redonkable
/
alistair23-linux
1
0
Fork 0
Code Releases Activity

[MIPS] Make csum_partial more readable 

Use standard o32 register name instead of T0, T1, etc, like memcpy.S.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
hifive-unleashed-5.1
Atsushi Nemoto 2006-12-08 01:04:31 +09:00 committed by Ralf Baechle
parent 14b36af46a
commit 52ffe760ea
1 changed files with 74 additions and 70 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

144
arch/mips/lib/csum_partial.S
View File

@ -12,19 +12,23 @@
#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
/*
* As we are sharing code base with the mips32 tree (which use the o32 ABI
* register definitions). We need to redefine the register definitions from
* the n64 ABI register naming to the o32 ABI register naming.
*/
#undef t0
#undef t1
#undef t2
#undef t3
#define t0 $8
#define t1 $9
#define t2 $10
#define t3 $11
#define t4 $12
#define t5 $13
#define t6 $14
#define t7 $15
#endif
#define ADDC(sum,reg) \
@ -64,37 +68,37 @@
/* unknown src alignment and < 8 bytes to go */
small_csumcpy:
move a1, T2
move a1, t2
andi T0, a1, 4
beqz T0, 1f
andi T0, a1, 2
andi t0, a1, 4
beqz t0, 1f
andi t0, a1, 2
/* Still a full word to go */
ulw T1, (src)
ulw t1, (src)
PTR_ADDIU src, 4
ADDC(sum, T1)
ADDC(sum, t1)
1: move T1, zero
beqz T0, 1f
andi T0, a1, 1
1: move t1, zero
beqz t0, 1f
andi t0, a1, 1
/* Still a halfword to go */
ulhu T1, (src)
ulhu t1, (src)
PTR_ADDIU src, 2
1: beqz T0, 1f
sll T1, T1, 16
1: beqz t0, 1f
sll t1, t1, 16
lbu T2, (src)
lbu t2, (src)
nop
#ifdef __MIPSEB__
sll T2, T2, 8
sll t2, t2, 8
#endif
or T1, T2
or t1, t2
1: ADDC(sum, T1)
1: ADDC(sum, t1)
/* fold checksum */
sll v1, sum, 16
@ -104,7 +108,7 @@ small_csumcpy:
addu sum, v1
/* odd buffer alignment? */
beqz T7, 1f
beqz t7, 1f
nop
sll v1, sum, 8
srl sum, sum, 8
@ -122,25 +126,25 @@ small_csumcpy:
.align 5
LEAF(csum_partial)
move sum, zero
move T7, zero
move t7, zero
sltiu t8, a1, 0x8
bnez t8, small_csumcpy /* < 8 bytes to copy */
move T2, a1
move t2, a1
beqz a1, out
andi T7, src, 0x1 /* odd buffer? */
andi t7, src, 0x1 /* odd buffer? */
hword_align:
beqz T7, word_align
beqz t7, word_align
andi t8, src, 0x2
lbu T0, (src)
lbu t0, (src)
LONG_SUBU a1, a1, 0x1
#ifdef __MIPSEL__
sll T0, T0, 8
sll t0, t0, 8
#endif
ADDC(sum, T0)
ADDC(sum, t0)
PTR_ADDU src, src, 0x1
andi t8, src, 0x2
@ -148,9 +152,9 @@ word_align:
beqz t8, dword_align
sltiu t8, a1, 56
lhu T0, (src)
lhu t0, (src)
LONG_SUBU a1, a1, 0x2
ADDC(sum, T0)
ADDC(sum, t0)
sltiu t8, a1, 56
PTR_ADDU src, src, 0x2
@ -162,9 +166,9 @@ dword_align:
beqz t8, qword_align
andi t8, src, 0x8
lw T0, 0x00(src)
lw t0, 0x00(src)
LONG_SUBU a1, a1, 0x4
ADDC(sum, T0)
ADDC(sum, t0)
PTR_ADDU src, src, 0x4
andi t8, src, 0x8
@ -172,11 +176,11 @@ qword_align:
beqz t8, oword_align
andi t8, src, 0x10
lw T0, 0x00(src)
lw T1, 0x04(src)
lw t0, 0x00(src)
lw t1, 0x04(src)
LONG_SUBU a1, a1, 0x8
ADDC(sum, T0)
ADDC(sum, T1)
ADDC(sum, t0)
ADDC(sum, t1)
PTR_ADDU src, src, 0x8
andi t8, src, 0x10
@ -184,46 +188,46 @@ 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)
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
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)
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
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)
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
beqz t2, do_end_words
andi t8, a1, 0x1c
move_32bytes:
CSUM_BIGCHUNK(src, 0x00, sum, T0, T1, T3, T4)
CSUM_BIGCHUNK(src, 0x00, sum, t0, t1, t3, t4)
andi t8, a1, 0x1c
PTR_ADDU src, src, 0x20
@ -232,22 +236,22 @@ do_end_words:
LONG_SRL t8, t8, 0x2
end_words:
lw T0, (src)
lw t0, (src)
LONG_SUBU t8, t8, 0x1
ADDC(sum, T0)
ADDC(sum, t0)
bnez t8, end_words
PTR_ADDU src, src, 0x4
maybe_end_cruft:
andi T2, a1, 0x3
andi t2, a1, 0x3
small_memcpy:
j small_csumcpy; move a1, T2 /* XXX ??? */
j small_csumcpy; move a1, t2 /* XXX ??? */
beqz t2, out
move a1, T2
move a1, t2
end_bytes:
lb T0, (src)
lb t0, (src)
LONG_SUBU a1, a1, 0x1
bnez a2, end_bytes
PTR_ADDU src, src, 0x1