s390x assembler pack.
This commit is contained in:
Andy Polyakov
parent
0a22e7446b
commit
87339c6290
@ -122,6 +122,7 @@ my $x86_out_asm="x86cpuid-out.o:bn86-out.o co86-out.o:dx86-out.o yx86-out.o:ax86
|
||||
|
||||
my $x86_64_asm="x86_64cpuid.o:x86_64-gcc.o::aes-x86_64.o::md5-x86_64.o:sha1-x86_64.o sha256-x86_64.o sha512-x86_64.o::rc4-x86_64.o::";
|
||||
my $ia64_asm=":bn-ia64.o::aes_core.o aes_cbc.o aes-ia64.o:::sha1-ia64.o sha256-ia64.o sha512-ia64.o::rc4-ia64.o::";
|
||||
my $s390x_asm=":bn-s390x.o::aes_cbc.o aes-s390x.o:::sha1-s390x.o sha256-s390x.o sha512-s390x.o::::";
|
||||
|
||||
my $no_asm="::::::::::";
|
||||
|
||||
@ -328,6 +329,7 @@ my %table=(
|
||||
"linux-ia64-ecc","ecc:-DL_ENDIAN -DTERMIO -O2 -Wall -no_cpprt::-D_REENTRANT::-ldl:SIXTY_FOUR_BIT_LONG RC4_CHUNK:${ia64_asm}:dlfcn:linux-shared:-fPIC::.so.\$(SHLIB_MAJOR).\$(SHLIB_MINOR)",
|
||||
"linux-ia64-icc","icc:-DL_ENDIAN -DTERMIO -O2 -Wall -no_cpprt::-D_REENTRANT::-ldl:SIXTY_FOUR_BIT_LONG RC4_CHUNK:${ia64_asm}:dlfcn:linux-shared:-fPIC::.so.\$(SHLIB_MAJOR).\$(SHLIB_MINOR)",
|
||||
"linux-x86_64", "gcc:-m64 -DL_ENDIAN -DTERMIO -O3 -Wall -DMD32_REG_T=int::-D_REENTRANT::-ldl:SIXTY_FOUR_BIT_LONG RC4_CHUNK BF_PTR2 DES_INT DES_UNROLL:${x86_64_asm}:dlfcn:linux-shared:-fPIC:-m64:.so.\$(SHLIB_MAJOR).\$(SHLIB_MINOR)",
|
||||
"linux-s390x", "gcc:-m64 -DB_ENDIAN -DTERMIO -O3 -Wall::-D_REENTRANT::-ldl:SIXTY_FOUR_BIT_LONG RC4_CHAR RC4_CHUNK DES_INT DES_UNROLL:${s390x_asm}:dlfcn:linux-shared:-fPIC:-m64:.so.\$(SHLIB_MAJOR).\$(SHLIB_MINOR)",
|
||||
#### SPARC Linux setups
|
||||
# Ray Miller <ray.miller@computing-services.oxford.ac.uk> has patiently
|
||||
# assisted with debugging of following two configs.
|
||||
|
||||
27
TABLE
27
TABLE
@ -2835,6 +2835,33 @@ $shared_extension = .so.$(SHLIB_MAJOR).$(SHLIB_MINOR)
|
||||
$ranlib =
|
||||
$arflags =
|
||||
|
||||
*** linux-s390x
|
||||
$cc = gcc
|
||||
$cflags = -m64 -DB_ENDIAN -DTERMIO -O3 -Wall
|
||||
$unistd =
|
||||
$thread_cflag = -D_REENTRANT
|
||||
$sys_id =
|
||||
$lflags = -ldl
|
||||
$bn_ops = SIXTY_FOUR_BIT_LONG RC4_CHAR RC4_CHUNK DES_INT DES_UNROLL
|
||||
$cpuid_obj =
|
||||
$bn_obj = bn-s390x.o
|
||||
$des_obj =
|
||||
$aes_obj = aes_cbc.o aes-s390x.o
|
||||
$bf_obj =
|
||||
$md5_obj =
|
||||
$sha1_obj = sha1-s390x.o sha256-s390x.o sha512-s390x.o
|
||||
$cast_obj =
|
||||
$rc4_obj =
|
||||
$rmd160_obj =
|
||||
$rc5_obj =
|
||||
$dso_scheme = dlfcn
|
||||
$shared_target= linux-shared
|
||||
$shared_cflag = -fPIC
|
||||
$shared_ldflag = -m64
|
||||
$shared_extension = .so.$(SHLIB_MAJOR).$(SHLIB_MINOR)
|
||||
$ranlib =
|
||||
$arflags =
|
||||
|
||||
*** linux-sparcv8
|
||||
$cc = gcc
|
||||
$cflags = -mv8 -DB_ENDIAN -DTERMIO -O3 -fomit-frame-pointer -Wall -DBN_DIV2W
|
||||
|
||||
2
config
2
config
@ -589,7 +589,7 @@ case "$GUESSOS" in
|
||||
sh*-*-linux2) OUT="linux-generic32"; options="$options -DL_ENDIAN" ;;
|
||||
m68k*-*-linux2) OUT="linux-generic32"; options="$options -DB_ENDIAN" ;;
|
||||
s390-*-linux2) OUT="linux-generic32"; options="$options -DB_ENDIAN -DNO_ASM" ;;
|
||||
s390x-*-linux2) OUT="linux-generic64"; options="$options -DB_ENDIAN" ;;
|
||||
s390x-*-linux2) OUT="linux-s390x" ;;
|
||||
x86_64-*-linux?) OUT="linux-x86_64" ;;
|
||||
*86-*-linux2) OUT="linux-elf"
|
||||
if [ "$GCCVER" -gt 28 ]; then
|
||||
|
||||
1071
crypto/aes/asm/aes-s390x.pl
Normal file
1071
crypto/aes/asm/aes-s390x.pl
Normal file
File diff suppressed because it is too large
Load Diff
@ -95,6 +95,9 @@ x86_64-gcc.o: asm/x86_64-gcc.c
|
||||
bn-ia64.s: asm/ia64.S
|
||||
$(CC) $(CFLAGS) -E asm/ia64.S > $@
|
||||
|
||||
bn-s390x.s: asm/s390x.S
|
||||
$(CC) $(CFLAGS) -E asm/s390x.S > $@
|
||||
|
||||
# GNU assembler fails to compile PA-RISC2 modules, insist on calling
|
||||
# vendor assembler...
|
||||
pa-risc2W.o: asm/pa-risc2W.s
|
||||
|
||||
678
crypto/bn/asm/s390x.S
Executable file
678
crypto/bn/asm/s390x.S
Executable file
@ -0,0 +1,678 @@
|
||||
.ident "s390x.S, version 1.0"
|
||||
// ====================================================================
|
||||
// Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
|
||||
// project.
|
||||
//
|
||||
// Rights for redistribution and usage in source and binary forms are
|
||||
// granted according to the OpenSSL license. Warranty of any kind is
|
||||
// disclaimed.
