diff --git a/mpi/amd64/mpih-add1.S b/mpi/amd64/mpih-add1.S index 39c00c52..833a43cb 100644 --- a/mpi/amd64/mpih-add1.S +++ b/mpi/amd64/mpih-add1.S @@ -1,63 +1,64 @@ /* AMD64 (x86_64) add_n -- Add two limb vectors of the same length > 0 and store * sum in a third limb vector. * * Copyright (C) 1992, 1994, 1995, 1998, * 2001, 2002, 2006 Free Software Foundation, Inc. * * This file is part of Libgcrypt. * * Libgcrypt is free software; you can redistribute it and/or modify * 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 option) any later version. * * Libgcrypt is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA * * Note: This code is heavily based on the GNU MP Library. * Actually it's the same code with only minor changes in the * way the data is stored; this is to support the abstraction * of an optional secure memory allocation which may be used * to avoid revealing of sensitive data due to paging etc. */ #include "sysdep.h" #include "asm-syntax.h" /******************* * mpi_limb_t * _gcry_mpih_add_n( mpi_ptr_t res_ptr, rdi * mpi_ptr_t s1_ptr, rsi * mpi_ptr_t s2_ptr, rdx * mpi_size_t size) rcx */ -.text + TEXT + ALIGN(4) .globl C_SYMBOL_NAME(_gcry_mpih_add_n) C_SYMBOL_NAME(_gcry_mpih_add_n:) FUNC_ENTRY() leaq (%rsi,%rcx,8), %rsi leaq (%rdi,%rcx,8), %rdi leaq (%rdx,%rcx,8), %rdx negq %rcx xorl %eax, %eax /* clear cy */ ALIGN(4) /* minimal alignment for claimed speed */ .Loop: movq (%rsi,%rcx,8), %rax movq (%rdx,%rcx,8), %r10 adcq %r10, %rax movq %rax, (%rdi,%rcx,8) incq %rcx jne .Loop movq %rcx, %rax /* zero %rax */ adcq %rax, %rax FUNC_EXIT() diff --git a/mpi/amd64/mpih-lshift.S b/mpi/amd64/mpih-lshift.S index a9c7d7e1..c11e808c 100644 --- a/mpi/amd64/mpih-lshift.S +++ b/mpi/amd64/mpih-lshift.S @@ -1,78 +1,79 @@ /* AMD64 (x86_64) lshift -- Left shift a limb vector and store * result in a second limb vector. * * Copyright (C) 1992, 1994, 1995, 1998, * 2001, 2002, 2006 Free Software Foundation, Inc. * * This file is part of Libgcrypt. * * Libgcrypt is free software; you can redistribute it and/or modify * 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 option) any later version. * * Libgcrypt is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA * * Note: This code is heavily based on the GNU MP Library. * Actually it's the same code with only minor changes in the * way the data is stored; this is to support the abstraction * of an optional secure memory allocation which may be used * to avoid revealing of sensitive data due to paging etc. */ #include "sysdep.h" #include "asm-syntax.h" /******************* * mpi_limb_t * _gcry_mpih_lshift( mpi_ptr_t wp, rdi * mpi_ptr_t up, rsi * mpi_size_t usize, rdx * unsigned cnt) rcx */ -.text + TEXT + ALIGN(4) .globl C_SYMBOL_NAME(_gcry_mpih_lshift) C_SYMBOL_NAME(_gcry_mpih_lshift:) FUNC_ENTRY() /* Note: %xmm6 and %xmm7 not used for WIN64 ABI compatibility. */ movq -8(%rsi,%rdx,8), %xmm4 movd %ecx, %xmm1 movl $64, %eax subl %ecx, %eax movd %eax, %xmm0 movdqa %xmm4, %xmm3 psrlq %xmm0, %xmm4 movd %xmm4, %rax subq $2, %rdx jl .Lendo ALIGN(4) /* minimal alignment for claimed speed */ .Loop: movq (%rsi,%rdx,8), %xmm5 movdqa %xmm5, %xmm2 psrlq %xmm0, %xmm5 psllq %xmm1, %xmm3 por %xmm5, %xmm3 movq %xmm3, 8(%rdi,%rdx,8) je .Lende movq -8(%rsi,%rdx,8), %xmm4 movdqa %xmm4, %xmm3 psrlq %xmm0, %xmm4 psllq %xmm1, %xmm2 por %xmm4, %xmm2 movq %xmm2, (%rdi,%rdx,8) subq $2, %rdx jge .Loop .Lendo: movdqa %xmm3, %xmm2 .Lende: psllq %xmm1, %xmm2 movq %xmm2, (%rdi) FUNC_EXIT() diff --git a/mpi/amd64/mpih-mul1.S b/mpi/amd64/mpih-mul1.S index dacb9d87..de5fa3ce 100644 --- a/mpi/amd64/mpih-mul1.S +++ b/mpi/amd64/mpih-mul1.S @@ -1,66 +1,66 @@ /* AMD64 mul_1 -- Multiply a limb vector with a limb and store * the result in a second limb vector. * Copyright (C) 1992, 1994, 1998, * 2001, 2002, 2006 Free Software Foundation, Inc. * * This file is part of Libgcrypt. * * Libgcrypt is free software; you can redistribute it and/or modify * 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 option) any later version. * * Libgcrypt is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA * * Note: This code is heavily based on the GNU MP Library. * Actually it's the same code with only minor changes in the * way the data is stored; this is to support the abstraction * of an optional secure memory allocation which may be used * to avoid revealing of sensitive data due to paging etc. */ #include "sysdep.h" #include "asm-syntax.h" /******************* * mpi_limb_t * _gcry_mpih_mul_1( mpi_ptr_t res_ptr, (rdi) * mpi_ptr_t s1_ptr, (rsi) * mpi_size_t s1_size, (rdx) * mpi_limb_t s2_limb) (rcx) */ TEXT - ALIGN(5) - .byte 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 + ALIGN(4) GLOBL C_SYMBOL_NAME(_gcry_mpih_mul_1) C_SYMBOL_NAME(_gcry_mpih_mul_1:) FUNC_ENTRY() movq %rdx, %r11 leaq (%rsi,%rdx,8), %rsi leaq (%rdi,%rdx,8), %rdi negq %r11 xorl %r8d, %r8d + ALIGN(4) .Loop: movq (%rsi,%r11,8), %rax mulq %rcx addq %r8, %rax movl $0, %r8d adcq %rdx, %r8 movq %rax, (%rdi,%r11,8) incq %r11 jne .Loop movq %r8, %rax FUNC_EXIT() diff --git a/mpi/amd64/mpih-mul2.S b/mpi/amd64/mpih-mul2.S index 07913586..0b3025d6 100644 --- a/mpi/amd64/mpih-mul2.S +++ b/mpi/amd64/mpih-mul2.S @@ -1,65 +1,66 @@ /* AMD64 addmul2 -- Multiply a limb vector with a limb and add * the result to a second limb vector. * * Copyright (C) 1992, 1994, 1998, * 2001, 2002, 2006 Free Software Foundation, Inc. * * This file is part of Libgcrypt. * * Libgcrypt is free software; you can redistribute it and/or modify * 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 option) any later version. * * Libgcrypt is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA * * Note: This code is heavily based on the GNU MP Library. * Actually it's the same code with only minor changes in the * way the data is stored; this is to support the abstraction * of an optional secure memory allocation which may be used * to avoid revealing of sensitive data due to paging etc. */ #include "sysdep.h" #include "asm-syntax.h" /******************* * mpi_limb_t * _gcry_mpih_addmul_1( mpi_ptr_t res_ptr, (rdi) * mpi_ptr_t s1_ptr, (rsi) * mpi_size_t s1_size, (rdx) * mpi_limb_t s2_limb) (rcx) */ TEXT + ALIGN(4) GLOBL C_SYMBOL_NAME(_gcry_mpih_addmul_1) C_SYMBOL_NAME(_gcry_mpih_addmul_1:) FUNC_ENTRY() movq %rdx, %r11 leaq (%rsi,%rdx,8), %rsi leaq (%rdi,%rdx,8), %rdi negq %r11 xorl %r8d, %r8d xorl %r10d, %r10d - ALIGN(3) /* minimal alignment for claimed speed */ + ALIGN(4) /* minimal alignment for claimed speed */ .Loop: movq (%rsi,%r11,8), %rax mulq %rcx addq (%rdi,%r11,8), %rax adcq %r10, %rdx addq %r8, %rax movq %r10, %r8 movq %rax, (%rdi,%r11,8) adcq %rdx, %r8 incq %r11 jne .Loop movq %r8, %rax FUNC_EXIT() diff --git a/mpi/amd64/mpih-mul3.S b/mpi/amd64/mpih-mul3.S index f8889eb2..7d3486e8 100644 --- a/mpi/amd64/mpih-mul3.S +++ b/mpi/amd64/mpih-mul3.S @@ -1,66 +1,67 @@ /* AMD64 submul_1 -- Multiply a limb vector with a limb and add * the result to a second limb vector. * * Copyright (C) 1992, 1994, 1998, * 2001, 2002, 2006 Free Software Foundation, Inc. * * This file is part of Libgcrypt. * * Libgcrypt is free software; you can redistribute it and/or modify * 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 option) any later version. * * Libgcrypt is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA * * Note: This code is heavily based on the GNU MP Library. * Actually it's the same code with only minor changes in the * way the data is stored; this is to support the abstraction * of an optional secure memory allocation which may be used * to avoid revealing of sensitive data due to paging etc. */ #include "sysdep.h" #include "asm-syntax.h" /******************* * mpi_limb_t * _gcry_mpih_submul_1( mpi_ptr_t res_ptr, (rdi) * mpi_ptr_t s1_ptr, (rsi) * mpi_size_t s1_size, (rdx) * mpi_limb_t s2_limb) (rcx) */ TEXT + ALIGN(4) GLOBL C_SYMBOL_NAME(_gcry_mpih_submul_1) C_SYMBOL_NAME(_gcry_mpih_submul_1:) FUNC_ENTRY() movq %rdx, %r11 leaq (%rsi,%r11,8), %rsi leaq (%rdi,%r11,8), %rdi negq %r11 xorl %r8d, %r8d - ALIGN(3) /* minimal alignment for claimed speed */ + ALIGN(4) /* minimal alignment for claimed speed */ .Loop: movq (%rsi,%r11,8), %rax movq (%rdi,%r11,8), %r10 mulq %rcx subq %r8, %r10 movl $0, %r8d adcl %r8d, %r8d subq %rax, %r10 adcq %rdx, %r8 movq %r10, (%rdi,%r11,8) incq %r11 jne .Loop movq %r8, %rax FUNC_EXIT() diff --git a/mpi/amd64/mpih-rshift.S b/mpi/amd64/mpih-rshift.S index 8ecf155f..430ba4b0 100644 --- a/mpi/amd64/mpih-rshift.S +++ b/mpi/amd64/mpih-rshift.S @@ -1,81 +1,82 @@ /* AMD64 (x86_64) rshift -- Right shift a limb vector and store * result in a second limb vector. * * Copyright (C) 1992, 1994, 1995, 1998, * 2001, 2002, 2006 Free Software Foundation, Inc. * * This file is part of Libgcrypt. * * Libgcrypt is free software; you can redistribute it and/or modify * 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 option) any later version. * * Libgcrypt is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA * * Note: This code is heavily based on the GNU MP Library. * Actually it's the same code with only minor changes in the * way the data is stored; this is to support the abstraction * of an optional secure memory allocation which may be used * to avoid revealing of sensitive data due to paging etc. */ #include "sysdep.h" #include "asm-syntax.h" /******************* * mpi_limb_t * _gcry_mpih_rshift( mpi_ptr_t wp, rdi * mpi_ptr_t up, rsi * mpi_size_t usize, rdx * unsigned cnt) rcx */ -.text + TEXT + ALIGN(4) .globl C_SYMBOL_NAME(_gcry_mpih_rshift) C_SYMBOL_NAME(_gcry_mpih_rshift:) FUNC_ENTRY() /* Note: %xmm6 and %xmm7 not used for WIN64 ABI compatibility. */ movq (%rsi), %xmm4 movd %ecx, %xmm1 movl $64, %eax subl %ecx, %eax movd %eax, %xmm0 movdqa %xmm4, %xmm3 psllq %xmm0, %xmm4 movd %xmm4, %rax leaq (%rsi,%rdx,8), %rsi leaq (%rdi,%rdx,8), %rdi negq %rdx addq $2, %rdx jg .Lendo ALIGN(4) /* minimal alignment for claimed speed */ .Loop: movq -8(%rsi,%rdx,8), %xmm5 movdqa %xmm5, %xmm2 psllq %xmm0, %xmm5 psrlq %xmm1, %xmm3 por %xmm5, %xmm3 movq %xmm3, -16(%rdi,%rdx,8) je .Lende movq (%rsi,%rdx,8), %xmm4 movdqa %xmm4, %xmm3 psllq %xmm0, %xmm4 psrlq %xmm1, %xmm2 por %xmm4, %xmm2 movq %xmm2, -8(%rdi,%rdx,8) addq $2, %rdx jle .Loop .Lendo: movdqa %xmm3, %xmm2 .Lende: psrlq %xmm1, %xmm2 movq %xmm2, -8(%rdi) FUNC_EXIT() diff --git a/mpi/amd64/mpih-sub1.S b/mpi/amd64/mpih-sub1.S index d60b58a5..8c61cb20 100644 --- a/mpi/amd64/mpih-sub1.S +++ b/mpi/amd64/mpih-sub1.S @@ -1,62 +1,63 @@ /* AMD64 (x86_64) sub_n -- Subtract two limb vectors of the same length > 0 and store * sum in a third limb vector. * * Copyright (C) 1992, 1994, 1995, 1998, * 2001, 2002, 2006 Free Software Foundation, Inc. * * This file is part of Libgcrypt. * * Libgcrypt is free software; you can redistribute it and/or modify * 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 option) any later version. * * Libgcrypt is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public * License along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA * * Note: This code is heavily based on the GNU MP Library. * Actually it's the same code with only minor changes in the * way the data is stored; this is to support the abstraction * of an optional secure memory allocation which may be used * to avoid revealing of sensitive data due to paging etc. */ #include "sysdep.h" #include "asm-syntax.h" /******************* * mpi_limb_t * _gcry_mpih_sub_n( mpi_ptr_t res_ptr, rdi * mpi_ptr_t s1_ptr, rsi * mpi_ptr_t s2_ptr, rdx * mpi_size_t size) rcx */ -.text + TEXT + ALIGN(4) .globl C_SYMBOL_NAME(_gcry_mpih_sub_n) C_SYMBOL_NAME(_gcry_mpih_sub_n:) FUNC_ENTRY() leaq (%rsi,%rcx,8), %rsi leaq (%rdi,%rcx,8), %rdi leaq (%rdx,%rcx,8), %rdx negq %rcx xorl %eax, %eax /* clear cy */ ALIGN(4) /* minimal alignment for claimed speed */ .Loop: movq (%rsi,%rcx,8), %rax movq (%rdx,%rcx,8), %r10 sbbq %r10, %rax movq %rax, (%rdi,%rcx,8) incq %rcx jne .Loop movq %rcx, %rax /* zero %rax */ adcq %rax, %rax FUNC_EXIT()