From 0314322d1303065568f33869cbd01d7f7367efc4 Mon Sep 17 00:00:00 2001 From: Michael Witten Date: Fri, 29 Jul 2011 12:53:20 +0000 Subject: Docs: Pedantry: [Cc]ordic -> CORDIC According to: http://en.wikipedia.org/wiki/CORDIC it stands for: *CO*ordinate *R*otation *DI*gital *C*omputer Signed-off-by: Michael Witten Signed-off-by: Jiri Kosina --- lib/Kconfig | 4 ++-- lib/cordic.c | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig b/lib/Kconfig index 6c695ff9cab..0bb69f67448 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -270,9 +270,9 @@ config AVERAGE If unsure, say N. config CORDIC - tristate "Cordic function" + tristate "CORDIC function" help - The option provides arithmetic function using cordic algorithm + The option provides arithmetic function using CORDIC algorithm so its calculations are in fixed point. Modules can select this when they require this function. Module will be called cordic. diff --git a/lib/cordic.c b/lib/cordic.c index aa27a88d7e0..09b34036534 100644 --- a/lib/cordic.c +++ b/lib/cordic.c @@ -96,6 +96,6 @@ struct cordic_iq cordic_calc_iq(s32 theta) } EXPORT_SYMBOL(cordic_calc_iq); -MODULE_DESCRIPTION("Cordic functions"); +MODULE_DESCRIPTION("CORDIC functions"); MODULE_AUTHOR("Broadcom Corporation"); MODULE_LICENSE("Dual BSD/GPL"); -- cgit v1.2.3-70-g09d2 From d89ce936b4d6905c2bdf7fa06778d89ae2ee2151 Mon Sep 17 00:00:00 2001 From: Michael Witten Date: Fri, 29 Jul 2011 12:59:51 +0000 Subject: Docs: wording: functions -> algorithm The code seems to provide a single function that implements the CORDIC algorithm. Signed-off-by: Michael Witten Signed-off-by: Jiri Kosina --- lib/Kconfig | 2 +- lib/cordic.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig b/lib/Kconfig index 0bb69f67448..d8c8a16e4ee 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -270,7 +270,7 @@ config AVERAGE If unsure, say N. config CORDIC - tristate "CORDIC function" + tristate "CORDIC algorithm" help The option provides arithmetic function using CORDIC algorithm so its calculations are in fixed point. Modules can select this diff --git a/lib/cordic.c b/lib/cordic.c index 09b34036534..6cf477839eb 100644 --- a/lib/cordic.c +++ b/lib/cordic.c @@ -96,6 +96,6 @@ struct cordic_iq cordic_calc_iq(s32 theta) } EXPORT_SYMBOL(cordic_calc_iq); -MODULE_DESCRIPTION("CORDIC functions"); +MODULE_DESCRIPTION("CORDIC algorithm"); MODULE_AUTHOR("Broadcom Corporation"); MODULE_LICENSE("Dual BSD/GPL"); -- cgit v1.2.3-70-g09d2 From 435a95c5c8b279668cea32014e5c30821581792d Mon Sep 17 00:00:00 2001 From: Michael Witten Date: Fri, 29 Jul 2011 13:36:04 +0000 Subject: Docs: Kconfig: CORDIC description This is just some copyediting. Signed-off-by: Michael Witten Signed-off-by: Jiri Kosina --- lib/Kconfig | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig b/lib/Kconfig index d8c8a16e4ee..abff69d4ff8 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -272,9 +272,8 @@ config AVERAGE config CORDIC tristate "CORDIC algorithm" help - The option provides arithmetic function using CORDIC algorithm - so its calculations are in fixed point. Modules can select this - when they require this function. Module will be called cordic. + This option provides an implementation of the CORDIC algorithm; + calculations are in fixed point. Module will be called cordic. config LLIST bool -- cgit v1.2.3-70-g09d2 From cdec9cb5167ab1113ba9c58e395f664d9d3f9acb Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Wed, 31 Aug 2011 14:05:16 +0300 Subject: crypto: GnuPG based MPI lib - source files (part 1) Adds the multi-precision-integer maths library which was originally taken from GnuPG and ported to the kernel by (among others) David Howells. This version is taken from Fedora kernel 2.6.32-71.14.1.el6. The difference is that checkpatch reported errors and warnings have been fixed. This library is used to implemenet RSA digital signature verification used in IMA/EVM integrity protection subsystem. Due to patch size limitation, the patch is divided into 4 parts. Signed-off-by: Dmitry Kasatkin --- lib/mpi/generic_mpih-add1.c | 61 +++++ lib/mpi/generic_mpih-lshift.c | 63 +++++ lib/mpi/generic_mpih-mul1.c | 57 +++++ lib/mpi/generic_mpih-mul2.c | 60 +++++ lib/mpi/generic_mpih-mul3.c | 61 +++++ lib/mpi/generic_mpih-rshift.c | 63 +++++ lib/mpi/generic_mpih-sub1.c | 60 +++++ lib/mpi/mpi-bit.c | 236 ++++++++++++++++++ lib/mpi/mpi-pow.c | 323 +++++++++++++++++++++++++ lib/mpi/mpicoder.c | 365 ++++++++++++++++++++++++++++ lib/mpi/mpih-cmp.c | 56 +++++ lib/mpi/mpih-div.c | 541 ++++++++++++++++++++++++++++++++++++++++++ lib/mpi/mpih-mul.c | 527 ++++++++++++++++++++++++++++++++++++++++ lib/mpi/mpiutil.c | 208 ++++++++++++++++ 14 files changed, 2681 insertions(+) create mode 100644 lib/mpi/generic_mpih-add1.c create mode 100644 lib/mpi/generic_mpih-lshift.c create mode 100644 lib/mpi/generic_mpih-mul1.c create mode 100644 lib/mpi/generic_mpih-mul2.c create mode 100644 lib/mpi/generic_mpih-mul3.c create mode 100644 lib/mpi/generic_mpih-rshift.c create mode 100644 lib/mpi/generic_mpih-sub1.c create mode 100644 lib/mpi/mpi-bit.c create mode 100644 lib/mpi/mpi-pow.c create mode 100644 lib/mpi/mpicoder.c create mode 100644 lib/mpi/mpih-cmp.c create mode 100644 lib/mpi/mpih-div.c create mode 100644 lib/mpi/mpih-mul.c create mode 100644 lib/mpi/mpiutil.c (limited to 'lib') diff --git a/lib/mpi/generic_mpih-add1.c b/lib/mpi/generic_mpih-add1.c new file mode 100644 index 00000000000..c94c7dd344b --- /dev/null +++ b/lib/mpi/generic_mpih-add1.c @@ -0,0 +1,61 @@ +/* mpihelp-add_1.c - MPI helper functions + * Copyright (C) 1994, 1996, 1997, 1998, + * 2000 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" +#include "longlong.h" + +mpi_limb_t +mpihelp_add_n(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_ptr_t s2_ptr, mpi_size_t size) +{ + mpi_limb_t x, y, cy; + mpi_size_t j; + + /* The loop counter and index J goes from -SIZE to -1. This way + the loop becomes faster. */ + j = -size; + + /* Offset the base pointers to compensate for the negative indices. */ + s1_ptr -= j; + s2_ptr -= j; + res_ptr -= j; + + cy = 0; + do { + y = s2_ptr[j]; + x = s1_ptr[j]; + y += cy; /* add previous carry to one addend */ + cy = y < cy; /* get out carry from that addition */ + y += x; /* add other addend */ + cy += y < x; /* get out carry from that add, combine */ + res_ptr[j] = y; + } while (++j); + + return cy; +} diff --git a/lib/mpi/generic_mpih-lshift.c b/lib/mpi/generic_mpih-lshift.c new file mode 100644 index 00000000000..86318927231 --- /dev/null +++ b/lib/mpi/generic_mpih-lshift.c @@ -0,0 +1,63 @@ +/* mpihelp-lshift.c - MPI helper functions + * Copyright (C) 1994, 1996, 1998, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" + +/* Shift U (pointed to by UP and USIZE digits long) CNT bits to the left + * and store the USIZE least significant digits of the result at WP. + * Return the bits shifted out from the most significant digit. + * + * Argument constraints: + * 1. 0 < CNT < BITS_PER_MP_LIMB + * 2. If the result is to be written over the input, WP must be >= UP. + */ + +mpi_limb_t +mpihelp_lshift(mpi_ptr_t wp, mpi_ptr_t up, mpi_size_t usize, unsigned int cnt) +{ + mpi_limb_t high_limb, low_limb; + unsigned sh_1, sh_2; + mpi_size_t i; + mpi_limb_t retval; + + sh_1 = cnt; + wp += 1; + sh_2 = BITS_PER_MPI_LIMB - sh_1; + i = usize - 1; + low_limb = up[i]; + retval = low_limb >> sh_2; + high_limb = low_limb; + while (--i >= 0) { + low_limb = up[i]; + wp[i] = (high_limb << sh_1) | (low_limb >> sh_2); + high_limb = low_limb; + } + wp[i] = high_limb << sh_1; + + return retval; +} diff --git a/lib/mpi/generic_mpih-mul1.c b/lib/mpi/generic_mpih-mul1.c new file mode 100644 index 00000000000..1668dfd9092 --- /dev/null +++ b/lib/mpi/generic_mpih-mul1.c @@ -0,0 +1,57 @@ +/* mpihelp-mul_1.c - MPI helper functions + * Copyright (C) 1994, 1996, 1997, 1998, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" +#include "longlong.h" + +mpi_limb_t +mpihelp_mul_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, mpi_size_t s1_size, + mpi_limb_t s2_limb) +{ + mpi_limb_t cy_limb; + mpi_size_t j; + mpi_limb_t prod_high, prod_low; + + /* The loop counter and index J goes from -S1_SIZE to -1. This way + * the loop becomes faster. */ + j = -s1_size; + + /* Offset the base pointers to compensate for the negative indices. */ + s1_ptr -= j; + res_ptr -= j; + + cy_limb = 0; + do { + umul_ppmm(prod_high, prod_low, s1_ptr[j], s2_limb); + prod_low += cy_limb; + cy_limb = (prod_low < cy_limb ? 1 : 0) + prod_high; + res_ptr[j] = prod_low; + } while (++j); + + return cy_limb; +} diff --git a/lib/mpi/generic_mpih-mul2.c b/lib/mpi/generic_mpih-mul2.c new file mode 100644 index 00000000000..8a7b29ee174 --- /dev/null +++ b/lib/mpi/generic_mpih-mul2.c @@ -0,0 +1,60 @@ +/* mpihelp-mul_2.c - MPI helper functions + * Copyright (C) 1994, 1996, 1997, 1998, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" +#include "longlong.h" + +mpi_limb_t +mpihelp_addmul_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb) +{ + mpi_limb_t cy_limb; + mpi_size_t j; + mpi_limb_t prod_high, prod_low; + mpi_limb_t x; + + /* The loop counter and index J goes from -SIZE to -1. This way + * the loop becomes faster. */ + j = -s1_size; + res_ptr -= j; + s1_ptr -= j; + + cy_limb = 0; + do { + umul_ppmm(prod_high, prod_low, s1_ptr[j], s2_limb); + + prod_low += cy_limb; + cy_limb = (prod_low < cy_limb ? 1 : 0) + prod_high; + + x = res_ptr[j]; + prod_low = x + prod_low; + cy_limb += prod_low < x ? 1 : 0; + res_ptr[j] = prod_low; + } while (++j); + return cy_limb; +} diff --git a/lib/mpi/generic_mpih-mul3.c b/lib/mpi/generic_mpih-mul3.c new file mode 100644 index 00000000000..f96df327be6 --- /dev/null +++ b/lib/mpi/generic_mpih-mul3.c @@ -0,0 +1,61 @@ +/* mpihelp-mul_3.c - MPI helper functions + * Copyright (C) 1994, 1996, 1997, 1998, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" +#include "longlong.h" + +mpi_limb_t +mpihelp_submul_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb) +{ + mpi_limb_t cy_limb; + mpi_size_t j; + mpi_limb_t prod_high, prod_low; + mpi_limb_t x; + + /* The loop counter and index J goes from -SIZE to -1. This way + * the loop becomes faster. */ + j = -s1_size; + res_ptr -= j; + s1_ptr -= j; + + cy_limb = 0; + do { + umul_ppmm(prod_high, prod_low, s1_ptr[j], s2_limb); + + prod_low += cy_limb; + cy_limb = (prod_low < cy_limb ? 1 : 0) + prod_high; + + x = res_ptr[j]; + prod_low = x - prod_low; + cy_limb += prod_low > x ? 1 : 0; + res_ptr[j] = prod_low; + } while (++j); + + return cy_limb; +} diff --git a/lib/mpi/generic_mpih-rshift.c b/lib/mpi/generic_mpih-rshift.c new file mode 100644 index 00000000000..ffa328818ca --- /dev/null +++ b/lib/mpi/generic_mpih-rshift.c @@ -0,0 +1,63 @@ +/* mpih-rshift.c - MPI helper functions + * Copyright (C) 1994, 1996, 1998, 1999, + * 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GNUPG + * + * GNUPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GNUPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" + +/* Shift U (pointed to by UP and USIZE limbs long) CNT bits to the right + * and store the USIZE least significant limbs of the result at WP. + * The bits shifted out to the right are returned. + * + * Argument constraints: + * 1. 0 < CNT < BITS_PER_MP_LIMB + * 2. If the result is to be written over the input, WP must be <= UP. + */ + +mpi_limb_t +mpihelp_rshift(mpi_ptr_t wp, mpi_ptr_t up, mpi_size_t usize, unsigned cnt) +{ + mpi_limb_t high_limb, low_limb; + unsigned sh_1, sh_2; + mpi_size_t i; + mpi_limb_t retval; + + sh_1 = cnt; + wp -= 1; + sh_2 = BITS_PER_MPI_LIMB - sh_1; + high_limb = up[0]; + retval = high_limb << sh_2; + low_limb = high_limb; + for (i = 1; i < usize; i++) { + high_limb = up[i]; + wp[i] = (low_limb >> sh_1) | (high_limb << sh_2); + low_limb = high_limb; + } + wp[i] = low_limb >> sh_1; + + return retval; +} diff --git a/lib/mpi/generic_mpih-sub1.c b/lib/mpi/generic_mpih-sub1.c new file mode 100644 index 00000000000..5d98ab7d685 --- /dev/null +++ b/lib/mpi/generic_mpih-sub1.c @@ -0,0 +1,60 @@ +/* mpihelp-add_2.c - MPI helper functions + * Copyright (C) 1994, 1996, 1997, 1998, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" +#include "longlong.h" + +mpi_limb_t +mpihelp_sub_n(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_ptr_t s2_ptr, mpi_size_t size) +{ + mpi_limb_t x, y, cy; + mpi_size_t j; + + /* The loop counter and index J goes from -SIZE to -1. This way + the loop becomes faster. */ + j = -size; + + /* Offset the base pointers to compensate for the negative indices. */ + s1_ptr -= j; + s2_ptr -= j; + res_ptr -= j; + + cy = 0; + do { + y = s2_ptr[j]; + x = s1_ptr[j]; + y += cy; /* add previous carry to subtrahend */ + cy = y < cy; /* get out carry from that addition */ + y = x - y; /* main subtract */ + cy += y > x; /* get out carry from the subtract, combine */ + res_ptr[j] = y; + } while (++j); + + return cy; +} diff --git a/lib/mpi/mpi-bit.c b/lib/mpi/mpi-bit.c new file mode 100644 index 00000000000..854c9c6da02 --- /dev/null +++ b/lib/mpi/mpi-bit.c @@ -0,0 +1,236 @@ +/* mpi-bit.c - MPI bit level fucntions + * Copyright (C) 1998, 1999 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +#include "mpi-internal.h" +#include "longlong.h" + +const unsigned char __clz_tab[] = { + 0, 1, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, 5, 5, 5, 5, 5, 5, 5, 5, + 5, 5, 5, 5, 5, 5, 5, 5, + 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, + 6, 6, 6, 6, 6, 6, 6, 6, + 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, + 7, 7, 7, 7, 7, 7, 7, 7, + 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, + 7, 7, 7, 7, 7, 7, 7, 7, + 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, + 8, 8, 8, 8, 8, 8, 8, 8, + 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, + 8, 8, 8, 8, 8, 8, 8, 8, + 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, + 8, 8, 8, 8, 8, 8, 8, 8, + 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, + 8, 8, 8, 8, 8, 8, 8, 8, +}; + +#define A_LIMB_1 ((mpi_limb_t) 1) + +/**************** + * Sometimes we have MSL (most significant limbs) which are 0; + * this is for some reasons not good, so this function removes them. + */ +void mpi_normalize(MPI a) +{ + for (; a->nlimbs && !a->d[a->nlimbs - 1]; a->nlimbs--) + ; +} + +/**************** + * Return the number of bits in A. + */ +unsigned mpi_get_nbits(MPI a) +{ + unsigned n; + + mpi_normalize(a); + + if (a->nlimbs) { + mpi_limb_t alimb = a->d[a->nlimbs - 1]; + if (alimb) + count_leading_zeros(n, alimb); + else + n = BITS_PER_MPI_LIMB; + n = BITS_PER_MPI_LIMB - n + (a->nlimbs - 1) * BITS_PER_MPI_LIMB; + } else + n = 0; + return n; +} +EXPORT_SYMBOL_GPL(mpi_get_nbits); + +/**************** + * Test whether bit N is set. + */ +int mpi_test_bit(MPI a, unsigned n) +{ + unsigned limbno, bitno; + mpi_limb_t limb; + + limbno = n / BITS_PER_MPI_LIMB; + bitno = n % BITS_PER_MPI_LIMB; + + if (limbno >= a->nlimbs) + return 0; /* too far left: this is a 0 */ + limb = a->d[limbno]; + return (limb & (A_LIMB_1 << bitno)) ? 1 : 0; +} + +/**************** + * Set bit N of A. + */ +int mpi_set_bit(MPI a, unsigned n) +{ + unsigned limbno, bitno; + + limbno = n / BITS_PER_MPI_LIMB; + bitno = n % BITS_PER_MPI_LIMB; + + if (limbno >= a->nlimbs) { /* resize */ + if (a->alloced >= limbno) + if (mpi_resize(a, limbno + 1) < 0) + return -ENOMEM; + a->nlimbs = limbno + 1; + } + a->d[limbno] |= (A_LIMB_1 << bitno); + return 0; +} + +/**************** + * Set bit N of A. and clear all bits above + */ +int mpi_set_highbit(MPI a, unsigned n) +{ + unsigned limbno, bitno; + + limbno = n / BITS_PER_MPI_LIMB; + bitno = n % BITS_PER_MPI_LIMB; + + if (limbno >= a->nlimbs) { /* resize */ + if (a->alloced >= limbno) + if (mpi_resize(a, limbno + 1) < 0) + return -ENOMEM; + a->nlimbs = limbno + 1; + } + a->d[limbno] |= (A_LIMB_1 << bitno); + for (bitno++; bitno < BITS_PER_MPI_LIMB; bitno++) + a->d[limbno] &= ~(A_LIMB_1 << bitno); + a->nlimbs = limbno + 1; + return 0; +} + +/**************** + * clear bit N of A and all bits above + */ +void mpi_clear_highbit(MPI a, unsigned n) +{ + unsigned limbno, bitno; + + limbno = n / BITS_PER_MPI_LIMB; + bitno = n % BITS_PER_MPI_LIMB; + + if (limbno >= a->nlimbs) + return; /* not allocated, so need to clear bits :-) */ + + for (; bitno < BITS_PER_MPI_LIMB; bitno++) + a->d[limbno] &= ~(A_LIMB_1 << bitno); + a->nlimbs = limbno + 1; +} + +/**************** + * Clear bit N of A. + */ +void mpi_clear_bit(MPI a, unsigned n) +{ + unsigned limbno, bitno; + + limbno = n / BITS_PER_MPI_LIMB; + bitno = n % BITS_PER_MPI_LIMB; + + if (limbno >= a->nlimbs) + return; /* don't need to clear this bit, it's to far to left */ + a->d[limbno] &= ~(A_LIMB_1 << bitno); +} + +/**************** + * Shift A by N bits to the right + * FIXME: should use alloc_limb if X and A are same. + */ +int mpi_rshift(MPI x, MPI a, unsigned n) +{ + mpi_ptr_t xp; + mpi_size_t xsize; + + xsize = a->nlimbs; + x->sign = a->sign; + if (RESIZE_IF_NEEDED(x, (size_t) xsize) < 0) + return -ENOMEM; + xp = x->d; + + if (xsize) { + mpihelp_rshift(xp, a->d, xsize, n); + MPN_NORMALIZE(xp, xsize); + } + x->nlimbs = xsize; + return 0; +} + +/**************** + * Shift A by COUNT limbs to the left + * This is used only within the MPI library + */ +int mpi_lshift_limbs(MPI a, unsigned int count) +{ + mpi_ptr_t ap = a->d; + int n = a->nlimbs; + int i; + + if (!count || !n) + return 0; + + if (RESIZE_IF_NEEDED(a, n + count) < 0) + return -ENOMEM; + + for (i = n - 1; i >= 0; i--) + ap[i + count] = ap[i]; + for (i = 0; i < count; i++) + ap[i] = 0; + a->nlimbs += count; + return 0; +} + +/**************** + * Shift A by COUNT limbs to the right + * This is used only within the MPI library + */ +void mpi_rshift_limbs(MPI a, unsigned int count) +{ + mpi_ptr_t ap = a->d; + mpi_size_t n = a->nlimbs; + unsigned int i; + + if (count >= n) { + a->nlimbs = 0; + return; + } + + for (i = 0; i < n - count; i++) + ap[i] = ap[i + count]; + ap[i] = 0; + a->nlimbs -= count; +} diff --git a/lib/mpi/mpi-pow.c b/lib/mpi/mpi-pow.c new file mode 100644 index 00000000000..b04a3cf8008 --- /dev/null +++ b/lib/mpi/mpi-pow.c @@ -0,0 +1,323 @@ +/* mpi-pow.c - MPI functions + * Copyright (C) 1994, 1996, 1998, 2000 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include +#include "mpi-internal.h" +#include "longlong.h" + +/**************** + * RES = BASE ^ EXP mod MOD + */ +int mpi_powm(MPI res, MPI base, MPI exp, MPI mod) +{ + mpi_ptr_t mp_marker = NULL, bp_marker = NULL, ep_marker = NULL; + mpi_ptr_t xp_marker = NULL; + mpi_ptr_t tspace = NULL; + mpi_ptr_t rp, ep, mp, bp; + mpi_size_t esize, msize, bsize, rsize; + int esign, msign, bsign, rsign; + mpi_size_t size; + int mod_shift_cnt; + int negative_result; + int assign_rp = 0; + mpi_size_t tsize = 0; /* to avoid compiler warning */ + /* fixme: we should check that the warning is void */ + int rc = -ENOMEM; + + esize = exp->nlimbs; + msize = mod->nlimbs; + size = 2 * msize; + esign = exp->sign; + msign = mod->sign; + + rp = res->d; + ep = exp->d; + + if (!msize) + msize = 1 / msize; /* provoke a signal */ + + if (!esize) { + /* Exponent is zero, result is 1 mod MOD, i.e., 1 or 0 + * depending on if MOD equals 1. */ + rp[0] = 1; + res->nlimbs = (msize == 1 && mod->d[0] == 1) ? 