|
||||
// ====================================================================
|
||||
|
||||
.text
|
||||
|
||||
#define zero %r0
|
||||
|
||||
// BN_ULONG bn_mul_add_words(BN_ULONG *r2,BN_ULONG *r3,int r4,BN_ULONG r5);
|
||||
.globl bn_mul_add_words
|
||||
.type bn_mul_add_words,@function
|
||||
.align 4
|
||||
bn_mul_add_words:
|
||||
lghi zero,0 // zero = 0
|
||||
la %r1,0(%r2) // put rp aside
|
||||
lghi %r2,0 // i=0;
|
||||
ltgfr %r4,%r4
|
||||
bler %r14 // if (len<=0) return 0;
|
||||
|
||||
stmg %r6,%r10,48(%r15)
|
||||
lghi %r8,0 // carry = 0
|
||||
srag %r10,%r4,2 // cnt=len/4
|
||||
jz .Loop1_madd
|
||||
|
||||
.Loop4_madd:
|
||||
lg %r7,0(%r2,%r3) // ap[i]
|
||||
mlgr %r6,%r5 // *=w
|
||||
algr %r7,%r8 // +=carry
|
||||
alcgr %r6,zero
|
||||
alg %r7,0(%r2,%r1) // +=rp[i]
|
||||
alcgr %r6,zero
|
||||
stg %r7,0(%r2,%r1) // rp[i]=
|
||||
|
||||
lg %r9,8(%r2,%r3)
|
||||
mlgr %r8,%r5
|
||||
algr %r9,%r6
|
||||
alcgr %r8,zero
|
||||
alg %r9,8(%r2,%r1)
|
||||
alcgr %r8,zero
|
||||
stg %r9,8(%r2,%r1)
|
||||
|
||||
lg %r7,16(%r2,%r3)
|
||||
mlgr %r6,%r5
|
||||
algr %r7,%r8
|
||||
alcgr %r6,zero
|
||||
alg %r7,16(%r2,%r1)
|
||||
alcgr %r6,zero
|
||||
stg %r7,16(%r2,%r1)
|
||||
|
||||
lg %r9,24(%r2,%r3)
|
||||
mlgr %r8,%r5
|
||||
algr %r9,%r6
|
||||
alcgr %r8,zero
|
||||
alg %r9,24(%r2,%r1)
|
||||
alcgr %r8,zero
|
||||
stg %r9,24(%r2,%r1)
|
||||
|
||||
la %r2,32(%r2) // i+=4
|
||||
brct %r10,.Loop4_madd
|
||||
|
||||
lghi %r10,3
|
||||
nr %r4,%r10 // cnt=len%4
|
||||
jz .Lend_madd
|
||||
|
||||
.Loop1_madd:
|
||||
lg %r7,0(%r2,%r3) // ap[i]
|
||||
mlgr %r6,%r5 // *=w
|
||||
algr %r7,%r8 // +=carry
|
||||
alcgr %r6,zero
|
||||
alg %r7,0(%r2,%r1) // +=rp[i]
|
||||
alcgr %r6,zero
|
||||
stg %r7,0(%r2,%r1) // rp[i]=
|
||||
|
||||
lgr %r8,%r6
|
||||
la %r2,8(%r2) // i++
|
||||
brct %r4,.Loop1_madd
|
||||
|
||||
.Lend_madd:
|
||||
lgr %r2,%r8
|
||||
lmg %r6,%r10,48(%r15)
|
||||
br %r14
|
||||
.size bn_mul_add_words,.-bn_mul_add_words
|
||||
|
||||
// BN_ULONG bn_mul_words(BN_ULONG *r2,BN_ULONG *r3,int r4,BN_ULONG r5);
|
||||
.globl bn_mul_words
|
||||
.type bn_mul_words,@function
|
||||
.align 4
|
||||
bn_mul_words:
|
||||
lghi zero,0 // zero = 0
|
||||
la %r1,0(%r2) // put rp aside
|
||||
lghi %r2,0 // i=0;
|
||||
ltgfr %r4,%r4
|
||||
bler %r14 // if (len<=0) return 0;
|
||||
|
||||
stmg %r6,%r10,48(%r15)
|
||||
lghi %r8,0 // carry = 0
|
||||
srag %r10,%r4,2 // cnt=len/4
|
||||
jz .Loop1_mul
|
||||
|
||||
.Loop4_mul:
|
||||
lg %r7,0(%r2,%r3) // ap[i]
|
||||
mlgr %r6,%r5 // *=w
|
||||
algr %r7,%r8 // +=carry
|
||||
alcgr %r6,zero
|
||||
stg %r7,0(%r2,%r1) // rp[i]=
|
||||
|
||||
lg %r9,8(%r2,%r3)
|
||||
mlgr %r8,%r5
|
||||
algr %r9,%r6
|
||||
alcgr %r8,zero
|
||||
stg %r9,8(%r2,%r1)
|
||||
|
||||
lg %r7,16(%r2,%r3)
|
||||
mlgr %r6,%r5
|
||||
algr %r7,%r8
|
||||
alcgr %r6,zero
|
||||
stg %r7,16(%r2,%r1)
|
||||
|
||||
lg %r9,24(%r2,%r3)
|
||||
mlgr %r8,%r5
|
||||
algr %r9,%r6
|
||||
alcgr %r8,zero
|
||||
stg %r9,24(%r2,%r1)
|
||||
|
||||
la %r2,32(%r2) // i+=4
|
||||
brct %r10,.Loop4_mul
|
||||
|
||||
lghi %r10,3
|
||||
nr %r4,%r10 // cnt=len%4
|
||||
jz .Lend_mul
|
||||
|
||||
.Loop1_mul:
|
||||
lg %r7,0(%r2,%r3) // ap[i]
|
||||
mlgr %r6,%r5 // *=w
|
||||
algr %r7,%r8 // +=carry
|
||||
alcgr %r6,zero
|
||||
stg %r7,0(%r2,%r1) // rp[i]=
|
||||
|
||||
lgr %r8,%r6
|
||||
la %r2,8(%r2) // i++
|
||||
brct %r4,.Loop1_mul
|
||||
|
||||
.Lend_mul:
|
||||
lgr %r2,%r8
|
||||
lmg %r6,%r10,48(%r15)
|
||||
br %r14
|
||||
.size bn_mul_words,.-bn_mul_words
|
||||
|
||||
// void bn_sqr_words(BN_ULONG *r2,BN_ULONG *r2,int r4)
|
||||
.globl bn_sqr_words
|
||||
.type bn_sqr_words,@function
|
||||
.align 4
|
||||
bn_sqr_words:
|
||||
ltgfr %r4,%r4
|
||||
bler %r14
|
||||
|
||||
stmg %r6,%r7,48(%r15)
|
||||
srag %r1,%r4,2 // cnt=len/4
|
||||
jz .Loop1_sqr
|
||||
|
||||
.Loop4_sqr:
|
||||
lg %r7,0(%r3)
|
||||
mlgr %r6,%r7
|
||||
stg %r7,0(%r2)
|
||||
stg %r6,8(%r2)
|
||||
|
||||
lg %r7,8(%r3)
|
||||
mlgr %r6,%r7
|
||||
stg %r7,16(%r2)
|
||||
stg %r6,24(%r2)
|
||||
|
||||
lg %r7,16(%r3)
|
||||
mlgr %r6,%r7
|
||||
stg %r7,32(%r2)
|
||||
stg %r6,40(%r2)
|
||||
|
||||
lg %r7,24(%r3)
|
||||
mlgr %r6,%r7
|
||||
stg %r7,48(%r2)
|
||||
stg %r6,56(%r2)
|
||||
|
||||
la %r3,32(%r3)
|
||||
la %r2,64(%r2)
|
||||
brct %r1,.Loop4_sqr
|
||||
|
||||
lghi %r1,3
|
||||
nr %r4,%r1 // cnt=len%4
|
||||
jz .Lend_sqr
|
||||
|
||||
.Loop1_sqr:
|
||||
lg %r7,0(%r3)
|
||||
mlgr %r6,%r7
|
||||
stg %r7,0(%r2)
|
||||
stg %r6,8(%r2)
|
||||
|
