|
| 1 | +2024-03-27 Matthias Kretz < [email protected]> |
| 2 | + |
| 3 | + * include/experimental/bits/simd_x86.h (_S_masked_unary): |
| 4 | + Cast inputs < 16 bytes to 16 byte vectors before calling the |
| 5 | + right subtraction builtin. Before returning, truncate to the |
| 6 | + return vector type. |
| 7 | + |
| 8 | +2024-03-27 Matthias Kretz < [email protected]> |
| 9 | + |
| 10 | + * include/experimental/bits/simd_x86.h (_S_masked_unary): Call |
| 11 | + the 4- and 8-byte variants of __builtin_ia32_subp[ds] without |
| 12 | + rounding direction argument. |
| 13 | + |
| 14 | +2024-03-27 Srinivas Yadav Singanaboina < [email protected]> |
| 15 | + |
| 16 | + * include/Makefile.am: Add simd_sve.h. |
| 17 | + * include/Makefile.in: Add simd_sve.h. |
| 18 | + * include/experimental/bits/simd.h: Add new SveAbi. |
| 19 | + * include/experimental/bits/simd_builtin.h: Use |
| 20 | + __no_sve_deduce_t to support existing Neon Abi. |
| 21 | + * include/experimental/bits/simd_converter.h: Convert |
| 22 | + sequentially when sve is available. |
| 23 | + * include/experimental/bits/simd_detail.h: Define sve |
| 24 | + specific macro. |
| 25 | + * include/experimental/bits/simd_math.h: Fallback frexp |
| 26 | + to execute sequntially when sve is available, to handle |
| 27 | + fixed_size_simd return type that always uses sve. |
| 28 | + * include/experimental/simd: Include bits/simd_sve.h. |
| 29 | + * testsuite/experimental/simd/tests/bits/main.h: Enable |
| 30 | + testing for sve128, sve256, sve512. |
| 31 | + * include/experimental/bits/simd_sve.h: New file. |
| 32 | + |
1 | 33 | 2024-03-26 Arsen Arsenović < [email protected]>
|
2 | 34 |
|
3 | 35 | * include/std/generator (generator::_Iterator::operator*): Fix
|
|
0 commit comments