0 : 1; + res->sign = 0; + goto leave; + } + + /* Normalize MOD (i.e. make its most significant bit set) as required by + * mpn_divrem. This will make the intermediate values in the calculation + * slightly larger, but the correct result is obtained after a final + * reduction using the original MOD value. */ + mp = mp_marker = mpi_alloc_limb_space(msize); + if (!mp) + goto enomem; + count_leading_zeros(mod_shift_cnt, mod->d[msize - 1]); + if (mod_shift_cnt) + mpihelp_lshift(mp, mod->d, msize, mod_shift_cnt); + else + MPN_COPY(mp, mod->d, msize); + + bsize = base->nlimbs; + bsign = base->sign; + if (bsize > msize) { /* The base is larger than the module. Reduce it. */ + /* Allocate (BSIZE + 1) with space for remainder and quotient. + * (The quotient is (bsize - msize + 1) limbs.) */ + bp = bp_marker = mpi_alloc_limb_space(bsize + 1); + if (!bp) + goto enomem; + MPN_COPY(bp, base->d, bsize); + /* We don't care about the quotient, store it above the remainder, + * at BP + MSIZE. */ + mpihelp_divrem(bp + msize, 0, bp, bsize, mp, msize); + bsize = msize; + /* Canonicalize the base, since we are going to multiply with it + * quite a few times. */ + MPN_NORMALIZE(bp, bsize); + } else + bp = base->d; + + if (!bsize) { + res->nlimbs = 0; + res->sign = 0; + goto leave; + } + + if (res->alloced < size) { + /* We have to allocate more space for RES. If any of the input + * parameters are identical to RES, defer deallocation of the old + * space. */ + if (rp == ep || rp == mp || rp == bp) { + rp = mpi_alloc_limb_space(size); + if (!rp) + goto enomem; + assign_rp = 1; + } else { + if (mpi_resize(res, size) < 0) + goto enomem; + rp = res->d; + } + } else { /* Make BASE, EXP and MOD not overlap with RES. */ + if (rp == bp) { + /* RES and BASE are identical. Allocate temp. space for BASE. */ + BUG_ON(bp_marker); + bp = bp_marker = mpi_alloc_limb_space(bsize); + if (!bp) + goto enomem; + MPN_COPY(bp, rp, bsize); + } + if (rp == ep) { + /* RES and EXP are identical. Allocate temp. space for EXP. */ + ep = ep_marker = mpi_alloc_limb_space(esize); + if (!ep) + goto enomem; + MPN_COPY(ep, rp, esize); + } + if (rp == mp) { + /* RES and MOD are identical. Allocate temporary space for MOD. */ + BUG_ON(mp_marker); + mp = mp_marker = mpi_alloc_limb_space(msize); + if (!mp) + goto enomem; + MPN_COPY(mp, rp, msize); + } + } + + MPN_COPY(rp, bp, bsize); + rsize = bsize; + rsign = bsign; + + { + mpi_size_t i; + mpi_ptr_t xp; + int c; + mpi_limb_t e; + mpi_limb_t carry_limb; + struct karatsuba_ctx karactx; + + xp = xp_marker = mpi_alloc_limb_space(2 * (msize + 1)); + if (!xp) + goto enomem; + + memset(&karactx, 0, sizeof karactx); + negative_result = (ep[0] & 1) && base->sign; + + i = esize - 1; + e = ep[i]; + count_leading_zeros(c, e); + e = (e << c) << 1; /* shift the exp bits to the left, lose msb */ + c = BITS_PER_MPI_LIMB - 1 - c; + + /* Main loop. + * + * Make the result be pointed to alternately by XP and RP. This + * helps us avoid block copying, which would otherwise be necessary + * with the overlap restrictions of mpihelp_divmod. With 50% probability + * the result after this loop will be in the area originally pointed + * by RP (==RES->d), and with 50% probability in the area originally + * pointed to by XP. + */ + + for (;;) { + while (c) { + mpi_ptr_t tp; + mpi_size_t xsize; + + /*if (mpihelp_mul_n(xp, rp, rp, rsize) < 0) goto enomem */ + if (rsize < KARATSUBA_THRESHOLD) + mpih_sqr_n_basecase(xp, rp, rsize); + else { + if (!tspace) { + tsize = 2 * rsize; + tspace = + mpi_alloc_limb_space(tsize); + if (!tspace) + goto enomem; + } else if (tsize < (2 * rsize)) { + mpi_free_limb_space(tspace); + tsize = 2 * rsize; + tspace = + mpi_alloc_limb_space(tsize); + if (!tspace) + goto enomem; + } + mpih_sqr_n(xp, rp, rsize, tspace); + } + + xsize = 2 * rsize; + if (xsize > msize) { + mpihelp_divrem(xp + msize, 0, xp, xsize, + mp, msize); + xsize = msize; + } + + tp = rp; + rp = xp; + xp = tp; + rsize = xsize; + + if ((mpi_limb_signed_t) e < 0) { + /*mpihelp_mul( xp, rp, rsize, bp, bsize ); */ + if (bsize < KARATSUBA_THRESHOLD) { + mpi_limb_t tmp; + if (mpihelp_mul + (xp, rp, rsize, bp, bsize, + &tmp) < 0) + goto enomem; + } else { + if (mpihelp_mul_karatsuba_case + (xp, rp, rsize, bp, bsize, + &karactx) < 0) + goto enomem; + } + + xsize = rsize + bsize; + if (xsize > msize) { + mpihelp_divrem(xp + msize, 0, + xp, xsize, mp, + msize); + xsize = msize; + } + + tp = rp; + rp = xp; + xp = tp; + rsize = xsize; + } + e <<= 1; + c--; + } + + i--; + if (i < 0) + break; + e = ep[i]; + c = BITS_PER_MPI_LIMB; + } + + /* We shifted MOD, the modulo reduction argument, left MOD_SHIFT_CNT + * steps. Adjust the result by reducing it with the original MOD. + * + * Also make sure the result is put in RES->d (where it already + * might be, see above). + */ + if (mod_shift_cnt) { + carry_limb = + mpihelp_lshift(res->d, rp, rsize, mod_shift_cnt); + rp = res->d; + if (carry_limb) { + rp[rsize] = carry_limb; + rsize++; + } + } else { + MPN_COPY(res->d, rp, rsize); + rp = res->d; + } + + if (rsize >= msize) { + mpihelp_divrem(rp + msize, 0, rp, rsize, mp, msize); + rsize = msize; + } + + /* Remove any leading zero words from the result. */ + if (mod_shift_cnt) + mpihelp_rshift(rp, rp, rsize, mod_shift_cnt); + MPN_NORMALIZE(rp, rsize); + + mpihelp_release_karatsuba_ctx(&karactx); + } + + if (negative_result && rsize) { + if (mod_shift_cnt) + mpihelp_rshift(mp, mp, msize, mod_shift_cnt); + mpihelp_sub(rp, mp, msize, rp, rsize); + rsize = msize; + rsign = msign; + MPN_NORMALIZE(rp, rsize); + } + res->nlimbs = rsize; + res->sign = rsign; + +leave: + rc = 0; +enomem: + if (assign_rp) + mpi_assign_limb_space(res, rp, size); + if (mp_marker) + mpi_free_limb_space(mp_marker); + if (bp_marker) + mpi_free_limb_space(bp_marker); + if (ep_marker) + mpi_free_limb_space(ep_marker); + if (xp_marker) + mpi_free_limb_space(xp_marker); + if (tspace) + mpi_free_limb_space(tspace); + return rc; +} +EXPORT_SYMBOL_GPL(mpi_powm); diff --git a/lib/mpi/mpicoder.c b/lib/mpi/mpicoder.c new file mode 100644 index 00000000000..fe84bb978e3 --- /dev/null +++ b/lib/mpi/mpicoder.c @@ -0,0 +1,365 @@ +/* mpicoder.c - Coder for the external representation of MPIs + * Copyright (C) 1998, 1999 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +#include "mpi-internal.h" + +#define DIM(v) (sizeof(v)/sizeof((v)[0])) +#define MAX_EXTERN_MPI_BITS 16384 + +static uint8_t asn[15] = /* Object ID is 1.3.14.3.2.26 */ +{ 0x30, 0x21, 0x30, 0x09, 0x06, 0x05, 0x2b, 0x0e, 0x03, + 0x02, 0x1a, 0x05, 0x00, 0x04, 0x14 +}; + +MPI do_encode_md(const void *sha_buffer, unsigned nbits) +{ + int nframe = (nbits + 7) / 8; + uint8_t *frame, *fr_pt; + int i = 0, n; + size_t asnlen = DIM(asn); + MPI a = MPI_NULL; + + if (SHA1_DIGEST_LENGTH + asnlen + 4 > nframe) + pr_info("MPI: can't encode a %d bit MD into a %d bits frame\n", + (int)(SHA1_DIGEST_LENGTH * 8), (int)nbits); + + /* We encode the MD in this way: + * + * 0 A PAD(n bytes) 0 ASN(asnlen bytes) MD(len bytes) + * + * PAD consists of FF bytes. + */ + frame = kmalloc(nframe, GFP_KERNEL); + if (!frame) + return MPI_NULL; + n = 0; + frame[n++] = 0; + frame[n++] = 1; /* block type */ + i = nframe - SHA1_DIGEST_LENGTH - asnlen - 3; + + if (i <= 1) { + pr_info("MPI: message digest encoding failed\n"); + kfree(frame); + return a; + } + + memset(frame + n, 0xff, i); + n += i; + frame[n++] = 0; + memcpy(frame + n, &asn, asnlen); + n += asnlen; + memcpy(frame + n, sha_buffer, SHA1_DIGEST_LENGTH); + n += SHA1_DIGEST_LENGTH; + + i = nframe; + fr_pt = frame; + + if (n != nframe) { + printk + ("MPI: message digest encoding failed, frame length is wrong\n"); + kfree(frame); + return a; + } + + a = mpi_alloc((nframe + BYTES_PER_MPI_LIMB - 1) / BYTES_PER_MPI_LIMB); + mpi_set_buffer(a, frame, nframe, 0); + kfree(frame); + + return a; +} + +MPI mpi_read_from_buffer(const void *xbuffer, unsigned *ret_nread) +{ + const uint8_t *buffer = xbuffer; + int i, j; + unsigned nbits, nbytes, nlimbs, nread = 0; + mpi_limb_t a; + MPI val = MPI_NULL; + + if (*ret_nread < 2) + goto leave; + nbits = buffer[0] << 8 | buffer[1]; + + if (nbits > MAX_EXTERN_MPI_BITS) { + pr_info("MPI: mpi too large (%u bits)\n", nbits); + goto leave; + } + buffer += 2; + nread = 2; + + nbytes = (nbits + 7) / 8; + nlimbs = (nbytes + BYTES_PER_MPI_LIMB - 1) / BYTES_PER_MPI_LIMB; + val = mpi_alloc(nlimbs); + if (!val) + return MPI_NULL; + i = BYTES_PER_MPI_LIMB - nbytes % BYTES_PER_MPI_LIMB; + i %= BYTES_PER_MPI_LIMB; + val->nbits = nbits; + j = val->nlimbs = nlimbs; + val->sign = 0; + for (; j > 0; j--) { + a = 0; + for (; i < BYTES_PER_MPI_LIMB; i++) { + if (++nread > *ret_nread) { + printk + ("MPI: mpi larger than buffer nread=%d ret_nread=%d\n", + nread, *ret_nread); + goto leave; + } + a <<= 8; + a |= *buffer++; + } + i = 0; + val->d[j - 1] = a; + } + +leave: + *ret_nread = nread; + return val; +} +EXPORT_SYMBOL_GPL(mpi_read_from_buffer); + +/**************** + * Make an mpi from a character string. + */ +int mpi_fromstr(MPI val, const char *str) +{ + int hexmode = 0, sign = 0, prepend_zero = 0, i, j, c, c1, c2; + unsigned nbits, nbytes, nlimbs; + mpi_limb_t a; + + if (*str == '-') { + sign = 1; + str++; + } + if (*str == '0' && str[1] == 'x') + hexmode = 1; + else + return -EINVAL; /* other bases are not yet supported */ + str += 2; + + nbits = strlen(str) * 4; + if (nbits % 8) + prepend_zero = 1; + nbytes = (nbits + 7) / 8; + nlimbs = (nbytes + BYTES_PER_MPI_LIMB - 1) / BYTES_PER_MPI_LIMB; + if (val->alloced < nlimbs) + if (!mpi_resize(val, nlimbs)) + return -ENOMEM; + i = BYTES_PER_MPI_LIMB - nbytes % BYTES_PER_MPI_LIMB; + i %= BYTES_PER_MPI_LIMB; + j = val->nlimbs = nlimbs; + val->sign = sign; + for (; j > 0; j--) { + a = 0; + for (; i < BYTES_PER_MPI_LIMB; i++) { + if (prepend_zero) { + c1 = '0'; + prepend_zero = 0; + } else + c1 = *str++; + assert(c1); + c2 = *str++; + assert(c2); + if (c1 >= '0' && c1 <= '9') + c = c1 - '0'; + else if (c1 >= 'a' && c1 <= 'f') + c = c1 - 'a' + 10; + else if (c1 >= 'A' && c1 <= 'F') + c = c1 - 'A' + 10; + else { + mpi_clear(val); + return 1; + } + c <<= 4; + if (c2 >= '0' && c2 <= '9') + c |= c2 - '0'; + else if (c2 >= 'a' && c2 <= 'f') + c |= c2 - 'a' + 10; + else if (c2 >= 'A' && c2 <= 'F') + c |= c2 - 'A' + 10; + else { + mpi_clear(val); + return 1; + } + a <<= 8; + a |= c; + } + i = 0; + + val->d[j - 1] = a; + } + + return 0; +} +EXPORT_SYMBOL_GPL(mpi_fromstr); + +/**************** + * Special function to get the low 8 bytes from an mpi. + * This can be used as a keyid; KEYID is an 2 element array. + * Return the low 4 bytes. + */ +u32 mpi_get_keyid(const MPI a, u32 *keyid) +{ +#if BYTES_PER_MPI_LIMB == 4 + if (keyid) { + keyid[0] = a->nlimbs >= 2 ? a->d[1] : 0; + keyid[1] = a->nlimbs >= 1 ? a->d[0] : 0; + } + return a->nlimbs >= 1 ? a->d[0] : 0; +#elif BYTES_PER_MPI_LIMB == 8 + if (keyid) { + keyid[0] = a->nlimbs ? (u32) (a->d[0] >> 32) : 0; + keyid[1] = a->nlimbs ? (u32) (a->d[0] & 0xffffffff) : 0; + } + return a->nlimbs ? (u32) (a->d[0] & 0xffffffff) : 0; +#else +#error Make this function work with other LIMB sizes +#endif +} + +/**************** + * Return an allocated buffer with the MPI (msb first). + * NBYTES receives the length of this buffer. Caller must free the + * return string (This function does return a 0 byte buffer with NBYTES + * set to zero if the value of A is zero. If sign is not NULL, it will + * be set to the sign of the A. + */ +void *mpi_get_buffer(MPI a, unsigned *nbytes, int *sign) +{ + uint8_t *p, *buffer; + mpi_limb_t alimb; + int i; + unsigned int n; + + if (sign) + *sign = a->sign; + *nbytes = n = a->nlimbs * BYTES_PER_MPI_LIMB; + if (!n) + n++; /* avoid zero length allocation */ + p = buffer = kmalloc(n, GFP_KERNEL); + + for (i = a->nlimbs - 1; i >= 0; i--) { + alimb = a->d[i]; +#if BYTES_PER_MPI_LIMB == 4 + *p++ = alimb >> 24; + *p++ = alimb >> 16; + *p++ = alimb >> 8; + *p++ = alimb; +#elif BYTES_PER_MPI_LIMB == 8 + *p++ = alimb >> 56; + *p++ = alimb >> 48; + *p++ = alimb >> 40; + *p++ = alimb >> 32; + *p++ = alimb >> 24; + *p++ = alimb >> 16; + *p++ = alimb >> 8; + *p++ = alimb; +#else +#error please implement for this limb size. +#endif + } + + /* this is sub-optimal but we need to do the shift operation + * because the caller has to free the returned buffer */ + for (p = buffer; !*p && *nbytes; p++, --*nbytes) + ; + if (p != buffer) + memmove(buffer, p, *nbytes); + + return buffer; +} +EXPORT_SYMBOL_GPL(mpi_get_buffer); + +/**************** + * Use BUFFER to update MPI. + */ +int mpi_set_buffer(MPI a, const void *xbuffer, unsigned nbytes, int sign) +{ + const uint8_t *buffer = xbuffer, *p; + mpi_limb_t alimb; + int nlimbs; + int i; + + nlimbs = (nbytes + BYTES_PER_MPI_LIMB - 1) / BYTES_PER_MPI_LIMB; + if (RESIZE_IF_NEEDED(a, nlimbs) < 0) + return -ENOMEM; + a->sign = sign; + + for (i = 0, p = buffer + nbytes - 1; p >= buffer + BYTES_PER_MPI_LIMB;) { +#if BYTES_PER_MPI_LIMB == 4 + alimb = (mpi_limb_t) *p--; + alimb |= (mpi_limb_t) *p-- << 8; + alimb |= (mpi_limb_t) *p-- << 16; + alimb |= (mpi_limb_t) *p-- << 24; +#elif BYTES_PER_MPI_LIMB == 8 + alimb = (mpi_limb_t) *p--; + alimb |= (mpi_limb_t) *p-- << 8; + alimb |= (mpi_limb_t) *p-- << 16; + alimb |= (mpi_limb_t) *p-- << 24; + alimb |= (mpi_limb_t) *p-- << 32; + alimb |= (mpi_limb_t) *p-- << 40; + alimb |= (mpi_limb_t) *p-- << 48; + alimb |= (mpi_limb_t) *p-- << 56; +#else +#error please implement for this limb size. +#endif + a->d[i++] = alimb; + } + if (p >= buffer) { +#if BYTES_PER_MPI_LIMB == 4 + alimb = *p--; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 8; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 16; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 24; +#elif BYTES_PER_MPI_LIMB == 8 + alimb = (mpi_limb_t) *p--; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 8; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 16; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 24; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 32; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 40; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 48; + if (p >= buffer) + alimb |= (mpi_limb_t) *p-- << 56; +#else +#error please implement for this limb size. +#endif + a->d[i++] = alimb; + } + a->nlimbs = i; + + if (i != nlimbs) { + pr_emerg("MPI: mpi_set_buffer: Assertion failed (%d != %d)", i, + nlimbs); + BUG(); + } + return 0; +} +EXPORT_SYMBOL_GPL(mpi_set_buffer); diff --git a/lib/mpi/mpih-cmp.c b/lib/mpi/mpih-cmp.c new file mode 100644 index 00000000000..b2fd39677f1 --- /dev/null +++ b/lib/mpi/mpih-cmp.c @@ -0,0 +1,56 @@ +/* mpihelp-sub.c - MPI helper functions + * Copyright (C) 1994, 1996 Free Software Foundation, Inc. + * Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" + +/**************** + * Compare OP1_PTR/OP1_SIZE with OP2_PTR/OP2_SIZE. + * There are no restrictions on the relative sizes of + * the two arguments. + * Return 1 if OP1 > OP2, 0 if they are equal, and -1 if OP1 < OP2. + */ +int mpihelp_cmp(mpi_ptr_t op1_ptr, mpi_ptr_t op2_ptr, mpi_size_t size) +{ + mpi_size_t i; + mpi_limb_t op1_word, op2_word; + + for (i = size - 1; i >= 0; i--) { + op1_word = op1_ptr[i]; + op2_word = op2_ptr[i]; + if (op1_word != op2_word) + goto diff; + } + return 0; + +diff: + /* This can *not* be simplified to + * op2_word - op2_word + * since that expression might give signed overflow. */ + return (op1_word > op2_word) ? 1 : -1; +} diff --git a/lib/mpi/mpih-div.c b/lib/mpi/mpih-div.c new file mode 100644 index 00000000000..87ede162dfa --- /dev/null +++ b/lib/mpi/mpih-div.c @@ -0,0 +1,541 @@ +/* mpihelp-div.c - MPI helper functions + * Copyright (C) 1994, 1996 Free Software Foundation, Inc. + * Copyright (C) 1998, 1999 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" +#include "longlong.h" + +#ifndef UMUL_TIME +#define UMUL_TIME 1 +#endif +#ifndef UDIV_TIME +#define UDIV_TIME UMUL_TIME +#endif + +/* FIXME: We should be using invert_limb (or invert_normalized_limb) + * here (not udiv_qrnnd). + */ + +mpi_limb_t +mpihelp_mod_1(mpi_ptr_t dividend_ptr, mpi_size_t dividend_size, + mpi_limb_t divisor_limb) +{ + mpi_size_t i; + mpi_limb_t n1, n0, r; + int dummy; + + /* Botch: Should this be handled at all? Rely on callers? */ + if (!dividend_size) + return 0; + + /* If multiplication is much faster than division, and the + * dividend is large, pre-invert the divisor, and use + * only multiplications in the inner loop. + * + * This test should be read: + * Does it ever help to use udiv_qrnnd_preinv? + * && Does what we save compensate for the inversion overhead? + */ + if (UDIV_TIME > (2 * UMUL_TIME + 6) + && (UDIV_TIME - (2 * UMUL_TIME + 6)) * dividend_size > UDIV_TIME) { + int normalization_steps; + + count_leading_zeros(normalization_steps, divisor_limb); + if (normalization_steps) { + mpi_limb_t divisor_limb_inverted; + + divisor_limb <<= normalization_steps; + + /* Compute (2**2N - 2**N * DIVISOR_LIMB) / DIVISOR_LIMB. The + * result is a (N+1)-bit approximation to 1/DIVISOR_LIMB, with the + * most significant bit (with weight 2**N) implicit. + * + * Special case for DIVISOR_LIMB == 100...000. + */ + if (!(divisor_limb << 1)) + divisor_limb_inverted = ~(mpi_limb_t) 0; + else + udiv_qrnnd(divisor_limb_inverted, dummy, + -divisor_limb, 0, divisor_limb); + + n1 = dividend_ptr[dividend_size - 1]; + r = n1 >> (BITS_PER_MPI_LIMB - normalization_steps); + + /* Possible optimization: + * if (r == 0 + * && divisor_limb > ((n1 << normalization_steps) + * | (dividend_ptr[dividend_size - 2] >> ...))) + * ...one division less... + */ + for (i = dividend_size - 2; i >= 0; i--) { + n0 = dividend_ptr[i]; + UDIV_QRNND_PREINV(dummy, r, r, + ((n1 << normalization_steps) + | (n0 >> + (BITS_PER_MPI_LIMB - + normalization_steps))), + divisor_limb, + divisor_limb_inverted); + n1 = n0; + } + UDIV_QRNND_PREINV(dummy, r, r, + n1 << normalization_steps, + divisor_limb, divisor_limb_inverted); + return r >> normalization_steps; + } else { + mpi_limb_t divisor_limb_inverted; + + /* Compute (2**2N - 2**N * DIVISOR_LIMB) / DIVISOR_LIMB. The + * result is a (N+1)-bit approximation to 1/DIVISOR_LIMB, with the + * most significant bit (with weight 2**N) implicit. + * + * Special case for DIVISOR_LIMB == 100...000. + */ + if (!(divisor_limb << 1)) + divisor_limb_inverted = ~(mpi_limb_t) 0; + else + udiv_qrnnd(divisor_limb_inverted, dummy, + -divisor_limb, 0, divisor_limb); + + i = dividend_size - 1; + r = dividend_ptr[i]; + + if (r >= divisor_limb) + r = 0; + else + i--; + + for (; i >= 0; i--) { + n0 = dividend_ptr[i]; + UDIV_QRNND_PREINV(dummy, r, r, + n0, divisor_limb, + divisor_limb_inverted); + } + return r; + } + } else { + if (UDIV_NEEDS_NORMALIZATION) { + int normalization_steps; + + count_leading_zeros(normalization_steps, divisor_limb); + if (normalization_steps) { + divisor_limb <<= normalization_steps; + + n1 = dividend_ptr[dividend_size - 1]; + r = n1 >> (BITS_PER_MPI_LIMB - + normalization_steps); + + /* Possible optimization: + * if (r == 0 + * && divisor_limb > ((n1 << normalization_steps) + * | (dividend_ptr[dividend_size - 2] >> ...))) + * ...one division less... + */ + for (i = dividend_size - 2; i >= 0; i--) { + n0 = dividend_ptr[i]; + udiv_qrnnd(dummy, r, r, + ((n1 << normalization_steps) + | (n0 >> + (BITS_PER_MPI_LIMB - + normalization_steps))), + divisor_limb); + n1 = n0; + } + udiv_qrnnd(dummy, r, r, + n1 << normalization_steps, + divisor_limb); + return r >> normalization_steps; + } + } + /* No normalization needed, either because udiv_qrnnd doesn't require + * it, or because DIVISOR_LIMB is already normalized. */ + i = dividend_size - 1; + r = dividend_ptr[i]; + + if (r >= divisor_limb) + r = 0; + else + i--; + + for (; i >= 0; i--) { + n0 = dividend_ptr[i]; + udiv_qrnnd(dummy, r, r, n0, divisor_limb); + } + return r; + } +} + +/* Divide num (NP/NSIZE) by den (DP/DSIZE) and write + * the NSIZE-DSIZE least significant quotient limbs at QP + * and the DSIZE long remainder at NP. If QEXTRA_LIMBS is + * non-zero, generate that many fraction bits and append them after the + * other quotient limbs. + * Return the most significant limb of the quotient, this is always 0 or 1. + * + * Preconditions: + * 0. NSIZE >= DSIZE. + * 1. The most significant bit of the divisor must be set. + * 2. QP must either not overlap with the input operands at all, or + * QP + DSIZE >= NP must hold true. (This means that it's + * possible to put the quotient in the high part of NUM, right after the + * remainder in NUM. + * 3. NSIZE >= DSIZE, even if QEXTRA_LIMBS is non-zero. + */ + +mpi_limb_t +mpihelp_divrem(mpi_ptr_t qp, mpi_size_t qextra_limbs, + mpi_ptr_t np, mpi_size_t nsize, mpi_ptr_t dp, mpi_size_t dsize) +{ + mpi_limb_t most_significant_q_limb = 0; + + switch (dsize) { + case 0: + /* We are asked to divide by zero, so go ahead and do it! (To make + the compiler not remove this statement, return the value.) */ + return 1 / dsize; + + case 1: + { + mpi_size_t i; + mpi_limb_t n1; + mpi_limb_t d; + + d = dp[0]; + n1 = np[nsize - 1]; + + if (n1 >= d) { + n1 -= d; + most_significant_q_limb = 1; + } + + qp += qextra_limbs; + for (i = nsize - 2; i >= 0; i--) + udiv_qrnnd(qp[i], n1, n1, np[i], d); + qp -= qextra_limbs; + + for (i = qextra_limbs - 1; i >= 0; i--) + udiv_qrnnd(qp[i], n1, n1, 0, d); + + np[0] = n1; + } + break; + + case 2: + { + mpi_size_t i; + mpi_limb_t n1, n0, n2; + mpi_limb_t d1, d0; + + np += nsize - 2; + d1 = dp[1]; + d0 = dp[0]; + n1 = np[1]; + n0 = np[0]; + + if (n1 >= d1 && (n1 > d1 || n0 >= d0)) { + sub_ddmmss(n1, n0, n1, n0, d1, d0); + most_significant_q_limb = 1; + } + + for (i = qextra_limbs + nsize - 2 - 1; i >= 0; i--) { + mpi_limb_t q; + mpi_limb_t r; + + if (i >= qextra_limbs) + np--; + else + np[0] = 0; + + if (n1 == d1) { + /* Q should be either 111..111 or 111..110. Need special + * treatment of this rare case as normal division would + * give overflow. */ + q = ~(mpi_limb_t) 0; + + r = n0 + d1; + if (r < d1) { /* Carry in the addition? */ + add_ssaaaa(n1, n0, r - d0, + np[0], 0, d0); + qp[i] = q; + continue; + } + n1 = d0 - (d0 != 0 ? 