head 1.1; branch 1.1.1; access; symbols netbsd-11-0-RC7:1.1.1.2 netbsd-11-0-RC6:1.1.1.2 netbsd-11-0-RC5:1.1.1.2 netbsd-11-0-RC4:1.1.1.2 netbsd-11-0-RC3:1.1.1.2 netbsd-11-0-RC2:1.1.1.2 netbsd-11-0-RC1:1.1.1.2 perseant-exfatfs-base-20250801:1.1.1.2 netbsd-11:1.1.1.2.0.6 netbsd-11-base:1.1.1.2 netbsd-10-1-RELEASE:1.1.1.2 perseant-exfatfs-base-20240630:1.1.1.2 perseant-exfatfs:1.1.1.2.0.4 perseant-exfatfs-base:1.1.1.2 netbsd-10-0-RELEASE:1.1.1.2 netbsd-10-0-RC6:1.1.1.2 netbsd-10-0-RC5:1.1.1.2 netbsd-10-0-RC4:1.1.1.2 netbsd-10-0-RC3:1.1.1.2 netbsd-10-0-RC2:1.1.1.2 netbsd-10-0-RC1:1.1.1.2 netbsd-10:1.1.1.2.0.2 netbsd-10-base:1.1.1.2 gmp-6-2-1:1.1.1.2 cjep_sun2x-base1:1.1.1.1 cjep_sun2x:1.1.1.1.0.4 cjep_sun2x-base:1.1.1.1 cjep_staticlib_x-base1:1.1.1.1 cjep_staticlib_x:1.1.1.1.0.2 cjep_staticlib_x-base:1.1.1.1 gmp-6-2-0:1.1.1.1 gmp:1.1.1; locks; strict; comment @;; @; 1.1 date 2020.09.27.00.27.05; author mrg; state Exp; branches 1.1.1.1; next ; commitid BWuUFuEU17KgrCpC; 1.1.1.1 date 2020.09.27.00.27.05; author mrg; state Exp; branches; next 1.1.1.2; commitid BWuUFuEU17KgrCpC; 1.1.1.2 date 2021.07.11.21.14.44; author mrg; state Exp; branches; next ; commitid saHZLNSlLcjF8C0D; desc @@ 1.1 log @Initial revision @ text @dnl Power9 mpn_mul_2. dnl Contributed to the GNU project by Torbjörn Granlund. dnl Copyright 2018 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 3 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. If not, see http://www.gnu.org/licenses/. include(`../config.m4') C cycles/limb C power9: 1.58 C STATUS C * Not written with any power9 pipeline understanding. C * The 4x unrolling was not motivated by any timing tests. C * No local scheduling for performance tweaking has been done. C * Decrease load scheduling! define(`rp', `r3') define(`up', `r4') define(`n', `r5') C Note: Reused as scratch define(`vp', `r6') C Note: Reused for v1 define(`v0', `r7') define(`v1', `r6') ASM_START() PROLOGUE(mpn_mul_2) std r28, -32(r1) std r29, -24(r1) std r30, -16(r1) std r31, -8(r1) subfic r0, n, 0 C clear CA subfo r0, r0, r0 C clear OV and r0 cmpdi cr7, n, 4 ld v0, 0(vp) ld v1, 8(vp) srdi r10, n, 2 mtctr r10 rldicl. r9, n, 0, 63 bne cr0, L(bx1) L(bx0): rldicl. r9, n, 63, 63 ld r8, 0(up) ld r9, 8(up) li r11, 0 mulld r28, r8, v0 mulhdu r31, r8, v0 blt cr7, L(2) mulld r5, r8, v1 mulhdu r10, r8, v1 bne cr0, L(b10) L(b00): addi up, up, -8 addi rp, rp, -24 b L(lo0) L(b10): addi up, up, 8 addi rp, rp, -8 b L(lo2) L(2): addi rp, rp, -8 mulld r5, r8, v1 mulhdu r10, r8, v1 b L(cj2) L(bx1): rldicl. r9, n, 63, 63 ld r9, 0(up) ld r8, 8(up) li r10, 0 mulld r29, r9, v0 mulhdu r30, r9, v0 mulld r12, r9, v1 mulhdu r11, r9, v1 bne cr0, L(b11) L(b01): addi rp, rp, -16 b L(lo1) L(b11): addi up, up, 16 blt cr7, L(end) L(top): ld r9, 0(up) maddld( r28, r8, v0, r10) C 0 4 -> adde maddhdu(r31, r8, v0, r10) C 1 5 adde r0, r29, r0 C 7 11 std r0, 0(rp) mulld r5, r8, v1 C 1 5 -> addex mulhdu r10, r8, v1 C 2 6 