||||
la %r3,8(%r3)
|
||||
la %r2,16(%r2)
|
||||
brct %r4,.Loop1_sqr
|
||||
|
||||
.Lend_sqr:
|
||||
lmg %r6,%r7,48(%r15)
|
||||
br %r14
|
||||
.size bn_sqr_words,.-bn_sqr_words
|
||||
|
||||
// BN_ULONG bn_div_words(BN_ULONG h,BN_ULONG l,BN_ULONG d);
|
||||
.globl bn_div_words
|
||||
.type bn_div_words,@function
|
||||
.align 4
|
||||
bn_div_words:
|
||||
dlgr %r2,%r4
|
||||
lgr %r2,%r3
|
||||
br %r14
|
||||
.size bn_div_words,.-bn_div_words
|
||||
|
||||
// BN_ULONG bn_add_words(BN_ULONG *r2,BN_ULONG *r3,BN_ULONG *r4,int r5);
|
||||
.globl bn_add_words
|
||||
.type bn_add_words,@function
|
||||
.align 4
|
||||
bn_add_words:
|
||||
la %r1,0(%r2) // put rp aside
|
||||
lghi %r2,0 // i=0
|
||||
ltgfr %r5,%r5
|
||||
bler %r14 // if (len<=0) return 0;
|
||||
|
||||
stg %r6,48(%r15)
|
||||
lghi %r6,3
|
||||
nr %r6,%r5 // len%4
|
||||
sra %r5,2 // len/4, use sra because it sets condition code
|
||||
jz .Loop1_add // carry is incidentally cleared if branch taken
|
||||
algr %r2,%r2 // clear carry
|
||||
|
||||
.Loop4_add:
|
||||
lg %r0,0(%r2,%r3)
|
||||
alcg %r0,0(%r2,%r4)
|
||||
stg %r0,0(%r2,%r1)
|
||||
lg %r0,8(%r2,%r3)
|
||||
alcg %r0,8(%r2,%r4)
|
||||
stg %r0,8(%r2,%r1)
|
||||
lg %r0,16(%r2,%r3)
|
||||
alcg %r0,16(%r2,%r4)
|
||||
stg %r0,16(%r2,%r1)
|
||||
lg %r0,24(%r2,%r3)
|
||||
alcg %r0,24(%r2,%r4)
|
||||
stg %r0,24(%r2,%r1)
|
||||
|
||||
la %r2,32(%r2) // i+=4
|
||||
brct %r5,.Loop4_add
|
||||
|
||||
la %r6,1(%r6) // see if len%4 is zero ...
|
||||
brct %r6,.Loop1_add // without touching condition code:-)
|
||||
|
||||
.Lexit_add:
|
||||
lghi %r2,0
|
||||
alcgr %r2,%r2
|
||||
lg %r6,48(%r15)
|
||||
br %r14
|
||||
|
||||
.Loop1_add:
|
||||
lg %r0,0(%r2,%r3)
|
||||
alcg %r0,0(%r2,%r4)
|
||||
stg %r0,0(%r2,%r1)
|
||||
|
||||
la %r2,8(%r2) // i++
|
||||
brct %r6,.Loop1_add
|
||||
|
||||
j .Lexit_add
|
||||
.size bn_add_words,.-bn_add_words
|
||||
|
||||
// BN_ULONG bn_sub_words(BN_ULONG *r2,BN_ULONG *r3,BN_ULONG *r4,int r5);
|
||||
.globl bn_sub_words
|
||||
.type bn_sub_words,@function
|
||||
.align 4
|
||||
bn_sub_words:
|
||||
la %r1,0(%r2) // put rp aside
|
||||
lghi %r2,0 // i=0
|
||||
ltgfr %r5,%r5
|
||||
bler %r14 // if (len<=0) return 0;
|
||||
|
||||
stg %r6,48(%r15)
|
||||
lghi %r6,3
|
||||
nr %r6,%r5 // len%4
|
||||
sra %r5,2 // len/4, use sra because it sets condition code
|
||||
jnz .Loop4_sub // borrow is incidentally cleared if branch taken
|
||||
slgr %r2,%r2 // clear borrow
|
||||
|
||||
.Loop1_sub:
|
||||
lg %r0,0(%r2,%r3)
|
||||
slbg %r0,0(%r2,%r4)
|
||||
stg %r0,0(%r2,%r1)
|
||||
|
||||
la %r2,8(%r2) // i++
|
||||
brct %r6,.Loop1_sub
|
||||
j .Lexit_sub
|
||||
|
||||
.Loop4_sub:
|
||||
lg %r0,0(%r2,%r3)
|
||||
slbg %r0,0(%r2,%r4)
|
||||
stg %r0,0(%r2,%r1)
|
||||
lg %r0,8(%r2,%r3)
|
||||
slbg %r0,8(%r2,%r4)
|
||||
stg %r0,8(%r2,%r1)
|
||||
lg %r0,16(%r2,%r3)
|
||||
slbg %r0,16(%r2,%r4)
|
||||
stg %r0,16(%r2,%r1)
|
||||
lg %r0,24(%r2,%r3)
|
||||
slbg %r0,24(%r2,%r4)
|
||||
stg %r0,24(%r2,%r1)
|
||||
|
||||
la %r2,32(%r2) // i+=4
|
||||
brct %r5,.Loop4_sub
|
||||
|
||||
la %r6,1(%r6) // see if len%4 is zero ...
|
||||
brct %r6,.Loop1_sub // without touching condition code:-)
|
||||
|
||||
.Lexit_sub:
|
||||
lghi %r2,0
|
||||
slbgr %r2,%r2
|
||||
lcgr %r2,%r2
|
||||
lg %r6,48(%r15)
|
||||
br %r14
|
||||
.size bn_sub_words,.-bn_sub_words
|
||||
|
||||
#define c1 %r1
|
||||
#define c2 %r5
|
||||
#define c3 %r8
|
||||
|
||||
#define mul_add_c(ai,bi,c1,c2,c3) \
|
||||
lg %r7,ai*8(%r3); \
|
||||
mlg %r6,bi*8(%r4); \
|
||||
algr c1,%r7; \
|
||||
alcgr c2,%r6; \
|
||||
alcgr c3,zero
|
||||
|
||||
// void bn_mul_comba8(BN_ULONG *r2,BN_ULONG *r3,BN_ULONG *r4);
|
||||
.globl bn_mul_comba8
|
||||
.type bn_mul_comba8,@function
|
||||
.align 4
|
||||
bn_mul_comba8:
|
||||
stmg %r6,%r8,48(%r15)
|
||||
|
||||
lghi c1,0
|
||||
lghi c2,0
|
||||
lghi c3,0
|
||||
lghi zero,0
|
||||
|
||||
mul_add_c(0,0,c1,c2,c3);
|
||||
stg c1,0*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
mul_add_c(0,1,c2,c3,c1);
|
||||
mul_add_c(1,0,c2,c3,c1);
|
||||
stg c2,1*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
mul_add_c(2,0,c3,c1,c2);
|
||||
mul_add_c(1,1,c3,c1,c2);
|
||||
mul_add_c(0,2,c3,c1,c2);
|
||||
stg c3,2*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
mul_add_c(0,3,c1,c2,c3);
|
||||
mul_add_c(1,2,c1,c2,c3);
|
||||
mul_add_c(2,1,c1,c2,c3);
|
||||
mul_add_c(3,0,c1,c2,c3);
|
||||
stg c1,3*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
mul_add_c(4,0,c2,c3,c1);
|
||||
