From 5c9ef58ec4bcb2def9e30f0b156f9cfcb1d0d163 Mon Sep 17 00:00:00 2001 From: Austin Foxley Date: Sun, 22 Nov 2009 12:17:38 -0800 Subject: sh: Add new optimisation to the SH4 memcpy This optimization is based on prefetching and 64bit data transfer via FPU (only for the little endianess) Tests shows that: ---------------------------------------- Memory bandwidth | Gain | sh4-300 | sh4-200 ---------------------------------------- 512 bytes to 16KiB | ~20% | ~25% from 32KiB to 16MiB | ~190% | ~5% ---------------------------------------- Signed-off-by: Austin Foxley Signed-off-by: Giuseppe Cavallaro Signed-off-by: Carmelo Amoroso --- libc/string/sh/memchr.S | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+) create mode 100644 libc/string/sh/memchr.S (limited to 'libc/string/sh/memchr.S') diff --git a/libc/string/sh/memchr.S b/libc/string/sh/memchr.S new file mode 100644 index 000000000..6b7142f69 --- /dev/null +++ b/libc/string/sh/memchr.S @@ -0,0 +1,30 @@ +/* $Id: memchr.S,v 1.1 2000/04/14 16:49:01 mjd Exp $ + * + * "memchr" implementation of SuperH + * + * Copyright (C) 1999 Niibe Yutaka + * + * Licensed under the LGPL v2.1, see the file COPYING.LIB in this tarball. + */ + +/* + * void *memchr(const void *s, int c, size_t n); + */ + +#include + +ENTRY(memchr) + tst r6,r6 + bt/s 2f + exts.b r5,r5 +1: mov.b @r4,r1 + cmp/eq r1,r5 + bt/s 3f + dt r6 + bf/s 1b + add #1,r4 +2: mov #0,r4 +3: rts + mov r4,r0 +END(memchr) +libc_hidden_def (memchr) -- cgit v1.2.3