2000-10-15 03:22:34 +00:00
|
|
|
;! HP-PA-1.1 __mpn_submul_1 -- Multiply a limb vector with a limb and
|
|
|
|
;! subtract the result from a second limb vector.
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2022-01-01 18:54:23 +00:00
|
|
|
;! Copyright (C) 1992-2022 Free Software Foundation, Inc.
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
;! This file is part of the GNU MP Library.
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
;! The GNU MP Library is free software; you can redistribute it and/or modify
|
2001-07-07 10:13:33 +00:00
|
|
|
;! it under the terms of the GNU Lesser General Public License as published by
|
|
|
|
;! the Free Software Foundation; either version 2.1 of the License, or (at your
|
2000-10-15 03:22:34 +00:00
|
|
|
;! option) any later version.
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
;! The GNU MP Library is distributed in the hope that it will be useful, but
|
|
|
|
;! WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
|
2001-07-07 10:13:33 +00:00
|
|
|
;! or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public
|
2000-10-15 03:22:34 +00:00
|
|
|
;! License for more details.
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2001-07-07 10:13:33 +00:00
|
|
|
;! You should have received a copy of the GNU Lesser General Public License
|
2012-03-09 23:56:38 +00:00
|
|
|
;! along with the GNU MP Library. If not, see
|
Prefer https to http for gnu.org and fsf.org URLs
Also, change sources.redhat.com to sourceware.org.
This patch was automatically generated by running the following shell
script, which uses GNU sed, and which avoids modifying files imported
from upstream:
sed -ri '
s,(http|ftp)(://(.*\.)?(gnu|fsf|sourceware)\.org($|[^.]|\.[^a-z])),https\2,g
s,(http|ftp)(://(.*\.)?)sources\.redhat\.com($|[^.]|\.[^a-z]),https\2sourceware.org\4,g
' \
$(find $(git ls-files) -prune -type f \
! -name '*.po' \
! -name 'ChangeLog*' \
! -path COPYING ! -path COPYING.LIB \
! -path manual/fdl-1.3.texi ! -path manual/lgpl-2.1.texi \
! -path manual/texinfo.tex ! -path scripts/config.guess \
! -path scripts/config.sub ! -path scripts/install-sh \
! -path scripts/mkinstalldirs ! -path scripts/move-if-change \
! -path INSTALL ! -path locale/programs/charmap-kw.h \
! -path po/libc.pot ! -path sysdeps/gnu/errlist.c \
! '(' -name configure \
-execdir test -f configure.ac -o -f configure.in ';' ')' \
! '(' -name preconfigure \
-execdir test -f preconfigure.ac ';' ')' \
-print)
and then by running 'make dist-prepare' to regenerate files built
from the altered files, and then executing the following to cleanup:
chmod a+x sysdeps/unix/sysv/linux/riscv/configure
# Omit irrelevant whitespace and comment-only changes,
# perhaps from a slightly-different Autoconf version.
git checkout -f \
sysdeps/csky/configure \
sysdeps/hppa/configure \
sysdeps/riscv/configure \
sysdeps/unix/sysv/linux/csky/configure
# Omit changes that caused a pre-commit check to fail like this:
# remote: *** error: sysdeps/powerpc/powerpc64/ppc-mcount.S: trailing lines
git checkout -f \
sysdeps/powerpc/powerpc64/ppc-mcount.S \
sysdeps/unix/sysv/linux/s390/s390-64/syscall.S
# Omit change that caused a pre-commit check to fail like this:
# remote: *** error: sysdeps/sparc/sparc64/multiarch/memcpy-ultra3.S: last line does not end in newline
git checkout -f sysdeps/sparc/sparc64/multiarch/memcpy-ultra3.S
2019-09-07 05:40:42 +00:00
|
|
|
;! <https://www.gnu.org/licenses/>.
|
1995-10-16 01:18:40 +00:00
|
|
|
|
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
;! INPUT PARAMETERS
|
|
|
|
;! res_ptr r26
|
|
|
|
;! s1_ptr r25
|
|
|
|
;! size r24
|
|
|
|
;! s2_limb r23
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
;! This runs at 12 cycles/limb on a PA7000. With the used instructions, it
|
|
|
|
;! can not become faster due to data cache contention after a store. On the
|
|
|
|
;! PA7100 it runs at 11 cycles/limb, and that can not be improved either,
|
|
|
|
;! since only the xmpyu does not need the integer pipeline, so the only
|
|
|
|
;! dual-issue we will get are addc+xmpyu. Unrolling could gain a cycle/limb
|
|
|
|
;! on the PA7100.
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
;! There are some ideas described in mul_1.s that applies to this code too.
