summaryrefslogtreecommitdiffstats
path: root/arch/sparc64/lib/checksum.S
diff options
context:
space:
mode:
Diffstat (limited to 'arch/sparc64/lib/checksum.S')
-rw-r--r--arch/sparc64/lib/checksum.S76
1 files changed, 42 insertions, 34 deletions
diff --git a/arch/sparc64/lib/checksum.S b/arch/sparc64/lib/checksum.S
index b63f0d6e8..10eebb8df 100644
--- a/arch/sparc64/lib/checksum.S
+++ b/arch/sparc64/lib/checksum.S
@@ -71,8 +71,9 @@ csum_partial_end_cruft:
or %o5, %o4, %o4 ! coalese with hword (if any)
6: addcc %o4, %o2, %o2 ! add to sum
1: sllx %g4, 32, %g4 ! give gfp back
+ addc %g0, %o2, %o0 ! add final carry into retval
retl ! get outta here
- addc %g0, %o2, %o0 ! add final carry into retval
+ srl %o0, 0, %o0
/* Also do alignment out of band to get better cache patterns. */
csum_partial_fix_alignment:
@@ -82,7 +83,9 @@ csum_partial_fix_alignment:
*/
.globl csum_partial
csum_partial: /* %o0=buf, %o1=len, %o2=sum */
+ srl %o1, 0, %o1 ! doof scheiss
andcc %o0, 0x7, %g0 ! alignment problems?
+ srl %o2, 0, %o2
be,pt %icc, csum_partial_fix_aligned ! yep, handle it
andn %o1, 0x7f, %o3 ! num loop iterations
cmp %o1, 6
@@ -154,31 +157,31 @@ __csum_partial_copy_start:
99: ba,pt %xcc, 30f; \
a, b, %o3; \
.section __ex_table,z##alloc; \
- .align 4; \
- .word 98b, 99b; \
+ .align 8; \
+ .xword 98b, 99b; \
.text; \
.align 4
#define EX2(x,y,z) \
98: x,y; \
.section __ex_table,z##alloc; \
- .align 4; \
- .word 98b, 30f; \
+ .align 8; \
+ .xword 98b, 30f; \
.text; \
.align 4
#define EX3(x,y,z) \
98: x,y; \
.section __ex_table,z##alloc; \
- .align 4; \
- .word 98b, 96f; \
+ .align 8; \
+ .xword 98b, 96f; \
.text; \
.align 4
#define EXT(start,end,handler,z) \
.section __ex_table,z##alloc; \
- .align 4; \
- .word start, 0, end, handler; \
+ .align 8; \
+ .xword start, 0, end, handler; \
.text; \
.align 4
@@ -189,12 +192,12 @@ __csum_partial_copy_start:
* please check the fixup code below as well.
*/
#define CSUMCOPY_BIGCHUNK_ALIGNED(src, dst, sum, off, t0, t1, t2, t3, t4, t5, t6, t7) \
- ldd [src + off + 0x00], t0; \
- ldd [src + off + 0x08], t2; \
+ ldda [src + off + 0x00] %asi, t0; \
+ ldda [src + off + 0x08] %asi, t2; \
addccc t0, sum, sum; \
- ldd [src + off + 0x10], t4; \
+ ldda [src + off + 0x10] %asi, t4; \
addccc t1, sum, sum; \
- ldd [src + off + 0x18], t6; \
+ ldda [src + off + 0x18] %asi, t6; \
addccc t2, sum, sum; \
std t0, [dst + off + 0x00]; \
addccc t3, sum, sum; \
@@ -211,10 +214,10 @@ __csum_partial_copy_start:
* Viking MXCC into streaming mode. Ho hum...
