aboutsummaryrefslogtreecommitdiff
path: root/arch/arm/lib/csumpartialcopygeneric.S
diff options
context:
space:
mode:
Diffstat (limited to 'arch/arm/lib/csumpartialcopygeneric.S')
-rw-r--r--arch/arm/lib/csumpartialcopygeneric.S169
1 files changed, 85 insertions, 84 deletions
diff --git a/arch/arm/lib/csumpartialcopygeneric.S b/arch/arm/lib/csumpartialcopygeneric.S
index d3a2f4667db..d6e742d2400 100644
--- a/arch/arm/lib/csumpartialcopygeneric.S
+++ b/arch/arm/lib/csumpartialcopygeneric.S
@@ -22,8 +22,8 @@ dst .req r1
len .req r2
sum .req r3
-.zero: mov r0, sum
- load_regs ea
+.Lzero: mov r0, sum
+ load_regs
/*
* Align an unaligned destination pointer. We know that
@@ -31,8 +31,9 @@ sum .req r3
* the length. Note that the source pointer hasn't been
* aligned yet.
*/
-.dst_unaligned: tst dst, #1
- beq .dst_16bit
+.Ldst_unaligned:
+ tst dst, #1
+ beq .Ldst_16bit
load1b ip
sub len, len, #1
@@ -41,7 +42,7 @@ sum .req r3
tst dst, #2
moveq pc, lr @ dst is now 32bit aligned
-.dst_16bit: load2b r8, ip
+.Ldst_16bit: load2b r8, ip
sub len, len, #2
adcs sum, sum, r8, put_byte_0
strb r8, [dst], #1
@@ -53,12 +54,12 @@ sum .req r3
* Handle 0 to 7 bytes, with any alignment of source and
* destination pointers. Note that when we get here, C = 0
*/
-.less8: teq len, #0 @ check for zero count
- beq .zero
+.Lless8: teq len, #0 @ check for zero count
+ beq .Lzero
/* we must have at least one byte. */
tst dst, #1 @ dst 16-bit aligned
- beq .less8_aligned
+ beq .Lless8_aligned
/* Align dst */
load1b ip
@@ -66,7 +67,7 @@ sum .req r3
adcs sum, sum, ip, put_byte_1 @ update checksum
strb ip, [dst], #1
tst len, #6
- beq .less8_byteonly
+ beq .Lless8_byteonly
1: load2b r8, ip
sub len, len, #2
@@ -74,27 +75,26 @@ sum .req r3
strb r8, [dst], #1
adcs sum, sum, ip, put_byte_1
strb ip, [dst], #1
-.less8_aligned: tst len, #6
+.Lless8_aligned:
+ tst len, #6
bne 1b
-.less8_byteonly:
+.Lless8_byteonly:
tst len, #1
- beq .done
+ beq .Ldone
load1b r8
adcs sum, sum, r8, put_byte_0 @ update checksum
strb r8, [dst], #1
- b .done
+ b .Ldone
FN_ENTRY
- mov ip, sp
save_regs
- sub fp, ip, #4
cmp len, #8 @ Ensure that we have at least
- blo .less8 @ 8 bytes to copy.
+ blo .Lless8 @ 8 bytes to copy.
adds sum, sum, #0 @ C = 0
tst dst, #3 @ Test destination alignment
- blne .dst_unaligned @ align destination, return here
+ blne .Ldst_unaligned @ align destination, return here
/*
* Ok, the dst pointer is now 32bit aligned, and we know
@@ -103,7 +103,7 @@ FN_ENTRY
*/
tst src, #3 @ Test source alignment
- bne .src_not_aligned
+ bne .Lsrc_not_aligned
/* Routine for src & dst aligned */
@@ -136,17 +136,17 @@ FN_ENTRY
adcs sum, sum, r4
4: ands len, len, #3
- beq .done
+ beq .Ldone
load1l r4
tst len, #2
mov r5, r4, get_byte_0
- beq .exit
- adcs sum, sum, r4, push #16
+ beq .Lexit
+ adcs sum, sum, r4, lspush #16
strb r5, [dst], #1
mov r5, r4, get_byte_1
strb r5, [dst], #1
mov r5, r4, get_byte_2
-.exit: tst len, #1
+.Lexit: tst len, #1
strneb r5, [dst], #1
andne r5, r5, #255
adcnes sum, sum, r5, put_byte_0
@@ -157,37 +157,37 @@ FN_ENTRY
* the inefficient byte manipulations in the
* architecture independent code.