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
;! It seems possible to make this run as fast as __mpn_addmul_1, if we use
|
|
|
|
;! sub,>>= %r29,%r20,%r22
|
|
|
|
;! addi 1,%r28,%r28
|
|
|
|
;! but that requires reworking the hairy software pipeline...
|
1995-10-16 01:18:40 +00:00
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
.text
|
1995-10-16 01:18:40 +00:00
|
|
|
.export __mpn_submul_1
|
2013-06-05 20:26:40 +00:00
|
|
|
__mpn_submul_1:
|
1995-10-16 01:18:40 +00:00
|
|
|
.proc
|
|
|
|
.callinfo frame=64,no_calls
|
|
|
|
.entry
|
|
|
|
|
|
|
|
ldo 64(%r30),%r30
|
|
|
|
fldws,ma 4(%r25),%fr5
|
2000-10-15 03:22:34 +00:00
|
|
|
stw %r23,-16(%r30) ;! move s2_limb ...
|
1995-10-16 01:18:40 +00:00
|
|
|
addib,= -1,%r24,L$just_one_limb
|
2000-10-15 03:22:34 +00:00
|
|
|
fldws -16(%r30),%fr4 ;! ... into fr4
|
|
|
|
add %r0,%r0,%r0 ;! clear carry
|
1995-10-16 01:18:40 +00:00
|
|
|
xmpyu %fr4,%fr5,%fr6
|
|
|
|
fldws,ma 4(%r25),%fr7
|
|
|
|
fstds %fr6,-16(%r30)
|
|
|
|
xmpyu %fr4,%fr7,%fr8
|
2000-10-15 03:22:34 +00:00
|
|
|
ldw -12(%r30),%r20 ;! least significant limb in product
|
1995-10-16 01:18:40 +00:00
|
|
|
ldw -16(%r30),%r28
|
|
|
|
|
|
|
|
fstds %fr8,-16(%r30)
|
|
|
|
addib,= -1,%r24,L$end
|
|
|
|
ldw -12(%r30),%r1
|
|
|
|
|
2000-10-15 03:22:34 +00:00
|
|
|
;! Main loop
|
2013-06-05 20:26:40 +00:00
|
|
|
L$loop:
|
2000-10-15 03:22:34 +00:00
|
|
|
ldws 0(%r26),%r29
|
1995-10-16 01:18:40 +00:00
|
|
|
fldws,ma 4(%r25),%fr5
|
2000-10-15 03:22:34 +00:00
|
|
|
sub %r29,%r20,%r22
|
|
|
|
add %r22,%r20,%r0
|
1995-10-16 01:18:40 +00:00
|
|
|
stws,ma %r22,4(%r26)
|
2000-10-15 03:22:34 +00:00
|
|
|
addc %r28,%r1,%r20
|
1995-10-16 01:18:40 +00:00
|
|
|
xmpyu %fr4,%fr5,%fr6
|
|
|
|
ldw -16(%r30),%r28
|
|
|
|
fstds %fr6,-16(%r30)
|
|
|
|
addc %r0,%r28,%r28
|
|
|
|
addib,<> -1,%r24,L$loop
|
|
|
|
ldw -12(%r30),%r1
|
|
|
|
|
2013-06-05 20:26:40 +00:00
|
|
|
L$end:
|
2000-10-15 03:22:34 +00:00
|
|
|
ldw 0(%r26),%r29
|
|
|
|
sub %r29,%r20,%r22
|
|
|
|
add %r22,%r20,%r0
|
1995-10-16 01:18:40 +00:00
|
|
|
stws,ma %r22,4(%r26)
|
2000-10-15 03:22:34 +00:00
|
|
|
addc %r28,%r1,%r20
|
1995-10-16 01:18:40 +00:00
|
|
|
ldw -16(%r30),%r28
|
|
|
|
ldws 0(%r26),%r29
|
|
|
|
addc %r0,%r28,%r28
|
2000-10-15 03:22:34 +00:00
|
|
|
sub %r29,%r20,%r22
|
|
|
|
add %r22,%r20,%r0
|
1995-10-16 01:18:40 +00:00
|
|
|
stws,ma %r22,4(%r26)
|
|
|
|
addc %r0,%r28,%r28
|
|
|
|
bv 0(%r2)
|
|
|
|
ldo -64(%r30),%r30
|
|
|
|
|
2013-06-05 20:26:40 +00:00
|
|
|
L$just_one_limb:
|
1995-10-16 01:18:40 +00:00
|
|
|
xmpyu %fr4,%fr5,%fr6
|
|
|
|
ldw 0(%r26),%r29
|
|
|
|
fstds %fr6,-16(%r30)
|
|
|
|
ldw -12(%r30),%r1
|
|
|
|
ldw -16(%r30),%r28
|
|
|
|
sub %r29,%r1,%r22
|
|
|
|
add %r22,%r1,%r0
|
|
|
|
stw %r22,0(%r26)
|
|
|
|
addc %r0,%r28,%r28
|
|
|
|
bv 0(%r2)
|
|
|
|
ldo -64(%r30),%r30
|
|
|
|
|
|
|
|
.exit
|
|
|
|
.procend
|