mpir/mpn/alpha/ev6/addmul_1.asm

391 lines
9.9 KiB
NASM

dnl Alpha ev6 mpn_addmul_1 -- Multiply a limb vector with a limb and add the
dnl result to a second limb vector.
dnl Copyright 2000, 2003, 2004, 2005 Free Software Foundation, Inc.
dnl This file is part of the GNU MP Library.
dnl The GNU MP Library is free software; you can redistribute it and/or modify
dnl it under the terms of the GNU Lesser General Public License as published
dnl by the Free Software Foundation; either version 2.1 of the License, or (at
dnl your option) any later version.
dnl The GNU MP Library is distributed in the hope that it will be useful, but
dnl WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public
dnl License for more details.
dnl You should have received a copy of the GNU Lesser General Public License
dnl along with the GNU MP Library; see the file COPYING.LIB. If not, write
dnl to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
dnl Boston, MA 02110-1301, USA.
include(`../config.m4')
C cycles/limb
C EV4: 42
C EV5: 18
C EV6: 3.5
C INPUT PARAMETERS
C rp r16
C up r17
C n r18
C vlimb r19
dnl This code was written in cooperation with ev6 pipeline expert Steve Root.
dnl The stores can issue a cycle late so we have paired no-op's to 'catch'
dnl them, so that further disturbance to the schedule is damped.
dnl We couldn't pair the loads, because the entangled schedule of the carry's
dnl has to happen on one side {0} of the machine.
dnl This is a great schedule for the d_cache, a poor schedule for the b_cache.
dnl The lockup on U0 means that any stall can't be recovered from. Consider a
dnl ldq in L1, say that load gets stalled because it collides with a fill from
dnl the b_cache. On the next cycle, this load gets priority. If first looks
dnl at L0, and goes there. The instruction we intended for L0 gets to look at
dnl L1, which is NOT where we want it. It either stalls 1, because it can't
dnl go in L0, or goes there, and causes a further instruction to stall.
dnl So for b_cache, we're likely going to want to put one or more cycles back
dnl into the code! And, of course, put in lds prefetch for the rp[] operand.
dnl At a place where we have an mt followed by a bookkeeping, put the
dnl bookkeeping in upper, and the prefetch into lower.
dnl Note, the ldq's and stq's are at the end of the quadpacks. Note, we'd
dnl like not to have an ldq or an stq to preceded a conditional branch in a
dnl quadpack. The conditional branch moves the retire pointer one cycle
dnl later.
ASM_START()
PROLOGUE(mpn_addmul_1)
ldq r3, 0(r17) C
and r18, 7, r20 C
lda r18, -9(r18) C
cmpeq r20, 1, r21 C
beq r21, $L1 C
$1mod8: ldq r5, 0(r16) C
mulq r19, r3, r7 C
umulh r19, r3, r8 C
addq r5, r7, r23 C
cmpult r23, r7, r20 C
addq r8, r20, r0 C
stq r23, 0(r16) C
bge r18, $ent1 C
ret r31, (r26), 1 C
$L1: lda r8, 0(r31) C zero carry reg
lda r24, 0(r31) C zero carry reg
cmpeq r20, 2, r21 C
bne r21, $2mod8 C
cmpeq r20, 3, r21 C
bne r21, $3mod8 C
cmpeq r20, 4, r21 C
bne r21, $4mod8 C
cmpeq r20, 5, r21 C
bne r21, $5mod8 C
cmpeq r20, 6, r21 C
bne r21, $6mod8 C
cmpeq r20, 7, r21 C
beq r21, $0mod8 C
$7mod8: ldq r5, 0(r16) C
lda r17, 8(r17) C
mulq r19, r3, r7 C
umulh r19, r3, r24 C
addq r5, r7, r23 C