*/
#define CSUMCOPY_BIGCHUNK(src, dst, sum, off, t0, t1, t2, t3, t4, t5, t6, t7) \
- ldd [src + off + 0x00], t0; \
- ldd [src + off + 0x08], t2; \
- ldd [src + off + 0x10], t4; \
- ldd [src + off + 0x18], t6; \
+ ldda [src + off + 0x00] %asi, t0; \
+ ldda [src + off + 0x08] %asi, t2; \
+ ldda [src + off + 0x10] %asi, t4; \
+ ldda [src + off + 0x18] %asi, t6; \
st t0, [dst + off + 0x00]; \
addccc t0, sum, sum; \
st t1, [dst + off + 0x04]; \
@@ -234,8 +237,8 @@ __csum_partial_copy_start:
/* Yuck, 6 superscalar cycles... */
#define CSUMCOPY_LASTCHUNK(src, dst, sum, off, t0, t1, t2, t3) \
- ldd [src - off - 0x08], t0; \
- ldd [src - off - 0x00], t2; \
+ ldda [src - off - 0x08] %asi, t0; \
+ ldda [src - off - 0x00] %asi, t2; \
addccc t0, sum, sum; \
st t0, [dst - off - 0x08]; \
addccc t1, sum, sum; \
@@ -250,7 +253,7 @@ cc_end_cruft:
andcc %o3, 8, %g0 ! begin checks for that code
be,pn %icc, 1f
and %o3, 4, %g5
- EX(ldd [%o0 + 0x00], %g2, and %o3, 0xf,#)
+ EX(ldda [%o0 + 0x00] %asi, %g2, and %o3, 0xf,#)
add %o1, 8, %o1
addcc %g2, %g7, %g7
add %o0, 8, %o0
@@ -260,7 +263,7 @@ cc_end_cruft:
EX2(st %g3, [%o1 - 0x04],#)
1: brz,pt %g5, 1f
andcc %o3, 3, %o3
- EX(ld [%o0 + 0x00], %g2, add %o3, 4,#)
+ EX(lda [%o0 + 0x00] %asi, %g2, add %o3, 4,#)
add %o1, 4, %o1
addcc %g2, %g7, %g7
EX2(st %g2, [%o1 - 0x04],#)
@@ -272,20 +275,21 @@ cc_end_cruft:
subcc %o3, 2, %o3
ba,pt %xcc, 4f
clr %o4
-2: EX(lduh [%o0 + 0x00], %o4, add %o3, 2,#)
+2: EX(lduha [%o0 + 0x00] %asi, %o4, add %o3, 2,#)
add %o0, 2, %o0
EX2(sth %o4, [%o1 + 0x00],#)
be,pn %icc, 6f
add %o1, 2, %o1
sll %o4, 16, %o4
-4: EX(ldub [%o0 + 0x00], %o5, add %g0, 1,#)
+4: EX(lduba [%o0 + 0x00] %asi, %o5, add %g0, 1,#)
EX2(stb %o5, [%o1 + 0x00],#)
sll %o5, 8, %o5
or %o5, %o4, %o4
6: addcc %o4, %g7, %g7
1: sllx %g4, 32, %g4
+ addc %g0, %g7, %o0
retl
- addc %g0, %g7, %o0
+ srl %o0, 0, %o0
/* Sun, you just can't beat me, you just can't. Stop trying,
* give up. I'm serious, I am going to kick the living shit
@@ -295,7 +299,9 @@ cc_end_cruft:
.globl __csum_partial_copy_sparc_generic
__csum_partial_copy_sparc_generic:
/* %o0=src, %o1=dest, %g1=len, %g7=sum */
+ srl %g7, 0, %g7 ! you neve know...
xor %o0, %o1, %o4 ! get changing bits
+ srl %g1, 0, %g1 ! doof scheiss
andcc %o4, 3, %g0 ! check for mismatched alignment
bne,pn %icc, ccslow ! better this than unaligned/fixups
andcc %o0, 7, %g0 ! need to align things?