mul_add_c(3,1,c2,c3,c1);
|
||||
mul_add_c(2,2,c2,c3,c1);
|
||||
mul_add_c(1,3,c2,c3,c1);
|
||||
mul_add_c(0,4,c2,c3,c1);
|
||||
stg c2,4*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
mul_add_c(0,5,c3,c1,c2);
|
||||
mul_add_c(1,4,c3,c1,c2);
|
||||
mul_add_c(2,3,c3,c1,c2);
|
||||
mul_add_c(3,2,c3,c1,c2);
|
||||
mul_add_c(4,1,c3,c1,c2);
|
||||
mul_add_c(5,0,c3,c1,c2);
|
||||
stg c3,5*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
mul_add_c(6,0,c1,c2,c3);
|
||||
mul_add_c(5,1,c1,c2,c3);
|
||||
mul_add_c(4,2,c1,c2,c3);
|
||||
mul_add_c(3,3,c1,c2,c3);
|
||||
mul_add_c(2,4,c1,c2,c3);
|
||||
mul_add_c(1,5,c1,c2,c3);
|
||||
mul_add_c(0,6,c1,c2,c3);
|
||||
stg c1,6*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
mul_add_c(0,7,c2,c3,c1);
|
||||
mul_add_c(1,6,c2,c3,c1);
|
||||
mul_add_c(2,5,c2,c3,c1);
|
||||
mul_add_c(3,4,c2,c3,c1);
|
||||
mul_add_c(4,3,c2,c3,c1);
|
||||
mul_add_c(5,2,c2,c3,c1);
|
||||
mul_add_c(6,1,c2,c3,c1);
|
||||
mul_add_c(7,0,c2,c3,c1);
|
||||
stg c2,7*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
mul_add_c(7,1,c3,c1,c2);
|
||||
mul_add_c(6,2,c3,c1,c2);
|
||||
mul_add_c(5,3,c3,c1,c2);
|
||||
mul_add_c(4,4,c3,c1,c2);
|
||||
mul_add_c(3,5,c3,c1,c2);
|
||||
mul_add_c(2,6,c3,c1,c2);
|
||||
mul_add_c(1,7,c3,c1,c2);
|
||||
stg c3,8*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
mul_add_c(2,7,c1,c2,c3);
|
||||
mul_add_c(3,6,c1,c2,c3);
|
||||
mul_add_c(4,5,c1,c2,c3);
|
||||
mul_add_c(5,4,c1,c2,c3);
|
||||
mul_add_c(6,3,c1,c2,c3);
|
||||
mul_add_c(7,2,c1,c2,c3);
|
||||
stg c1,9*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
mul_add_c(7,3,c2,c3,c1);
|
||||
mul_add_c(6,4,c2,c3,c1);
|
||||
mul_add_c(5,5,c2,c3,c1);
|
||||
mul_add_c(4,6,c2,c3,c1);
|
||||
mul_add_c(3,7,c2,c3,c1);
|
||||
stg c2,10*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
mul_add_c(4,7,c3,c1,c2);
|
||||
mul_add_c(5,6,c3,c1,c2);
|
||||
mul_add_c(6,5,c3,c1,c2);
|
||||
mul_add_c(7,4,c3,c1,c2);
|
||||
stg c3,11*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
mul_add_c(7,5,c1,c2,c3);
|
||||
mul_add_c(6,6,c1,c2,c3);
|
||||
mul_add_c(5,7,c1,c2,c3);
|
||||
stg c1,12*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
|
||||
mul_add_c(6,7,c2,c3,c1);
|
||||
mul_add_c(7,6,c2,c3,c1);
|
||||
stg c2,13*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
mul_add_c(7,7,c3,c1,c2);
|
||||
stg c3,14*8(%r2)
|
||||
stg c1,15*8(%r2)
|
||||
|
||||
lmg %r6,%r8,48(%r15)
|
||||
br %r14
|
||||
.size bn_mul_comba8,.-bn_mul_comba8
|
||||
|
||||
// void bn_mul_comba4(BN_ULONG *r2,BN_ULONG *r3,BN_ULONG *r4);
|
||||
.globl bn_mul_comba4
|
||||
.type bn_mul_comba4,@function
|
||||
.align 4
|
||||
bn_mul_comba4:
|
||||
stmg %r6,%r8,48(%r15)
|
||||
|
||||
lghi c1,0
|
||||
lghi c2,0
|
||||
lghi c3,0
|
||||
lghi zero,0
|
||||
|
||||
mul_add_c(0,0,c1,c2,c3);
|
||||
stg c1,0*8(%r3)
|
||||
lghi c1,0
|
||||
|
||||
mul_add_c(0,1,c2,c3,c1);
|
||||
mul_add_c(1,0,c2,c3,c1);
|
||||
stg c2,1*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
mul_add_c(2,0,c3,c1,c2);
|
||||
mul_add_c(1,1,c3,c1,c2);
|
||||
mul_add_c(0,2,c3,c1,c2);
|
||||
stg c3,2*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
mul_add_c(0,3,c1,c2,c3);
|
||||
mul_add_c(1,2,c1,c2,c3);
|
||||
mul_add_c(2,1,c1,c2,c3);
|
||||
mul_add_c(3,0,c1,c2,c3);
|
||||
stg c1,3*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
mul_add_c(3,1,c2,c3,c1);
|
||||
mul_add_c(2,2,c2,c3,c1);
|
||||
mul_add_c(1,3,c2,c3,c1);
|
||||
stg c2,4*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
mul_add_c(2,3,c3,c1,c2);
|
||||
mul_add_c(3,2,c3,c1,c2);
|
||||
stg c3,5*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
mul_add_c(3,3,c1,c2,c3);
|
||||
stg c1,6*8(%r2)
|
||||
stg c2,7*8(%r2)
|
||||
|
||||
stmg %r6,%r8,48(%r15)
|
||||
br %r14
|
||||
.size bn_mul_comba4,.-bn_mul_comba4
|
||||
|
||||
#define sqr_add_c(ai,c1,c2,c3) \
|
||||
lg %r7,ai*8(%r3); \
|
||||
mlgr %r6,%r7; \
|
||||
algr c1,%r7; \
|
||||
alcgr c2,%r6; \
|
||||
alcgr c3,zero
|
||||
|
||||
#define sqr_add_c2(ai,aj,c1,c2,c3) \
|
||||
lg %r7,ai*8(%r3); \
|
||||
mlg %r6,aj*8(%r3); \
|
||||
algr c1,%r7; \
|
||||
alcgr c2,%r6; \
|
||||
alcgr c3,zero; \
|
||||
algr c1,%r7; \
|
||||
alcgr c2,%r6; \
|
||||
alcgr c3,zero
|
||||
|
||||
// void bn_sqr_comba8(BN_ULONG *r2,BN_ULONG *r3);
|
||||
.globl bn_sqr_comba8
|
||||
.type bn_sqr_comba8,@function
|
||||
.align 4
|
||||
bn_sqr_comba8:
|
||||
stmg %r6,%r8,48(%r15)
|
||||
|
||||
lghi c1,0
|
||||
lghi c2,0
|
||||
lghi c3,0
|
||||
lghi zero,0
|
||||
|
||||