1 : 0); + n0 = -d0; + } else { + udiv_qrnnd(q, r, n1, n0, d1); + umul_ppmm(n1, n0, d0, q); + } + + n2 = np[0]; +q_test: + if (n1 > r || (n1 == r && n0 > n2)) { + /* The estimated Q was too large. */ + q--; + sub_ddmmss(n1, n0, n1, n0, 0, d0); + r += d1; + if (r >= d1) /* If not carry, test Q again. */ + goto q_test; + } + + qp[i] = q; + sub_ddmmss(n1, n0, r, n2, n1, n0); + } + np[1] = n1; + np[0] = n0; + } + break; + + default: + { + mpi_size_t i; + mpi_limb_t dX, d1, n0; + + np += nsize - dsize; + dX = dp[dsize - 1]; + d1 = dp[dsize - 2]; + n0 = np[dsize - 1]; + + if (n0 >= dX) { + if (n0 > dX + || mpihelp_cmp(np, dp, dsize - 1) >= 0) { + mpihelp_sub_n(np, np, dp, dsize); + n0 = np[dsize - 1]; + most_significant_q_limb = 1; + } + } + + for (i = qextra_limbs + nsize - dsize - 1; i >= 0; i--) { + mpi_limb_t q; + mpi_limb_t n1, n2; + mpi_limb_t cy_limb; + + if (i >= qextra_limbs) { + np--; + n2 = np[dsize]; + } else { + n2 = np[dsize - 1]; + MPN_COPY_DECR(np + 1, np, dsize - 1); + np[0] = 0; + } + + if (n0 == dX) { + /* This might over-estimate q, but it's probably not worth + * the extra code here to find out. */ + q = ~(mpi_limb_t) 0; + } else { + mpi_limb_t r; + + udiv_qrnnd(q, r, n0, np[dsize - 1], dX); + umul_ppmm(n1, n0, d1, q); + + while (n1 > r + || (n1 == r + && n0 > np[dsize - 2])) { + q--; + r += dX; + if (r < dX) /* I.e. "carry in previous addition?" */ + break; + n1 -= n0 < d1; + n0 -= d1; + } + } + + /* Possible optimization: We already have (q * n0) and (1 * n1) + * after the calculation of q. Taking advantage of that, we + * could make this loop make two iterations less. */ + cy_limb = mpihelp_submul_1(np, dp, dsize, q); + + if (n2 != cy_limb) { + mpihelp_add_n(np, np, dp, dsize); + q--; + } + + qp[i] = q; + n0 = np[dsize - 1]; + } + } + } + + return most_significant_q_limb; +} + +/**************** + * Divide (DIVIDEND_PTR,,DIVIDEND_SIZE) by DIVISOR_LIMB. + * Write DIVIDEND_SIZE limbs of quotient at QUOT_PTR. + * Return the single-limb remainder. + * There are no constraints on the value of the divisor. + * + * QUOT_PTR and DIVIDEND_PTR might point to the same limb. + */ + +mpi_limb_t +mpihelp_divmod_1(mpi_ptr_t quot_ptr, + mpi_ptr_t dividend_ptr, mpi_size_t dividend_size, + mpi_limb_t divisor_limb) +{ + mpi_size_t i; + mpi_limb_t n1, n0, r; + int dummy; + + if (!dividend_size) + return 0; + + /* If multiplication is much faster than division, and the + * dividend is large, pre-invert the divisor, and use + * only multiplications in the inner loop. + * + * This test should be read: + * Does it ever help to use udiv_qrnnd_preinv? + * && Does what we save compensate for the inversion overhead? + */ + if (UDIV_TIME > (2 * UMUL_TIME + 6) + && (UDIV_TIME - (2 * UMUL_TIME + 6)) * dividend_size > UDIV_TIME) { + int normalization_steps; + + count_leading_zeros(normalization_steps, divisor_limb); + if (normalization_steps) { + mpi_limb_t divisor_limb_inverted; + + divisor_limb <<= normalization_steps; + + /* Compute (2**2N - 2**N * DIVISOR_LIMB) / DIVISOR_LIMB. The + * result is a (N+1)-bit approximation to 1/DIVISOR_LIMB, with the + * most significant bit (with weight 2**N) implicit. + */ + /* Special case for DIVISOR_LIMB == 100...000. */ + if (!(divisor_limb << 1)) + divisor_limb_inverted = ~(mpi_limb_t) 0; + else + udiv_qrnnd(divisor_limb_inverted, dummy, + -divisor_limb, 0, divisor_limb); + + n1 = dividend_ptr[dividend_size - 1]; + r = n1 >> (BITS_PER_MPI_LIMB - normalization_steps); + + /* Possible optimization: + * if (r == 0 + * && divisor_limb > ((n1 << normalization_steps) + * | (dividend_ptr[dividend_size - 2] >> ...))) + * ...one division less... + */ + for (i = dividend_size - 2; i >= 0; i--) { + n0 = dividend_ptr[i]; + UDIV_QRNND_PREINV(quot_ptr[i + 1], r, r, + ((n1 << normalization_steps) + | (n0 >> + (BITS_PER_MPI_LIMB - + normalization_steps))), + divisor_limb, + divisor_limb_inverted); + n1 = n0; + } + UDIV_QRNND_PREINV(quot_ptr[0], r, r, + n1 << normalization_steps, + divisor_limb, divisor_limb_inverted); + return r >> normalization_steps; + } else { + mpi_limb_t divisor_limb_inverted; + + /* Compute (2**2N - 2**N * DIVISOR_LIMB) / DIVISOR_LIMB. The + * result is a (N+1)-bit approximation to 1/DIVISOR_LIMB, with the + * most significant bit (with weight 2**N) implicit. + */ + /* Special case for DIVISOR_LIMB == 100...000. */ + if (!(divisor_limb << 1)) + divisor_limb_inverted = ~(mpi_limb_t) 0; + else + udiv_qrnnd(divisor_limb_inverted, dummy, + -divisor_limb, 0, divisor_limb); + + i = dividend_size - 1; + r = dividend_ptr[i]; + + if (r >= divisor_limb) + r = 0; + else + quot_ptr[i--] = 0; + + for (; i >= 0; i--) { + n0 = dividend_ptr[i]; + UDIV_QRNND_PREINV(quot_ptr[i], r, r, + n0, divisor_limb, + divisor_limb_inverted); + } + return r; + } + } else { + if (UDIV_NEEDS_NORMALIZATION) { + int normalization_steps; + + count_leading_zeros(normalization_steps, divisor_limb); + if (normalization_steps) { + divisor_limb <<= normalization_steps; + + n1 = dividend_ptr[dividend_size - 1]; + r = n1 >> (BITS_PER_MPI_LIMB - + normalization_steps); + + /* Possible optimization: + * if (r == 0 + * && divisor_limb > ((n1 << normalization_steps) + * | (dividend_ptr[dividend_size - 2] >> ...))) + * ...one division less... + */ + for (i = dividend_size - 2; i >= 0; i--) { + n0 = dividend_ptr[i]; + udiv_qrnnd(quot_ptr[i + 1], r, r, + ((n1 << normalization_steps) + | (n0 >> + (BITS_PER_MPI_LIMB - + normalization_steps))), + divisor_limb); + n1 = n0; + } + udiv_qrnnd(quot_ptr[0], r, r, + n1 << normalization_steps, + divisor_limb); + return r >> normalization_steps; + } + } + /* No normalization needed, either because udiv_qrnnd doesn't require + * it, or because DIVISOR_LIMB is already normalized. */ + i = dividend_size - 1; + r = dividend_ptr[i]; + + if (r >= divisor_limb) + r = 0; + else + quot_ptr[i--] = 0; + + for (; i >= 0; i--) { + n0 = dividend_ptr[i]; + udiv_qrnnd(quot_ptr[i], r, r, n0, divisor_limb); + } + return r; + } +} diff --git a/lib/mpi/mpih-mul.c b/lib/mpi/mpih-mul.c new file mode 100644 index 00000000000..c69c5eef233 --- /dev/null +++ b/lib/mpi/mpih-mul.c @@ -0,0 +1,527 @@ +/* mpihelp-mul.c - MPI helper functions + * Copyright (C) 1994, 1996, 1998, 1999, + * 2000 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include +#include "mpi-internal.h" +#include "longlong.h" + +#define MPN_MUL_N_RECURSE(prodp, up, vp, size, tspace) \ + do { \ + if ((size) < KARATSUBA_THRESHOLD) \ + mul_n_basecase(prodp, up, vp, size); \ + else \ + mul_n(prodp, up, vp, size, tspace); \ + } while (0); + +#define MPN_SQR_N_RECURSE(prodp, up, size, tspace) \ + do { \ + if ((size) < KARATSUBA_THRESHOLD) \ + mpih_sqr_n_basecase(prodp, up, size); \ + else \ + mpih_sqr_n(prodp, up, size, tspace); \ + } while (0); + +/* Multiply the natural numbers u (pointed to by UP) and v (pointed to by VP), + * both with SIZE limbs, and store the result at PRODP. 2 * SIZE limbs are + * always stored. Return the most significant limb. + * + * Argument constraints: + * 1. PRODP != UP and PRODP != VP, i.e. the destination + * must be distinct from the multiplier and the multiplicand. + * + * + * Handle simple cases with traditional multiplication. + * + * This is the most critical code of multiplication. All multiplies rely + * on this, both small and huge. Small ones arrive here immediately. Huge + * ones arrive here as this is the base case for Karatsuba's recursive + * algorithm below. + */ + +static mpi_limb_t +mul_n_basecase(mpi_ptr_t prodp, mpi_ptr_t up, mpi_ptr_t vp, mpi_size_t size) +{ + mpi_size_t i; + mpi_limb_t cy; + mpi_limb_t v_limb; + + /* Multiply by the first limb in V separately, as the result can be + * stored (not added) to PROD. We also avoid a loop for zeroing. */ + v_limb = vp[0]; + if (v_limb <= 1) { + if (v_limb == 1) + MPN_COPY(prodp, up, size); + else + MPN_ZERO(prodp, size); + cy = 0; + } else + cy = mpihelp_mul_1(prodp, up, size, v_limb); + + prodp[size] = cy; + prodp++; + + /* For each iteration in the outer loop, multiply one limb from + * U with one limb from V, and add it to PROD. */ + for (i = 1; i < size; i++) { + v_limb = vp[i]; + if (v_limb <= 1) { + cy = 0; + if (v_limb == 1) + cy = mpihelp_add_n(prodp, prodp, up, size); + } else + cy = mpihelp_addmul_1(prodp, up, size, v_limb); + + prodp[size] = cy; + prodp++; + } + + return cy; +} + +static void +mul_n(mpi_ptr_t prodp, mpi_ptr_t up, mpi_ptr_t vp, + mpi_size_t size, mpi_ptr_t tspace) +{ + if (size & 1) { + /* The size is odd, and the code below doesn't handle that. + * Multiply the least significant (size - 1) limbs with a recursive + * call, and handle the most significant limb of S1 and S2 + * separately. + * A slightly faster way to do this would be to make the Karatsuba + * code below behave as if the size were even, and let it check for + * odd size in the end. I.e., in essence move this code to the end. + * Doing so would save us a recursive call, and potentially make the + * stack grow a lot less. + */ + mpi_size_t esize = size - 1; /* even size */ + mpi_limb_t cy_limb; + + MPN_MUL_N_RECURSE(prodp, up, vp, esize, tspace); + cy_limb = mpihelp_addmul_1(prodp + esize, up, esize, vp[esize]); + prodp[esize + esize] = cy_limb; + cy_limb = mpihelp_addmul_1(prodp + esize, vp, size, up[esize]); + prodp[esize + size] = cy_limb; + } else { + /* Anatolij Alekseevich Karatsuba's divide-and-conquer algorithm. + * + * Split U in two pieces, U1 and U0, such that + * U = U0 + U1*(B**n), + * and V in V1 and V0, such that + * V = V0 + V1*(B**n). + * + * UV is then computed recursively using the identity + * + * 2n n n n + * UV = (B + B )U V + B (U -U )(V -V ) + (B + 1)U V + * 1 1 1 0 0 1 0 0 + * + * Where B = 2**BITS_PER_MP_LIMB. + */ + mpi_size_t hsize = size >> 1; + mpi_limb_t cy; + int negflg; + + /* Product H. ________________ ________________ + * |_____U1 x V1____||____U0 x V0_____| + * Put result in upper part of PROD and pass low part of TSPACE + * as new TSPACE. + */ + MPN_MUL_N_RECURSE(prodp + size, up + hsize, vp + hsize, hsize, + tspace); + + /* Product M. ________________ + * |_(U1-U0)(V0-V1)_| + */ + if (mpihelp_cmp(up + hsize, up, hsize) >= 0) { + mpihelp_sub_n(prodp, up + hsize, up, hsize); + negflg = 0; + } else { + mpihelp_sub_n(prodp, up, up + hsize, hsize); + negflg = 1; + } + if (mpihelp_cmp(vp + hsize, vp, hsize) >= 0) { + mpihelp_sub_n(prodp + hsize, vp + hsize, vp, hsize); + negflg ^= 1; + } else { + mpihelp_sub_n(prodp + hsize, vp, vp + hsize, hsize); + /* No change of NEGFLG. */ + } + /* Read temporary operands from low part of PROD. + * Put result in low part of TSPACE using upper part of TSPACE + * as new TSPACE. + */ + MPN_MUL_N_RECURSE(tspace, prodp, prodp + hsize, hsize, + tspace + size); + + /* Add/copy product H. */ + MPN_COPY(prodp + hsize, prodp + size, hsize); + cy = mpihelp_add_n(prodp + size, prodp + size, + prodp + size + hsize, hsize); + + /* Add product M (if NEGFLG M is a negative number) */ + if (negflg) + cy -= + mpihelp_sub_n(prodp + hsize, prodp + hsize, tspace, + size); + else + cy += + mpihelp_add_n(prodp + hsize, prodp + hsize, tspace, + size); + + /* Product L. ________________ ________________ + * |________________||____U0 x V0_____| + * Read temporary operands from low part of PROD. + * Put result in low part of TSPACE using upper part of TSPACE + * as new TSPACE. + */ + MPN_MUL_N_RECURSE(tspace, up, vp, hsize, tspace + size); + + /* Add/copy Product L (twice) */ + + cy += mpihelp_add_n(prodp + hsize, prodp + hsize, tspace, size); + if (cy) + mpihelp_add_1(prodp + hsize + size, + prodp + hsize + size, hsize, cy); + + MPN_COPY(prodp, tspace, hsize); + cy = mpihelp_add_n(prodp + hsize, prodp + hsize, tspace + hsize, + hsize); + if (cy) + mpihelp_add_1(prodp + size, prodp + size, size, 1); + } +} + +void mpih_sqr_n_basecase(mpi_ptr_t prodp, mpi_ptr_t up, mpi_size_t size) +{ + mpi_size_t i; + mpi_limb_t cy_limb; + mpi_limb_t v_limb; + + /* Multiply by the first limb in V separately, as the result can be + * stored (not added) to PROD. We also avoid a loop for zeroing. */ + v_limb = up[0]; + if (v_limb <= 1) { + if (v_limb == 1) + MPN_COPY(prodp, up, size); + else + MPN_ZERO(prodp, size); + cy_limb = 0; + } else + cy_limb = mpihelp_mul_1(prodp, up, size, v_limb); + + prodp[size] = cy_limb; + prodp++; + + /* For each iteration in the outer loop, multiply one limb from + * U with one limb from V, and add it to PROD. */ + for (i = 1; i < size; i++) { + v_limb = up[i]; + if (v_limb <= 1) { + cy_limb = 0; + if (v_limb == 1) + cy_limb = mpihelp_add_n(prodp, prodp, up, size); + } else + cy_limb = mpihelp_addmul_1(prodp, up, size, v_limb); + + prodp[size] = cy_limb; + prodp++; + } +} + +void +mpih_sqr_n(mpi_ptr_t prodp, mpi_ptr_t up, mpi_size_t size, mpi_ptr_t tspace) +{ + if (size & 1) { + /* The size is odd, and the code below doesn't handle that. + * Multiply the least significant (size - 1) limbs with a recursive + * call, and handle the most significant limb of S1 and S2 + * separately. + * A slightly faster way to do this would be to make the Karatsuba + * code below behave as if the size were even, and let it check for + * odd size in the end. I.e., in essence move this code to the end. + * Doing so would save us a recursive call, and potentially make the + * stack grow a lot less. + */ + mpi_size_t esize = size - 1; /* even size */ + mpi_limb_t cy_limb; + + MPN_SQR_N_RECURSE(prodp, up, esize, tspace); + cy_limb = mpihelp_addmul_1(prodp + esize, up, esize, up[esize]); + prodp[esize + esize] = cy_limb; + cy_limb = mpihelp_addmul_1(prodp + esize, up, size, up[esize]); + + prodp[esize + size] = cy_limb; + } else { + mpi_size_t hsize = size >> 1; + mpi_limb_t cy; + + /* Product H. ________________ ________________ + * |_____U1 x U1____||____U0 x U0_____| + * Put result in upper part of PROD and pass low part of TSPACE + * as new TSPACE. + */ + MPN_SQR_N_RECURSE(prodp + size, up + hsize, hsize, tspace); + + /* Product M. ________________ + * |_(U1-U0)(U0-U1)_| + */ + if (mpihelp_cmp(up + hsize, up, hsize) >= 0) + mpihelp_sub_n(prodp, up + hsize, up, hsize); + else + mpihelp_sub_n(prodp, up, up + hsize, hsize); + + /* Read temporary operands from low part of PROD. + * Put result in low part of TSPACE using upper part of TSPACE + * as new TSPACE. */ + MPN_SQR_N_RECURSE(tspace, prodp, hsize, tspace + size); + + /* Add/copy product H */ + MPN_COPY(prodp + hsize, prodp + size, hsize); + cy = mpihelp_add_n(prodp + size, prodp + size, + prodp + size + hsize, hsize); + + /* Add product M (if NEGFLG M is a negative number). */ + cy -= mpihelp_sub_n(prodp + hsize, prodp + hsize, tspace, size); + + /* Product L. ________________ ________________ + * |________________||____U0 x U0_____| + * Read temporary operands from low part of PROD. + * Put result in low part of TSPACE using upper part of TSPACE + * as new TSPACE. */ + MPN_SQR_N_RECURSE(tspace, up, hsize, tspace + size); + + /* Add/copy Product L (twice). */ + cy += mpihelp_add_n(prodp + hsize, prodp + hsize, tspace, size); + if (cy) + mpihelp_add_1(prodp + hsize + size, + prodp + hsize + size, hsize, cy); + + MPN_COPY(prodp, tspace, hsize); + cy = mpihelp_add_n(prodp + hsize, prodp + hsize, tspace + hsize, + hsize); + if (cy) + mpihelp_add_1(prodp + size, prodp + size, size, 1); + } +} + +/* This should be made into an inline function in gmp.h. */ +int mpihelp_mul_n(mpi_ptr_t prodp, mpi_ptr_t up, mpi_ptr_t vp, mpi_size_t size) +{ + if (up == vp) { + if (size < KARATSUBA_THRESHOLD) + mpih_sqr_n_basecase(prodp, up, size); + else { + mpi_ptr_t tspace; + tspace = mpi_alloc_limb_space(2 * size); + if (!tspace) + return -ENOMEM; + mpih_sqr_n(prodp, up, size, tspace); + mpi_free_limb_space(tspace); + } + } else { + if (size < KARATSUBA_THRESHOLD) + mul_n_basecase(prodp, up, vp, size); + else { + mpi_ptr_t tspace; + tspace = mpi_alloc_limb_space(2 * size); + if (!tspace) + return -ENOMEM; + mul_n(prodp, up, vp, size, tspace); + mpi_free_limb_space(tspace); + } + } + + return 0; +} + +int +mpihelp_mul_karatsuba_case(mpi_ptr_t prodp, + mpi_ptr_t up, mpi_size_t usize, + mpi_ptr_t vp, mpi_size_t vsize, + struct karatsuba_ctx *ctx) +{ + mpi_limb_t cy; + + if (!ctx->tspace || ctx->tspace_size < vsize) { + if (ctx->tspace) + mpi_free_limb_space(ctx->tspace); + ctx->tspace = mpi_alloc_limb_space(2 * vsize); + if (!ctx->tspace) + return -ENOMEM; + ctx->tspace_size = vsize; + } + + MPN_MUL_N_RECURSE(prodp, up, vp, vsize, ctx->tspace); + + prodp += vsize; + up += vsize; + usize -= vsize; + if (usize >= vsize) { + if (!ctx->tp || ctx->tp_size < vsize) { + if (ctx->tp) + mpi_free_limb_space(ctx->tp); + ctx->tp = mpi_alloc_limb_space(2 * vsize); + if (!ctx->tp) { + if (ctx->tspace) + mpi_free_limb_space(ctx->tspace); + ctx->tspace = NULL; + return -ENOMEM; + } + ctx->tp_size = vsize; + } + + do { + MPN_MUL_N_RECURSE(ctx->tp, up, vp, vsize, ctx->tspace); + cy = mpihelp_add_n(prodp, prodp, ctx->tp, vsize); + mpihelp_add_1(prodp + vsize, ctx->tp + vsize, vsize, + cy); + prodp += vsize; + up += vsize; + usize -= vsize; + } while (usize >= vsize); + } + + if (usize) { + if (usize < KARATSUBA_THRESHOLD) { + mpi_limb_t tmp; + if (mpihelp_mul(ctx->tspace, vp, vsize, up, usize, &tmp) + < 0) + return -ENOMEM; + } else { + if (!ctx->next) { + ctx->next = kzalloc(sizeof *ctx, GFP_KERNEL); + if (!ctx->next) + return -ENOMEM; + } + if (mpihelp_mul_karatsuba_case(ctx->tspace, + vp, vsize, + up, usize, + ctx->next) < 0) + return -ENOMEM; + } + + cy = mpihelp_add_n(prodp, prodp, ctx->tspace, vsize); + mpihelp_add_1(prodp + vsize, ctx->tspace + vsize, usize, cy); + } + + return 0; +} + +void mpihelp_release_karatsuba_ctx(struct karatsuba_ctx *ctx) +{ + struct karatsuba_ctx *ctx2; + + if (ctx->tp) + mpi_free_limb_space(ctx->tp); + if (ctx->tspace) + mpi_free_limb_space(ctx->tspace); + for (ctx = ctx->next; ctx; ctx = ctx2) { + ctx2 = ctx->next; + if (ctx->tp) + mpi_free_limb_space(ctx->tp); + if (ctx->tspace) + mpi_free_limb_space(ctx->tspace); + kfree(ctx); + } +} + +/* Multiply the natural numbers u (pointed to by UP, with USIZE limbs) + * and v (pointed to by VP, with VSIZE limbs), and store the result at + * PRODP. USIZE + VSIZE limbs are always stored, but if the input + * operands are normalized. Return the most significant limb of the + * result. + * + * NOTE: The space pointed to by PRODP is overwritten before finished + * with U and V, so overlap is an error. + * + * Argument constraints: + * 1. USIZE >= VSIZE. + * 2. PRODP != UP and PRODP != VP, i.e. the destination + * must be distinct from the multiplier and the multiplicand. + */ + +int +mpihelp_mul(mpi_ptr_t prodp, mpi_ptr_t up, mpi_size_t usize, + mpi_ptr_t vp, mpi_size_t vsize, mpi_limb_t *_result) +{ + mpi_ptr_t prod_endp = prodp + usize + vsize - 1; + mpi_limb_t cy; + struct karatsuba_ctx ctx; + + if (vsize < KARATSUBA_THRESHOLD) { + mpi_size_t i; + mpi_limb_t v_limb; + + if (!vsize) { + *_result = 0; + return 0; + } + + /* Multiply by the first limb in V separately, as the result can be + * stored (not added) to PROD. We also avoid a loop for zeroing. */ + v_limb = vp[0]; + if (v_limb <= 1) { + if (v_limb == 1) + MPN_COPY(prodp, up, usize); + else + MPN_ZERO(prodp, usize); + cy = 0; + } else + cy = mpihelp_mul_1(prodp, up, usize, v_limb); + + prodp[usize] = cy; + prodp++; + + /* For each iteration in the outer loop, multiply one limb from + * U with one limb from V, and add it to PROD. */ + for (i = 1; i < vsize; i++) { + v_limb = vp[i]; + if (v_limb <= 1) { + cy = 0; + if (v_limb == 1) + cy = mpihelp_add_n(prodp, prodp, up, + usize); + } else + cy = mpihelp_addmul_1(prodp, up, usize, v_limb); + + prodp[usize] = cy; + prodp++; + } + + *_result = cy; + return 0; + } + + memset(&ctx, 0, sizeof ctx); + if (mpihelp_mul_karatsuba_case(prodp, up, usize, vp, vsize, &ctx) < 0) + return -ENOMEM; + mpihelp_release_karatsuba_ctx(&ctx); + *_result = *prod_endp; + return 0; +} diff --git a/lib/mpi/mpiutil.c b/lib/mpi/mpiutil.c new file mode 100644 index 00000000000..eefc55d6b7f --- /dev/null +++ b/lib/mpi/mpiutil.c @@ -0,0 +1,208 @@ +/* mpiutil.ac - Utility functions for MPI + * Copyright (C) 1998, 1999 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +#include "mpi-internal.h" + +/**************** + * Note: It was a bad idea to use the number of limbs to allocate + * because on a alpha the limbs are large but we normally need + * integers of n bits - So we should chnage this to bits (or bytes). + * + * But mpi_alloc is used in a lot of places :-) + */ +MPI mpi_alloc(unsigned nlimbs) +{ + MPI a; + + a = kmalloc(sizeof *a, GFP_KERNEL); + if (!a) + return a; + + if (nlimbs) { + a->d = mpi_alloc_limb_space(nlimbs); + if (!a->d) { + kfree(a); + return NULL; + } + } else { + a->d = NULL; + } + + a->alloced = nlimbs; + a->nlimbs = 0; + a->sign = 0; + a->flags = 0; + a->nbits = 0; + return a; +} +EXPORT_SYMBOL_GPL(mpi_alloc); + +mpi_ptr_t mpi_alloc_limb_space(unsigned nlimbs) +{ + size_t len = nlimbs * sizeof(mpi_limb_t); + + return kmalloc(len, GFP_KERNEL); +} + +void mpi_free_limb_space(mpi_ptr_t a) +{ + if (!a) + return; + + kfree(a); +} + +void mpi_assign_limb_space(MPI a, mpi_ptr_t ap, unsigned nlimbs) +{ + mpi_free_limb_space(a->d); + a->d = ap; + a->alloced = nlimbs; +} + +/**************** + * Resize the array of A to NLIMBS. the additional space is cleared + * (set to 0) [done by m_realloc()] + */ +int mpi_resize(MPI a, unsigned nlimbs) +{ + void *p; + + if (nlimbs <= a->alloced) + return 0; /* no need to do it */ + + if (a->d) { + p = kmalloc(nlimbs * sizeof(mpi_limb_t), GFP_KERNEL); + if (!p) + return -ENOMEM; + memcpy(p, a->d, a->alloced * sizeof(mpi_limb_t)); + kfree(a->d); + a->d = p; + } else { + a->d = kzalloc(nlimbs * sizeof(mpi_limb_t), GFP_KERNEL); + if (!a->d) + return -ENOMEM; + } + a->alloced = nlimbs; + return 0; +} + +void mpi_clear(MPI a) +{ + a->nlimbs = 0; + a->nbits = 0; + a->flags = 0; +} + +void mpi_free(MPI a) +{ + if (!a) + return; + + if (a->flags & 4) + kfree(a->d); + else + mpi_free_limb_space(a->d); + + if (a->flags & ~7) + pr_info("invalid flag value in mpi\n"); + kfree(a); +} +EXPORT_SYMBOL_GPL(mpi_free); + +/**************** + * Note: This copy function should not interpret the MPI + * but copy it transparently. + */ +int mpi_copy(MPI *copied, const MPI a) +{ + size_t i; + MPI b; + + *copied = MPI_NULL; + + if (a) { + b = mpi_alloc(a->nlimbs); + if (!b) + return -ENOMEM; + + b->nlimbs = a->nlimbs; + b->sign = a->sign; + b->flags = a->flags; + b->nbits = a->nbits; + + for (i = 0; i < b->nlimbs; i++) + b->d[i] = a->d[i]; + + *copied = b; + } + + return 0; +} + +int mpi_set(MPI w, const MPI u) +{ + mpi_ptr_t wp, up; + mpi_size_t usize = u->nlimbs; + int usign = u->sign; + + if (RESIZE_IF_NEEDED(w, (size_t) usize) < 0) + return -ENOMEM; + + wp = w->d; + up = u->d; + MPN_COPY(wp, up, usize); + w->nlimbs = usize; + w->nbits = u->nbits; + w->flags = u->flags; + w->sign = usign; + return 0; +} + +int mpi_set_ui(MPI w, unsigned long u) +{ + if (RESIZE_IF_NEEDED(w, 1) < 0) + return -ENOMEM; + w->d[0] = u; + w->nlimbs = u ? 1 : 0; + w->sign = 0; + w->nbits = 0; + w->flags = 0; + return 0; +} + +MPI mpi_alloc_set_ui(unsigned long u) +{ + MPI w = mpi_alloc(1); + if (!w) + return w; + w->d[0] = u; + w->nlimbs = u ? 1 : 0; + w->sign = 0; + return w; +} + +void mpi_swap(MPI a, MPI b) +{ + struct gcry_mpi tmp; + + tmp = *a; + *a = *b; + *b = tmp; +} -- cgit v1.2.3-70-g09d2 From 5ce3e312ec5c11abce13215be70700778bf601f0 Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Wed, 31 Aug 2011 14:05:16 +0300 Subject: crypto: GnuPG based MPI lib - header files (part 2) Adds the multi-precision-integer maths library which was originally taken from GnuPG and ported to the kernel by (among others) David Howells. This version is taken from Fedora kernel 2.6.32-71.14.1.el6. The difference is that checkpatch reported errors and warnings have been fixed. This library is used to implemenet RSA digital signature verification used in IMA/EVM integrity protection subsystem. Due to patch size limitation, the patch is divided into 4 parts. Signed-off-by: Dmitry Kasatkin --- include/linux/mpi.h | 146 +++++ lib/mpi/longlong.h | 1478 ++++++++++++++++++++++++++++++++++++++++++++++++ lib/mpi/mpi-inline.h | 122 ++++ lib/mpi/mpi-internal.h | 261 +++++++++ 4 files changed, 2007 insertions(+) create mode 100644 include/linux/mpi.h create mode 100644 lib/mpi/longlong.h create mode 100644 lib/mpi/mpi-inline.h create mode 100644 lib/mpi/mpi-internal.h (limited to 'lib') diff --git a/include/linux/mpi.h b/include/linux/mpi.h new file mode 100644 index 00000000000..06f88994cca --- /dev/null +++ b/include/linux/mpi.h @@ -0,0 +1,146 @@ +/* mpi.h - Multi Precision Integers + * Copyright (C) 1994, 1996, 1998, 1999, + * 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GNUPG. + * + * GNUPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GNUPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#ifndef G10_MPI_H +#define G10_MPI_H + +#include + +/* DSI defines */ + +#define SHA1_DIGEST_LENGTH 20 + +/*end of DSI defines */ + +#define BYTES_PER_MPI_LIMB (BITS_PER_LONG / 8) +#define BITS_PER_MPI_LIMB BITS_PER_LONG + +typedef unsigned long int mpi_limb_t; +typedef signed long int mpi_limb_signed_t; + +struct gcry_mpi { + int alloced; /* array size (# of allocated limbs) */ + int nlimbs; /* number of valid limbs */ + int nbits; /* the real number of valid bits (info only) */ + int sign; /* indicates a negative number */ + unsigned flags; /* bit 0: array must be allocated in secure memory space */ + /* bit 1: not used */ + /* bit 2: the limb is a pointer to some m_alloced data */ + mpi_limb_t *d; /* array with the limbs */ +}; + +typedef struct gcry_mpi *MPI; + +#define MPI_NULL NULL + +#define mpi_get_nlimbs(a) ((a)->nlimbs) +#define mpi_is_neg(a) ((a)->sign) + +/*-- mpiutil.c --*/ +MPI mpi_alloc(unsigned nlimbs); +MPI mpi_alloc_secure(unsigned nlimbs); +MPI mpi_alloc_like(MPI a); +void mpi_free(MPI a); +int mpi_resize(MPI a, unsigned nlimbs); +int mpi_copy(MPI *copy, const MPI a); +void mpi_clear(MPI a); +int mpi_set(MPI w, MPI u); +int mpi_set_ui(MPI w, ulong u); +MPI mpi_alloc_set_ui(unsigned long u); +void mpi_m_check(MPI a); +void mpi_swap(MPI a, MPI b); + +/*-- mpicoder.c --*/ +MPI do_encode_md(const void *sha_buffer, unsigned nbits); +MPI mpi_read_from_buffer(const void *buffer, unsigned *ret_nread); +int mpi_fromstr(MPI val, const char *str); +u32 mpi_get_keyid(MPI a, u32 *keyid); +void *mpi_get_buffer(MPI a, unsigned *nbytes, int *sign); +void *mpi_get_secure_buffer(MPI a, unsigned *nbytes, int *sign); +int mpi_set_buffer(MPI a, const void *buffer, unsigned nbytes, int sign); + +#define log_mpidump g10_log_mpidump + +/*-- mpi-add.c --*/ +int mpi_add_ui(MPI w, MPI u, ulong v); +int mpi_add(MPI w, MPI u, MPI v); +int mpi_addm(MPI w, MPI u, MPI v, MPI m); +int mpi_sub_ui(MPI w, MPI u, ulong v); +int mpi_sub(MPI w, MPI u, MPI v); +int mpi_subm(MPI w, MPI u, MPI v, MPI m); + +/*-- mpi-mul.c --*/ +int mpi_mul_ui(MPI w, MPI u, ulong v); +int mpi_mul_2exp(MPI w, MPI u, ulong cnt); +int mpi_mul(MPI w, MPI u, MPI v); +int mpi_mulm(MPI w, MPI u, MPI v, MPI m); + +/*-- mpi-div.c --*/ +ulong mpi_fdiv_r_ui(MPI rem, MPI dividend, ulong divisor); +int mpi_fdiv_r(MPI rem, MPI dividend, MPI divisor); +int mpi_fdiv_q(MPI quot, MPI dividend, MPI divisor); +int mpi_fdiv_qr(MPI quot, MPI rem, MPI dividend, MPI divisor); +int mpi_tdiv_r(MPI rem, MPI num, MPI den); +int mpi_tdiv_qr(MPI quot, MPI rem, MPI num, MPI den); +int mpi_tdiv_q_2exp(MPI w, MPI u, unsigned count); +int mpi_divisible_ui(const MPI dividend, ulong divisor); + +/*-- mpi-gcd.c --*/ +int mpi_gcd(MPI g, const MPI a, const MPI b); + +/*-- mpi-pow.c --*/ +int mpi_pow(MPI w, MPI u, MPI v); +int mpi_powm(MPI res, MPI base, MPI exp, MPI mod); + +/*-- mpi-mpow.c --*/ +int mpi_mulpowm(MPI res, MPI *basearray, MPI *exparray, MPI mod); + +/*-- mpi-cmp.c --*/ +int mpi_cmp_ui(MPI u, ulong v); +int mpi_cmp(MPI u, MPI v); + +/*-- mpi-scan.c --*/ +int mpi_getbyte(MPI a, unsigned idx); +void mpi_putbyte(MPI a, unsigned idx, int value); +unsigned mpi_trailing_zeros(MPI a); + +/*-- mpi-bit.c --*/ +void mpi_normalize(MPI a); +unsigned mpi_get_nbits(MPI a); +int mpi_test_bit(MPI a, unsigned n); +int mpi_set_bit(MPI a, unsigned n); +int mpi_set_highbit(MPI a, unsigned n); +void mpi_clear_highbit(MPI a, unsigned n); +void mpi_clear_bit(MPI a, unsigned n); +int mpi_rshift(MPI x, MPI a, unsigned n); + +/*-- mpi-inv.c --*/ +int mpi_invm(MPI x, MPI u, MPI v); + +#endif /*G10_MPI_H */ diff --git a/lib/mpi/longlong.h b/lib/mpi/longlong.h new file mode 100644 index 00000000000..b87487b40a8 --- /dev/null +++ b/lib/mpi/longlong.h @@ -0,0 +1,1478 @@ +/* longlong.h -- definitions for mixed size 32/64 bit arithmetic. + * Note: I added some stuff for use with gnupg + * + * Copyright (C) 1991, 1992, 1993, 1994, 1996, 1998, + * 2000, 2001, 2002, 2003 Free Software Foundation, Inc. + * + * This file is free software; you can redistribute it and/or modify + * it under the terms of the GNU Library General Public License as published by + * the Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * This file 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 Library General Public + * License for more details. + * + * You should have received a copy of the GNU Library General Public License + * along with this file; see the file COPYING.LIB. If not, write to + * the Free Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, + * MA 02111-1307, USA. */ + +/* You have to define the following before including this file: + * + * UWtype -- An unsigned type, default type for operations (typically a "word") + * UHWtype -- An unsigned type, at least half the size of UWtype. + * UDWtype -- An unsigned type, at least twice as large a UWtype + * W_TYPE_SIZE -- size in bits of UWtype + * + * SItype, USItype -- Signed and unsigned 32 bit types. + * DItype, UDItype -- Signed and unsigned 64 bit types. + * + * On a 32 bit machine UWtype should typically be USItype; + * on a 64 bit machine, UWtype should typically be UDItype. +*/ + +#define __BITS4 (W_TYPE_SIZE / 4) +#define __ll_B ((UWtype) 1 << (W_TYPE_SIZE / 2)) +#define __ll_lowpart(t) ((UWtype) (t) & (__ll_B - 1)) +#define __ll_highpart(t) ((UWtype) (t) >> (W_TYPE_SIZE / 2)) + +/* This is used to make sure no undesirable sharing between different libraries + that use this file takes place. */ +#ifndef __MPN +#define __MPN(x) __##x +#endif + +/* Define auxiliary asm macros. + * + * 1) umul_ppmm(high_prod, low_prod, multipler, multiplicand) multiplies two + * UWtype integers MULTIPLER and MULTIPLICAND, and generates a two UWtype + * word product in HIGH_PROD and LOW_PROD. + * + * 2) __umulsidi3(a,b) multiplies two UWtype integers A and B, and returns a + * UDWtype product. This is just a variant of umul_ppmm. + + * 3) udiv_qrnnd(quotient, remainder, high_numerator, low_numerator, + * denominator) divides a UDWtype, composed by the UWtype integers + * HIGH_NUMERATOR and LOW_NUMERATOR, by DENOMINATOR and places the quotient + * in QUOTIENT and the remainder in REMAINDER. HIGH_NUMERATOR must be less + * than DENOMINATOR for correct operation. If, in addition, the most + * significant bit of DENOMINATOR must be 1, then the pre-processor symbol + * UDIV_NEEDS_NORMALIZATION is defined to 1. + * 4) sdiv_qrnnd(quotient, remainder, high_numerator, low_numerator, + * denominator). Like udiv_qrnnd but the numbers are signed. The quotient + * is rounded towards 0. + * + * 5) count_leading_zeros(count, x) counts the number of zero-bits from the + * msb to the first non-zero bit in the UWtype X. This is the number of + * steps X needs to be shifted left to set the msb. Undefined for X == 0, + * unless the symbol COUNT_LEADING_ZEROS_0 is defined to some value. + * + * 6) count_trailing_zeros(count, x) like count_leading_zeros, but counts + * from the least significant end. + * + * 7) add_ssaaaa(high_sum, low_sum, high_addend_1, low_addend_1, + * high_addend_2, low_addend_2) adds two UWtype integers, composed by + * HIGH_ADDEND_1 and LOW_ADDEND_1, and HIGH_ADDEND_2 and LOW_ADDEND_2 + * respectively. The result is placed in HIGH_SUM and LOW_SUM. Overflow + * (i.e. carry out) is not stored anywhere, and is lost. + * + * 8) sub_ddmmss(high_difference, low_difference, high_minuend, low_minuend, + * high_subtrahend, low_subtrahend) subtracts two two-word UWtype integers, + * composed by HIGH_MINUEND_1 and LOW_MINUEND_1, and HIGH_SUBTRAHEND_2 and + * LOW_SUBTRAHEND_2 respectively. The result is placed in HIGH_DIFFERENCE + * and LOW_DIFFERENCE. Overflow (i.e. carry out) is not stored anywhere, + * and is lost. + * + * If any of these macros are left undefined for a particular CPU, + * C macros are used. */ + +/* The CPUs come in alphabetical order below. + * + * Please add support for more CPUs here, or improve the current support + * for the CPUs below! */ + +#if defined(__GNUC__) && !defined(NO_ASM) + +/* We sometimes need to clobber "cc" with gcc2, but that would not be + understood by gcc1. Use cpp to avoid major code duplication. */ +#if __GNUC__ < 2 +#define __CLOBBER_CC +#define __AND_CLOBBER_CC +#else /* __GNUC__ >= 2 */ +#define __CLOBBER_CC : "cc" +#define __AND_CLOBBER_CC , "cc" +#endif /* __GNUC__ < 2 */ + +/*************************************** + ************** A29K ***************** + ***************************************/ +#if (defined(__a29k__) || defined(_AM29K)) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("add %1,%4,%5\n" \ + "addc %0,%2,%3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%r" ((USItype)(ah)), \ + "rI" ((USItype)(bh)), \ + "%r" ((USItype)(al)), \ + "rI" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("sub %1,%4,%5\n" \ + "subc %0,%2,%3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "r" ((USItype)(ah)), \ + "rI" ((USItype)(bh)), \ + "r" ((USItype)(al)), \ + "rI" ((USItype)(bl))) +#define umul_ppmm(xh, xl, m0, m1) \ +do { \ + USItype __m0 = (m0), __m1 = (m1); \ + __asm__ ("multiplu %0,%1,%2" \ + : "=r" ((USItype)(xl)) \ + : "r" (__m0), \ + "r" (__m1)); \ + __asm__ ("multmu %0,%1,%2" \ + : "=r" ((USItype)(xh)) \ + : "r" (__m0), \ + "r" (__m1)); \ +} while (0) +#define udiv_qrnnd(q, r, n1, n0, d) \ + __asm__ ("dividu %0,%3,%4" \ + : "=r" ((USItype)(q)), \ + "=q" ((USItype)(r)) \ + : "1" ((USItype)(n1)), \ + "r" ((USItype)(n0)), \ + "r" ((USItype)(d))) + +#define count_leading_zeros(count, x) \ + __asm__ ("clz %0,%1" \ + : "=r" ((USItype)(count)) \ + : "r" ((USItype)(x))) +#define COUNT_LEADING_ZEROS_0 32 +#endif /* __a29k__ */ + +#if defined(__alpha) && W_TYPE_SIZE == 64 +#define umul_ppmm(ph, pl, m0, m1) \ +do { \ + UDItype __m0 = (m0), __m1 = (m1); \ + __asm__ ("umulh %r1,%2,%0" \ + : "=r" ((UDItype) ph) \ + : "%rJ" (__m0), \ + "rI" (__m1)); \ + (pl) = __m0 * __m1; \ + } while (0) +#define UMUL_TIME 46 +#ifndef LONGLONG_STANDALONE +#define udiv_qrnnd(q, r, n1, n0, d) \ +do { UDItype __r; \ + (q) = __udiv_qrnnd(&__r, (n1), (n0), (d)); \ + (r) = __r; \ +} while (0) +extern UDItype __udiv_qrnnd(); +#define UDIV_TIME 220 +#endif /* LONGLONG_STANDALONE */ +#endif /* __alpha */ + +/*************************************** + ************** ARM ****************** + ***************************************/ +#if defined(__arm__) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("adds %1, %4, %5\n" \ + "adc %0, %2, %3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%r" ((USItype)(ah)), \ + "rI" ((USItype)(bh)), \ + "%r" ((USItype)(al)), \ + "rI" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("subs %1, %4, %5\n" \ + "sbc %0, %2, %3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "r" ((USItype)(ah)), \ + "rI" ((USItype)(bh)), \ + "r" ((USItype)(al)), \ + "rI" ((USItype)(bl))) +#if defined __ARM_ARCH_2__ || defined __ARM_ARCH_3__ +#define umul_ppmm(xh, xl, a, b) \ + __asm__ ("%@ Inlined umul_ppmm\n" \ + "mov %|r0, %2, lsr #16 @ AAAA\n" \ + "mov %|r2, %3, lsr #16 @ BBBB\n" \ + "bic %|r1, %2, %|r0, lsl #16 @ aaaa\n" \ + "bic %0, %3, %|r2, lsl #16 @ bbbb\n" \ + "mul %1, %|r1, %|r2 @ aaaa * BBBB\n" \ + "mul %|r2, %|r0, %|r2 @ AAAA * BBBB\n" \ + "mul %|r1, %0, %|r1 @ aaaa * bbbb\n" \ + "mul %0, %|r0, %0 @ AAAA * bbbb\n" \ + "adds %|r0, %1, %0 @ central sum\n" \ + "addcs %|r2, %|r2, #65536\n" \ + "adds %1, %|r1, %|r0, lsl #16\n" \ + "adc %0, %|r2, %|r0, lsr #16" \ + : "=&r" ((USItype)(xh)), \ + "=r" ((USItype)(xl)) \ + : "r" ((USItype)(a)), \ + "r" ((USItype)(b)) \ + : "r0", "r1", "r2") +#else +#define umul_ppmm(xh, xl, a, b) \ + __asm__ ("%@ Inlined umul_ppmm\n" \ + "umull %r1, %r0, %r2, %r3" \ + : "=&r" ((USItype)(xh)), \ + "=r" ((USItype)(xl)) \ + : "r" ((USItype)(a)), \ + "r" ((USItype)(b)) \ + : "r0", "r1") +#endif +#define UMUL_TIME 20 +#define UDIV_TIME 100 +#endif /* __arm__ */ + +/*************************************** + ************** CLIPPER ************** + ***************************************/ +#if defined(__clipper__) && W_TYPE_SIZE == 32 +#define umul_ppmm(w1, w0, u, v) \ + ({union {UDItype __ll; \ + struct {USItype __l, __h; } __i; \ + } __xx; \ + __asm__ ("mulwux %2,%0" \ + : "=r" (__xx.__ll) \ + : "%0" ((USItype)(u)), \ + "r" ((USItype)(v))); \ + (w1) = __xx.__i.__h; (w0) = __xx.__i.__l; }) +#define smul_ppmm(w1, w0, u, v) \ + ({union {DItype __ll; \ + struct {SItype __l, __h; } __i; \ + } __xx; \ + __asm__ ("mulwx %2,%0" \ + : "=r" (__xx.__ll) \ + : "%0" ((SItype)(u)), \ + "r" ((SItype)(v))); \ + (w1) = __xx.__i.__h; (w0) = __xx.__i.__l; }) +#define __umulsidi3(u, v) \ + ({UDItype __w; \ + __asm__ ("mulwux %2,%0" \ + : "=r" (__w) \ + : "%0" ((USItype)(u)), \ + "r" ((USItype)(v))); \ + __w; }) +#endif /* __clipper__ */ + +/*************************************** + ************** GMICRO *************** + ***************************************/ +#if defined(__gmicro__) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("add.w %5,%1\n" \ + "addx %3,%0" \ + : "=g" ((USItype)(sh)), \ + "=&g" ((USItype)(sl)) \ + : "%0" ((USItype)(ah)), \ + "g" ((USItype)(bh)), \ + "%1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("sub.w %5,%1\n" \ + "subx %3,%0" \ + : "=g" ((USItype)(sh)), \ + "=&g" ((USItype)(sl)) \ + : "0" ((USItype)(ah)), \ + "g" ((USItype)(bh)), \ + "1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#define umul_ppmm(ph, pl, m0, m1) \ + __asm__ ("mulx %3,%0,%1" \ + : "=g" ((USItype)(ph)), \ + "=r" ((USItype)(pl)) \ + : "%0" ((USItype)(m0)), \ + "g" ((USItype)(m1))) +#define udiv_qrnnd(q, r, nh, nl, d) \ + __asm__ ("divx %4,%0,%1" \ + : "=g" ((USItype)(q)), \ + "=r" ((USItype)(r)) \ + : "1" ((USItype)(nh)), \ + "0" ((USItype)(nl)), \ + "g" ((USItype)(d))) +#define count_leading_zeros(count, x) \ + __asm__ ("bsch/1 %1,%0" \ + : "=g" (count) \ + : "g" ((USItype)(x)), \ + "0" ((USItype)0)) +#endif + +/*************************************** + ************** HPPA ***************** + ***************************************/ +#if defined(__hppa) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("add %4,%5,%1\n" \ + "addc %2,%3,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%rM" ((USItype)(ah)), \ + "rM" ((USItype)(bh)), \ + "%rM" ((USItype)(al)), \ + "rM" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("sub %4,%5,%1\n" \ + "subb %2,%3,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "rM" ((USItype)(ah)), \ + "rM" ((USItype)(bh)), \ + "rM" ((USItype)(al)), \ + "rM" ((USItype)(bl))) +#if defined(_PA_RISC1_1) +#define umul_ppmm(wh, wl, u, v) \ +do { \ + union {UDItype __ll; \ + struct {USItype __h, __l; } __i; \ + } __xx; \ + __asm__ ("xmpyu %1,%2,%0" \ + : "=*f" (__xx.__ll) \ + : "*f" ((USItype)(u)), \ + "*f" ((USItype)(v))); \ + (wh) = __xx.__i.__h; \ + (wl) = __xx.__i.