@@ -309,7 +315,7 @@ __csum_partial_copy_sparc_generic:
andcc %o0, 0x2, %g0
be,pn %icc, 1f
andcc %o0, 0x4, %g0
- EX(lduh [%o0 + 0x00], %g4, add %g1, 0,#)
+ EX(lduha [%o0 + 0x00] %asi, %g4, add %g1, 0,#)
sub %g1, 2, %g1
EX2(sth %g4, [%o1 + 0x00],#)
add %o0, 2, %o0
@@ -325,7 +331,7 @@ __csum_partial_copy_sparc_generic:
or %g3, %g7, %g7
1: be,pt %icc, 3f
andn %g1, 0x7f, %g2
- EX(ld [%o0 + 0x00], %g4, add %g1, 0,#)
+ EX(lda [%o0 + 0x00] %asi, %g4, add %g1, 0,#)
sub %g1, 4, %g1
EX2(st %g4, [%o1 + 0x00],#)
add %o0, 4, %o0
@@ -372,8 +378,9 @@ cctbl: CSUMCOPY_LASTCHUNK(%o0,%o1,%g7,0x68,%g2,%g3,%g4,%g5)
ccte: bne,pn %icc, cc_end_cruft ! something left, handle it out of band
sethi %uhi(KERNBASE), %g4 ! restore gfp
mov %g7, %o0 ! give em the computed checksum
+ sllx %g4, 32, %g4 ! finish gfp restoration
retl ! return
- sllx %g4, 32, %g4 ! finish gfp restoration
+ srl %o0, 0, %o0
ccdbl: CSUMCOPY_BIGCHUNK_ALIGNED(%o0,%o1,%g7,0x00,%o4,%o5,%g2,%g3,%g4,%g5,%o2,%o3)
CSUMCOPY_BIGCHUNK_ALIGNED(%o0,%o1,%g7,0x20,%o4,%o5,%g2,%g3,%g4,%g5,%o2,%o3)
CSUMCOPY_BIGCHUNK_ALIGNED(%o0,%o1,%g7,0x40,%o4,%o5,%g2,%g3,%g4,%g5,%o2,%o3)
@@ -394,7 +401,7 @@ ccslow: mov 0, %g5
be,a,pt %icc, 1f
srl %g1, 1, %o3
sub %g1, 1, %g1
- EX(ldub [%o0], %g5, add %g1, 1,#)
+ EX(lduba [%o0] %asi, %g5, add %g1, 1,#)
add %o0, 1, %o0
EX2(stb %g5, [%o1],#)
srl %g1, 1, %o3
@@ -404,7 +411,7 @@ ccslow: mov 0, %g5
andcc %o0, 2, %g0
be,a,pt %icc, 1f
srl %o3, 1, %o3
- EX(lduh [%o0], %o4, add %g1, 0,#)
+ EX(lduha [%o0] %asi, %o4, add %g1, 0,#)
sub %g1, 2, %g1
srl %o4, 8, %g2
sub %o3, 1, %o3
@@ -416,7 +423,7 @@ ccslow: mov 0, %g5
add %o1, 2, %o1
1: brz,a,pn %o3, 2f
andcc %g1, 2, %g0
- EX3(ld [%o0], %o4,#)
+ EX3(lda [%o0] %asi, %o4,#)
5: srl %o4, 24, %g2
srl %o4, 16, %g3
EX2(stb %g2, [%o1],#)
@@ -430,7 +437,7 @@ ccslow: mov 0, %g5
add %o1, 4, %o1 ! is worthy). Maybe some day - with the sll/srl
subcc %o3, 1, %o3 ! tricks
bne,a,pt %icc, 5b
- EX3(ld [%o0], %o4,#)
+ EX3(lda [%o0] %asi, %o4,#)
sll %g5, 16, %g2
srl %g5, 16, %g5
srl %g2, 16, %g2
@@ -438,7 +445,7 @@ ccslow: mov 0, %g5
add %g2, %g5, %g5
2: be,a,pt %icc, 3f
andcc %g1, 1, %g0
- EX(lduh [%o0], %o4, and %g1, 3,#)
+ EX(lduha [%o0] %asi, %o4, and %g1, 3,#)
andcc %g1, 1, %g0
srl %o4, 8, %g2
add %o0, 2, %o0
@@ -448,7 +455,7 @@ ccslow: mov 0, %g5
add %o1, 2, %o1
3: be,a,pt %icc, 1f
sll %g5, 16, %o4
- EX(ldub [%o0], %g2, add %g0, 1,#)
+ EX(lduba [%o0] %asi, %g2, add %g0, 1,#)
sll %g2, 8, %o4
EX2(stb %g2, [%o1],#)
add %g5, %o4, %g5
@@ -463,8 +470,9 @@ ccslow: mov 0, %g5
sll %g2, 8, %g2
or %g2, %o4, %g5
4: addcc %g7, %g5, %g7
+ addc %g0, %g7, %o0
retl
- addc %g0, %g7, %o0
+ srl %o0, 0, %o0
__csum_partial_copy_end:
.section .fixup,#alloc,#execinstr
generated by cgit v1.2.3 (git 2.25.1) at 2024-09-20 08:39:48 +0000