sqr_add_c(0,c1,c2,c3);
|
||||
stg c1,0*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
sqr_add_c2(1,0,c2,c3,c1);
|
||||
stg c2,1*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
sqr_add_c(1,c3,c1,c2);
|
||||
sqr_add_c2(2,0,c3,c1,c2);
|
||||
stg c3,2*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
sqr_add_c2(3,0,c1,c2,c3);
|
||||
sqr_add_c2(2,1,c1,c2,c3);
|
||||
stg c1,3*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
sqr_add_c(2,c2,c3,c1);
|
||||
sqr_add_c2(3,1,c2,c3,c1);
|
||||
sqr_add_c2(4,0,c2,c3,c1);
|
||||
stg c2,4*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
sqr_add_c2(5,0,c3,c1,c2);
|
||||
sqr_add_c2(4,1,c3,c1,c2);
|
||||
sqr_add_c2(3,2,c3,c1,c2);
|
||||
stg c3,5*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
sqr_add_c(3,c1,c2,c3);
|
||||
sqr_add_c2(4,2,c1,c2,c3);
|
||||
sqr_add_c2(5,1,c1,c2,c3);
|
||||
sqr_add_c2(6,0,c1,c2,c3);
|
||||
stg c1,6*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
sqr_add_c2(7,0,c2,c3,c1);
|
||||
sqr_add_c2(6,1,c2,c3,c1);
|
||||
sqr_add_c2(5,2,c2,c3,c1);
|
||||
sqr_add_c2(4,3,c2,c3,c1);
|
||||
stg c2,7*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
sqr_add_c(4,c3,c1,c2);
|
||||
sqr_add_c2(5,3,c3,c1,c2);
|
||||
sqr_add_c2(6,2,c3,c1,c2);
|
||||
sqr_add_c2(7,1,c3,c1,c2);
|
||||
stg c3,8*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
sqr_add_c2(7,2,c1,c2,c3);
|
||||
sqr_add_c2(6,3,c1,c2,c3);
|
||||
sqr_add_c2(5,4,c1,c2,c3);
|
||||
stg c1,9*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
sqr_add_c(5,c2,c3,c1);
|
||||
sqr_add_c2(6,4,c2,c3,c1);
|
||||
sqr_add_c2(7,3,c2,c3,c1);
|
||||
stg c2,10*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
sqr_add_c2(7,4,c3,c1,c2);
|
||||
sqr_add_c2(6,5,c3,c1,c2);
|
||||
stg c3,11*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
sqr_add_c(6,c1,c2,c3);
|
||||
sqr_add_c2(7,5,c1,c2,c3);
|
||||
stg c1,12*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
sqr_add_c2(7,6,c2,c3,c1);
|
||||
stg c2,13*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
sqr_add_c(7,c3,c1,c2);
|
||||
stg c3,14*8(%r2)
|
||||
stg c1,15*8(%r2)
|
||||
|
||||
lmg %r6,%r8,48(%r15)
|
||||
br %r14
|
||||
.size bn_sqr_comba8,.-bn_sqr_comba8
|
||||
|
||||
// void bn_sqr_comba4(BN_ULONG *r2,BN_ULONG *r3);
|
||||
.globl bn_sqr_comba4
|
||||
.type bn_sqr_comba4,@function
|
||||
.align 4
|
||||
bn_sqr_comba4:
|
||||
stmg %r6,%r8,48(%r15)
|
||||
|
||||
lghi c1,0
|
||||
lghi c2,0
|
||||
lghi c3,0
|
||||
lghi zero,0
|
||||
|
||||
sqr_add_c(0,c1,c2,c3);
|
||||
stg c1,0*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
sqr_add_c2(1,0,c2,c3,c1);
|
||||
stg c2,1*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
sqr_add_c(1,c3,c1,c2);
|
||||
sqr_add_c2(2,0,c3,c1,c2);
|
||||
stg c3,2*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
sqr_add_c2(3,0,c1,c2,c3);
|
||||
sqr_add_c2(2,1,c1,c2,c3);
|
||||
stg c1,3*8(%r2)
|
||||
lghi c1,0
|
||||
|
||||
sqr_add_c(2,c2,c3,c1);
|
||||
sqr_add_c2(3,1,c2,c3,c1);
|
||||
stg c2,4*8(%r2)
|
||||
lghi c2,0
|
||||
|
||||
sqr_add_c2(3,2,c3,c1,c2);
|
||||
stg c3,5*8(%r2)
|
||||
lghi c3,0
|
||||
|
||||
sqr_add_c(3,c1,c2,c3);
|
||||
stg c1,6*8(%r2)
|
||||
stg c2,7*8(%r2)
|
||||
|
||||
lmg %r6,%r8,48(%r15)
|
||||
br %r14
|
||||
.size bn_sqr_comba4,.-bn_sqr_comba4
|
||||
223
crypto/sha/asm/sha1-s390x.pl
Normal file
223
crypto/sha/asm/sha1-s390x.pl
Normal file
@ -0,0 +1,223 @@
|
||||
#!/usr/bin/env perl
|
||||
|
||||
# ====================================================================
|
||||
# Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
|
||||
# project. The module is, however, dual licensed under OpenSSL and
|
||||
# CRYPTOGAMS licenses depending on where you obtain it. For further
|
||||
# details see http://www.openssl.org/~appro/cryptogams/.
|
||||
# ====================================================================
|
||||
|
||||
# SHA1 block procedure for s390x.
|
||||
|
||||
# April 2007.
|
||||
#
|
||||
# Performance is >30% better than gcc 3.3 generated code. But the real
|
||||
# twist is that SHA1 hardware support is detected and utilized. In
|
||||
# which case performance can reach further >4.5x for larger chunks.
|
||||
|
||||
$kimdfunc=0; # magic function code for kimd instruction,
|
||||
# 0 disables hardware support
|
||||
|
||||
$output=shift;
|
||||
open STDOUT,">$output";
|
||||
|
||||
$t0="%r0";
|
||||
$t1="%r1";
|
||||
$ctx="%r2";
|
||||
$inp="%r3";
|
||||
$len="%r4";
|
||||
|
||||
$A="%r5";
|
||||
$B="%r6";
|
||||
$C="%r7";
|
||||
