diff options
| author | Xi Ruoyao <xry111@xry111.site> | 2025-03-01 11:46:51 +0800 |
|---|---|---|
| committer | Lulu Cheng <chenglulu@loongson.cn> | 2025-08-18 09:09:36 +0800 |
| commit | 89c8638deb9ccb1f1e1a90f18f2e6b984b70fbe8 (patch) | |
| tree | 5bfcb9ab6046904466c0175cfd356eef6f280b58 /libgfortran/generated | |
| parent | c9374dbf9d16b35886711a91993bcdec490722e7 (diff) | |
| download | gcc-89c8638deb9ccb1f1e1a90f18f2e6b984b70fbe8.zip gcc-89c8638deb9ccb1f1e1a90f18f2e6b984b70fbe8.tar.gz gcc-89c8638deb9ccb1f1e1a90f18f2e6b984b70fbe8.tar.bz2 | |
LoongArch: Implement 16-byte atomic load with LSX
If the vector is naturally aligned, it cannot cross cache lines so the
LSX load is guaranteed to be atomic. Thus we can use LSX to do the
lock-free atomic load, instead of using a lock.
gcc/ChangeLog:
* config/loongarch/sync.md (atomic_loadti_lsx): New define_insn.
(atomic_loadti): New define_expand.
Diffstat (limited to 'libgfortran/generated')
0 files changed, 0 insertions, 0 deletions