[MIPS] Unify csum_partial.S

The 32-bit version and 64-bit version are almost equal.  Unify them.  This
makes further improvements (for example, copying with parallel, supporting
PREFETCH, etc.) easier.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>

2 files changed, 260 insertions, 2 deletions

diff --git a/arch/mips/lib/Makefile b/arch/mips/lib/Makefile
index b225543f530..888b61ea12f 100644
--- a/arch/mips/lib/Makefile
+++ b/arch/mips/lib/Makefile
@@ -2,8 +2,8 @@
 # Makefile for MIPS-specific library files..
 #
 
-lib-y	+= csum_partial_copy.o memcpy.o promlib.o strlen_user.o strncpy_user.o \
-	   strnlen_user.o uncached.o
+lib-y	+= csum_partial.o csum_partial_copy.o memcpy.o promlib.o \
+	   strlen_user.o strncpy_user.o strnlen_user.o uncached.o
 
 obj-y	+= iomap.o
 
diff --git a/arch/mips/lib/csum_partial.S b/arch/mips/lib/csum_partial.S
new file mode 100644
index 00000000000..15611d9df7a
--- /dev/null
+++ b/arch/mips/lib/csum_partial.S
@@ -0,0 +1,258 @@
+/*
+ * 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.
+ *
+ * Quick'n'dirty IP checksum ...
+ *
+ * Copyright (C) 1998, 1999 Ralf Baechle
+ * Copyright (C) 1999 Silicon Graphics, Inc.
+ */
+#include <asm/asm.h>
+#include <asm/regdef.h>
+
+#ifdef CONFIG_64BIT
+#define T0	ta0
+#define T1	ta1
+#define T2	ta2
+#define T3	ta3
+#define T4	t0
+#define T7	t3
+#else
+#define T0	t0
+#define T1	t1
+#define T2	t2
+#define T3	t3
+#define T4	t4
+#define T7	t7
+#endif
+
+#define ADDC(sum,reg)						\
+	addu	sum, reg;					\
+	sltu	v1, sum, reg;					\
+	addu	sum, v1
+
+#define CSUM_BIGCHUNK(src, offset, sum, _t0, _t1, _t2, _t3)	\
+	lw	_t0, (offset + 0x00)(src);			\
+	lw	_t1, (offset + 0x04)(src);			\
+	lw	_t2, (offset + 0x08)(src); 			\
+	lw	_t3, (offset + 0x0c)(src); 			\
+	ADDC(sum, _t0);						\
+	ADDC(sum, _t1);						\
+	ADDC(sum, _t2);						\
+	ADDC(sum, _t3);						\
+	lw	_t0, (offset + 0x10)(src);			\
+	lw	_t1, (offset + 0x14)(src);			\
+	lw	_t2, (offset + 0x18)(src);			\
+	lw	_t3, (offset + 0x1c)(src);			\
+	ADDC(sum, _t0);						\
+	ADDC(sum, _t1);						\
+	ADDC(sum, _t2);						\
+	ADDC(sum, _t3);						\
+
+/*
+ * a0: source address
+ * a1: length of the area to checksum
+ * a2: partial checksum
+ */
+
+#define src a0
+#define sum v0
+
+	.text
+	.set	noreorder
+
+/* unknown src alignment and < 8 bytes to go  */
+small_csumcpy:
+	move	a1, T2
+
+	andi	T0, a1, 4
+	beqz	T0, 1f
+	 andi	T0, a1, 2
+
+	/* Still a full word to go  */
+	ulw	T1, (src)
+	PTR_ADDIU	src, 4
+	ADDC(sum, T1)
+
+1:	move	T1, zero
+	beqz	T0, 1f
+	 andi	T0, a1, 1
+
+	/* Still a halfword to go  */
+	ulhu	T1, (src)
+	PTR_ADDIU	src, 2
+
+1:	beqz	T0, 1f
+	 sll	T1, T1, 16
+
+	lbu	T2, (src)
+	 nop
+
+#ifdef __MIPSEB__
+	sll	T2, T2, 8
+#endif
+	or	T1, T2
+
+1:	ADDC(sum, T1)
+
+	/* fold checksum */
+	sll	v1, sum, 16
+	addu	sum, v1
