1/* Intel Pentium-4 mpn_sub_n -- mpn subtraction. 2 * 3 * Copyright 2001, 2002 Free Software Foundation, Inc. 4 * 5 * This file is part of Libgcrypt. 6 * 7 * Libgcrypt is free software; you can redistribute it and/or modify 8 * it under the terms of the GNU Lesser General Public License as 9 * published by the Free Software Foundation; either version 2.1 of 10 * the License, or (at your option) any later version. 11 * 12 * Libgcrypt is distributed in the hope that it will be useful, 13 * but WITHOUT ANY WARRANTY; without even the implied warranty of 14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 15 * GNU Lesser General Public License for more details. 16 * 17 * You should have received a copy of the GNU Lesser General Public 18 * License along with this program; if not, write to the Free Software 19 * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA 20 * 21 * Note: This code is heavily based on the GNU MP Library. 22 * Actually it's the same code with only minor changes in the 23 * way the data is stored; this is to support the abstraction 24 * of an optional secure memory allocation which may be used 25 * to avoid revealing of sensitive data due to paging etc. 26 */ 27 28 29#include "sysdep.h" 30#include "asm-syntax.h" 31 32 33/******************* 34 * mpi_limb_t 35 * _gcry_mpih_sub_n( mpi_ptr_t res_ptr, (sp + 4) 36 * mpi_ptr_t s1_ptr, (sp + 8) 37 * mpi_ptr_t s2_ptr, (sp + 12) 38 * mpi_size_t size) (sp + 16) 39 * 40 * P4 Willamette, Northwood: 4.0 cycles/limb if dst!=src1 and dst!=src2 41 * 6.0 cycles/limb if dst==src1 or dst==src2 42 * P4 Prescott: >= 5 cycles/limb 43 * 44 * The main loop code is 2x unrolled so that the carry bit can alternate 45 * between mm0 and mm1. 46 */ 47 48 49.text 50 ALIGN (3) 51 .globl C_SYMBOL_NAME(_gcry_mpih_sub_n) 52C_SYMBOL_NAME(_gcry_mpih_sub_n:) 53 54 pxor %mm0, %mm0 55.Lstart_nc: 56 movl 8(%esp), %eax 57 movl %ebx, 8(%esp) 58 movl 12(%esp), %ebx 59 movl 4(%esp), %edx 60 movl 16(%esp), %ecx 61 62 leal (%eax,%ecx,4), %eax 63 leal (%ebx,%ecx,4), %ebx 64 leal (%edx,%ecx,4), %edx 65 negl %ecx 66 67.Ltop: 68/* 69 C eax src1 end 70 C ebx src2 end 71 C ecx counter, limbs, negative 72 C edx dst end 73 C mm0 carry bit 74*/ 75 76 movd (%eax,%ecx,4), %mm1 77 movd (%ebx,%ecx,4), %mm2 78 psubq %mm2, %mm1 79 80 psubq %mm0, %mm1 81 movd %mm1, (%edx,%ecx,4) 82 83 psrlq $63, %mm1 84 85 addl $1, %ecx 86 jz .Ldone_mm1 87 88 movd (%eax,%ecx,4), %mm0 89 movd (%ebx,%ecx,4), %mm2 90 psubq %mm2, %mm0 91 92 psubq %mm1, %mm0 93 movd %mm0, (%edx,%ecx,4) 94 95 psrlq $63, %mm0 96 97 addl $1, %ecx 98 jnz .Ltop 99 100 101 movd %mm0, %eax 102 movl 8(%esp), %ebx 103 emms 104 ret 105 106 107 108.Ldone_mm1: 109 movd %mm1, %eax 110 movl 8(%esp), %ebx 111 emms 112 ret 113