mirror of
git://sourceware.org/git/glibc.git
synced 2024-11-27 03:41:23 +08:00
2f1f7a5f8a
Qualcom's new core, oryon-1, has a different characteristics for memset than the current versions of memset. For non-zero, larger sizes, using GPRs rather than the SIMD stores is ~30% faster. For even larger sizes, using the nontemporal stores is needed not to polute the L1/L2 caches. For zero values, using `dc zva` should be used. Since we know the size will always be 64 bytes, we don't need to figure out the size there. I started with the emag memset and added back the `dc zva` code. Changes since v1: * v3: Fix comment formating Signed-off-by: Andrew Pinski <quic_apinski@quicinc.com> Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org> |
||
|---|---|---|
| .. | ||
| dl-symbol-redir-ifunc.h | ||
| ifunc-impl-list.c | ||
| init-arch.h | ||
| Makefile | ||
| memchr_generic.S | ||
| memchr_nosimd.S | ||
| memchr.c | ||
| memcpy_a64fx.S | ||
| memcpy_generic.S | ||
| memcpy_mops.S | ||
| memcpy_oryon1.S | ||
| memcpy_sve.S | ||
| memcpy_thunderx2.S | ||
| memcpy_thunderx.S | ||
| memcpy.c | ||
| memmove_mops.S | ||
| memmove.c | ||
| memset_a64fx.S | ||
| memset_emag.S | ||
| memset_generic.S | ||
| memset_kunpeng.S | ||
| memset_mops.S | ||
| memset_oryon1.S | ||
| memset_zva64.S | ||
| memset.c | ||
| strlen_asimd.S | ||
| strlen_generic.S | ||
| strlen.c | ||