*/
-.done: adc r0, sum, #0
+.Ldone: adc r0, sum, #0
ldr sum, [sp, #0] @ dst
tst sum, #1
movne r0, r0, ror #8
- load_regs ea
+ load_regs
-.src_not_aligned:
+.Lsrc_not_aligned:
adc sum, sum, #0 @ include C from dst alignment
and ip, src, #3
bic src, src, #3
load1l r5
cmp ip, #2
- beq .src2_aligned
- bhi .src3_aligned
- mov r4, r5, pull #8 @ C = 0
+ beq .Lsrc2_aligned
+ bhi .Lsrc3_aligned
+ mov r4, r5, lspull #8 @ C = 0
bics ip, len, #15
beq 2f
1: load4l r5, r6, r7, r8
- orr r4, r4, r5, push #24
- mov r5, r5, pull #8
- orr r5, r5, r6, push #24
- mov r6, r6, pull #8
- orr r6, r6, r7, push #24
- mov r7, r7, pull #8
- orr r7, r7, r8, push #24
+ orr r4, r4, r5, lspush #24
+ mov r5, r5, lspull #8
+ orr r5, r5, r6, lspush #24
+ mov r6, r6, lspull #8
+ orr r6, r6, r7, lspush #24
+ mov r7, r7, lspull #8
+ orr r7, r7, r8, lspush #24
stmia dst!, {r4, r5, r6, r7}
adcs sum, sum, r4
adcs sum, sum, r5
adcs sum, sum, r6
adcs sum, sum, r7
- mov r4, r8, pull #8
+ mov r4, r8, lspull #8
sub ip, ip, #16
teq ip, #0
bne 1b
@@ -196,50 +196,50 @@ FN_ENTRY
tst ip, #8
beq 3f
load2l r5, r6
- orr r4, r4, r5, push #24
- mov r5, r5, pull #8
- orr r5, r5, r6, push #24
+ orr r4, r4, r5, lspush #24
+ mov r5, r5, lspull #8
+ orr r5, r5, r6, lspush #24
stmia dst!, {r4, r5}
adcs sum, sum, r4
adcs sum, sum, r5
- mov r4, r6, pull #8
+ mov r4, r6, lspull #8
tst ip, #4
beq 4f
3: load1l r5
- orr r4, r4, r5, push #24
+ orr r4, r4, r5, lspush #24
str r4, [dst], #4
adcs sum, sum, r4
- mov r4, r5, pull #8
+ mov r4, r5, lspull #8
4: ands len, len, #3
- beq .done
+ beq .Ldone
mov r5, r4, get_byte_0
tst len, #2
- beq .exit
- adcs sum, sum, r4, push #16
+ beq .Lexit
+ adcs sum, sum, r4, lspush #16
strb r5, [dst], #1
mov r5, r4, get_byte_1
strb r5, [dst], #1
mov r5, r4, get_byte_2
- b .exit
+ b .Lexit
-.src2_aligned: mov r4, r5, pull #16
+.Lsrc2_aligned: mov r4, r5, lspull #16
adds sum, sum, #0
bics ip, len, #15
beq 2f
1: load4l r5, r6, r7, r8
- orr r4, r4, r5, push #16
- mov r5, r5, pull #16
- orr r5, r5, r6, push #16
- mov r6, r6, pull #16
- orr r6, r6, r7, push #16
- mov r7, r7, pull #16
- orr r7, r7, r8, push #16
+ orr r4, r4, r5, lspush #16
+ mov r5, r5, lspull #16
+ orr r5, r5, r6, lspush #16
+ mov r6, r6, lspull #16
+ orr r6, r6, r7, lspush #16
+ mov r7, r7, lspull #16
+ orr r7, r7, r8, lspush #16
stmia dst!, {r4, r5, r6, r7}
adcs sum, sum, r4
adcs sum, sum, r5
adcs sum, sum, r6
adcs sum, sum, r7