__l; \ +} while (0) +#define UMUL_TIME 8 +#define UDIV_TIME 60 +#else +#define UMUL_TIME 40 +#define UDIV_TIME 80 +#endif +#ifndef LONGLONG_STANDALONE +#define udiv_qrnnd(q, r, n1, n0, d) \ +do { USItype __r; \ + (q) = __udiv_qrnnd(&__r, (n1), (n0), (d)); \ + (r) = __r; \ +} while (0) +extern USItype __udiv_qrnnd(); +#endif /* LONGLONG_STANDALONE */ +#define count_leading_zeros(count, x) \ +do { \ + USItype __tmp; \ + __asm__ ( \ + "ldi 1,%0\n" \ + "extru,= %1,15,16,%%r0 ; Bits 31..16 zero?\n" \ + "extru,tr %1,15,16,%1 ; No. Shift down, skip add.\n" \ + "ldo 16(%0),%0 ; Yes. Perform add.\n" \ + "extru,= %1,23,8,%%r0 ; Bits 15..8 zero?\n" \ + "extru,tr %1,23,8,%1 ; No. Shift down, skip add.\n" \ + "ldo 8(%0),%0 ; Yes. Perform add.\n" \ + "extru,= %1,27,4,%%r0 ; Bits 7..4 zero?\n" \ + "extru,tr %1,27,4,%1 ; No. Shift down, skip add.\n" \ + "ldo 4(%0),%0 ; Yes. Perform add.\n" \ + "extru,= %1,29,2,%%r0 ; Bits 3..2 zero?\n" \ + "extru,tr %1,29,2,%1 ; No. Shift down, skip add.\n" \ + "ldo 2(%0),%0 ; Yes. Perform add.\n" \ + "extru %1,30,1,%1 ; Extract bit 1.\n" \ + "sub %0,%1,%0 ; Subtract it. " \ + : "=r" (count), "=r" (__tmp) : "1" (x)); \ +} while (0) +#endif /* hppa */ + +/*************************************** + ************** I370 ***************** + ***************************************/ +#if (defined(__i370__) || defined(__mvs__)) && W_TYPE_SIZE == 32 +#define umul_ppmm(xh, xl, m0, m1) \ +do { \ + union {UDItype __ll; \ + struct {USItype __h, __l; } __i; \ + } __xx; \ + USItype __m0 = (m0), __m1 = (m1); \ + __asm__ ("mr %0,%3" \ + : "=r" (__xx.__i.__h), \ + "=r" (__xx.__i.__l) \ + : "%1" (__m0), \ + "r" (__m1)); \ + (xh) = __xx.__i.__h; (xl) = __xx.__i.__l; \ + (xh) += ((((SItype) __m0 >> 31) & __m1) \ + + (((SItype) __m1 >> 31) & __m0)); \ +} while (0) +#define smul_ppmm(xh, xl, m0, m1) \ +do { \ + union {DItype __ll; \ + struct {USItype __h, __l; } __i; \ + } __xx; \ + __asm__ ("mr %0,%3" \ + : "=r" (__xx.__i.__h), \ + "=r" (__xx.__i.__l) \ + : "%1" (m0), \ + "r" (m1)); \ + (xh) = __xx.__i.__h; (xl) = __xx.__i.__l; \ +} while (0) +#define sdiv_qrnnd(q, r, n1, n0, d) \ +do { \ + union {DItype __ll; \ + struct {USItype __h, __l; } __i; \ + } __xx; \ + __xx.__i.__h = n1; __xx.__i.__l = n0; \ + __asm__ ("dr %0,%2" \ + : "=r" (__xx.__ll) \ + : "0" (__xx.__ll), "r" (d)); \ + (q) = __xx.__i.__l; (r) = __xx.__i.__h; \ +} while (0) +#endif + +/*************************************** + ************** I386 ***************** + ***************************************/ +#undef __i386__ +#if (defined(__i386__) || defined(__i486__)) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("addl %5,%1\n" \ + "adcl %3,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%0" ((USItype)(ah)), \ + "g" ((USItype)(bh)), \ + "%1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("subl %5,%1\n" \ + "sbbl %3,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "0" ((USItype)(ah)), \ + "g" ((USItype)(bh)), \ + "1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("mull %3" \ + : "=a" ((USItype)(w0)), \ + "=d" ((USItype)(w1)) \ + : "%0" ((USItype)(u)), \ + "rm" ((USItype)(v))) +#define udiv_qrnnd(q, r, n1, n0, d) \ + __asm__ ("divl %4" \ + : "=a" ((USItype)(q)), \ + "=d" ((USItype)(r)) \ + : "0" ((USItype)(n0)), \ + "1" ((USItype)(n1)), \ + "rm" ((USItype)(d))) +#define count_leading_zeros(count, x) \ +do { \ + USItype __cbtmp; \ + __asm__ ("bsrl %1,%0" \ + : "=r" (__cbtmp) : "rm" ((USItype)(x))); \ + (count) = __cbtmp ^ 31; \ +} while (0) +#define count_trailing_zeros(count, x) \ + __asm__ ("bsfl %1,%0" : "=r" (count) : "rm" ((USItype)(x))) +#ifndef UMUL_TIME +#define UMUL_TIME 40 +#endif +#ifndef UDIV_TIME +#define UDIV_TIME 40 +#endif +#endif /* 80x86 */ + +/*************************************** + ************** I860 ***************** + ***************************************/ +#if defined(__i860__) && W_TYPE_SIZE == 32 +#define rshift_rhlc(r, h, l, c) \ + __asm__ ("shr %3,r0,r0\n" \ + "shrd %1,%2,%0" \ + "=r" (r) : "r" (h), "r" (l), "rn" (c)) +#endif /* i860 */ + +/*************************************** + ************** I960 ***************** + ***************************************/ +#if defined(__i960__) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("cmpo 1,0\n" \ + "addc %5,%4,%1\n" \ + "addc %3,%2,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%dI" ((USItype)(ah)), \ + "dI" ((USItype)(bh)), \ + "%dI" ((USItype)(al)), \ + "dI" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("cmpo 0,0\n" \ + "subc %5,%4,%1\n" \ + "subc %3,%2,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "dI" ((USItype)(ah)), \ + "dI" ((USItype)(bh)), \ + "dI" ((USItype)(al)), \ + "dI" ((USItype)(bl))) +#define umul_ppmm(w1, w0, u, v) \ + ({union {UDItype __ll; \ + struct {USItype __l, __h; } __i; \ + } __xx; \ + __asm__ ("emul %2,%1,%0" \ + : "=d" (__xx.__ll) \ + : "%dI" ((USItype)(u)), \ + "dI" ((USItype)(v))); \ + (w1) = __xx.__i.__h; (w0) = __xx.__i.__l; }) +#define __umulsidi3(u, v) \ + ({UDItype __w; \ + __asm__ ("emul %2,%1,%0" \ + : "=d" (__w) \ + : "%dI" ((USItype)(u)), \ + "dI" ((USItype)(v))); \ + __w; }) +#define udiv_qrnnd(q, r, nh, nl, d) \ +do { \ + union {UDItype __ll; \ + struct {USItype __l, __h; } __i; \ + } __nn; \ + __nn.__i.__h = (nh); __nn.__i.__l = (nl); \ + __asm__ ("ediv %d,%n,%0" \ + : "=d" (__rq.__ll) \ + : "dI" (__nn.__ll), \ + "dI" ((USItype)(d))); \ + (r) = __rq.__i.__l; (q) = __rq.__i.__h; \ +} while (0) +#define count_leading_zeros(count, x) \ +do { \ + USItype __cbtmp; \ + __asm__ ("scanbit %1,%0" \ + : "=r" (__cbtmp) \ + : "r" ((USItype)(x))); \ + (count) = __cbtmp ^ 31; \ +} while (0) +#define COUNT_LEADING_ZEROS_0 (-32) /* sic */ +#if defined(__i960mx) /* what is the proper symbol to test??? */ +#define rshift_rhlc(r, h, l, c) \ +do { \ + union {UDItype __ll; \ + struct {USItype __l, __h; } __i; \ + } __nn; \ + __nn.__i.__h = (h); __nn.__i.__l = (l); \ + __asm__ ("shre %2,%1,%0" \ + : "=d" (r) : "dI" (__nn.__ll), "dI" (c)); \ +} +#endif /* i960mx */ +#endif /* i960 */ + +/*************************************** + ************** 68000 **************** + ***************************************/ +#if (defined(__mc68000__) || defined(__mc68020__) || defined(__NeXT__) || defined(mc68020)) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("add%.l %5,%1\n" \ + "addx%.l %3,%0" \ + : "=d" ((USItype)(sh)), \ + "=&d" ((USItype)(sl)) \ + : "%0" ((USItype)(ah)), \ + "d" ((USItype)(bh)), \ + "%1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("sub%.l %5,%1\n" \ + "subx%.l %3,%0" \ + : "=d" ((USItype)(sh)), \ + "=&d" ((USItype)(sl)) \ + : "0" ((USItype)(ah)), \ + "d" ((USItype)(bh)), \ + "1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#if (defined(__mc68020__) || defined(__NeXT__) || defined(mc68020)) +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("mulu%.l %3,%1:%0" \ + : "=d" ((USItype)(w0)), \ + "=d" ((USItype)(w1)) \ + : "%0" ((USItype)(u)), \ + "dmi" ((USItype)(v))) +#define UMUL_TIME 45 +#define udiv_qrnnd(q, r, n1, n0, d) \ + __asm__ ("divu%.l %4,%1:%0" \ + : "=d" ((USItype)(q)), \ + "=d" ((USItype)(r)) \ + : "0" ((USItype)(n0)), \ + "1" ((USItype)(n1)), \ + "dmi" ((USItype)(d))) +#define UDIV_TIME 90 +#define sdiv_qrnnd(q, r, n1, n0, d) \ + __asm__ ("divs%.l %4,%1:%0" \ + : "=d" ((USItype)(q)), \ + "=d" ((USItype)(r)) \ + : "0" ((USItype)(n0)), \ + "1" ((USItype)(n1)), \ + "dmi" ((USItype)(d))) +#define count_leading_zeros(count, x) \ + __asm__ ("bfffo %1{%b2:%b2},%0" \ + : "=d" ((USItype)(count)) \ + : "od" ((USItype)(x)), "n" (0)) +#define COUNT_LEADING_ZEROS_0 32 +#else /* not mc68020 */ +#define umul_ppmm(xh, xl, a, b) \ +do { USItype __umul_tmp1, __umul_tmp2; \ + __asm__ ("| Inlined umul_ppmm\n" \ + "move%.l %5,%3\n" \ + "move%.l %2,%0\n" \ + "move%.w %3,%1\n" \ + "swap %3\n" \ + "swap %0\n" \ + "mulu %2,%1\n" \ + "mulu %3,%0\n" \ + "mulu %2,%3\n" \ + "swap %2\n" \ + "mulu %5,%2\n" \ + "add%.l %3,%2\n" \ + "jcc 1f\n" \ + "add%.l %#0x10000,%0\n" \ + "1: move%.l %2,%3\n" \ + "clr%.w %2\n" \ + "swap %2\n" \ + "swap %3\n" \ + "clr%.w %3\n" \ + "add%.l %3,%1\n" \ + "addx%.l %2,%0\n" \ + "| End inlined umul_ppmm" \ + : "=&d" ((USItype)(xh)), "=&d" ((USItype)(xl)), \ + "=d" (__umul_tmp1), "=&d" (__umul_tmp2) \ + : "%2" ((USItype)(a)), "d" ((USItype)(b))); \ +} while (0) +#define UMUL_TIME 100 +#define UDIV_TIME 400 +#endif /* not mc68020 */ +#endif /* mc68000 */ + +/*************************************** + ************** 88000 **************** + ***************************************/ +#if defined(__m88000__) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("addu.co %1,%r4,%r5\n" \ + "addu.ci %0,%r2,%r3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%rJ" ((USItype)(ah)), \ + "rJ" ((USItype)(bh)), \ + "%rJ" ((USItype)(al)), \ + "rJ" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("subu.co %1,%r4,%r5\n" \ + "subu.ci %0,%r2,%r3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "rJ" ((USItype)(ah)), \ + "rJ" ((USItype)(bh)), \ + "rJ" ((USItype)(al)), \ + "rJ" ((USItype)(bl))) +#define count_leading_zeros(count, x) \ +do { \ + USItype __cbtmp; \ + __asm__ ("ff1 %0,%1" \ + : "=r" (__cbtmp) \ + : "r" ((USItype)(x))); \ + (count) = __cbtmp ^ 31; \ +} while (0) +#define COUNT_LEADING_ZEROS_0 63 /* sic */ +#if defined(__m88110__) +#define umul_ppmm(wh, wl, u, v) \ +do { \ + union {UDItype __ll; \ + struct {USItype __h, __l; } __i; \ + } __x; \ + __asm__ ("mulu.d %0,%1,%2" : "=r" (__x.__ll) : "r" (u), "r" (v)); \ + (wh) = __x.__i.__h; \ + (wl) = __x.__i.__l; \ +} while (0) +#define udiv_qrnnd(q, r, n1, n0, d) \ + ({union {UDItype __ll; \ + struct {USItype __h, __l; } __i; \ + } __x, __q; \ + __x.__i.__h = (n1); __x.__i.__l = (n0); \ + __asm__ ("divu.d %0,%1,%2" \ + : "=r" (__q.__ll) : "r" (__x.__ll), "r" (d)); \ + (r) = (n0) - __q.__l * (d); (q) = __q.__l; }) +#define UMUL_TIME 5 +#define UDIV_TIME 25 +#else +#define UMUL_TIME 17 +#define UDIV_TIME 150 +#endif /* __m88110__ */ +#endif /* __m88000__ */ + +/*************************************** + ************** MIPS ***************** + ***************************************/ +#if defined(__mips__) && W_TYPE_SIZE == 32 +#if __GNUC__ > 2 || __GNUC_MINOR__ >= 7 +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("multu %2,%3" \ + : "=l" ((USItype)(w0)), \ + "=h" ((USItype)(w1)) \ + : "d" ((USItype)(u)), \ + "d" ((USItype)(v))) +#else +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("multu %2,%3\n" \ + "mflo %0\n" \ + "mfhi %1" \ + : "=d" ((USItype)(w0)), \ + "=d" ((USItype)(w1)) \ + : "d" ((USItype)(u)), \ + "d" ((USItype)(v))) +#endif +#define UMUL_TIME 10 +#define UDIV_TIME 100 +#endif /* __mips__ */ + +/*************************************** + ************** MIPS/64 ************** + ***************************************/ +#if (defined(__mips) && __mips >= 3) && W_TYPE_SIZE == 64 +#if __GNUC__ > 2 || __GNUC_MINOR__ >= 7 +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("dmultu %2,%3" \ + : "=l" ((UDItype)(w0)), \ + "=h" ((UDItype)(w1)) \ + : "d" ((UDItype)(u)), \ + "d" ((UDItype)(v))) +#else +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("dmultu %2,%3\n" \ + "mflo %0\n" \ + "mfhi %1" \ + : "=d" ((UDItype)(w0)), \ + "=d" ((UDItype)(w1)) \ + : "d" ((UDItype)(u)), \ + "d" ((UDItype)(v))) +#endif +#define UMUL_TIME 20 +#define UDIV_TIME 140 +#endif /* __mips__ */ + +/*************************************** + ************** 32000 **************** + ***************************************/ +#if defined(__ns32000__) && W_TYPE_SIZE == 32 +#define umul_ppmm(w1, w0, u, v) \ + ({union {UDItype __ll; \ + struct {USItype __l, __h; } __i; \ + } __xx; \ + __asm__ ("meid %2,%0" \ + : "=g" (__xx.__ll) \ + : "%0" ((USItype)(u)), \ + "g" ((USItype)(v))); \ + (w1) = __xx.__i.__h; (w0) = __xx.__i.__l; }) +#define __umulsidi3(u, v) \ + ({UDItype __w; \ + __asm__ ("meid %2,%0" \ + : "=g" (__w) \ + : "%0" ((USItype)(u)), \ + "g" ((USItype)(v))); \ + __w; }) +#define udiv_qrnnd(q, r, n1, n0, d) \ + ({union {UDItype __ll; \ + struct {USItype __l, __h; } __i; \ + } __xx; \ + __xx.__i.__h = (n1); __xx.__i.__l = (n0); \ + __asm__ ("deid %2,%0" \ + : "=g" (__xx.__ll) \ + : "0" (__xx.__ll), \ + "g" ((USItype)(d))); \ + (r) = __xx.__i.__l; (q) = __xx.__i.__h; }) +#define count_trailing_zeros(count, x) \ +do { \ + __asm__("ffsd %2,%0" \ + : "=r"((USItype) (count)) \ + : "0"((USItype) 0), "r"((USItype) (x))); \ + } while (0) +#endif /* __ns32000__ */ + +/*************************************** + ************** PPC ****************** + ***************************************/ +#if (defined(_ARCH_PPC) || defined(_IBMR2)) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ +do { \ + if (__builtin_constant_p(bh) && (bh) == 0) \ + __asm__ ("{a%I4|add%I4c} %1,%3,%4\n\t{aze|addze} %0,%2" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%r" ((USItype)(ah)), \ + "%r" ((USItype)(al)), \ + "rI" ((USItype)(bl))); \ + else if (__builtin_constant_p(bh) && (bh) == ~(USItype) 0) \ + __asm__ ("{a%I4|add%I4c} %1,%3,%4\n\t{ame|addme} %0,%2" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%r" ((USItype)(ah)), \ + "%r" ((USItype)(al)), \ + "rI" ((USItype)(bl))); \ + else \ + __asm__ ("{a%I5|add%I5c} %1,%4,%5\n\t{ae|adde} %0,%2,%3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%r" ((USItype)(ah)), \ + "r" ((USItype)(bh)), \ + "%r" ((USItype)(al)), \ + "rI" ((USItype)(bl))); \ +} while (0) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ +do { \ + if (__builtin_constant_p(ah) && (ah) == 0) \ + __asm__ ("{sf%I3|subf%I3c} %1,%4,%3\n\t{sfze|subfze} %0,%2" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "r" ((USItype)(bh)), \ + "rI" ((USItype)(al)), \ + "r" ((USItype)(bl))); \ + else if (__builtin_constant_p(ah) && (ah) == ~(USItype) 0) \ + __asm__ ("{sf%I3|subf%I3c} %1,%4,%3\n\t{sfme|subfme} %0,%2" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "r" ((USItype)(bh)), \ + "rI" ((USItype)(al)), \ + "r" ((USItype)(bl))); \ + else if (__builtin_constant_p(bh) && (bh) == 0) \ + __asm__ ("{sf%I3|subf%I3c} %1,%4,%3\n\t{ame|addme} %0,%2" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "r" ((USItype)(ah)), \ + "rI" ((USItype)(al)), \ + "r" ((USItype)(bl))); \ + else if (__builtin_constant_p(bh) && (bh) == ~(USItype) 0) \ + __asm__ ("{sf%I3|subf%I3c} %1,%4,%3\n\t{aze|addze} %0,%2" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "r" ((USItype)(ah)), \ + "rI" ((USItype)(al)), \ + "r" ((USItype)(bl))); \ + else \ + __asm__ ("{sf%I4|subf%I4c} %1,%5,%4\n\t{sfe|subfe} %0,%3,%2" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "r" ((USItype)(ah)), \ + "r" ((USItype)(bh)), \ + "rI" ((USItype)(al)), \ + "r" ((USItype)(bl))); \ +} while (0) +#define count_leading_zeros(count, x) \ + __asm__ ("{cntlz|cntlzw} %0,%1" \ + : "=r" ((USItype)(count)) \ + : "r" ((USItype)(x))) +#define COUNT_LEADING_ZEROS_0 32 +#if defined(_ARCH_PPC) +#define umul_ppmm(ph, pl, m0, m1) \ +do { \ + USItype __m0 = (m0), __m1 = (m1); \ + __asm__ ("mulhwu %0,%1,%2" \ + : "=r" ((USItype) ph) \ + : "%r" (__m0), \ + "r" (__m1)); \ + (pl) = __m0 * __m1; \ +} while (0) +#define UMUL_TIME 15 +#define smul_ppmm(ph, pl, m0, m1) \ +do { \ + SItype __m0 = (m0), __m1 = (m1); \ + __asm__ ("mulhw %0,%1,%2" \ + : "=r" ((SItype) ph) \ + : "%r" (__m0), \ + "r" (__m1)); \ + (pl) = __m0 * __m1; \ +} while (0) +#define SMUL_TIME 14 +#define UDIV_TIME 120 +#else +#define umul_ppmm(xh, xl, m0, m1) \ +do { \ + USItype __m0 = (m0), __m1 = (m1); \ + __asm__ ("mul %0,%2,%3" \ + : "=r" ((USItype)(xh)), \ + "=q" ((USItype)(xl)) \ + : "r" (__m0), \ + "r" (__m1)); \ + (xh) += ((((SItype) __m0 >> 31) & __m1) \ + + (((SItype) __m1 >> 31) & __m0)); \ +} while (0) +#define UMUL_TIME 8 +#define smul_ppmm(xh, xl, m0, m1) \ + __asm__ ("mul %0,%2,%3" \ + : "=r" ((SItype)(xh)), \ + "=q" ((SItype)(xl)) \ + : "r" (m0), \ + "r" (m1)) +#define SMUL_TIME 4 +#define sdiv_qrnnd(q, r, nh, nl, d) \ + __asm__ ("div %0,%2,%4" \ + : "=r" ((SItype)(q)), "=q" ((SItype)(r)) \ + : "r" ((SItype)(nh)), "1" ((SItype)(nl)), "r" ((SItype)(d))) +#define UDIV_TIME 100 +#endif +#endif /* Power architecture variants. */ + +/*************************************** + ************** PYR ****************** + ***************************************/ +#if defined(__pyr__) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("addw %5,%1\n" \ + "addwc %3,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%0" ((USItype)(ah)), \ + "g" ((USItype)(bh)), \ + "%1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("subw %5,%1\n" \ + "subwb %3,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "0" ((USItype)(ah)), \ + "g" ((USItype)(bh)), \ + "1" ((USItype)(al)), \ + "g" ((USItype)(bl))) + /* This insn works on Pyramids with AP, XP, or MI CPUs, but not with SP. */ +#define umul_ppmm(w1, w0, u, v) \ + ({union {UDItype __ll; \ + struct {USItype __h, __l; } __i; \ + } __xx; \ + __asm__ ("movw %1,%R0\n" \ + "uemul %2,%0" \ + : "=&r" (__xx.__ll) \ + : "g" ((USItype) (u)), \ + "g" ((USItype)(v))); \ + (w1) = __xx.__i.__h; (w0) = __xx.__i.__l; }) +#endif /* __pyr__ */ + +/*************************************** + ************** RT/ROMP ************** + ***************************************/ +#if defined(__ibm032__) /* RT/ROMP */ && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("a %1,%5\n" \ + "ae %0,%3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%0" ((USItype)(ah)), \ + "r" ((USItype)(bh)), \ + "%1" ((USItype)(al)), \ + "r" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("s %1,%5\n" \ + "se %0,%3" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "0" ((USItype)(ah)), \ + "r" ((USItype)(bh)), \ + "1" ((USItype)(al)), \ + "r" ((USItype)(bl))) +#define umul_ppmm(ph, pl, m0, m1) \ +do { \ + USItype __m0 = (m0), __m1 = (m1); \ + __asm__ ( \ + "s r2,r2\n" \ + "mts r10,%2\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "m r2,%3\n" \ + "cas %0,r2,r0\n" \ + "mfs r10,%1" \ + : "=r" ((USItype)(ph)), \ + "=r" ((USItype)(pl)) \ + : "%r" (__m0), \ + "r" (__m1) \ + : "r2"); \ + (ph) += ((((SItype) __m0 >> 31) & __m1) \ + + (((SItype) __m1 >> 31) & __m0)); \ +} while (0) +#define UMUL_TIME 20 +#define UDIV_TIME 200 +#define count_leading_zeros(count, x) \ +do { \ + if ((x) >= 0x10000) \ + __asm__ ("clz %0,%1" \ + : "=r" ((USItype)(count)) \ + : "r" ((USItype)(x) >> 16)); \ + else { \ + __asm__ ("clz %0,%1" \ + : "=r" ((USItype)(count)) \ + : "r" ((USItype)(x))); \ + (count) += 16; \ + } \ +} while (0) +#endif /* RT/ROMP */ + +/*************************************** + ************** SH2 ****************** + ***************************************/ +#if (defined(__sh2__) || defined(__sh3__) || defined(__SH4__)) \ + && W_TYPE_SIZE == 32 +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ( \ + "dmulu.l %2,%3\n" \ + "sts macl,%1\n" \ + "sts mach,%0" \ + : "=r" ((USItype)(w1)), \ + "=r" ((USItype)(w0)) \ + : "r" ((USItype)(u)), \ + "r" ((USItype)(v)) \ + : "macl", "mach") +#define UMUL_TIME 5 +#endif + +/*************************************** + ************** SPARC **************** + ***************************************/ +#if defined(__sparc__) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("addcc %r4,%5,%1\n" \ + "addx %r2,%3,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "%rJ" ((USItype)(ah)), \ + "rI" ((USItype)(bh)), \ + "%rJ" ((USItype)(al)), \ + "rI" ((USItype)(bl)) \ + __CLOBBER_CC) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("subcc %r4,%5,%1\n" \ + "subx %r2,%3,%0" \ + : "=r" ((USItype)(sh)), \ + "=&r" ((USItype)(sl)) \ + : "rJ" ((USItype)(ah)), \ + "rI" ((USItype)(bh)), \ + "rJ" ((USItype)(al)), \ + "rI" ((USItype)(bl)) \ + __CLOBBER_CC) +#if defined(__sparc_v8__) +/* Don't match immediate range because, 1) it is not often useful, + 2) the 'I' flag thinks of the range as a 13 bit signed interval, + while we want to match a 13 bit interval, sign extended to 32 bits, + but INTERPRETED AS UNSIGNED. */ +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("umul %2,%3,%1;rd %%y,%0" \ + : "=r" ((USItype)(w1)), \ + "=r" ((USItype)(w0)) \ + : "r" ((USItype)(u)), \ + "r" ((USItype)(v))) +#define UMUL_TIME 5 +#ifndef SUPERSPARC /* SuperSPARC's udiv only handles 53 bit dividends */ +#define udiv_qrnnd(q, r, n1, n0, d) \ +do { \ + USItype __q; \ + __asm__ ("mov %1,%%y;nop;nop;nop;udiv %2,%3,%0" \ + : "=r" ((USItype)(__q)) \ + : "r" ((USItype)(n1)), \ + "r" ((USItype)(n0)), \ + "r" ((USItype)(d))); \ + (r) = (n0) - __q * (d); \ + (q) = __q; \ +} while (0) +#define UDIV_TIME 25 +#endif /* SUPERSPARC */ +#else /* ! __sparc_v8__ */ +#if defined(__sparclite__) +/* This has hardware multiply but not divide. It also has two additional + instructions scan (ffs from high bit) and divscc. */ +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("umul %2,%3,%1;rd %%y,%0" \ + : "=r" ((USItype)(w1)), \ + "=r" ((USItype)(w0)) \ + : "r" ((USItype)(u)), \ + "r" ((USItype)(v))) +#define UMUL_TIME 5 +#define udiv_qrnnd(q, r, n1, n0, d) \ + __asm__ ("! Inlined udiv_qrnnd\n" \ + "wr %%g0,%2,%%y ! Not a delayed write for sparclite\n" \ + "tst %%g0\n" \ + "divscc %3,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%%g1\n" \ + "divscc %%g1,%4,%0\n" \ + "rd %%y,%1\n" \ + "bl,a 1f\n" \ + "add %1,%4,%1\n" \ + "1: ! End of inline udiv_qrnnd" \ + : "=r" ((USItype)(q)), \ + "=r" ((USItype)(r)) \ + : "r" ((USItype)(n1)), \ + "r" ((USItype)(n0)), \ + "rI" ((USItype)(d)) \ + : "%g1" __AND_CLOBBER_CC) +#define UDIV_TIME 37 +#define count_leading_zeros(count, x) \ + __asm__ ("scan %1,0,%0" \ + : "=r" ((USItype)(x)) \ + : "r" ((USItype)(count))) +/* Early sparclites return 63 for an argument of 0, but they warn that future + implementations might change this. Therefore, leave COUNT_LEADING_ZEROS_0 + undefined. */ +#endif /* __sparclite__ */ +#endif /* __sparc_v8__ */ + /* Default to sparc v7 versions of umul_ppmm and udiv_qrnnd. */ +#ifndef umul_ppmm +#define umul_ppmm(w1, w0, u, v) \ + __asm__ ("! Inlined umul_ppmm\n" \ + "wr %%g0,%2,%%y ! SPARC has 0-3 delay insn after a wr\n" \ + "sra %3,31,%%g2 ! Don't move this insn\n" \ + "and %2,%%g2,%%g2 ! Don't move this insn\n" \ + "andcc %%g0,0,%%g1 ! Don't move this insn\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,%3,%%g1\n" \ + "mulscc %%g1,0,%%g1\n" \ + "add %%g1,%%g2,%0\n" \ + "rd %%y,%1" \ + : "=r" ((USItype)(w1)), \ + "=r" ((USItype)(w0)) \ + : "%rI" ((USItype)(u)), \ + "r" ((USItype)(v)) \ + : "%g1", "%g2" __AND_CLOBBER_CC) +#define UMUL_TIME 39 /* 39 instructions */ +#endif +#ifndef udiv_qrnnd +#ifndef LONGLONG_STANDALONE +#define udiv_qrnnd(q, r, n1, n0, d) \ +do { USItype __r; \ + (q) = __udiv_qrnnd(&__r, (n1), (n0), (d)); \ + (r) = __r; \ +} while (0) + extern USItype __udiv_qrnnd(); +#define UDIV_TIME 140 +#endif /* LONGLONG_STANDALONE */ +#endif /* udiv_qrnnd */ +#endif /* __sparc__ */ + +/*************************************** + ************** VAX ****************** + ***************************************/ +#if defined(__vax__) && W_TYPE_SIZE == 32 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("addl2 %5,%1\n" \ + "adwc %3,%0" \ + : "=g" ((USItype)(sh)), \ + "=&g" ((USItype)(sl)) \ + : "%0" ((USItype)(ah)), \ + "g" ((USItype)(bh)), \ + "%1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("subl2 %5,%1\n" \ + "sbwc %3,%0" \ + : "=g" ((USItype)(sh)), \ + "=&g" ((USItype)(sl)) \ + : "0" ((USItype)(ah)), \ + "g" ((USItype)(bh)), \ + "1" ((USItype)(al)), \ + "g" ((USItype)(bl))) +#define umul_ppmm(xh, xl, m0, m1) \ +do { \ + union {UDItype __ll; \ + struct {USItype __l, __h; } __i; \ + } __xx; \ + USItype __m0 = (m0), __m1 = (m1); \ + __asm__ ("emul %1,%2,$0,%0" \ + : "=g" (__xx.