MIPS: import optimized string functions from Linux

10x performance gain according to simple test on QEMU malta:
barebox:/ time memcpy 0xa0000000 0xa0001000 0x100000

Signed-off-by: Aleksey Kuleshov <rndfax@yandex.ru>
Acked-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

1 files changed, 146 insertions, 0 deletions

diff --git a/arch/mips/lib/memset.S b/arch/mips/lib/memset.S
new file mode 100644
index 0000000000..d3c1c72393
--- /dev/null
+++ b/arch/mips/lib/memset.S
@@ -0,0 +1,146 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1998, 1999, 2000 by Ralf Baechle
+ * Copyright (C) 1999, 2000 Silicon Graphics, Inc.
+ * Copyright (C) 2007 by Maciej W. Rozycki
+ * Copyright (C) 2011, 2012 MIPS Technologies, Inc.
+ *
+ * Kernel-mode memset function without exceptions
+ * by Aleksey Kuleshov (rndfax@yandex.ru), 2015
+ */
+#include <asm/asm.h>
+#include <asm/asm-offsets.h>
+#include <asm/regdef.h>
+
+#if LONGSIZE == 4
+#define LONG_S_L swl
+#define LONG_S_R swr
+#else
+#define LONG_S_L sdl
+#define LONG_S_R sdr
+#endif
+
+#define STORSIZE LONGSIZE
+#define STORMASK LONGMASK
+#define FILL64RG a1
+#define FILLPTRG t0
+
+/*
+ * memset(void *s, int c, size_t n)
+ *
+ * a0: start of area to clear
+ * a1: char to fill with
+ * a2: size of area to clear
+ */
+
+LEAF(memset)
+	beqz		a1, 1f
+	move		v0, a0			/* result */
+
+	andi		a1, 0xff		/* spread fillword */
+	LONG_SLL		t1, a1, 8
+	or		a1, t1
+	LONG_SLL		t1, a1, 16
+#if LONGSIZE == 8
+	or		a1, t1
+	LONG_SLL		t1, a1, 32
+#endif
+	or		a1, t1
+
+	.macro	f_fill64 dst, offset, val
+	LONG_S	\val, (\offset +  0 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset +  1 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset +  2 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset +  3 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset +  4 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset +  5 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset +  6 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset +  7 * STORSIZE)(\dst)
+#if (LONGSIZE == 4)
+	LONG_S	\val, (\offset +  8 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset +  9 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset + 10 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset + 11 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset + 12 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset + 13 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset + 14 * STORSIZE)(\dst)
+	LONG_S	\val, (\offset + 15 * STORSIZE)(\dst)
+#endif
+	.endm
+
+	.set	noreorder
+	.align	5
+
+1:
+	sltiu		t0, a2, STORSIZE	/* very small region? */
+	bnez		t0, .Lsmall_memset
+	andi		t0, a0, STORMASK	/* aligned? */
+
+	beqz		t0, 1f
+	PTR_SUBU	t0, STORSIZE		/* alignment in bytes */
+
+#ifdef __MIPSEB__
+	LONG_S_L	a1, (a0)		/* make word/dword aligned */
+#else
+	LONG_S_R	a1, (a0)		/* make word/dword aligned */
+#endif
+	PTR_SUBU	a0, t0			/* long align ptr */
+	PTR_ADDU	a2, t0			/* correct size */
+
+1:	ori		t1, a2, 0x3f		/* # of full blocks */
+	xori		t1, 0x3f
+	beqz		t1, .Lmemset_partial	/* no block to fill */
+	andi		t0, a2, 0x40-STORSIZE
+
+	PTR_ADDU	t1, a0			/* end address */
+	.set		reorder
+1:	PTR_ADDIU	a0, 64
+	f_fill64 a0, -64, FILL64RG
+	bne		t1, a0, 1b
+	.set		noreorder
+
+.Lmemset_partial:
+	PTR_LA		t1, 2f			/* where to start */
+#if LONGSIZE == 4
+	PTR_SUBU	t1, FILLPTRG
+#else
+	.set		noat
+	LONG_SRL	AT, FILLPTRG, 1
+	PTR_SUBU	t1, AT
+	.set		at
+#endif
+	jr		t1
+	PTR_ADDU	a0, t0			/* dest ptr */
+
+	.set		push
+	.set		noreorder
+	.set		nomacro
+	/* ... but first do longs ... */
+	f_fill64 a0, -64, FILL64RG
+2:	.set		pop
+	andi		a2, STORMASK		/* At most one long to go */
+
+	beqz		a2, 1f
+	PTR_ADDU	a0, a2			/* What's left */
+#ifdef __MIPSEB__
+	LONG_S_R	a1, -1(a0)
+#else
+	LONG_S_L	a1, -1(a0)
+#endif
+1:	jr		ra
+	move		a2, zero
+
+.Lsmall_memset:
+	beqz		a2, 2f
+	PTR_ADDU	t1, a0, a2
+
+1:	PTR_ADDIU	a0, 1			/* fill bytewise */
+	bne		t1, a0, 1b
+	sb		a1, -1(a0)
+
+2:	jr		ra			/* done */
+	move		a2, zero
+	END(memset)