schouten
fddd05bd7f
some clippy fixes
2025-07-07 23:28:47 +00:00
Jakub Beránek
b535003608
Remove std-detect dev dependency in core_arch
2025-07-07 21:50:45 +00:00
okaneco
600497f3ed
Mark avx broadcast intrinsics as safe
2025-07-07 21:50:01 +00:00
ltdk
e46d2207bd
Use simd_extract! macro instead of calling simd_extract directly
2025-07-07 21:48:32 +00:00
Amanieu d'Antras
8ae4e6a566
Update stabilization version of certain x86 intrinsics to 1.89
...
These were left as `CURRENT_RUSTC_VERSION` in the submodule.
2025-07-07 21:47:33 +00:00
Folkert de Vries
bb2d169700
use is_multiple_of to check if an addr is aligned
2025-07-07 17:08:03 +00:00
Folkert de Vries
26b7dd4304
the generic_arg_infer feature is now stable on nightly
2025-06-19 15:31:16 +00:00
satiscugcat
003516cd44
Test
2025-06-17 19:28:07 +00:00
satiscugcat
dcd35573f7
Ran cargo fmt, removed unexpected character
2025-06-17 19:28:07 +00:00
satiscugcat
fd1b9c72fd
Moved const fn outside unsafe block
2025-06-17 19:28:07 +00:00
satiscugcat
64dd3e4489
Changed implementation of _mm{256,512}_alignr_epi8 to match that of _mm_alignr_epi8 in ssse3.rs, also removed the import of the unreachable unchecked hint as it was no longer necessary
2025-06-17 19:28:07 +00:00
satiscugcat
48dafaa1c7
Changed implementation of bsrli in avx512.rs to match that of bslli, fixing a bug in the process.
2025-06-17 19:28:07 +00:00
satiscugcat
09475721a8
Changed the implementation of bsrli to match bslli in avx2 intrinsics
2025-06-17 19:28:07 +00:00
Aniket Mishra
1c5b2a659d
Fixing mistake in previous commit
2025-06-17 19:28:07 +00:00
Aniket Mishra
58db094d9e
Removal of redundant mod operation in _mm256_alignr_epi8
2025-06-17 19:28:07 +00:00
Aniket Mishra
8fb78f3f69
Fix in erroneous implementation of _mm256_bsrli_epi128
...
Fixing the issue mentioned in issue #1822 of rust-lang/stdarch.
2025-06-17 19:28:07 +00:00
Amanieu d'Antras
341889647e
Fixes to compile with latest Rust nightly
2025-06-11 14:11:23 +00:00
Amanieu d'Antras
68c0308cbe
Stabilize AVX512 intrinsics
...
FCP completed in tracking issue rust-lang/rust#111137
2025-06-02 19:45:44 +00:00
sayantn
7a9e01ad92
Fix incorrect intrinsic name in WASM
2025-06-02 19:45:08 +00:00
sayantn
77174a3f43
Fix incorrect intrinsic name in X86
2025-06-02 19:45:08 +00:00
sayantn
068cc378a6
Remove uses of deprecated type-specific pointers from ARM
2025-06-02 19:45:08 +00:00
sayantn
5d47941854
Use correct LLVM intrinsic for vmull and vaddv
2025-06-02 19:45:08 +00:00
sayantn
2665fe5729
Use rust intrinsics for more ARM intrinsics
2025-06-02 19:45:08 +00:00
Adam Gemmell
3fe6f7616d
Mark Neon f16 vectors as unstable
2025-06-02 18:34:18 +00:00
Adam Gemmell
d2587282eb
Revert vbsl[q]_f16 to unstable
2025-06-02 18:34:18 +00:00
Folkert de Vries
52115419a3
use simd_bitreverse on aarch64
2025-06-01 22:58:15 +00:00
sayantn
1ca4587841
Stabilize sha512, sm3 and sm4 intrinsics and runtime detection
2025-05-31 09:40:04 +00:00
sayantn
ba641e58a2
Stabilize keylocker intrinsics and runtime detection
2025-05-31 09:39:49 +00:00
sayantn
cd63f26d86
Upgrade more intrinsics to the new version
2025-05-30 18:02:17 +00:00
sayantn
3de76d47a2
Use the new definition of rdtscp intrinsic
...
- add `-Zverify-llvm-ir` in testsuite
2025-05-30 18:02:17 +00:00
sayantn
28960ece72
Fix s390x intrinsics
...
- use correct intrinsic for unpackl
- fix invalid use of `simd_{or,and,xor}` on floating point vectors
- `vec_search_string` should require `vector-enhancements-2`
2025-05-30 18:02:17 +00:00
sayantn
7afd613941
Fix PPC shift and rotate intrinsics
2025-05-30 18:02:17 +00:00
sayantn
79654521ee
Fix ldpte and lddir signature
...
- The 2nd argument of the LLVM intrinsic should be IMMARG
2025-05-30 18:02:17 +00:00
usamoi
a4a7184370
mark gfni, vaes, vpclmulqdq intrinsics as safe
2025-05-30 17:35:18 +00:00
Ralf Jung
9561e8e0a3
cmpxchg16b: use atomic_compare_exchange from libcore
2025-05-30 17:14:51 +00:00
Ralf Jung
b58ce625a2
allow aarch64_softfloat_neon for backwards compatibility
2025-05-21 10:39:23 +00:00
Folkert de Vries
7b995254d4
use a tuple to return the condition code
2025-05-20 16:39:24 +00:00
Folkert de Vries
42e6ad3ee8
use the right load instruction
2025-05-20 14:26:46 +00:00
Folkert de Vries
a850e8738a
avx512_target_feature is now stable on nightly
2025-05-20 14:26:46 +00:00
WANG Rui
63235a6ba5
Partially stabilize LoongArch target features
2025-05-12 19:47:49 +00:00