__ll) \ + : "g" (__m0), \ + "g" (__m1)); \ + (xh) = __xx.__i.__h; (xl) = __xx.__i.__l; \ + (xh) += ((((SItype) __m0 >> 31) & __m1) \ + + (((SItype) __m1 >> 31) & __m0)); \ +} while (0) +#define sdiv_qrnnd(q, r, n1, n0, d) \ +do { \ + union {DItype __ll; \ + struct {SItype __l, __h; } __i; \ + } __xx; \ + __xx.__i.__h = n1; __xx.__i.__l = n0; \ + __asm__ ("ediv %3,%2,%0,%1" \ + : "=g" (q), "=g" (r) \ + : "g" (__xx.__ll), "g" (d)); \ +} while (0) +#endif /* __vax__ */ + +/*************************************** + ************** Z8000 **************** + ***************************************/ +#if defined(__z8000__) && W_TYPE_SIZE == 16 +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ + __asm__ ("add %H1,%H5\n\tadc %H0,%H3" \ + : "=r" ((unsigned int)(sh)), \ + "=&r" ((unsigned int)(sl)) \ + : "%0" ((unsigned int)(ah)), \ + "r" ((unsigned int)(bh)), \ + "%1" ((unsigned int)(al)), \ + "rQR" ((unsigned int)(bl))) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ + __asm__ ("sub %H1,%H5\n\tsbc %H0,%H3" \ + : "=r" ((unsigned int)(sh)), \ + "=&r" ((unsigned int)(sl)) \ + : "0" ((unsigned int)(ah)), \ + "r" ((unsigned int)(bh)), \ + "1" ((unsigned int)(al)), \ + "rQR" ((unsigned int)(bl))) +#define umul_ppmm(xh, xl, m0, m1) \ +do { \ + union {long int __ll; \ + struct {unsigned int __h, __l; } __i; \ + } __xx; \ + unsigned int __m0 = (m0), __m1 = (m1); \ + __asm__ ("mult %S0,%H3" \ + : "=r" (__xx.__i.__h), \ + "=r" (__xx.__i.__l) \ + : "%1" (__m0), \ + "rQR" (__m1)); \ + (xh) = __xx.__i.__h; (xl) = __xx.__i.__l; \ + (xh) += ((((signed int) __m0 >> 15) & __m1) \ + + (((signed int) __m1 >> 15) & __m0)); \ +} while (0) +#endif /* __z8000__ */ + +#endif /* __GNUC__ */ + +/*************************************** + *********** Generic Versions ******** + ***************************************/ +#if !defined(umul_ppmm) && defined(__umulsidi3) +#define umul_ppmm(ph, pl, m0, m1) \ +{ \ + UDWtype __ll = __umulsidi3(m0, m1); \ + ph = (UWtype) (__ll >> W_TYPE_SIZE); \ + pl = (UWtype) __ll; \ +} +#endif + +#if !defined(__umulsidi3) +#define __umulsidi3(u, v) \ + ({UWtype __hi, __lo; \ + umul_ppmm(__hi, __lo, u, v); \ + ((UDWtype) __hi << W_TYPE_SIZE) | __lo; }) +#endif + + /* If this machine has no inline assembler, use C macros. */ + +#if !defined(add_ssaaaa) +#define add_ssaaaa(sh, sl, ah, al, bh, bl) \ +do { \ + UWtype __x; \ + __x = (al) + (bl); \ + (sh) = (ah) + (bh) + (__x < (al)); \ + (sl) = __x; \ +} while (0) +#endif + +#if !defined(sub_ddmmss) +#define sub_ddmmss(sh, sl, ah, al, bh, bl) \ +do { \ + UWtype __x; \ + __x = (al) - (bl); \ + (sh) = (ah) - (bh) - (__x > (al)); \ + (sl) = __x; \ +} while (0) +#endif + +#if !defined(umul_ppmm) +#define umul_ppmm(w1, w0, u, v) \ +do { \ + UWtype __x0, __x1, __x2, __x3; \ + UHWtype __ul, __vl, __uh, __vh; \ + UWtype __u = (u), __v = (v); \ + \ + __ul = __ll_lowpart(__u); \ + __uh = __ll_highpart(__u); \ + __vl = __ll_lowpart(__v); \ + __vh = __ll_highpart(__v); \ + \ + __x0 = (UWtype) __ul * __vl; \ + __x1 = (UWtype) __ul * __vh; \ + __x2 = (UWtype) __uh * __vl; \ + __x3 = (UWtype) __uh * __vh; \ + \ + __x1 += __ll_highpart(__x0);/* this can't give carry */ \ + __x1 += __x2; /* but this indeed can */ \ + if (__x1 < __x2) /* did we get it? */ \ + __x3 += __ll_B; /* yes, add it in the proper pos. */ \ + \ + (w1) = __x3 + __ll_highpart(__x1); \ + (w0) = (__ll_lowpart(__x1) << W_TYPE_SIZE/2) + __ll_lowpart(__x0); \ +} while (0) +#endif + +#if !defined(umul_ppmm) +#define smul_ppmm(w1, w0, u, v) \ +do { \ + UWtype __w1; \ + UWtype __m0 = (u), __m1 = (v); \ + umul_ppmm(__w1, w0, __m0, __m1); \ + (w1) = __w1 - (-(__m0 >> (W_TYPE_SIZE - 1)) & __m1) \ + - (-(__m1 >> (W_TYPE_SIZE - 1)) & __m0); \ +} while (0) +#endif + + /* Define this unconditionally, so it can be used for debugging. */ +#define __udiv_qrnnd_c(q, r, n1, n0, d) \ +do { \ + UWtype __d1, __d0, __q1, __q0, __r1, __r0, __m; \ + __d1 = __ll_highpart(d); \ + __d0 = __ll_lowpart(d); \ + \ + __r1 = (n1) % __d1; \ + __q1 = (n1) / __d1; \ + __m = (UWtype) __q1 * __d0; \ + __r1 = __r1 * __ll_B | __ll_highpart(n0); \ + if (__r1 < __m) { \ + __q1--, __r1 += (d); \ + if (__r1 >= (d)) /* i.e. we didn't get carry when adding to __r1 */ \ + if (__r1 < __m) \ + __q1--, __r1 += (d); \ + } \ + __r1 -= __m; \ + \ + __r0 = __r1 % __d1; \ + __q0 = __r1 / __d1; \ + __m = (UWtype) __q0 * __d0; \ + __r0 = __r0 * __ll_B | __ll_lowpart(n0); \ + if (__r0 < __m) { \ + __q0--, __r0 += (d); \ + if (__r0 >= (d)) \ + if (__r0 < __m) \ + __q0--, __r0 += (d); \ + } \ + __r0 -= __m; \ + \ + (q) = (UWtype) __q1 * __ll_B | __q0; \ + (r) = __r0; \ +} while (0) + +/* If the processor has no udiv_qrnnd but sdiv_qrnnd, go through + __udiv_w_sdiv (defined in libgcc or elsewhere). */ +#if !defined(udiv_qrnnd) && defined(sdiv_qrnnd) +#define udiv_qrnnd(q, r, nh, nl, d) \ +do { \ + UWtype __r; \ + (q) = __MPN(udiv_w_sdiv) (&__r, nh, nl, d); \ + (r) = __r; \ +} while (0) +#endif + + /* If udiv_qrnnd was not defined for this processor, use __udiv_qrnnd_c. */ +#if !defined(udiv_qrnnd) +#define UDIV_NEEDS_NORMALIZATION 1 +#define udiv_qrnnd __udiv_qrnnd_c +#endif + +#undef count_leading_zeros +#if !defined(count_leading_zeros) + extern +#ifdef __STDC__ + const +#endif + unsigned char __clz_tab[]; +#define count_leading_zeros(count, x) \ +do { \ + UWtype __xr = (x); \ + UWtype __a; \ + \ + if (W_TYPE_SIZE <= 32) { \ + __a = __xr < ((UWtype) 1 << 2*__BITS4) \ + ? (__xr < ((UWtype) 1 << __BITS4) ? 0 : __BITS4) \ + : (__xr < ((UWtype) 1 << 3*__BITS4) ? 2*__BITS4 : 3*__BITS4); \ + } \ + else { \ + for (__a = W_TYPE_SIZE - 8; __a > 0; __a -= 8) \ + if (((__xr >> __a) & 0xff) != 0) \ + break; \ + } \ + \ + (count) = W_TYPE_SIZE - (__clz_tab[__xr >> __a] + __a); \ +} while (0) + /* This version gives a well-defined value for zero. */ +#define COUNT_LEADING_ZEROS_0 W_TYPE_SIZE +#endif + +#if !defined(count_trailing_zeros) +/* Define count_trailing_zeros using count_leading_zeros. The latter might be + defined in asm, but if it is not, the C version above is good enough. */ +#define count_trailing_zeros(count, x) \ +do { \ + UWtype __ctz_x = (x); \ + UWtype __ctz_c; \ + count_leading_zeros(__ctz_c, __ctz_x & -__ctz_x); \ + (count) = W_TYPE_SIZE - 1 - __ctz_c; \ +} while (0) +#endif + +#ifndef UDIV_NEEDS_NORMALIZATION +#define UDIV_NEEDS_NORMALIZATION 0 +#endif diff --git a/lib/mpi/mpi-inline.h b/lib/mpi/mpi-inline.h new file mode 100644 index 00000000000..e2b39852b30 --- /dev/null +++ b/lib/mpi/mpi-inline.h @@ -0,0 +1,122 @@ +/* mpi-inline.h - Internal to the Multi Precision Integers + * Copyright (C) 1994, 1996, 1998, 1999 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#ifndef G10_MPI_INLINE_H +#define G10_MPI_INLINE_H + +#ifndef G10_MPI_INLINE_DECL +#define G10_MPI_INLINE_DECL extern inline +#endif + +G10_MPI_INLINE_DECL mpi_limb_t +mpihelp_add_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb) +{ + mpi_limb_t x; + + x = *s1_ptr++; + s2_limb += x; + *res_ptr++ = s2_limb; + if (s2_limb < x) { /* sum is less than the left operand: handle carry */ + while (--s1_size) { + x = *s1_ptr++ + 1; /* add carry */ + *res_ptr++ = x; /* and store */ + if (x) /* not 0 (no overflow): we can stop */ + goto leave; + } + return 1; /* return carry (size of s1 to small) */ + } + +leave: + if (res_ptr != s1_ptr) { /* not the same variable */ + mpi_size_t i; /* copy the rest */ + for (i = 0; i < s1_size - 1; i++) + res_ptr[i] = s1_ptr[i]; + } + return 0; /* no carry */ +} + +G10_MPI_INLINE_DECL mpi_limb_t +mpihelp_add(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, mpi_size_t s1_size, + mpi_ptr_t s2_ptr, mpi_size_t s2_size) +{ + mpi_limb_t cy = 0; + + if (s2_size) + cy = mpihelp_add_n(res_ptr, s1_ptr, s2_ptr, s2_size); + + if (s1_size - s2_size) + cy = mpihelp_add_1(res_ptr + s2_size, s1_ptr + s2_size, + s1_size - s2_size, cy); + return cy; +} + +G10_MPI_INLINE_DECL mpi_limb_t +mpihelp_sub_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb) +{ + mpi_limb_t x; + + x = *s1_ptr++; + s2_limb = x - s2_limb; + *res_ptr++ = s2_limb; + if (s2_limb > x) { + while (--s1_size) { + x = *s1_ptr++; + *res_ptr++ = x - 1; + if (x) + goto leave; + } + return 1; + } + +leave: + if (res_ptr != s1_ptr) { + mpi_size_t i; + for (i = 0; i < s1_size - 1; i++) + res_ptr[i] = s1_ptr[i]; + } + return 0; +} + +G10_MPI_INLINE_DECL mpi_limb_t +mpihelp_sub(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, mpi_size_t s1_size, + mpi_ptr_t s2_ptr, mpi_size_t s2_size) +{ + mpi_limb_t cy = 0; + + if (s2_size) + cy = mpihelp_sub_n(res_ptr, s1_ptr, s2_ptr, s2_size); + + if (s1_size - s2_size) + cy = mpihelp_sub_1(res_ptr + s2_size, s1_ptr + s2_size, + s1_size - s2_size, cy); + return cy; +} + +#endif /*G10_MPI_INLINE_H */ diff --git a/lib/mpi/mpi-internal.h b/lib/mpi/mpi-internal.h new file mode 100644 index 00000000000..77adcf6bc25 --- /dev/null +++ b/lib/mpi/mpi-internal.h @@ -0,0 +1,261 @@ +/* mpi-internal.h - Internal to the Multi Precision Integers + * Copyright (C) 1994, 1996 Free Software Foundation, Inc. + * Copyright (C) 1998, 2000 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#ifndef G10_MPI_INTERNAL_H +#define G10_MPI_INTERNAL_H + +#include +#include +#include +#include +#include +#include + +#define log_debug printk +#define log_bug printk + +#define assert(x) \ + do { \ + if (!x) \ + log_bug("failed assertion\n"); \ + } while (0); + +/* If KARATSUBA_THRESHOLD is not already defined, define it to a + * value which is good on most machines. */ + +/* tested 4, 16, 32 and 64, where 16 gave the best performance when + * checking a 768 and a 1024 bit ElGamal signature. + * (wk 22.12.97) */ +#ifndef KARATSUBA_THRESHOLD +#define KARATSUBA_THRESHOLD 16 +#endif + +/* The code can't handle KARATSUBA_THRESHOLD smaller than 2. */ +#if KARATSUBA_THRESHOLD < 2 +#undef KARATSUBA_THRESHOLD +#define KARATSUBA_THRESHOLD 2 +#endif + +typedef mpi_limb_t *mpi_ptr_t; /* pointer to a limb */ +typedef int mpi_size_t; /* (must be a signed type) */ + +#define ABS(x) (x >= 0 ? x : -x) +#define MIN(l, o) ((l) < (o) ? (l) : (o)) +#define MAX(h, i) ((h) > (i) ? (h) : (i)) + +static inline int RESIZE_IF_NEEDED(MPI a, unsigned b) +{ + if (a->alloced < b) + return mpi_resize(a, b); + return 0; +} + +/* Copy N limbs from S to D. */ +#define MPN_COPY(d, s, n) \ + do { \ + mpi_size_t _i; \ + for (_i = 0; _i < (n); _i++) \ + (d)[_i] = (s)[_i]; \ + } while (0) + +#define MPN_COPY_INCR(d, s, n) \ + do { \ + mpi_size_t _i; \ + for (_i = 0; _i < (n); _i++) \ + (d)[_i] = (d)[_i]; \ + } while (0) + +#define MPN_COPY_DECR(d, s, n) \ + do { \ + mpi_size_t _i; \ + for (_i = (n)-1; _i >= 0; _i--) \ + (d)[_i] = (s)[_i]; \ + } while (0) + +/* Zero N limbs at D */ +#define MPN_ZERO(d, n) \ + do { \ + int _i; \ + for (_i = 0; _i < (n); _i++) \ + (d)[_i] = 0; \ + } while (0) + +#define MPN_NORMALIZE(d, n) \ + do { \ + while ((n) > 0) { \ + if ((d)[(n)-1]) \ + break; \ + (n)--; \ + } \ + } while (0) + +#define MPN_NORMALIZE_NOT_ZERO(d, n) \ + do { \ + for (;;) { \ + if ((d)[(n)-1]) \ + break; \ + (n)--; \ + } \ + } while (0) + +#define MPN_MUL_N_RECURSE(prodp, up, vp, size, tspace) \ + do { \ + if ((size) < KARATSUBA_THRESHOLD) \ + mul_n_basecase(prodp, up, vp, size); \ + else \ + mul_n(prodp, up, vp, size, tspace); \ + } while (0); + +/* Divide the two-limb number in (NH,,NL) by D, with DI being the largest + * limb not larger than (2**(2*BITS_PER_MP_LIMB))/D - (2**BITS_PER_MP_LIMB). + * If this would yield overflow, DI should be the largest possible number + * (i.e., only ones). For correct operation, the most significant bit of D + * has to be set. Put the quotient in Q and the remainder in R. + */ +#define UDIV_QRNND_PREINV(q, r, nh, nl, d, di) \ + do { \ + mpi_limb_t _q, _ql, _r; \ + mpi_limb_t _xh, _xl; \ + umul_ppmm(_q, _ql, (nh), (di)); \ + _q += (nh); /* DI is 2**BITS_PER_MPI_LIMB too small */ \ + umul_ppmm(_xh, _xl, _q, (d)); \ + sub_ddmmss(_xh, _r, (nh), (nl), _xh, _xl); \ + if (_xh) { \ + sub_ddmmss(_xh, _r, _xh, _r, 0, (d)); \ + _q++; \ + if (_xh) { \ + sub_ddmmss(_xh, _r, _xh, _r, 0, (d)); \ + _q++; \ + } \ + } \ + if (_r >= (d)) { \ + _r -= (d); \ + _q++; \ + } \ + (r) = _r; \ + (q) = _q; \ + } while (0) + +/*-- mpiutil.c --*/ +mpi_ptr_t mpi_alloc_limb_space(unsigned nlimbs); +void mpi_free_limb_space(mpi_ptr_t a); +void mpi_assign_limb_space(MPI a, mpi_ptr_t ap, unsigned nlimbs); + +/*-- mpi-bit.c --*/ +void mpi_rshift_limbs(MPI a, unsigned int count); +int mpi_lshift_limbs(MPI a, unsigned int count); + +/*-- mpihelp-add.c --*/ +mpi_limb_t mpihelp_add_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb); +mpi_limb_t mpihelp_add_n(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_ptr_t s2_ptr, mpi_size_t size); +mpi_limb_t mpihelp_add(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, mpi_size_t s1_size, + mpi_ptr_t s2_ptr, mpi_size_t s2_size); + +/*-- mpihelp-sub.c --*/ +mpi_limb_t mpihelp_sub_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb); +mpi_limb_t mpihelp_sub_n(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_ptr_t s2_ptr, mpi_size_t size); +mpi_limb_t mpihelp_sub(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, mpi_size_t s1_size, + mpi_ptr_t s2_ptr, mpi_size_t s2_size); + +/*-- mpihelp-cmp.c --*/ +int mpihelp_cmp(mpi_ptr_t op1_ptr, mpi_ptr_t op2_ptr, mpi_size_t size); + +/*-- mpihelp-mul.c --*/ + +struct karatsuba_ctx { + struct karatsuba_ctx *next; + mpi_ptr_t tspace; + mpi_size_t tspace_size; + mpi_ptr_t tp; + mpi_size_t tp_size; +}; + +void mpihelp_release_karatsuba_ctx(struct karatsuba_ctx *ctx); + +mpi_limb_t mpihelp_addmul_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb); +mpi_limb_t mpihelp_submul_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb); +int mpihelp_mul_n(mpi_ptr_t prodp, mpi_ptr_t up, mpi_ptr_t vp, mpi_size_t size); +int mpihelp_mul(mpi_ptr_t prodp, mpi_ptr_t up, mpi_size_t usize, + mpi_ptr_t vp, mpi_size_t vsize, mpi_limb_t *_result); +void mpih_sqr_n_basecase(mpi_ptr_t prodp, mpi_ptr_t up, mpi_size_t size); +void mpih_sqr_n(mpi_ptr_t prodp, mpi_ptr_t up, mpi_size_t size, + mpi_ptr_t tspace); + +int mpihelp_mul_karatsuba_case(mpi_ptr_t prodp, + mpi_ptr_t up, mpi_size_t usize, + mpi_ptr_t vp, mpi_size_t vsize, + struct karatsuba_ctx *ctx); + +/*-- mpihelp-mul_1.c (or xxx/cpu/ *.S) --*/ +mpi_limb_t mpihelp_mul_1(mpi_ptr_t res_ptr, mpi_ptr_t s1_ptr, + mpi_size_t s1_size, mpi_limb_t s2_limb); + +/*-- mpihelp-div.c --*/ +mpi_limb_t mpihelp_mod_1(mpi_ptr_t dividend_ptr, mpi_size_t dividend_size, + mpi_limb_t divisor_limb); +mpi_limb_t mpihelp_divrem(mpi_ptr_t qp, mpi_size_t qextra_limbs, + mpi_ptr_t np, mpi_size_t nsize, + mpi_ptr_t dp, mpi_size_t dsize); +mpi_limb_t mpihelp_divmod_1(mpi_ptr_t quot_ptr, + mpi_ptr_t dividend_ptr, mpi_size_t dividend_size, + mpi_limb_t divisor_limb); + +/*-- mpihelp-shift.c --*/ +mpi_limb_t mpihelp_lshift(mpi_ptr_t wp, mpi_ptr_t up, mpi_size_t usize, + unsigned cnt); +mpi_limb_t mpihelp_rshift(mpi_ptr_t wp, mpi_ptr_t up, mpi_size_t usize, + unsigned cnt); + +/* Define stuff for longlong.h. */ +#define W_TYPE_SIZE BITS_PER_MPI_LIMB +typedef mpi_limb_t UWtype; +typedef unsigned int UHWtype; +#if defined(__GNUC__) +typedef unsigned int UQItype __attribute__ ((mode(QI))); +typedef int SItype __attribute__ ((mode(SI))); +typedef unsigned int USItype __attribute__ ((mode(SI))); +typedef int DItype __attribute__ ((mode(DI))); +typedef unsigned int UDItype __attribute__ ((mode(DI))); +#else +typedef unsigned char UQItype; +typedef long SItype; +typedef unsigned long USItype; +#endif + +#ifdef __GNUC__ +#include "mpi-inline.h" +#endif + +#endif /*G10_MPI_INTERNAL_H */ -- cgit v1.2.3-70-g09d2 From d9c46b184fcfd33c85a7dc48a653435a08e21f56 Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Wed, 31 Aug 2011 14:05:16 +0300 Subject: crypto: GnuPG based MPI lib - make files (part 3) Adds the multi-precision-integer maths library which was originally taken from GnuPG and ported to the kernel by (among others) David Howells. This version is taken from Fedora kernel 2.6.32-71.14.1.el6. The difference is that checkpatch reported errors and warnings have been fixed. This library is used to implemenet RSA digital signature verification used in IMA/EVM integrity protection subsystem. Due to patch size limitation, the patch is divided into 4 parts. Signed-off-by: Dmitry Kasatkin --- lib/Kconfig | 7 +++++++ lib/Makefile | 2 ++ lib/mpi/Makefile | 21 +++++++++++++++++++++ 3 files changed, 30 insertions(+) create mode 100644 lib/mpi/Makefile (limited to 'lib') diff --git a/lib/Kconfig b/lib/Kconfig index 32f3e5ae2be..4af07ec7f2d 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -276,4 +276,11 @@ config CORDIC so its calculations are in fixed point. Modules can select this when they require this function. Module will be called cordic. +config MPILIB + tristate "Multiprecision maths library" + help + Multiprecision maths library from GnuPG. + It is used to implement RSA digital signature verification, + which is used by IMA/EVM digital signature extension. + endmenu diff --git a/lib/Makefile b/lib/Makefile index a4da283f5dc..6ba8cbf4c72 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -115,6 +115,8 @@ obj-$(CONFIG_CPU_RMAP) += cpu_rmap.o obj-$(CONFIG_CORDIC) += cordic.o +obj-$(CONFIG_MPILIB) += mpi/ + hostprogs-y := gen_crc32table