+	sltu	v1, sum, v1
+	srl	sum, sum, 16
+	addu	sum, v1
+
+	/* odd buffer alignment? */
+	beqz	T7, 1f
+	 nop
+	sll	v1, sum, 8
+	srl	sum, sum, 8
+	or	sum, v1
+	andi	sum, 0xffff
+1:
+	.set	reorder
+	/* Add the passed partial csum.  */
+	ADDC(sum, a2)
+	jr	ra
+	.set	noreorder
+
+/* ------------------------------------------------------------------------- */
+
+	.align	5
+LEAF(csum_partial)
+	move	sum, zero
+	move	T7, zero
+
+	sltiu	t8, a1, 0x8
+	bnez	t8, small_csumcpy		/* < 8 bytes to copy */
+	 move	T2, a1
+
+	beqz	a1, out
+	 andi	T7, src, 0x1			/* odd buffer? */
+
+hword_align:
+	beqz	T7, word_align
+	 andi	t8, src, 0x2
+
+	lbu	T0, (src)
+	LONG_SUBU	a1, a1, 0x1
+#ifdef __MIPSEL__
+	sll	T0, T0, 8
+#endif
+	ADDC(sum, T0)
+	PTR_ADDU	src, src, 0x1
+	andi	t8, src, 0x2
+
+word_align:
+	beqz	t8, dword_align
+	 sltiu	t8, a1, 56
+
+	lhu	T0, (src)
+	LONG_SUBU	a1, a1, 0x2
+	ADDC(sum, T0)
+	sltiu	t8, a1, 56
+	PTR_ADDU	src, src, 0x2
+
+dword_align:
+	bnez	t8, do_end_words
+	 move	t8, a1
+
+	andi	t8, src, 0x4
+	beqz	t8, qword_align
+	 andi	t8, src, 0x8
+
+	lw	T0, 0x00(src)
+	LONG_SUBU	a1, a1, 0x4
+	ADDC(sum, T0)
+	PTR_ADDU	src, src, 0x4
+	andi	t8, src, 0x8
+
+qword_align:
+	beqz	t8, oword_align
+	 andi	t8, src, 0x10
+
+	lw	T0, 0x00(src)
+	lw	T1, 0x04(src)
+	LONG_SUBU	a1, a1, 0x8
+	ADDC(sum, T0)
+	ADDC(sum, T1)
+	PTR_ADDU	src, src, 0x8
+	andi	t8, src, 0x10
+
+oword_align:
+	beqz	t8, begin_movement
+	 LONG_SRL	t8, a1, 0x7
+
+	lw	T3, 0x08(src)
+	lw	T4, 0x0c(src)
+	lw	T0, 0x00(src)
+	lw	T1, 0x04(src)
+	ADDC(sum, T3)
+	ADDC(sum, T4)
+	ADDC(sum, T0)
+	ADDC(sum, T1)
+	LONG_SUBU	a1, a1, 0x10
+	PTR_ADDU	src, src, 0x10
+	LONG_SRL	t8, a1, 0x7
+
+begin_movement:
+	beqz	t8, 1f
+	 andi	T2, a1, 0x40
+
+move_128bytes:
+	CSUM_BIGCHUNK(src, 0x00, sum, T0, T1, T3, T4)
+	CSUM_BIGCHUNK(src, 0x20, sum, T0, T1, T3, T4)
+	CSUM_BIGCHUNK(src, 0x40, sum, T0, T1, T3, T4)
+	CSUM_BIGCHUNK(src, 0x60, sum, T0, T1, T3, T4)
+	LONG_SUBU	t8, t8, 0x01
+	bnez	t8, move_128bytes
+	 PTR_ADDU	src, src, 0x80
+
+1:
+	beqz	T2, 1f
+	 andi	T2, a1, 0x20
+
+move_64bytes:
+	CSUM_BIGCHUNK(src, 0x00, sum, T0, T1, T3, T4)
+	CSUM_BIGCHUNK(src, 0x20, sum, T0, T1, T3, T4)
+	PTR_ADDU	src, src, 0x40
+
+1:
+	beqz	T2, do_end_words
+	 andi	t8, a1, 0x1c
+
+move_32bytes:
+	CSUM_BIGCHUNK(src, 0x00, sum, T0, T1, T3, T4)
+	andi	t8, a1, 0x1c
+	PTR_ADDU	src, src, 0x20
+
+do_end_words:
+	beqz	t8, maybe_end_cruft
+	 LONG_SRL	t8, t8, 0x2
+
+end_words:
+	lw	T0, (src)
+	LONG_SUBU	t8, t8, 0x1
+	ADDC(sum, T0)
+	bnez	t8, end_words
+	 PTR_ADDU	src, src, 0x4
+
+maybe_end_cruft:
+	andi	T2, a1, 0x3
+
+small_memcpy:
+ j small_csumcpy; move a1, T2		/* XXX ??? */
+	beqz	t2, out
+	 move	a1, T2
+
+end_bytes:
+	lb	T0, (src)
+	LONG_SUBU	a1, a1, 0x1
+	bnez	a2, end_bytes
+	 PTR_ADDU	src, src, 0x1
+
+out:
+	jr	ra
+	 move	v0, sum
+	END(csum_partial)