The 1st patch superseds [v1]. More tweaks to vec_perm follow.
Bootstrapped and regtested on loongarch64-linux-gnu. Ok for trunk? And
is the 1st one ok for releases/gcc-{15,14}?
[v1]:https://gcc.gnu.org/pipermail/gcc-patches/2025-November/700812.html
*** BLURB HERE ***
Xi Ruoyao (7):
LoongArch: Fix wrong code from loongarch_expand_vec_perm_1 [PR 122695]
LoongArch: Avoid memory load when use xvperm.w for vec_permv4di
LoongArch: NFC: Simplify vec_permv8sf logic
LoongArch: NFC: Move [x]vshuf.* to simd.md
LoongArch: Micro-optimize the blend step for vec_perm<LASX>
LoongArch: NFC: Simplify logic of vec_perm{v32qi,v16hi}
LoongArch: NFC: Drop loongarch_expand_vec_perm
gcc/config/loongarch/lasx.md | 36 +--
gcc/config/loongarch/loongarch-protos.h | 1 -
gcc/config/loongarch/loongarch.cc | 269 ++++--------------
gcc/config/loongarch/lsx.md | 54 +---
gcc/config/loongarch/simd.md | 75 ++++-
.../gcc.target/loongarch/pr122695-1.c | 22 ++
.../gcc.target/loongarch/pr122695-2.c | 22 ++
7 files changed, 188 insertions(+), 291 deletions(-)
create mode 100644 gcc/testsuite/gcc.target/loongarch/pr122695-1.c
create mode 100644 gcc/testsuite/gcc.target/loongarch/pr122695-2.c
--
2.51.2