cmpult r23, r7, r20 C
addq r24, r20, r24 C
stq r23, 0(r16) C
lda r16, 8(r16) C
ldq r3, 0(r17) C
$6mod8: ldq r1, 8(r17) C
mulq r19, r3, r25 C
umulh r19, r3, r3 C
mulq r19, r1, r28 C
ldq r0, 16(r17) C
ldq r4, 0(r16) C
umulh r19, r1, r8 C
ldq r1, 24(r17) C
lda r17, 48(r17) C L1 bookkeeping
mulq r19, r0, r2 C
ldq r5, 8(r16) C
lda r16, -32(r16) C L1 bookkeeping
umulh r19, r0, r6 C
addq r4, r25, r4 C lo + acc
mulq r19, r1, r7 C
br r31, $ent6 C
$ent1: lda r17, 8(r17) C
lda r16, 8(r16) C
lda r8, 0(r0) C
ldq r3, 0(r17) C
$0mod8: ldq r1, 8(r17) C
mulq r19, r3, r2 C
umulh r19, r3, r6 C
mulq r19, r1, r7 C
ldq r0, 16(r17) C
ldq r4, 0(r16) C
umulh r19, r1, r24 C
ldq r1, 24(r17) C
mulq r19, r0, r25 C
ldq r5, 8(r16) C
umulh r19, r0, r3 C
addq r4, r2, r4 C lo + acc
mulq r19, r1, r28 C
lda r16, -16(r16) C
br r31, $ent0 C
$3mod8: ldq r5, 0(r16) C
lda r17, 8(r17) C
mulq r19, r3, r7 C
umulh r19, r3, r8 C
addq r5, r7, r23 C
cmpult r23, r7, r20 C
addq r8, r20, r24 C
stq r23, 0(r16) C
lda r16, 8(r16) C
ldq r3, 0(r17) C
$2mod8: ldq r1, 8(r17) C
mulq r19, r3, r25 C
umulh r19, r3, r3 C
mulq r19, r1, r28 C
ble r18, $n23 C
ldq r0, 16(r17) C
ldq r4, 0(r16) C
umulh r19, r1, r8 C
ldq r1, 24(r17) C
lda r17, 16(r17) C L1 bookkeeping
mulq r19, r0, r2 C
ldq r5, 8(r16) C
lda r16, 0(r16) C L1 bookkeeping
umulh r19, r0, r6 C
addq r4, r25, r4 C lo + acc
mulq r19, r1, r7 C
br r31, $ent2 C
$5mod8: ldq r5, 0(r16) C
lda r17, 8(r17) C
mulq r19, r3, r7 C
umulh r19, r3, r24 C
addq r5, r7, r23 C
cmpult r23, r7, r20 C
addq r24, r20, r8 C
stq r23, 0(r16) C
lda r16, 8(r16) C
ldq r3, 0(r17) C
$4mod8: ldq r1, 8(r17) C
mulq r19, r3, r2 C
umulh r19, r3, r6 C
mulq r19, r1, r7 C
ldq r0, 16(r17) C
ldq r4, 0(r16) C
umulh r19, r1, r24 C
ldq r1, 24(r17) C
lda r17, 32(r17) C L1 bookkeeping
mulq r19, r0, r25 C
ldq r5, 8(r16) C
lda r16, 16(r16) C L1 bookkeeping
umulh r19, r0, r3 C
addq r4, r2, r4 C lo + acc
mulq r19, r1, r28 C
cmpult r4, r2, r20 C L0 lo add => carry
addq r4, r8, r22 C U0 hi add => answer
ble r18, $Lend C
ALIGN(16)
$Loop:
bis r31, r31, r31 C U1 mt
cmpult r22, r8, r21 C L0 hi add => carry
addq r6, r20, r6 C U0 hi mul + carry
ldq r0, 0(r17) C
bis r31, r31, r31 C U1 mt
addq r5, r7, r23 C L0 lo + acc
addq r6, r21, r6 C U0 hi mul + carry
ldq r4, 0(r16) C L1
umulh r19, r1, r8 C U1
cmpult r23, r7, r20 C L0 lo add => carry
addq r23, r6, r23 C U0 hi add => answer
ldq r1, 8(r17) C L1
mulq r19, r0, r2 C U1
cmpult r23, r6, r21 C L0 hi add => carry
addq r24, r20, r24 C U0 hi mul + carry
ldq r5, 8(r16) C L1
umulh r19, r0, r6 C U1
addq r4, r25, r4 C U0 lo + acc
stq r22, -16(r16) C L0
stq r23, -8(r16) C L1
bis r31, r31, r31 C L0 st slosh
mulq r19, r1, r7 C U1
bis r31, r31, r31 C L1 st slosh
addq r24, r21, r24 C U0 hi mul + carry
$ent2:
cmpult r4, r25, r20 C L0 lo add => carry
bis r31, r31, r31 C U1 mt
lda r18, -8(r18) C L1 bookkeeping
addq r4, r24, r22 C U0 hi add => answer
bis r31, r31, r31 C U1 mt
cmpult r22, r24, r21 C L0 hi add => carry
addq r3, r20, r3 C U0 hi mul + carry
ldq r0, 16(r17) C L1
bis r31, r31, r31 C U1 mt
addq r5, r28, r23 C L0 lo + acc
addq r3, r21, r3 C U0 hi mul + carry
ldq r4, 16(r16) C L1
umulh r19, r1, r24 C U1
cmpult r23, r28, r20 C L0 lo add => carry
addq r23, r3, r23 C U0 hi add => answer
