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 <austinf@cetoncorp.com>
Signed-off-by: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: Carmelo Amoroso <carmelo.amoroso@st.com>

1 files changed, 43 insertions, 0 deletions

diff --git a/libc/string/sh/sh4/strncpy.S b/libc/string/sh/sh4/strncpy.S
new file mode 100644
index 000000000..8a16f39d4
--- /dev/null
+++ b/libc/string/sh/sh4/strncpy.S
@@ -0,0 +1,43 @@
+/* strncpy implementation for SUPERH
+ *
+ * Copyright (C) 2009 STMicroelectronics Ltd.
+ *
+ * Author: Giuseppe Cavallaro <peppe.cavallaro@st.com>
+ *
+ * Licensed under the LGPL v2.1, see the file COPYING.LIB in this tarball.
+ */
+
+/*
+	char *strncpy(char *dest, const char *src, size_t n);
+ */
+
+#include <sysdep.h>
+
+ENTRY(strncpy)
+	mov	#0,r0
+	bra	2f
+	 mov	r4,r2
+1:
+	mov.b	r1,@(r0,r2)
+	add	#1,r0
+2:
+	cmp/hs	r6,r0
+	bt	5f
+	mov.b	@(r0,r5),r1
+	tst	r1,r1
+	bf/s	1b
+	 cmp/hs	r6,r0
+	bra	4f
+	 nop
+3:
+	mov.b	r1,@(r0,r2)
+	add	#1,r0
+	cmp/hs	r6,r0
+4:
+	bf/s	3b
+	 mov	#0,r1
+5:
+	rts
+	 mov     r2,r0
+END(strncpy)
+libc_hidden_def(strncpy)