$D="%r8";
|
||||
$E="%r9"; @V=($A,$B,$C,$D,$E);
|
||||
$K_00_19="%r10";
|
||||
$K_20_39="%r11";
|
||||
$K_40_59="%r12";
|
||||
$K_60_79="%r13";
|
||||
$Xi="%r14";
|
||||
$sp="%r15";
|
||||
|
||||
$frame=160+16*4;
|
||||
|
||||
sub BODY_00_15 {
|
||||
my ($i,$a,$b,$c,$d,$e)=@_;
|
||||
my $xi=($i&1)?$Xi:$t1;
|
||||
|
||||
$code.=<<___ if ($i<16 && !($i&1));
|
||||
lg $Xi,`$i*4`($inp)
|
||||
___
|
||||
$code.=<<___;
|
||||
alr $e,$K_00_19 ### $i
|
||||
rll $t0,$a,5
|
||||
alr $e,$t0
|
||||
lr $t0,$d
|
||||
xr $t0,$c
|
||||
nr $t0,$b
|
||||
xr $t0,$d
|
||||
alr $e,$t0
|
||||
rll $b,$b,30
|
||||
___
|
||||
$code.=<<___ if ($i<16 && !($i&1));
|
||||
srlg $xi,$Xi,32
|
||||
stg $Xi,`160+$i*4`($sp)
|
||||
___
|
||||
$code.=<<___;
|
||||
alr $e,$xi
|
||||
___
|
||||
}
|
||||
|
||||
sub Xupdate {
|
||||
my $i=shift;
|
||||
|
||||
return if ($i&1); # Xupdate is vectorized and executed every 2nd cycle
|
||||
$code.=<<___;
|
||||
lg $Xi,`160+4*($i%16)`($sp) ### Xupdate($i)
|
||||
xg $Xi,`160+4*(($i+2)%16)`($sp)
|
||||
xg $Xi,`160+4*(($i+8)%16)`($sp)
|
||||
___
|
||||
if ((($i+13)%16)==15) {
|
||||
$code.=<<___;
|
||||
llgf $t0,`160+4*15`($sp)
|
||||
x $Xi,`160+0`($sp)
|
||||
sllg $t0,$t0,32
|
||||
xgr $Xi,$t0
|
||||
___
|
||||
} else {
|
||||
$code.=<<___;
|
||||
xg $Xi,`160+4*(($i+13)%16)`($sp)
|
||||
___
|
||||
}
|
||||
$code.=<<___;
|
||||
rll $Xi,$Xi,1
|
||||
rllg $t1,$Xi,32
|
||||
rll $t1,$t1,1
|
||||
rllg $Xi,$t1,32
|
||||
stg $Xi,`160+4*($i%16)`($sp)
|
||||
___
|
||||
}
|
||||
|
||||
sub BODY_16_19 {
|
||||
&Xupdate(@_[0]);
|
||||
&BODY_00_15(@_);
|
||||
}
|
||||
|
||||
sub BODY_20_39 {
|
||||
my ($i,$a,$b,$c,$d,$e)=@_;
|
||||
my $xi=($i&1)?$Xi:$t1;
|
||||
my $K_XX_XX=($i<40)?$K_20_39:$K_60_79;
|
||||
|
||||
&Xupdate($i);
|
||||
$code.=<<___;
|
||||
alr $e,$K_XX_XX ### $i
|
||||
rll $t0,$a,5
|
||||
alr $e,$t0
|
||||
lr $t0,$b
|
||||
xr $t0,$c
|
||||
xr $t0,$d
|
||||
alr $e,$t0
|
||||
rll $b,$b,30
|
||||
alr $e,$xi
|
||||
___
|
||||
}
|
||||
|
||||
sub BODY_40_59 {
|
||||
my ($i,$a,$b,$c,$d,$e)=@_;
|
||||
my $xi=($i&1)?$Xi:$t1;
|
||||
|
||||
&Xupdate($i);
|
||||
$code.=<<___;
|
||||
alr $e,$K_40_59 ### $i
|
||||
rll $t0,$a,5
|
||||
alr $e,$t0
|
||||
lr $t0,$b
|
||||
or $t0,$c
|
||||
nr $t0,$d
|
||||
alr $e,$xi
|
||||
lr $t1,$b
|
||||
nr $t1,$c
|
||||
or $t0,$t1
|
||||
alr $e,$t0
|
||||
rll $b,$b,30
|
||||
___
|
||||
}
|
||||
|
||||
$code.=<<___;
|
||||
.text
|
||||
.globl sha1_block_data_order
|
||||
.type sha1_block_data_order,\@function
|
||||
sha1_block_data_order:
|
||||
___
|
||||
$code.=<<___ if ($kimdfunc);
|
||||
lghi %r0,0
|
||||
la %r1,16($sp)
|
||||
.long 0xb93e0002 # kimd %r0,%r2
|
||||
lg %r0,16($sp)
|
||||
tmhh %r0,`0x8000>>$kimdfunc`
|
||||
jz .Lsoftware
|
||||
lghi %r0,$kimdfunc
|
||||
lgr %r1,$ctx
|
||||
lgr %r2,$inp
|
||||
sllg %r3,$len,6
|
||||
.long 0xb93e0002 # kimd %r0,%r2
|
||||
brc 1,.-4 # pay attention to "partial completion"
|
||||
br %r14
|
||||
.Lsoftware:
|
||||
___
|
||||
$code.=<<___;
|
||||
stmg %r6,%r15,48($sp)
|
||||
lgr %r0,$sp
|
||||
aghi $sp,-$frame
|
||||
stg %r0,0($sp)
|
||||
|
||||
sllg $len,$len,6
|
||||
la $len,0($inp,$len)
|
||||
|
||||
llgf $A,0($ctx)
|
||||
llgf $B,4($ctx)
|
||||
llgf $C,8($ctx)
|
||||
llgf $D,12($ctx)
|
||||
llgf $E,16($ctx)
|
||||
|
||||
llilh $K_00_19,0x5a82
|
||||
oill $K_00_19,0x7999
|
||||
llilh $K_20_39,0x6ed9
|
||||
oill $K_20_39,0xeba1
|
||||
llilh $K_40_59,0x8f1b
|
||||
oill $K_40_59,0xbcdc
|
||||
llilh $K_60_79,0xca62
|
||||
oill $K_60_79,0xc1d6
|
||||
.Lloop:
|
||||
___
|
||||
for ($i=0;$i<16;$i++) { &BODY_00_15($i,@V); unshift(@V,pop(@V)); }
|
||||
for (;$i<20;$i++) { &BODY_16_19($i,@V); unshift(@V,pop(@V)); }
|
||||
for (;$i<40;$i++) { &BODY_20_39($i,@V); unshift(@V,pop(@V)); }
|
||||
for (;$i<60;$i++) { &BODY_40_59($i,@V); unshift(@V,pop(@V)); }
|
||||
for (;$i<80;$i++) { &BODY_20_39($i,@V); unshift(@V,pop(@V)); }
|
||||
$code.=<<___;
|
||||
|
||||
al $A,0($ctx)
|
||||
al $B,4($ctx)
|
||||
al $C,8($ctx)
|
||||
al $D,12($ctx)
|
||||
al $E,16($ctx)
|
||||
st $A,0($ctx)
|
||||
st $B,4($ctx)
|
||||
st $C,8($ctx)
|
||||
st $D,12($ctx)
|
||||
st $E,16($ctx)
|
||||
la $inp,64($inp)
|
||||
clgr $inp,$len
|
||||
jne .Lloop
|
||||
|
||||
lmg %r6,%r15,`$frame+48`($sp)
|
||||
br %r14
|
||||
.size sha1_block_data_order,.-sha1_block_data_order
|
||||
.string "SHA1 block transform for s390x, CRYPTOGAMS by <appro\@openssl.org>"
|
||||
___
|
||||
|
||||
$code =~ s/\`([^\`]*)\`/eval $1/gem;
|
||||
|
||||
print $code;
|
||||
close STDOUT;