ldq r1, 24(r17) C L1
mulq r19, r0, r25 C U1
cmpult r23, r3, r21 C L0 hi add => carry
addq r8, r20, r8 C U0 hi mul + carry
ldq r5, 24(r16) C L1
umulh r19, r0, r3 C U1
addq r4, r2, r4 C U0 lo + acc
stq r22, 0(r16) C L0
stq r23, 8(r16) C L1
bis r31, r31, r31 C L0 st slosh
mulq r19, r1, r28 C U1
bis r31, r31, r31 C L1 st slosh
addq r8, r21, r8 C L0 hi mul + carry
$ent0:
cmpult r4, r2, r20 C L0 lo add => carry
bis r31, r31, r31 C U1 mt
lda r17, 64(r17) C L1 bookkeeping
addq r4, r8, r22 C U0 hi add => answer
bis r31, r31, r31 C U1 mt
cmpult r22, r8, r21 C L0 hi add => carry
addq r6, r20, r6 C U0 hi mul + carry
ldq r0, -32(r17) C L1
bis r31, r31, r31 C U1 mt
addq r5, r7, r23 C L0 lo + acc
addq r6, r21, r6 C U0 hi mul + carry
ldq r4, 32(r16) C L1
umulh r19, r1, r8 C U1
cmpult r23, r7, r20 C L0 lo add => carry
addq r23, r6, r23 C U0 hi add => answer
ldq r1, -24(r17) C L1
mulq r19, r0, r2 C U1
cmpult r23, r6, r21 C L0 hi add => carry
addq r24, r20, r24 C U0 hi mul + carry
ldq r5, 40(r16) C L1
umulh r19, r0, r6 C U1
addq r4, r25, r4 C U0 lo + acc
stq r22, 16(r16) C L0
stq r23, 24(r16) C L1
bis r31, r31, r31 C L0 st slosh
mulq r19, r1, r7 C U1
bis r31, r31, r31 C L1 st slosh
addq r24, r21, r24 C U0 hi mul + carry
$ent6:
cmpult r4, r25, r20 C L0 lo add => carry
bis r31, r31, r31 C U1 mt
lda r16, 64(r16) C L1 bookkeeping
addq r4, r24, r22 C U0 hi add => answer
bis r31, r31, r31 C U1 mt
cmpult r22, r24, r21 C L0 hi add => carry
addq r3, r20, r3 C U0 hi mul + carry
ldq r0, -16(r17) C L1
bis r31, r31, r31 C U1 mt
addq r5, r28, r23 C L0 lo + acc
addq r3, r21, r3 C U0 hi mul + carry
ldq r4, -16(r16) C L1
umulh r19, r1, r24 C U1
cmpult r23, r28, r20 C L0 lo add => carry
addq r23, r3, r23 C U0 hi add => answer
ldq r1, -8(r17) C L1
mulq r19, r0, r25 C U1
cmpult r23, r3, r21 C L0 hi add => carry
addq r8, r20, r8 C U0 hi mul + carry
ldq r5, -8(r16) C L1
umulh r19, r0, r3 C U1
addq r4, r2, r4 C L0 lo + acc
stq r22, -32(r16) C L0
stq r23, -24(r16) C L1
bis r31, r31, r31 C L0 st slosh
mulq r19, r1, r28 C U1
bis r31, r31, r31 C L1 st slosh
addq r8, r21, r8 C U0 hi mul + carry
cmpult r4, r2, r20 C L0 lo add => carry
addq r4, r8, r22 C U0 hi add => answer
ldl r31, 256(r17) C prefetch up[]
bgt r18, $Loop C U1 bookkeeping
$Lend: cmpult r22, r8, r21 C
addq r6, r20, r6 C
addq r5, r7, r23 C
addq r6, r21, r6 C
ldq r4, 0(r16) C
umulh r19, r1, r8 C
cmpult r23, r7, r20 C
addq r23, r6, r23 C
cmpult r23, r6, r21 C
addq r24, r20, r24 C
ldq r5, 8(r16) C
addq r4, r25, r4 C
stq r22, -16(r16) C
stq r23, -8(r16) C
addq r24, r21, r24 C
cmpult r4, r25, r20 C
addq r4, r24, r22 C
cmpult r22, r24, r21 C
addq r3, r20, r3 C
addq r5, r28, r23 C
addq r3, r21, r3 C
cmpult r23, r28, r20 C
addq r23, r3, r23 C
cmpult r23, r3, r21 C
addq r8, r20, r8 C
stq r22, 0(r16) C
stq r23, 8(r16) C
addq r8, r21, r0 C
ret r31, (r26), 1 C
$n23: ldq r4, 0(r16) C
ldq r5, 8(r16) C
umulh r19, r1, r8 C
addq r4, r25, r4 C
cmpult r4, r25, r20 C
addq r4, r24, r22 C
cmpult r22, r24, r21 C
addq r3, r20, r3 C
addq r5, r28, r23 C
addq r3, r21, r3 C
cmpult r23, r28, r20 C
addq r23, r3, r23 C
cmpult r23, r3, r21 C
addq r8, r20, r8 C
stq r22, 0(r16) C
stq r23, 8(r16) C
addq r8, r21, r0 C
ret r31, (r26), 1 C
EPILOGUE()
ASM_END()