Return to mul_1.asm CVS log

Up to [local] / OpenXM_contrib / gmp / mpn / alpha

Annotation of OpenXM_contrib/gmp/mpn/alpha/mul_1.asm, Revision 1.1.1.2

1.1       maekawa     1: dnl  Alpha __gmpn_mul_1 -- Multiply a limb vector with a limb and store
                      2: dnl  the result in a second limb vector.
                      3:
1.1.1.2 ! ohara       4: dnl  Copyright 1992, 1994, 1995, 2000, 2002 Free Software Foundation, Inc.
1.1       maekawa     5:
                      6: dnl  This file is part of the GNU MP Library.
                      7:
                      8: dnl  The GNU MP Library is free software; you can redistribute it and/or modify
1.1.1.2 ! ohara       9: dnl  it under the terms of the GNU Lesser General Public License as published
        !            10: dnl  by the Free Software Foundation; either version 2.1 of the License, or (at
        !            11: dnl  your option) any later version.
1.1       maekawa    12:
                     13: dnl  The GNU MP Library is distributed in the hope that it will be useful, but
                     14: dnl  WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
                     15: dnl  or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public
                     16: dnl  License for more details.
                     17:
                     18: dnl  You should have received a copy of the GNU Lesser General Public License
                     19: dnl  along with the GNU MP Library; see the file COPYING.LIB.  If not, write to
                     20: dnl  the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston,
                     21: dnl  MA 02111-1307, USA.
                     22:
                     23: include(`../config.m4')
                     24:
1.1.1.2 ! ohara      25: C  INPUT PARAMETERS
        !            26: C  rp  r16
        !            27: C  up  r17
        !            28: C  n   r18
        !            29: C  vl  r19
        !            30: C  cl  r20
1.1       maekawa    31:
1.1.1.2 ! ohara      32: C  This code runs at 42 cycles/limb on EV4, 18 cycles/limb on EV5, and 7
        !            33: C  cycles/limb on EV6.
1.1       maekawa    34:
                     35: ASM_START()
1.1.1.2 ! ohara      36: PROLOGUE(mpn_mul_1c)
        !            37:        ldq     r2,0(r17)       C r2 = s1_limb
        !            38:        lda     r18,-1(r18)     C size--
        !            39:        mulq    r2,r19,r3       C r3 = prod_low
        !            40:        umulh   r2,r19,r4       C r4 = prod_high
        !            41:        beq     r18,$Le1c       C jump if size was == 1
        !            42:        ldq     r2,8(r17)       C r2 = s1_limb
        !            43:        lda     r18,-1(r18)     C size--
        !            44:        addq    r3,r20,r3       C r3 = cy_limb + cl
        !            45:        stq     r3,0(r16)
        !            46:        cmpult  r3,r20,r0       C r0 = carry from (cy_limb + cl)
        !            47:        bne     r18,$Loop       C jump if size was == 2
        !            48:        br      r31,$Le2
        !            49: $Le1c: addq    r3,r20,r3       C r3 = cy_limb + cl
        !            50:        cmpult  r3,r20,r0       C r0 = carry from (cy_limb + cl)
        !            51: $Le1:  stq     r3,0(r16)
        !            52:        addq    r4,r0,r0
        !            53:        ret     r31,(r26),1
        !            54: EPILOGUE(mpn_mul_1c)
        !            55:
1.1       maekawa    56: PROLOGUE(mpn_mul_1)
                     57:        ldq     r2,0(r17)       C r2 = s1_limb
1.1.1.2 ! ohara      58:        lda     r18,-1(r18)     C size--
1.1       maekawa    59:        mulq    r2,r19,r3       C r3 = prod_low
1.1.1.2 ! ohara      60:        bic     r31,r31,r0      C clear cy_limb
        !            61:        umulh   r2,r19,r4       C r4 = prod_high
        !            62:        beq     r18,$Le1        C jump if size was == 1
1.1       maekawa    63:        ldq     r2,8(r17)       C r2 = s1_limb
1.1.1.2 ! ohara      64:        lda     r18,-1(r18)     C size--
1.1       maekawa    65:        stq     r3,0(r16)
1.1.1.2 ! ohara      66:        beq     r18,$Le2        C jump if size was == 2
1.1       maekawa    67:
                     68:        ALIGN(8)
                     69: $Loop: mulq    r2,r19,r3       C r3 = prod_low
                     70:        addq    r4,r0,r0        C cy_limb = cy_limb + 'cy'
1.1.1.2 ! ohara      71:        lda     r18,-1(r18)     C size--
        !            72:        umulh   r2,r19,r4       C r4 = prod_high
1.1       maekawa    73:        ldq     r2,16(r17)      C r2 = s1_limb
1.1.1.2 ! ohara      74:        lda     r17,8(r17)      C s1_ptr++
1.1       maekawa    75:        addq    r3,r0,r3        C r3 = cy_limb + prod_low
                     76:        stq     r3,8(r16)
                     77:        cmpult  r3,r0,r0        C r0 = carry from (cy_limb + prod_low)
1.1.1.2 ! ohara      78:        lda     r16,8(r16)      C res_ptr++
1.1       maekawa    79:        bne     r18,$Loop
                     80:
1.1.1.2 ! ohara      81: $Le2:  mulq    r2,r19,r3       C r3 = prod_low
1.1       maekawa    82:        addq    r4,r0,r0        C cy_limb = cy_limb + 'cy'
1.1.1.2 ! ohara      83:        umulh   r2,r19,r4       C r4 = prod_high
1.1       maekawa    84:        addq    r3,r0,r3        C r3 = cy_limb + prod_low
                     85:        cmpult  r3,r0,r0        C r0 = carry from (cy_limb + prod_low)
                     86:        stq     r3,8(r16)
                     87:        addq    r4,r0,r0        C cy_limb = prod_high + cy
                     88:        ret     r31,(r26),1
                     89: EPILOGUE(mpn_mul_1)
                     90: ASM_END()