|
||||
290
crypto/sha/asm/sha512-s390x.pl
Normal file
290
crypto/sha/asm/sha512-s390x.pl
Normal file
@ -0,0 +1,290 @@
|
||||
#!/usr/bin/env perl
|
||||
|
||||
# ====================================================================
|
||||
# Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
|
||||
# project. The module is, however, dual licensed under OpenSSL and
|
||||
# CRYPTOGAMS licenses depending on where you obtain it. For further
|
||||
# details see http://www.openssl.org/~appro/cryptogams/.
|
||||
# ====================================================================
|
||||
|
||||
# SHA256/512 block procedures for s390x.
|
||||
|
||||
# April 2007.
|
||||
#
|
||||
# sha256_block_data_order is reportedly >3 times faster than gcc 3.3
|
||||
# generated code (must be a bug in compiler, as improvement is
|
||||
# "pathologically" high, in particular in comparison to other SHA
|
||||
# modules). But the real twist is that it detects if hardware support
|
||||
# for SHA256 is available and in such case utilizes it. Then the
|
||||
# performance can reach >6.5x of assembler one for larger chunks.
|
||||
#
|
||||
# sha512_block_data_order is ~70% faster than gcc 3.3 generated code.
|
||||
|
||||
$t0="%r0";
|
||||
$t1="%r1";
|
||||
$ctx="%r2";
|
||||
$inp="%r3";
|
||||
$len="%r4"; # used as index in inner loop
|
||||
|
||||
$A="%r5";
|
||||
$B="%r6";
|
||||
$C="%r7";
|
||||
$D="%r8";
|
||||
$E="%r9";
|
||||
$F="%r10";
|
||||
$G="%r11";
|
||||
$H="%r12"; @V=($A,$B,$C,$D,$E,$F,$G,$H);
|
||||
$tbl="%r13";
|
||||
$T1="%r14";
|
||||
$sp="%r15";
|
||||
|
||||
$output=shift;
|
||||
open STDOUT,">$output";
|
||||
|
||||
if ($output =~ /512/) {
|
||||
$label="512";
|
||||
$SZ=8;
|
||||
$LD="lg"; # load from memory
|
||||
$ST="stg"; # store to memory
|
||||
$ADD="alg"; # add with memory operand
|
||||
$ROT="rllg"; # rotate left
|
||||
$SHR="srlg"; # logical right shift [see even at the end]
|
||||
@Sigma0=(25,30,36);
|
||||
@Sigma1=(23,46,50);
|
||||
@sigma0=(56,63, 7);
|
||||
@sigma1=( 3,45, 6);
|
||||
$rounds=80;
|
||||
$kimdfunc=0; # 0 means unknown/unsupported/unimplemented
|
||||
} else {
|
||||
$label="256";
|
||||
$SZ=4;
|
||||
$LD="llgf"; # load from memory
|
||||
$ST="st"; # store to memory
|
||||
$ADD="al"; # add with memory operand
|
||||
$ROT="rll"; # rotate left
|
||||
$SHR="srl"; # logical right shift
|
||||
@Sigma0=(10,19,30);
|
||||
@Sigma1=( 7,21,26);
|
||||
@sigma0=(14,25, 3);
|
||||
@sigma1=(13,15,10);
|
||||
$rounds=64;
|
||||
$kimdfunc=0; # magic function code for kimd instruction,
|
||||
# 0 disables hardware support
|
||||
}
|
||||
$Func="sha${label}_block_data_order";
|
||||
$Table="K${label}";
|
||||
$frame=160+16*$SZ;
|
||||
|
||||
sub BODY_00_15 {
|
||||
my ($i,$a,$b,$c,$d,$e,$f,$g,$h) = @_;
|
||||
|
||||
$code.=<<___ if ($i<16);
|
||||
$LD $T1,`$i*$SZ`($inp) ### $i
|
||||
___
|
||||
$code.=<<___;
|
||||
$ROT $t0,$e,$Sigma1[0]
|
||||
$ROT $t1,$e,$Sigma1[1]
|
||||
xgr $t0,$t1
|
||||
$ROT $t1,$t1,`$Sigma1[2]-$Sigma1[1]`
|
||||
xgr $t0,$t1 # Sigma1(e)
|
||||
$ST $T1,`160+$SZ*($i%16)`($sp)
|
||||
algr $T1,$t0 # T1+=Sigma1(e)
|
||||
algr $T1,$h # T1+=h
|
||||
$ADD $T1,`$i*$SZ`($len,$tbl) # T1+=K[i]
|
||||
lgr $t0,$f
|
||||
xgr $t0,$g
|
||||
ngr $t0,$e
|
||||
xgr $t0,$g # Ch(e,f,g)
|
||||
algr $T1,$t0 # T1+=Ch(e,f,g)
|
||||
$ROT $h,$a,$Sigma0[0]
|
||||
$ROT $t0,$a,$Sigma0[1]
|
||||
xgr $h,$t0
|
||||
$ROT $t0,$t0,`$Sigma0[2]-$Sigma0[1]`
|
||||
xgr $h,$t0 # h=Sigma0(a)
|
||||
lgr $t0,$a
|
||||
ogr $t0,$b
|
||||
ngr $t0,$c
|
||||
lgr $t1,$a
|
||||
ngr $t1,$b
|
||||
ogr $t0,$t1 # Maj(a,b,c)
|
||||
algr $h,$t0 # h+=Maj(a,b,c)
|
||||
algr $d,$T1 # d+=T1
|
||||
algr $h,$T1 # h+=T1
|
||||
___
|
||||
}
|
||||
|
||||
sub BODY_16_XX {
|
||||
my ($i,$a,$b,$c,$d,$e,$f,$g,$h) = @_;
|
||||
|
||||
$code.=<<___;
|
||||
$LD $T1,`160+$SZ*(($i+1)%16)`($sp) ### $i
|
||||
$LD $t1,`160+$SZ*(($i+14)%16)`($sp)
|
||||
$ROT $t0,$T1,$sigma0[0]
|
||||
$SHR $T1,$sigma0[2]
|
||||
xgr $T1,$t0
|
||||
$ROT $t0,$t0,`$sigma0[1]-$sigma0[0]`
|
||||
xgr $T1,$t0 # sigma0(X[i+1])
|
||||
$ROT $t0,$t1,$sigma1[0]
|
||||
$ADD $T1,`160+$SZ*($i%16)`($sp) # +=X[i]
|
||||
$SHR $t1,$sigma1[2]
|
||||
xgr $t1,$t0
|
||||
$ADD $T1,`160+$SZ*(($i+9)%16)`($sp) # +=X[i+9]
|
||||
$ROT $t0,$t0,`$sigma1[1]-$sigma1[0]`
|
||||
xgr $t1,$t0 # sigma1(X[i+14])
|
||||
algr $T1,$t1 # +=sigma1(X[i+14])
|
||||
___
|
||||
&BODY_00_15(@_);
|
||||
}
|
||||
|
||||
$code.=<<___;
|
||||
.text
|
||||
.align 64
|
||||
.type $Table,\@object
|
||||
$Table:
|
||||
___
|
||||
$code.=<<___ if ($SZ==4);
|
||||
.long 0x428a2f98,0x71374491,0xb5c0fbcf,0xe9b5dba5
|
||||
.long 0x3956c25b,0x59f111f1,0x923f82a4,0xab1c5ed5
|
||||
.long 0xd807aa98,0x12835b01,0x243185be,0x550c7dc3
|
||||
.long 0x72be5d74,0x80deb1fe,0x9bdc06a7,0xc19bf174
|
||||
.long 0xe49b69c1,0xefbe4786,0x0fc19dc6,0x240ca1cc
|
||||
.long 0x2de92c6f,0x4a7484aa,0x5cb0a9dc,0x76f988da
|
||||
.long 0x983e5152,0xa831c66d,0xb00327c8,0xbf597fc7
|
||||
.long 0xc6e00bf3,0xd5a79147,0x06ca6351,0x14292967