addex( r0, r12, r30, 0) C 8 12 L(lo2): ld r8, 8(up) maddld( r29, r9, v0, r11) C 1 5 -> adde maddhdu(r30, r9, v0, r11) C 2 6 adde r0, r28, r0 C 8 12 std r0, 8(rp) mulld r12, r9, v1 C 2 6 -> addex mulhdu r11, r9, v1 C 3 7 addex( r0, r5, r31, 0) C 5 9 13 L(lo1): ld r9, 16(up) maddld( r28, r8, v0, r10) C 2 6 -> adde maddhdu(r31, r8, v0, r10) C 3 7 adde r0, r29, r0 C 5 9 13 std r0, 16(rp) mulld r5, r8, v1 C 3 7 -> addex mulhdu r10, r8, v1 C 4 8 addex( r0, r12, r30, 0) C 6 10 L(lo0): ld r8, 24(up) maddld( r29, r9, v0, r11) C 3 7 -> adde maddhdu(r30, r9, v0, r11) C 4 8 adde r0, r28, r0 C 6 10 std r0, 24(rp) mulld r12, r9, v1 C 4 8 -> addex mulhdu r11, r9, v1 C 5 9 addex( r0, r5, r31, 0) C 7 11 addi up, up, 32 addi rp, rp, 32 bdnz L(top) L(end): ld r9, 0(up) maddld( r28, r8, v0, r10) C 0 4 maddhdu(r31, r8, v0, r10) C 1 5 adde r0, r29, r0 C 7 11 std r0, 0(rp) C -4 mulld r5, r8, v1 C 1 5 mulhdu r10, r8, v1 C 2 6 addex( r0, r12, r30, 0) C 8 12 L(cj2): maddld( r29, r9, v0, r11) C 1 5 -2 maddhdu(r30, r9, v0, r11) C 2 6 -1 adde r0, r28, r0 C 8 12 -3 std r0, 8(rp) C -3 mulld r12, r9, v1 C 2 6 -1 mulhdu r11, r9, v1 C 3 7 0 = return limb addex( r0, r5, r31, 0) C 5 9 13 adde r0, r29, r0 C 5 9 13 -2 std r0, 16(rp) C -2 addex( r0, r12, r30, 0) C 6 10 -1 adde r0, r0, r10 C -1 std r0, 24(rp) C -1 li r4, 0 addze r3, r11 addex( r3, r3, r4, 0) L(ret): ld r28, -32(r1) ld r29, -24(r1) ld r30, -16(r1) ld r31, -8(r1) blr EPILOGUE() ASM_END() @ 1.1.1.1 log @initial import of GMP 6.2.0. changes include: - Bug fixes to gmp_snprintf, conversion to double, mpz_powm, and mpf_set_str. - New functions for factorial, primorial, fibonacci, mpz_2fac_ui, and mpz_mfac_uiui. - MIPS r6 cores are now supported. - Various speeds ups. @ text @@ 1.1.1.2 log @initial import of GMP 6.2.1. from their NEWS: Changes between GMP version 6.2.0 and 6.2.1 BUGS FIXED * A possible overflow of type int is avoided for mpz_cmp on huge operands. * Overflows are more carefully detected and reported for mpz_pow_ui. * A bug in longlong.h for aarch64 sub_ddmmss, not affecting GMP, was healed. FEATURES * C90 compliance. * Initial support for Darwin on arm64, and improved portability. * Support for more processors. @ text @d8 1 a8 1 dnl d10 4 a13 14 dnl it under the terms of either: dnl dnl * the GNU Lesser General Public License as published by the Free dnl Software Foundation; either version 3 of the License, or (at your dnl option) any later version. dnl dnl or dnl dnl * the GNU General Public License as published by the Free Software dnl Foundation; either version 2 of the License, or (at your option) any dnl later version. dnl dnl or both in parallel, as here. dnl d16 5 a20 6 dnl or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License dnl for more details. dnl dnl You should have received copies of the GNU General Public License and the dnl GNU Lesser General Public License along with the GNU MP Library. If not, dnl see https://www.gnu.org/licenses/. @