diff options
| author | Wilco Dijkstra <wdijkstr@arm.com> | 2020-08-28 17:51:40 +0100 |
|---|---|---|
| committer | Wilco Dijkstra <wdijkstr@arm.com> | 2020-10-12 18:28:42 +0100 |
| commit | 64458aabeb7f6d15b389cb49b9faf4925db354fa (patch) | |
| tree | 8732dddbbe462de56a8d6a0b54c25086d962c7e7 | |
| parent | 58c6a7ae53c647390a3057c247d34643e1201aac (diff) | |
| download | glibc-64458aabeb7f6d15b389cb49b9faf4925db354fa.zip glibc-64458aabeb7f6d15b389cb49b9faf4925db354fa.tar.gz glibc-64458aabeb7f6d15b389cb49b9faf4925db354fa.tar.bz2 | |
AArch64: Improve backwards memmove performance
On some microarchitectures performance of the backwards memmove improves if
the stores use STR with decreasing addresses. So change the memmove loop
in memcpy_advsimd.S to use 2x STR rather than STP.
Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org>
(cherry picked from commit bd394d131c10c9ec22c6424197b79410042eed99)
| -rw-r--r-- | sysdeps/aarch64/multiarch/memcpy_advsimd.S | 7 |
1 files changed, 4 insertions, 3 deletions
diff --git a/sysdeps/aarch64/multiarch/memcpy_advsimd.S b/sysdeps/aarch64/multiarch/memcpy_advsimd.S index d4ba747..48bb6d7 100644 --- a/sysdeps/aarch64/multiarch/memcpy_advsimd.S +++ b/sysdeps/aarch64/multiarch/memcpy_advsimd.S @@ -223,12 +223,13 @@ L(copy_long_backwards): b.ls L(copy64_from_start) L(loop64_backwards): - stp A_q, B_q, [dstend, -32] + str B_q, [dstend, -16] + str A_q, [dstend, -32] ldp A_q, B_q, [srcend, -96] - stp C_q, D_q, [dstend, -64] + str D_q, [dstend, -48] + str C_q, [dstend, -64]! ldp C_q, D_q, [srcend, -128] sub srcend, srcend, 64 - sub dstend, dstend, 64 subs count, count, 64 b.hi L(loop64_backwards) |