|
||||
.long 0x27b70a85,0x2e1b2138,0x4d2c6dfc,0x53380d13
|
||||
.long 0x650a7354,0x766a0abb,0x81c2c92e,0x92722c85
|
||||
.long 0xa2bfe8a1,0xa81a664b,0xc24b8b70,0xc76c51a3
|
||||
.long 0xd192e819,0xd6990624,0xf40e3585,0x106aa070
|
||||
.long 0x19a4c116,0x1e376c08,0x2748774c,0x34b0bcb5
|
||||
.long 0x391c0cb3,0x4ed8aa4a,0x5b9cca4f,0x682e6ff3
|
||||
.long 0x748f82ee,0x78a5636f,0x84c87814,0x8cc70208
|
||||
.long 0x90befffa,0xa4506ceb,0xbef9a3f7,0xc67178f2
|
||||
___
|
||||
$code.=<<___ if ($SZ==8);
|
||||
.quad 0x428a2f98d728ae22,0x7137449123ef65cd
|
||||
.quad 0xb5c0fbcfec4d3b2f,0xe9b5dba58189dbbc
|
||||
.quad 0x3956c25bf348b538,0x59f111f1b605d019
|
||||
.quad 0x923f82a4af194f9b,0xab1c5ed5da6d8118
|
||||
.quad 0xd807aa98a3030242,0x12835b0145706fbe
|
||||
.quad 0x243185be4ee4b28c,0x550c7dc3d5ffb4e2
|
||||
.quad 0x72be5d74f27b896f,0x80deb1fe3b1696b1
|
||||
.quad 0x9bdc06a725c71235,0xc19bf174cf692694
|
||||
.quad 0xe49b69c19ef14ad2,0xefbe4786384f25e3
|
||||
.quad 0x0fc19dc68b8cd5b5,0x240ca1cc77ac9c65
|
||||
.quad 0x2de92c6f592b0275,0x4a7484aa6ea6e483
|
||||
.quad 0x5cb0a9dcbd41fbd4,0x76f988da831153b5
|
||||
.quad 0x983e5152ee66dfab,0xa831c66d2db43210
|
||||
.quad 0xb00327c898fb213f,0xbf597fc7beef0ee4
|
||||
.quad 0xc6e00bf33da88fc2,0xd5a79147930aa725
|
||||
.quad 0x06ca6351e003826f,0x142929670a0e6e70
|
||||
.quad 0x27b70a8546d22ffc,0x2e1b21385c26c926
|
||||
.quad 0x4d2c6dfc5ac42aed,0x53380d139d95b3df
|
||||
.quad 0x650a73548baf63de,0x766a0abb3c77b2a8
|
||||
.quad 0x81c2c92e47edaee6,0x92722c851482353b
|
||||
.quad 0xa2bfe8a14cf10364,0xa81a664bbc423001
|
||||
.quad 0xc24b8b70d0f89791,0xc76c51a30654be30
|
||||
.quad 0xd192e819d6ef5218,0xd69906245565a910
|
||||
.quad 0xf40e35855771202a,0x106aa07032bbd1b8
|
||||
.quad 0x19a4c116b8d2d0c8,0x1e376c085141ab53
|
||||
.quad 0x2748774cdf8eeb99,0x34b0bcb5e19b48a8
|
||||
.quad 0x391c0cb3c5c95a63,0x4ed8aa4ae3418acb
|
||||
.quad 0x5b9cca4f7763e373,0x682e6ff3d6b2b8a3
|
||||
.quad 0x748f82ee5defb2fc,0x78a5636f43172f60
|
||||
.quad 0x84c87814a1f0ab72,0x8cc702081a6439ec
|
||||
.quad 0x90befffa23631e28,0xa4506cebde82bde9
|
||||
.quad 0xbef9a3f7b2c67915,0xc67178f2e372532b
|
||||
.quad 0xca273eceea26619c,0xd186b8c721c0c207
|
||||
.quad 0xeada7dd6cde0eb1e,0xf57d4f7fee6ed178
|
||||
.quad 0x06f067aa72176fba,0x0a637dc5a2c898a6
|
||||
.quad 0x113f9804bef90dae,0x1b710b35131c471b
|
||||
.quad 0x28db77f523047d84,0x32caab7b40c72493
|
||||
.quad 0x3c9ebe0a15c9bebc,0x431d67c49c100d4c
|
||||
.quad 0x4cc5d4becb3e42b6,0x597f299cfc657e2a
|
||||
.quad 0x5fcb6fab3ad6faec,0x6c44198c4a475817
|
||||
___
|
||||
$code.=<<___;
|
||||
.size $Table,.-$Table
|
||||
.globl $Func
|
||||
.type $Func,\@function
|
||||
$Func:
|
||||
___
|
||||
$code.=<<___ if ($kimdfunc);
|
||||
lghi %r0,0
|
||||
la %r1,16($sp)
|
||||
.long 0xb93e0002 # kimd %r0,%r2
|
||||
lg %r0,16($sp)
|
||||
tmhh %r0,`0x8000>>$kimdfunc`
|
||||
jz .Lsoftware
|
||||
lghi %r0,$kimdfunc
|
||||
lgr %r1,$ctx
|
||||
lgr %r2,$inp
|
||||
sllg %r3,$len,`log(16*$SZ)/log(2)`
|
||||
.long 0xb93e0002 # kimd %r0,%r2
|
||||
brc 1,.-4 # pay attention to "partial completion"
|
||||
br %r14
|
||||
.Lsoftware:
|
||||
___
|
||||
$code.=<<___;
|
||||
sllg $len,$len,`log(16*$SZ)/log(2)`
|
||||
la $len,0($inp,$len)
|
||||
stmg $len,%r15,32($sp)
|
||||
lgr %r0,$sp
|
||||
aghi $sp,-$frame
|
||||
stg %r0,0($sp)
|
||||
|
||||
bras $tbl,.Lpic
|
||||
.Lpic: aghi $tbl,$Table-.Lpic
|
||||
|
||||
$LD $A,`0*$SZ`($ctx)
|
||||
$LD $B,`1*$SZ`($ctx)
|
||||
$LD $C,`2*$SZ`($ctx)
|
||||
$LD $D,`3*$SZ`($ctx)
|
||||
$LD $E,`4*$SZ`($ctx)
|
||||
$LD $F,`5*$SZ`($ctx)
|
||||
$LD $G,`6*$SZ`($ctx)
|
||||
$LD $H,`7*$SZ`($ctx)
|
||||
|
||||
.Lloop:
|
||||
lghi $len,0
|
||||
___
|
||||
for ($i=0;$i<16;$i++) { &BODY_00_15($i,@V); unshift(@V,pop(@V)); }
|
||||
$code.=".Lrounds_16_xx:\n";
|
||||
for (;$i<32;$i++) { &BODY_16_XX($i,@V); unshift(@V,pop(@V)); }
|
||||
$code.=<<___;
|
||||
aghi $len,`16*$SZ`
|
||||
lghi $t0,`($rounds-16)*$SZ`
|
||||
clgr $len,$t0
|
||||
jne .Lrounds_16_xx
|
||||
|
||||
$ADD $A,`0*$SZ`($ctx)
|
||||
$ADD $B,`1*$SZ`($ctx)
|
||||
$ADD $C,`2*$SZ`($ctx)
|
||||
$ADD $D,`3*$SZ`($ctx)
|
||||
$ADD $E,`4*$SZ`($ctx)
|
||||
$ADD $F,`5*$SZ`($ctx)
|
||||
$ADD $G,`6*$SZ`($ctx)
|
||||
$ADD $H,`7*$SZ`($ctx)
|
||||
$ST $A,`0*$SZ`($ctx)
|
||||
$ST $B,`1*$SZ`($ctx)
|
||||
$ST $C,`2*$SZ`($ctx)
|
||||
$ST $D,`3*$SZ`($ctx)
|
||||
$ST $E,`4*$SZ`($ctx)
|
||||
$ST $F,`5*$SZ`($ctx)
|
||||
$ST $G,`6*$SZ`($ctx)
|
||||
$ST $H,`7*$SZ`($ctx)
|
||||
la $inp,`16*$SZ`($inp)
|
||||
clg $inp,`$frame+32`($sp)
|
||||
jne .Lloop
|
||||
|
||||
lmg %r6,%r15,`$frame+48`($sp)
|
||||
br %r14
|
||||
.size $Func,.-$Func
|
||||
.string "SHA${label} block transform for s390x, CRYPTOGAMS by <appro\@openssl.org>"
|
||||
___
|
||||
|
||||
$code =~ s/\`([^\`]*)\`/eval $1/gem;
|
||||
# unlike 32-bit shift 64-bit one takes three arguments
|
||||
$code =~ s/(srlg\s+)(%r[0-9]+),/$1$2,$2,/gm;
|
||||
|
||||
print $code;
|
||||
close STDOUT;
|
||||
Loading…
x
Reference in New Issue
Block a user