- mov r4, r8, pull #16
+ mov r4, r8, lspull #16
sub ip, ip, #16
teq ip, #0
bne 1b
@@ -248,52 +248,52 @@ FN_ENTRY
tst ip, #8
beq 3f
load2l r5, r6
- orr r4, r4, r5, push #16
- mov r5, r5, pull #16
- orr r5, r5, r6, push #16
+ orr r4, r4, r5, lspush #16
+ mov r5, r5, lspull #16
+ orr r5, r5, r6, lspush #16
stmia dst!, {r4, r5}
adcs sum, sum, r4
adcs sum, sum, r5
- mov r4, r6, pull #16
+ mov r4, r6, lspull #16
tst ip, #4
beq 4f
3: load1l r5
- orr r4, r4, r5, push #16
+ orr r4, r4, r5, lspush #16
str r4, [dst], #4
adcs sum, sum, r4
- mov r4, r5, pull #16
+ mov r4, r5, lspull #16
4: ands len, len, #3
- beq .done
+ beq .Ldone
mov r5, r4, get_byte_0
tst len, #2
- beq .exit
+ beq .Lexit
adcs sum, sum, r4
strb r5, [dst], #1
mov r5, r4, get_byte_1
strb r5, [dst], #1
tst len, #1
- beq .done
+ beq .Ldone
load1b r5
- b .exit
+ b .Lexit
-.src3_aligned: mov r4, r5, pull #24
+.Lsrc3_aligned: mov r4, r5, lspull #24
adds sum, sum, #0
bics ip, len, #15
beq 2f
1: load4l r5, r6, r7, r8
- orr r4, r4, r5, push #8
- mov r5, r5, pull #24
- orr r5, r5, r6, push #8
- mov r6, r6, pull #24
- orr r6, r6, r7, push #8
- mov r7, r7, pull #24
- orr r7, r7, r8, push #8
+ orr r4, r4, r5, lspush #8
+ mov r5, r5, lspull #24
+ orr r5, r5, r6, lspush #8
+ mov r6, r6, lspull #24
+ orr r6, r6, r7, lspush #8
+ mov r7, r7, lspull #24
+ orr r7, r7, r8, lspush #8
stmia dst!, {r4, r5, r6, r7}
adcs sum, sum, r4
adcs sum, sum, r5
adcs sum, sum, r6
adcs sum, sum, r7
- mov r4, r8, pull #24
+ mov r4, r8, lspull #24
sub ip, ip, #16
teq ip, #0
bne 1b
@@ -302,30 +302,31 @@ FN_ENTRY
tst ip, #8
beq 3f
load2l r5, r6
- orr r4, r4, r5, push #8
- mov r5, r5, pull #24
- orr r5, r5, r6, push #8
+ orr r4, r4, r5, lspush #8
+ mov r5, r5, lspull #24
+ orr r5, r5, r6, lspush #8
stmia dst!, {r4, r5}
adcs sum, sum, r4
adcs sum, sum, r5
- mov r4, r6, pull #24
+ mov r4, r6, lspull #24
tst ip, #4
beq 4f
3: load1l r5
- orr r4, r4, r5, push #8
+ orr r4, r4, r5, lspush #8
str r4, [dst], #4
adcs sum, sum, r4
- mov r4, r5, pull #24
+ mov r4, r5, lspull #24
4: ands len, len, #3
- beq .done
+ beq .Ldone
mov r5, r4, get_byte_0
tst len, #2
- beq .exit
+ beq .Lexit
strb r5, [dst], #1
adcs sum, sum, r4
load1l r4
mov r5, r4, get_byte_0
strb r5, [dst], #1
- adcs sum, sum, r4, push #24
+ adcs sum, sum, r4, lspush #24
mov r5, r4, get_byte_1
- b .exit
+ b .Lexit
+FN_EXIT
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-12-05 22:09:01 +0000