clean-files := crc32table.h diff --git a/lib/mpi/Makefile b/lib/mpi/Makefile new file mode 100644 index 00000000000..45ca90a8639 --- /dev/null +++ b/lib/mpi/Makefile @@ -0,0 +1,21 @@ +# +# MPI multiprecision maths library (from gpg) +# + +obj-$(CONFIG_MPILIB) = mpi.o + +mpi-y = \ + generic_mpih-lshift.o \ + generic_mpih-mul1.o \ + generic_mpih-mul2.o \ + generic_mpih-mul3.o \ + generic_mpih-rshift.o \ + generic_mpih-sub1.o \ + generic_mpih-add1.o \ + mpicoder.o \ + mpi-bit.o \ + mpih-cmp.o \ + mpih-div.o \ + mpih-mul.o \ + mpi-pow.o \ + mpiutil.o -- cgit v1.2.3-70-g09d2 From 7e8dec918ef8e0f68b4937c3c50fa57002077a4d Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Mon, 7 Nov 2011 15:16:37 +0200 Subject: crypto: GnuPG based MPI lib - additional sources (part 4) Adds the multi-precision-integer maths library which was originally taken from GnuPG and ported to the kernel by (among others) David Howells. This version is taken from Fedora kernel 2.6.32-71.14.1.el6. The difference is that checkpatch reported errors and warnings have been fixed. This library is used to implemenet RSA digital signature verification used in IMA/EVM integrity protection subsystem. Due to patch size limitation, the patch is divided into 4 parts. This code is unnecessary for RSA digital signature verification, but for completeness it is included here and can be compiled, if CONFIG_MPILIB_EXTRA is enabled. Signed-off-by: Dmitry Kasatkin --- lib/Kconfig | 10 ++ lib/mpi/Makefile | 11 ++ lib/mpi/generic_mpi-asm-defs.h | 4 + lib/mpi/mpi-add.c | 234 +++++++++++++++++++++++++++++ lib/mpi/mpi-cmp.c | 68 +++++++++ lib/mpi/mpi-div.c | 333 +++++++++++++++++++++++++++++++++++++++++ lib/mpi/mpi-gcd.c | 59 ++++++++ lib/mpi/mpi-inline.c | 31 ++++ lib/mpi/mpi-inv.c | 187 +++++++++++++++++++++++ lib/mpi/mpi-mpow.c | 133 ++++++++++++++++ lib/mpi/mpi-mul.c | 194 ++++++++++++++++++++++++ lib/mpi/mpi-scan.c | 136 +++++++++++++++++ 12 files changed, 1400 insertions(+) create mode 100644 lib/mpi/generic_mpi-asm-defs.h create mode 100644 lib/mpi/mpi-add.c create mode 100644 lib/mpi/mpi-cmp.c create mode 100644 lib/mpi/mpi-div.c create mode 100644 lib/mpi/mpi-gcd.c create mode 100644 lib/mpi/mpi-inline.c create mode 100644 lib/mpi/mpi-inv.c create mode 100644 lib/mpi/mpi-mpow.c create mode 100644 lib/mpi/mpi-mul.c create mode 100644 lib/mpi/mpi-scan.c (limited to 'lib') diff --git a/lib/Kconfig b/lib/Kconfig index 4af07ec7f2d..d411de53be4 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -283,4 +283,14 @@ config MPILIB It is used to implement RSA digital signature verification, which is used by IMA/EVM digital signature extension. +config MPILIB_EXTRA + bool "Multiprecision maths library - additional sources" + depends on MPILIB + help + Multiprecision maths library from GnuPG. + It is used to implement RSA digital signature verification, + which is used by IMA/EVM digital signature extension. + This code in unnecessary for RSA digital signature verification, + and can be compiled if needed. + endmenu diff --git a/lib/mpi/Makefile b/lib/mpi/Makefile index 45ca90a8639..567d52e74d7 100644 --- a/lib/mpi/Makefile +++ b/lib/mpi/Makefile @@ -19,3 +19,14 @@ mpi-y = \ mpih-mul.o \ mpi-pow.o \ mpiutil.o + +mpi-$(CONFIG_MPILIB_EXTRA) += \ + mpi-add.o \ + mpi-div.o \ + mpi-cmp.o \ + mpi-gcd.o \ + mpi-inline.o \ + mpi-inv.o \ + mpi-mpow.o \ + mpi-mul.o \ + mpi-scan.o diff --git a/lib/mpi/generic_mpi-asm-defs.h b/lib/mpi/generic_mpi-asm-defs.h new file mode 100644 index 00000000000..047d1f5a724 --- /dev/null +++ b/lib/mpi/generic_mpi-asm-defs.h @@ -0,0 +1,4 @@ +/* This file defines some basic constants for the MPI machinery. We + * need to define the types on a per-CPU basis, so it is done with + * this file here. */ +#define BYTES_PER_MPI_LIMB (SIZEOF_UNSIGNED_LONG) diff --git a/lib/mpi/mpi-add.c b/lib/mpi/mpi-add.c new file mode 100644 index 00000000000..f56b9ba295e --- /dev/null +++ b/lib/mpi/mpi-add.c @@ -0,0 +1,234 @@ +/* mpi-add.c - MPI functions + * Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + * Copyright (C) 1994, 1996 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include "mpi-internal.h" + +/**************** + * Add the unsigned integer V to the mpi-integer U and store the + * result in W. U and V may be the same. + */ +int mpi_add_ui(MPI w, const MPI u, unsigned long v) +{ + mpi_ptr_t wp, up; + mpi_size_t usize, wsize; + int usign, wsign; + + usize = u->nlimbs; + usign = u->sign; + wsign = 0; + + /* If not space for W (and possible carry), increase space. */ + wsize = usize + 1; + if (w->alloced < wsize) + if (mpi_resize(w, wsize) < 0) + return -ENOMEM; + + /* These must be after realloc (U may be the same as W). */ + up = u->d; + wp = w->d; + + if (!usize) { /* simple */ + wp[0] = v; + wsize = v ? 1 : 0; + } else if (!usign) { /* mpi is not negative */ + mpi_limb_t cy; + cy = mpihelp_add_1(wp, up, usize, v); + wp[usize] = cy; + wsize = usize + cy; + } else { /* The signs are different. Need exact comparison to determine + * which operand to subtract from which. */ + if (usize == 1 && up[0] < v) { + wp[0] = v - up[0]; + wsize = 1; + } else { + mpihelp_sub_1(wp, up, usize, v); + /* Size can decrease with at most one limb. */ + wsize = usize - (wp[usize - 1] == 0); + wsign = 1; + } + } + + w->nlimbs = wsize; + w->sign = wsign; + return 0; +} + +int mpi_add(MPI w, MPI u, MPI v) +{ + mpi_ptr_t wp, up, vp; + mpi_size_t usize, vsize, wsize; + int usign, vsign, wsign; + + if (u->nlimbs < v->nlimbs) { /* Swap U and V. */ + usize = v->nlimbs; + usign = v->sign; + vsize = u->nlimbs; + vsign = u->sign; + wsize = usize + 1; + if (RESIZE_IF_NEEDED(w, wsize) < 0) + return -ENOMEM; + /* These must be after realloc (u or v may be the same as w). */ + up = v->d; + vp = u->d; + } else { + usize = u->nlimbs; + usign = u->sign; + vsize = v->nlimbs; + vsign = v->sign; + wsize = usize + 1; + if (RESIZE_IF_NEEDED(w, wsize) < 0) + return -ENOMEM; + /* These must be after realloc (u or v may be the same as w). */ + up = u->d; + vp = v->d; + } + wp = w->d; + wsign = 0; + + if (!vsize) { /* simple */ + MPN_COPY(wp, up, usize); + wsize = usize; + wsign = usign; + } else if (usign != vsign) { /* different sign */ + /* This test is right since USIZE >= VSIZE */ + if (usize != vsize) { + mpihelp_sub(wp, up, usize, vp, vsize); + wsize = usize; + MPN_NORMALIZE(wp, wsize); + wsign = usign; + } else if (mpihelp_cmp(up, vp, usize) < 0) { + mpihelp_sub_n(wp, vp, up, usize); + wsize = usize; + MPN_NORMALIZE(wp, wsize); + if (!usign) + wsign = 1; + } else { + mpihelp_sub_n(wp, up, vp, usize); + wsize = usize; + MPN_NORMALIZE(wp, wsize); + if (usign) + wsign = 1; + } + } else { /* U and V have same sign. Add them. */ + mpi_limb_t cy = mpihelp_add(wp, up, usize, vp, vsize); + wp[usize] = cy; + wsize = usize + cy; + if (usign) + wsign = 1; + } + + w->nlimbs = wsize; + w->sign = wsign; + return 0; +} + +/**************** + * Subtract the unsigned integer V from the mpi-integer U and store the + * result in W. + */ +int mpi_sub_ui(MPI w, MPI u, unsigned long v) +{ + mpi_ptr_t wp, up; + mpi_size_t usize, wsize; + int usign, wsign; + + usize = u->nlimbs; + usign = u->sign; + wsign = 0; + + /* If not space for W (and possible carry), increase space. */ + wsize = usize + 1; + if (w->alloced < wsize) + if (mpi_resize(w, wsize) < 0) + return -ENOMEM; + + /* These must be after realloc (U may be the same as W). */ + up = u->d; + wp = w->d; + + if (!usize) { /* simple */ + wp[0] = v; + wsize = v ? 1 : 0; + wsign = 1; + } else if (usign) { /* mpi and v are negative */ + mpi_limb_t cy; + cy = mpihelp_add_1(wp, up, usize, v); + wp[usize] = cy; + wsize = usize + cy; + } else { /* The signs are different. Need exact comparison to determine + * which operand to subtract from which. */ + if (usize == 1 && up[0] < v) { + wp[0] = v - up[0]; + wsize = 1; + wsign = 1; + } else { + mpihelp_sub_1(wp, up, usize, v); + /* Size can decrease with at most one limb. */ + wsize = usize - (wp[usize - 1] == 0); + } + } + + w->nlimbs = wsize; + w->sign = wsign; + return 0; +} + +int mpi_sub(MPI w, MPI u, MPI v) +{ + int rc; + + if (w == v) { + MPI vv; + if (mpi_copy(&vv, v) < 0) + return -ENOMEM; + vv->sign = !vv->sign; + rc = mpi_add(w, u, vv); + mpi_free(vv); + } else { + /* fixme: this is not thread-save (we temp. modify v) */ + v->sign = !v->sign; + rc = mpi_add(w, u, v); + v->sign = !v->sign; + } + return rc; +} + +int mpi_addm(MPI w, MPI u, MPI v, MPI m) +{ + if (mpi_add(w, u, v) < 0 || mpi_fdiv_r(w, w, m) < 0) + return -ENOMEM; + return 0; +} + +int mpi_subm(MPI w, MPI u, MPI v, MPI m) +{ + if (mpi_sub(w, u, v) < 0 || mpi_fdiv_r(w, w, m) < 0) + return -ENOMEM; + return 0; +} diff --git a/lib/mpi/mpi-cmp.c b/lib/mpi/mpi-cmp.c new file mode 100644 index 00000000000..914bc42a8a8 --- /dev/null +++ b/lib/mpi/mpi-cmp.c @@ -0,0 +1,68 @@ +/* mpi-cmp.c - MPI functions + * Copyright (C) 1998, 1999 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +#include "mpi-internal.h" + +int mpi_cmp_ui(MPI u, unsigned long v) +{ + mpi_limb_t limb = v; + + mpi_normalize(u); + if (!u->nlimbs && !limb) + return 0; + if (u->sign) + return -1; + if (u->nlimbs > 1) + return 1; + + if (u->d[0] == limb) + return 0; + else if (u->d[0] > limb) + return 1; + else + return -1; +} + +int mpi_cmp(MPI u, MPI v) +{ + mpi_size_t usize, vsize; + int cmp; + + mpi_normalize(u); + mpi_normalize(v); + usize = u->nlimbs; + vsize = v->nlimbs; + if (!u->sign && v->sign) + return 1; + if (u->sign && !v->sign) + return -1; + if (usize != vsize && !u->sign && !v->sign) + return usize - vsize; + if (usize != vsize && u->sign && v->sign) + return vsize + usize; + if (!usize) + return 0; + cmp = mpihelp_cmp(u->d, v->d, usize); + if (!cmp) + return 0; + if ((cmp < 0 ? 1 : 0) == (u->sign ? 1 : 0)) + return 1; + return -1; +} diff --git a/lib/mpi/mpi-div.c b/lib/mpi/mpi-div.c new file mode 100644 index 00000000000..c3087d1390c --- /dev/null +++ b/lib/mpi/mpi-div.c @@ -0,0 +1,333 @@ +/* mpi-div.c - MPI functions + * Copyright (C) 1994, 1996 Free Software Foundation, Inc. + * Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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. + * The GNU MP Library itself is published under the LGPL; + * however I decided to publish this code under the plain GPL. + */ + +#include +#include "mpi-internal.h" +#include "longlong.h" + +int mpi_fdiv_r(MPI rem, MPI dividend, MPI divisor) +{ + int rc = -ENOMEM; + int divisor_sign = divisor->sign; + MPI temp_divisor = NULL; + + /* We need the original value of the divisor after the remainder has been + * preliminary calculated. We have to copy it to temporary space if it's + * the same variable as REM. */ + if (rem == divisor) { + if (mpi_copy(&temp_divisor, divisor) < 0) + goto nomem; + divisor = temp_divisor; + } + + if (mpi_tdiv_qr(NULL, rem, dividend, divisor) < 0) + goto nomem; + if (((divisor_sign ? 1 : 0) ^ (dividend->sign ? 1 : 0)) && rem->nlimbs) + if (mpi_add(rem, rem, divisor) < 0) + goto nomem; + + rc = 0; + +nomem: + if (temp_divisor) + mpi_free(temp_divisor); + return rc; +} + +/**************** + * Division rounding the quotient towards -infinity. + * The remainder gets the same sign as the denominator. + * rem is optional + */ + +ulong mpi_fdiv_r_ui(MPI rem, MPI dividend, ulong divisor) +{ + mpi_limb_t rlimb; + + rlimb = mpihelp_mod_1(dividend->d, dividend->nlimbs, divisor); + if (rlimb && dividend->sign) + rlimb = divisor - rlimb; + + if (rem) { + rem->d[0] = rlimb; + rem->nlimbs = rlimb ? 1 : 0; + } + return rlimb; +} + +int mpi_fdiv_q(MPI quot, MPI dividend, MPI divisor) +{ + MPI tmp = mpi_alloc(mpi_get_nlimbs(quot)); + if (!tmp) + return -ENOMEM; + mpi_fdiv_qr(quot, tmp, dividend, divisor); + mpi_free(tmp); + return 0; +} + +int mpi_fdiv_qr(MPI quot, MPI rem, MPI dividend, MPI divisor) +{ + int divisor_sign = divisor->sign; + MPI temp_divisor = NULL; + + if (quot == divisor || rem == divisor) { + if (mpi_copy(&temp_divisor, divisor) < 0) + return -ENOMEM; + divisor = temp_divisor; + } + + if (mpi_tdiv_qr(quot, rem, dividend, divisor) < 0) + goto nomem; + + if ((divisor_sign ^ dividend->sign) && rem->nlimbs) { + if (mpi_sub_ui(quot, quot, 1) < 0) + goto nomem; + if (mpi_add(rem, rem, divisor) < 0) + goto nomem; + } + + if (temp_divisor) + mpi_free(temp_divisor); + + return 0; + +nomem: + mpi_free(temp_divisor); + return -ENOMEM; +} + +/* If den == quot, den needs temporary storage. + * If den == rem, den needs temporary storage. + * If num == quot, num needs temporary storage. + * If den has temporary storage, it can be normalized while being copied, + * i.e no extra storage should be allocated. + */ + +int mpi_tdiv_r(MPI rem, MPI num, MPI den) +{ + return mpi_tdiv_qr(NULL, rem, num, den); +} + +int mpi_tdiv_qr(MPI quot, MPI rem, MPI num, MPI den) +{ + int rc = -ENOMEM; + mpi_ptr_t np, dp; + mpi_ptr_t qp, rp; + mpi_size_t nsize = num->nlimbs; + mpi_size_t dsize = den->nlimbs; + mpi_size_t qsize, rsize; + mpi_size_t sign_remainder = num->sign; + mpi_size_t sign_quotient = num->sign ^ den->sign; + unsigned normalization_steps; + mpi_limb_t q_limb; + mpi_ptr_t marker[5]; + int markidx = 0; + + memset(marker, 0, sizeof(marker)); + + /* Ensure space is enough for quotient and remainder. + * We need space for an extra limb in the remainder, because it's + * up-shifted (normalized) below. */ + rsize = nsize + 1; + if (mpi_resize(rem, rsize) < 0) + goto nomem; + + qsize = rsize - dsize; /* qsize cannot be bigger than this. */ + if (qsize <= 0) { + if (num != rem) { + rem->nlimbs = num->nlimbs; + rem->sign = num->sign; + MPN_COPY(rem->d, num->d, nsize); + } + if (quot) { + /* This needs to follow the assignment to rem, in case the + * numerator and quotient are the same. */ + quot->nlimbs = 0; + quot->sign = 0; + } + return 0; + } + + if (quot) + if (mpi_resize(quot, qsize) < 0) + goto nomem; + + /* Read pointers here, when reallocation is finished. */ + np = num->d; + dp = den->d; + rp = rem->d; + + /* Optimize division by a single-limb divisor. */ + if (dsize == 1) { + mpi_limb_t rlimb; + if (quot) { + qp = quot->d; + rlimb = mpihelp_divmod_1(qp, np, nsize, dp[0]); + qsize -= qp[qsize - 1] == 0; + quot->nlimbs = qsize; + quot->sign = sign_quotient; + } else + rlimb = mpihelp_mod_1(np, nsize, dp[0]); + rp[0] = rlimb; + rsize = rlimb != 0 ? 1 : 0; + rem->nlimbs = rsize; + rem->sign = sign_remainder; + return 0; + } + + if (quot) { + qp = quot->d; + /* Make sure QP and NP point to different objects. Otherwise the + * numerator would be gradually overwritten by the quotient limbs. */ + if (qp == np) { /* Copy NP object to temporary space. */ + np = marker[markidx++] = mpi_alloc_limb_space(nsize); + MPN_COPY(np, qp, nsize); + } + } else /* Put quotient at top of remainder. */ + qp = rp + dsize; + + count_leading_zeros(normalization_steps, dp[dsize - 1]); + + /* Normalize the denominator, i.e. make its most significant bit set by + * shifting it NORMALIZATION_STEPS bits to the left. Also shift the + * numerator the same number of steps (to keep the quotient the same!). + */ + if (normalization_steps) { + mpi_ptr_t tp; + mpi_limb_t nlimb; + + /* Shift up the denominator setting the most significant bit of + * the most significant word. Use temporary storage not to clobber + * the original contents of the denominator. */ + tp = marker[markidx++] = mpi_alloc_limb_space(dsize); + if (!tp) + goto nomem; + mpihelp_lshift(tp, dp, dsize, normalization_steps); + dp = tp; + + /* Shift up the numerator, possibly introducing a new most + * significant word. Move the shifted numerator in the remainder + * meanwhile. */ + nlimb = mpihelp_lshift(rp, np, nsize, normalization_steps); + if (nlimb) { + rp[nsize] = nlimb; + rsize = nsize + 1; + } else + rsize = nsize; + } else { + /* The denominator is already normalized, as required. Copy it to + * temporary space if it overlaps with the quotient or remainder. */ + if (dp == rp || (quot && (dp == qp))) { + mpi_ptr_t tp; + + tp = marker[markidx++] = mpi_alloc_limb_space(dsize); + if (!tp) + goto nomem; + MPN_COPY(tp, dp, dsize); + dp = tp; + } + + /* Move the numerator to the remainder. */ + if (rp != np) + MPN_COPY(rp, np, nsize); + + rsize = nsize; + } + + q_limb = mpihelp_divrem(qp, 0, rp, rsize, dp, dsize); + + if (quot) { + qsize = rsize - dsize; + if (q_limb) { + qp[qsize] = q_limb; + qsize += 1; + } + + quot->nlimbs = qsize; + quot->sign = sign_quotient; + } + + rsize = dsize; + MPN_NORMALIZE(rp, rsize); + + if (normalization_steps && rsize) { + mpihelp_rshift(rp, rp, rsize, normalization_steps); + rsize -= rp[rsize - 1] == 0 ? 1 : 0; + } + + rem->nlimbs = rsize; + rem->sign = sign_remainder; + + rc = 0; +nomem: + while (markidx) + mpi_free_limb_space(marker[--markidx]); + return rc; +} + +int mpi_tdiv_q_2exp(MPI w, MPI u, unsigned count) +{ + mpi_size_t usize, wsize; + mpi_size_t limb_cnt; + + usize = u->nlimbs; + limb_cnt = count / BITS_PER_MPI_LIMB; + wsize = usize - limb_cnt; + if (limb_cnt >= usize) + w->nlimbs = 0; + else { + mpi_ptr_t wp; + mpi_ptr_t up; + + if (RESIZE_IF_NEEDED(w, wsize) < 0) + return -ENOMEM; + wp = w->d; + up = u->d; + + count %= BITS_PER_MPI_LIMB; + if (count) { + mpihelp_rshift(wp, up + limb_cnt, wsize, count); + wsize -= !wp[wsize - 1]; + } else { + MPN_COPY_INCR(wp, up + limb_cnt, wsize); + } + + w->nlimbs = wsize; + } + return 0; +} + +/**************** + * Check whether dividend is divisible by divisor + * (note: divisor must fit into a limb) + */ +int mpi_divisible_ui(MPI dividend, ulong divisor) +{ + return !mpihelp_mod_1(dividend->d, dividend->nlimbs, divisor); +} diff --git a/lib/mpi/mpi-gcd.c b/lib/mpi/mpi-gcd.c new file mode 100644 index 00000000000..13c48aef9c4 --- /dev/null +++ b/lib/mpi/mpi-gcd.c @@ -0,0 +1,59 @@ +/* mpi-gcd.c - MPI functions + * Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +#include "mpi-internal.h" + +/**************** + * Find the greatest common divisor G of A and B. + * Return: true if this 1, false in all other cases + */ +int mpi_gcd(MPI g, const MPI xa, const MPI xb) +{ + MPI a = NULL, b = NULL; + + if (mpi_copy(&a, xa) < 0) + goto nomem; + + if (mpi_copy(&b, xb) < 0) + goto nomem; + + /* TAOCP Vol II, 4.5.2, Algorithm A */ + a->sign = 0; + b->sign = 0; + while (mpi_cmp_ui(b, 0)) { + if (mpi_fdiv_r(g, a, b) < 0) /* g used as temorary variable */ + goto nomem; + if (mpi_set(a, b) < 0) + goto nomem; + if (mpi_set(b, g) < 0) + goto nomem; + } + if (mpi_set(g, a) < 0) + goto nomem; + + mpi_free(a); + mpi_free(b); + return !mpi_cmp_ui(g, 1); + +nomem: + mpi_free(a); + mpi_free(b); + return -ENOMEM; +} diff --git a/lib/mpi/mpi-inline.c b/lib/mpi/mpi-inline.c new file mode 100644 index 00000000000..654f68aeed8 --- /dev/null +++ b/lib/mpi/mpi-inline.c @@ -0,0 +1,31 @@ +/* mpi-inline.c + * Copyright (C) 1999, 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +/* put the inline functions as real functions into the lib */ +#define G10_MPI_INLINE_DECL + +#include "mpi-internal.h" + +/* always include the header becuase it is only + * included by mpi-internal if __GCC__ is defined but we + * need it here in all cases and the above definition of + * of the macro allows us to do so + */ +#include "mpi-inline.h" diff --git a/lib/mpi/mpi-inv.c b/lib/mpi/mpi-inv.c new file mode 100644 index 00000000000..0951f984774 --- /dev/null +++ b/lib/mpi/mpi-inv.c @@ -0,0 +1,187 @@ +/* mpi-inv.c - MPI functions + * Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +#include "mpi-internal.h" + +/**************** + * Calculate the multiplicative inverse X of A mod N + * That is: Find the solution x for + * 1 = (a*x) mod n + */ +int mpi_invm(MPI x, const MPI a, const MPI n) +{ + /* Extended Euclid's algorithm (See TAOPC Vol II, 4.5.2, Alg X) + * modified according to Michael Penk's solution for Exercice 35 + * with further enhancement */ + MPI u = NULL, v = NULL; + MPI u1 = NULL, u2 = NULL, u3 = NULL; + MPI v1 = NULL, v2 = NULL, v3 = NULL; + MPI t1 = NULL, t2 = NULL, t3 = NULL; + unsigned k; + int sign; + int odd = 0; + int rc = -ENOMEM; + + if (mpi_copy(&u, a) < 0) + goto cleanup; + if (mpi_copy(&v, n) < 0) + goto cleanup; + + for (k = 0; !mpi_test_bit(u, 0) && !mpi_test_bit(v, 0); k++) { + if (mpi_rshift(u, u, 1) < 0) + goto cleanup; + if (mpi_rshift(v, v, 1) < 0) + goto cleanup; + } + odd = mpi_test_bit(v, 0); + + u1 = mpi_alloc_set_ui(1); + if (!u1) + goto cleanup; + if (!odd) { + u2 = mpi_alloc_set_ui(0); + if (!u2) + goto cleanup; + } + if (mpi_copy(&u3, u) < 0) + goto cleanup; + if (mpi_copy(&v1, v) < 0) + goto cleanup; + if (!odd) { + v2 = mpi_alloc(mpi_get_nlimbs(u)); + if (!v2) + goto cleanup; + if (mpi_sub(v2, u1, u) < 0) + goto cleanup; /* U is used as const 1 */ + } + if (mpi_copy(&v3, v) < 0) + goto cleanup; + if (mpi_test_bit(u, 0)) { /* u is odd */ + t1 = mpi_alloc_set_ui(0); + if (!t1) + goto cleanup; + if (!odd) { + t2 = mpi_alloc_set_ui(1); + if (!t2) + goto cleanup; + t2->sign = 1; + } + if (mpi_copy(&t3, v) < 0) + goto cleanup; + t3->sign = !t3->sign; + goto Y4; + } else { + t1 = mpi_alloc_set_ui(1); + if (!t1) + goto cleanup; + if (!odd) { + t2 = mpi_alloc_set_ui(0); + if (!t2) + goto cleanup; + } + if (mpi_copy(&t3, u) < 0) + goto cleanup; + } + do { + do { + if (!odd) { + if (mpi_test_bit(t1, 0) || mpi_test_bit(t2, 0)) { /* one is odd */ + if (mpi_add(t1, t1, v) < 0) + goto cleanup; + if (mpi_sub(t2, t2, u) < 0) + goto cleanup; + } + if (mpi_rshift(t1, t1, 1) < 0) + goto cleanup; + if (mpi_rshift(t2, t2, 1) < 0) + goto cleanup; + if (mpi_rshift(t3, t3, 1) < 0) + goto cleanup; + } else { + if (mpi_test_bit(t1, 0)) + if (mpi_add(t1, t1, v) < 0) + goto cleanup; + if (mpi_rshift(t1, t1, 1) < 0) + goto cleanup; + if (mpi_rshift(t3, t3, 1) < 0) + goto cleanup; + } +Y4: + ; + } while (!mpi_test_bit(t3, 0)); /* while t3 is even */ + + if (!t3->sign) { + if (mpi_set(u1, t1) < 0) + goto cleanup; + if (!odd) + if (mpi_set(u2, t2) < 0) + goto cleanup; + if (mpi_set(u3, t3) < 0) + goto cleanup; + } else { + if (mpi_sub(v1, v, t1) < 0) + goto cleanup; + sign = u->sign; + u->sign = !u->sign; + if (!odd) + if (mpi_sub(v2, u, t2) < 0) + goto cleanup; + u->sign = sign; + sign = t3->sign; + t3->sign = !t3->sign; + if (mpi_set(v3, t3) < 0) + goto cleanup; + t3->sign = sign; + } + if (mpi_sub(t1, u1, v1) < 0) + goto cleanup; + if (!odd) + if (mpi_sub(t2, u2, v2) < 0) + goto cleanup; + if (mpi_sub(t3, u3, v3) < 0) + goto cleanup; + if (t1->sign) { + if (mpi_add(t1, t1, v) < 0) + goto cleanup; + if (!odd) + if (mpi_sub(t2, t2, u) < 0) + goto cleanup; + } + } while (mpi_cmp_ui(t3, 0)); /* while t3 != 0 */ + /* mpi_lshift( u3, k ); */ + rc = mpi_set(x, u1); + +cleanup: + mpi_free(u1); + mpi_free(v1); + mpi_free(t1); + if (!odd) { + mpi_free(u2); + mpi_free(v2); + mpi_free(t2); + } + mpi_free(u3); + mpi_free(v3); + mpi_free(t3); + + mpi_free(u); + mpi_free(v); + return rc; +} diff --git a/lib/mpi/mpi-mpow.c b/lib/mpi/mpi-mpow.c new file mode 100644 index 00000000000..4cc75933c5a --- /dev/null +++ b/lib/mpi/mpi-mpow.c @@ -0,0 +1,133 @@ +/* mpi-mpow.c - MPI functions + * Copyright (C) 1998, 1999, 2000 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +#include "mpi-internal.h" +#include "longlong.h" + +static int build_index(const MPI *exparray, int k, int i, int t) +{ + int j, bitno; + int index = 0; + + bitno = t - i; + for (j = k - 1; j >= 0; j--) { + index <<= 1; + if (mpi_test_bit(exparray[j], bitno)) + index |= 1; + } + return index; +} + +/**************** + * RES = (BASE[0] ^ EXP[0]) * (BASE[1] ^ EXP[1]) * ... * mod M + */ +int mpi_mulpowm(MPI res, MPI *basearray, MPI *exparray, MPI m) +{ + int rc = -ENOMEM; + int k; /* number of elements */ + int t; /* bit size of largest exponent */ + int i, j, idx; + MPI *G = NULL; /* table with precomputed values of size 2^k */ + MPI tmp = NULL; + + for (k = 0; basearray[k]; k++) + ; + if (!k) { + pr_emerg("mpi_mulpowm: assert(k) failed\n"); + BUG(); + } + for (t = 0, i = 0; (tmp = exparray[i]); i++) { + j = mpi_get_nbits(tmp); + if (j > t) + t = j; + } + if (i != k) { + pr_emerg("mpi_mulpowm: assert(i==k) failed\n"); + BUG(); + } + if (!t) { + pr_emerg("mpi_mulpowm: assert(t) failed\n"); + BUG(); + } + if (k >= 10) { + pr_emerg("mpi_mulpowm: assert(k<10) failed\n"); + BUG(); + } + + G = kzalloc((1 << k) * sizeof *G, GFP_KERNEL); + if (!G) + goto nomem; + + /* and calculate */ + tmp = mpi_alloc(mpi_get_nlimbs(m) + 1); + if (!tmp) + goto nomem; + if (mpi_set_ui(res, 1) < 0) + goto nomem; + for (i = 1; i <= t; i++) { + if (mpi_mulm(tmp, res, res, m) < 0) + goto nomem; + idx = build_index(exparray, k, i, t); + if (!(idx >= 0 && idx < (1 << k))) { + pr_emerg("mpi_mulpowm: assert(idx >= 0 && idx < (1<nlimbs; + sign = mult->sign; + + if (!size || !small_mult) { + prod->nlimbs = 0; + prod->sign = 0; + return 0; + } + + prod_size = size + 1; + if (prod->alloced < prod_size) + if (mpi_resize(prod, prod_size) < 0) + return -ENOMEM; + prod_ptr = prod->d; + + cy = mpihelp_mul_1(prod_ptr, mult->d, size, (mpi_limb_t) small_mult); + if (cy) + prod_ptr[size++] = cy; + prod->nlimbs = size; + prod->sign = sign; + return 0; +} + +int mpi_mul_2exp(MPI w, MPI u, unsigned long cnt) +{ + mpi_size_t usize, wsize, limb_cnt; + mpi_ptr_t wp; + mpi_limb_t wlimb; + int usign, wsign; + + usize = u->nlimbs; + usign = u->sign; + + if (!usize) { + w->nlimbs = 0; + w->sign = 0; + return 0; + } + + limb_cnt = cnt / BITS_PER_MPI_LIMB; + wsize = usize + limb_cnt + 1; + if (w->alloced < wsize) + if (mpi_resize(w, wsize) < 0) + return -ENOMEM; + wp = w->d; + wsize = usize + limb_cnt; + wsign = usign; + + cnt %= BITS_PER_MPI_LIMB; + if (cnt) { + wlimb = mpihelp_lshift(wp + limb_cnt, u->d, usize, cnt); + if (wlimb) { + wp[wsize] = wlimb; + wsize++; + } + } else { + MPN_COPY_DECR(wp + limb_cnt, u->d, usize); + } + + /* Zero all whole limbs at low end. Do it here and not before calling + * mpn_lshift, not to lose for U == W. */ + MPN_ZERO(wp, limb_cnt); + + w->nlimbs = wsize; + w->sign = wsign; + return 0; +} + +int mpi_mul(MPI w, MPI u, MPI v) +{ + int rc = -ENOMEM; + mpi_size_t usize, vsize, wsize; + mpi_ptr_t up, vp, wp; + mpi_limb_t cy; + int usign, vsign, sign_product; + int assign_wp = 0; + mpi_ptr_t tmp_limb = NULL; + + if (u->nlimbs < v->nlimbs) { /* Swap U and V. */ + usize = v->nlimbs; + usign = v->sign; + up = v->d; + vsize = u->nlimbs; + vsign = u->sign; + vp = u->d; + } else { + usize = u->nlimbs; + usign = u->sign; + up = u->d; + vsize = v->nlimbs; + vsign = v->sign; + vp = v->d; + } + sign_product = usign ^ vsign; + wp = w->d; + + /* Ensure W has space enough to store the result. */ + wsize = usize + vsize; + if (w->alloced < (size_t) wsize) { + if (wp == up || wp == vp) { + wp = mpi_alloc_limb_space(wsize); + if (!wp) + goto nomem; + assign_wp = 1; + } else { + if (mpi_resize(w, wsize) < 0) + goto nomem; + wp = w->d; + } + } else { /* Make U and V not overlap with W. */ + if (wp == up) { + /* W and U are identical. Allocate temporary space for U. */ + up = tmp_limb = mpi_alloc_limb_space(usize); + if (!up) + goto nomem; + /* Is V identical too? Keep it identical with U. */ + if (wp == vp) + vp = up; + /* Copy to the temporary space. */ + MPN_COPY(up, wp, usize); + } else if (wp == vp) { + /* W and V are identical. Allocate temporary space for V. */ + vp = tmp_limb = mpi_alloc_limb_space(vsize); + if (!vp) + goto nomem; + /* Copy to the temporary space. */ + MPN_COPY(vp, wp, vsize); + } + } + + if (!vsize) + wsize = 0; + else { + if (mpihelp_mul(wp, up, usize, vp, vsize, &cy) < 0) + goto nomem; + wsize -= cy ? 0 : 1; + } + + if (assign_wp) + mpi_assign_limb_space(w, wp, wsize); + + w->nlimbs = wsize; + w->sign = sign_product; + rc = 0; +nomem: + if (tmp_limb) + mpi_free_limb_space(tmp_limb); + return rc; +} + +int mpi_mulm(MPI w, MPI u, MPI v, MPI m) +{ + if (mpi_mul(w, u, v) < 0) + return -ENOMEM; + return mpi_fdiv_r(w, w, m); +} diff --git a/lib/mpi/mpi-scan.c b/lib/mpi/mpi-scan.c new file mode 100644 index 00000000000..b2da5ad9619 --- /dev/null +++ b/lib/mpi/mpi-scan.c @@ -0,0 +1,136 @@ +/* mpi-scan.c - MPI functions + * Copyright (C) 1998, 1999, 2000, 2001 Free Software Foundation, Inc. + * + * This file is part of GnuPG. + * + * GnuPG is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * GnuPG 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 General Public License for more details. + * + * You should have received a copy of the GNU 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 + */ + +#include "mpi-internal.h" +#include "longlong.h" + +/**************** + * Scan through an mpi and return byte for byte. a -1 is returned to indicate + * the end of the mpi. Scanning is done from the lsb to the msb, returned + * values are in the range of 0 .. 255. + * + * FIXME: This code is VERY ugly! + */ +int mpi_getbyte(const MPI a, unsigned idx) +{ + int i, j; + unsigned n; + mpi_ptr_t ap; + mpi_limb_t limb; + + ap = a->d; + for (n = 0, i = 0; i < a->nlimbs; i++) { + limb = ap[i]; + for (j = 0; j < BYTES_PER_MPI_LIMB; j++, n++) + if (n == idx) + return (limb >> j * 8) & 0xff; + } + return -1; +} + +/**************** + * Put a value at position IDX into A. idx counts from lsb to msb + */ +void mpi_putbyte(MPI a, unsigned idx, int xc) +{ + int i, j; + unsigned n; + mpi_ptr_t ap; + mpi_limb_t limb, c; + + c = xc & 0xff; + ap = a->d; + for (n = 0, i = 0; i < a->alloced; i++) { + limb = ap[i]; + for (j = 0; j < BYTES_PER_MPI_LIMB; j++, n++) + if (n == idx) { +#if BYTES_PER_MPI_LIMB == 4 + if (j == 0) + limb = (limb & 0xffffff00) | c; + else if (j == 1) + limb = (limb & 0xffff00ff) | (c << 8); + else if (j == 2) + limb = (limb & 0xff00ffff) | (c << 16); + else + limb = (limb & 0x00ffffff) | (c << 24); +#elif BYTES_PER_MPI_LIMB == 8 + if (j == 0) + limb = (limb & 0xffffffffffffff00) | c; + else if (j == 1) + limb = + (limb & 0xffffffffffff00ff) | (c << + 8); + else if (j == 2) + limb = + (limb & 0xffffffffff00ffff) | (c << + 16); + else if (j == 3) + limb = + (limb & 0xffffffff00ffffff) | (c << + 24); + else if (j == 4) + limb = + (limb & 0xffffff00ffffffff) | (c << + 32); + else if (j == 5) + limb = + (limb & 0xffff00ffffffffff) | (c << + 40); + else if (j == 6) + limb = + (limb & 0xff00ffffffffffff) | (c << + 48); + else + limb = + (limb & 0x00ffffffffffffff) | (c << + 56); +#else +#error please enhance this function, its ugly - i know. +#endif + if (a->nlimbs <= i) + a->nlimbs = i + 1; + ap[i] = limb; + return; + } + } + log_bug("index out of range\n"); +} + +/**************** + * Count the number of zerobits at the low end of A + */ +unsigned mpi_trailing_zeros(const MPI a) +{ + unsigned n, count = 0; + + for (n = 0; n < a->nlimbs; n++) { + if (a->d[n]) { + unsigned nn; + mpi_limb_t alimb = a->d[n]; + + count_trailing_zeros(nn, alimb); + count += nn; + break; + } + count += BITS_PER_MPI_LIMB; + } + return count; + +} -- cgit v1.2.3-70-g09d2 From 051dbb918c7fb7da8e64a2cd0d804ba73399709f Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Fri, 14 Oct 2011 15:25:16 +0300 Subject: crypto: digital signature verification support This patch implements RSA digital signature verification using GnuPG library. The format of the signature and the public key is defined by their respective headers. The signature header contains version information, algorithm, and keyid, which was used to generate the signature. The key header contains version and algorythim type. The payload of the signature and the key are multi-precision integers. The signing and key management utilities evm-utils provide functionality to generate signatures and load keys into the kernel keyring. When the key is added to the kernel keyring, the keyid defines the name of the key. Signed-off-by: Dmitry Kasatkin Acked-by: Mimi Zohar --- Documentation/digsig.txt | 96 ++++++++++++++++ include/linux/digsig.h | 64 +++++++++++ lib/Kconfig | 8 ++ lib/Makefile | 1 + lib/digsig.c | 284 +++++++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 453 insertions(+) create mode 100644 Documentation/digsig.txt create mode 100644 include/linux/digsig.h create mode 100644 lib/digsig.c (limited to 'lib') diff --git a/Documentation/digsig.txt b/Documentation/digsig.txt new file mode 100644 index 00000000000..3f682889068 --- /dev/null +++ b/Documentation/digsig.txt @@ -0,0 +1,96 @@ +Digital Signature Verification API + +CONTENTS + +1. Introduction +2. API +3. User-space utilities + + +1. Introduction + +Digital signature verification API provides a method to verify digital signature. +Currently digital signatures are used by the IMA/EVM integrity protection subsystem. + +Digital signature verification is implemented using cut-down kernel port of +GnuPG multi-precision integers (MPI) library. The kernel port provides +memory allocation errors handling, has been refactored according to kernel +coding style, and checkpatch.pl reported errors and warnings have been fixed. + +Public key and signature consist of header and MPIs. + +struct pubkey_hdr { + uint8_t version; /* key format version */ + time_t timestamp; /* key made, always 0 for now */ + uint8_t algo; + uint8_t nmpi; + char mpi[0]; +} __packed; + +struct signature_hdr { + uint8_t version; /* signature format version */ + time_t timestamp; /* signature made */ + uint8_t algo; + uint8_t hash; + uint8_t keyid[8]; + uint8_t nmpi; + char mpi[0]; +} __packed; + +keyid equals to SHA1[12-19] over the total key content. +Signature header is used as an input to generate a signature. +Such approach insures that key or signature header could not be changed. +It protects timestamp from been changed and can be used for rollback +protection. + +2. API + +API currently includes only 1 function: + + digsig_verify() - digital signature verification with public key + + +/** + * digsig_verify() - digital signature verification with public key + * @keyring: keyring to search key in + * @sig: digital signature + * @sigen: length of the signature + * @data: data + * @datalen: length of the data + * @return: 0 on success, -EINVAL otherwise + * + * Verifies data integrity against digital signature. + * Currently only RSA is supported. + * Normally hash of the content is used as a data for this function. + * + */ +int digsig_verify(struct key *keyring, const char *sig, int siglen, + const char *data, int datalen); + +3. User-space utilities + +The signing and key management utilities evm-utils provide functionality +to generate signatures, to load keys into the kernel keyring. +Keys can be in PEM or converted to the kernel format. +When the key is added to the kernel keyring, the keyid defines the name +of the key: 5D2B05FC633EE3E8 in the example bellow. + +Here is example output of the keyctl utility. + +$ keyctl show +Session Keyring + -3 --alswrv 0 0 keyring: _ses +603976250 --alswrv 0 -1 \_ keyring: _uid.0 +817777377 --alswrv 0 0 \_ user: kmk +891974900 --alswrv 0 0 \_ encrypted: evm-key +170323636 --alswrv 0 0 \_ keyring: _module +548221616 --alswrv 0 0 \_ keyring: _ima +128198054 --alswrv 0 0 \_ keyring: _evm + +$ keyctl list 128198054 +1 key in keyring: +620789745: --alswrv 0 0 user: 5D2B05FC633EE3E8 + + +Dmitry Kasatkin +06.10.2011 diff --git a/include/linux/digsig.h b/include/linux/digsig.h new file mode 100644 index 00000000000..efae755017d --- /dev/null +++ b/include/linux/digsig.h @@ -0,0 +1,64 @@ +/* + * Copyright (C) 2011 Nokia Corporation + * Copyright (C) 2011 Intel Corporation + * + * Author: + * Dmitry Kasatkin + * + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, version 2 of the License. + * + */ + +#ifndef _DIGSIG_H +#define _DIGSIG_H + +#include + +enum pubkey_algo { + PUBKEY_ALGO_RSA, + PUBKEY_ALGO_MAX, +}; + +enum digest_algo { + DIGEST_ALGO_SHA1, + DIGEST_ALGO_SHA256, + DIGEST_ALGO_MAX +}; + +struct pubkey_hdr { + uint8_t version; /* key format version */ + time_t timestamp; /* key made, always 0 for now */ + uint8_t algo; + uint8_t nmpi; + char mpi[0]; +} __packed; + +struct signature_hdr { + uint8_t version; /* signature format version */ + time_t timestamp; /* signature made */ + uint8_t algo; + uint8_t hash; + uint8_t keyid[8]; + uint8_t nmpi; + char mpi[0]; +} __packed; + +#if defined(CONFIG_DIGSIG) || defined(CONFIG_DIGSIG_MODULE) + +int digsig_verify(struct key *keyring, const char *sig, int siglen, + const char *digest, int digestlen); + +#else + +static inline int digsig_verify(struct key *keyring, const char *sig, + int siglen, const char *digest, int digestlen) +{ + return -EOPNOTSUPP; +} + +#endif /* CONFIG_DIGSIG */ + +#endif /* _DIGSIG_H */ diff --git a/lib/Kconfig b/lib/Kconfig index d411de53be4..c1a89185fe6 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -293,4 +293,12 @@ config MPILIB_EXTRA This code in unnecessary for RSA digital signature verification, and can be compiled if needed. +config DIGSIG + tristate "In-kernel signature checker" + depends on CRYPTO + select MPILIB + help + Digital signature verification. Currently only RSA is supported. + Implementation is done using GnuPG MPI library + endmenu diff --git a/lib/Makefile b/lib/Makefile index 6ba8cbf4c72..0f5cff267af 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -116,6 +116,7 @@ obj-$(CONFIG_CPU_RMAP) += cpu_rmap.o obj-$(CONFIG_CORDIC) += cordic.o obj-$(CONFIG_MPILIB) += mpi/ +obj-$(CONFIG_DIGSIG) += digsig.o hostprogs-y := gen_crc32table clean-files := crc32table.h diff --git a/lib/digsig.c b/lib/digsig.c new file mode 100644 index 00000000000..fd2402f67f8 --- /dev/null +++ b/lib/digsig.c @@ -0,0 +1,284 @@ +/* + * Copyright (C) 2011 Nokia Corporation + * Copyright (C) 2011 Intel Corporation + * + * Author: + * Dmitry Kasatkin + * + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation, version 2 of the License. + * + * File: sign.c + * implements signature (RSA) verification + * pkcs decoding is based on LibTomCrypt code + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +static struct crypto_shash *shash; + +static int pkcs_1_v1_5_decode_emsa(const unsigned char *msg, + unsigned long msglen, + unsigned long modulus_bitlen, + unsigned char *out, + unsigned long *outlen, + int *is_valid) +{ + unsigned long modulus_len, ps_len, i; + int result; + + /* default to invalid packet */ + *is_valid = 0; + + modulus_len = (modulus_bitlen >> 3) + (modulus_bitlen & 7 ? 1 : 0); + + /* test message size */ + if ((msglen > modulus_len) || (modulus_len < 11)) + return -EINVAL; + + /* separate encoded message */ + if ((msg[0] != 0x00) || (msg[1] != (unsigned char)1)) { + result = -EINVAL; + goto bail; + } + + for (i = 2; i < modulus_len - 1; i++) + if (msg[i] != 0xFF) + break; + + /* separator check */ + if (msg[i] != 0) { + /* There was no octet with hexadecimal value 0x00 + to separate ps from m. */ + result = -EINVAL; + goto bail; + } + + ps_len = i - 2; + + if (*outlen < (msglen - (2 + ps_len + 1))) { + *outlen = msglen - (2 + ps_len + 1); + result = -EOVERFLOW; + goto bail; + } + + *outlen = (msglen - (2 + ps_len + 1)); + memcpy(out, &msg[2 + ps_len + 1], *outlen); + + /* valid packet */ + *is_valid = 1; + result = 0; +bail: + return result; +} + +/* + * RSA Signature verification with public key + */ +static int digsig_verify_rsa(struct key *key, + const char *sig, int siglen, + const char *h, int hlen) +{ + int err = -EINVAL; + unsigned long len; + unsigned long mlen, mblen; + unsigned nret, l; + int valid, head, i; + unsigned char *out1 = NULL, *out2 = NULL; + MPI in = NULL, res = NULL, pkey[2]; + uint8_t *p, *datap, *endp; + struct user_key_payload *ukp; + struct pubkey_hdr *pkh; + + down_read(&key->sem); + ukp = key->payload.data; + pkh = (struct pubkey_hdr *)ukp->data; + + if (pkh->version != 1) + goto err1; + + if (pkh->algo != PUBKEY_ALGO_RSA) + goto err1; + + if (pkh->nmpi != 2) + goto err1; + + datap = pkh->mpi; + endp = datap + ukp->datalen; + + for (i = 0; i < pkh->nmpi; i++) { + unsigned int remaining = endp - datap; + pkey[i] = mpi_read_from_buffer(datap, &remaining); + datap += remaining; + } + + mblen = mpi_get_nbits(pkey[0]); + mlen = (mblen + 7)/8; + + err = -ENOMEM; + + out1 = kzalloc(mlen, GFP_KERNEL); + if (!out1) + goto err; + + out2 = kzalloc(mlen, GFP_KERNEL); + if (!out2) + goto err; + + nret = siglen; + in = mpi_read_from_buffer(sig, &nret); + if (!in) + goto err; + + res = mpi_alloc(mpi_get_nlimbs(in) * 2); + if (!res) + goto err; + + err = mpi_powm(res, in, pkey[1], pkey[0]); + if (err) + goto err; + + if (mpi_get_nlimbs(res) * BYTES_PER_MPI_LIMB > mlen) { + err = -EINVAL; + goto err; + } + + p = mpi_get_buffer(res, &l, NULL); + if (!p) { + err = -EINVAL; + goto err; + } + + len = mlen; + head = len - l; + memset(out1, 0, head); + memcpy(out1 + head, p, l); + + err = -EINVAL; + pkcs_1_v1_5_decode_emsa(out1, len, mblen, out2, &len, &valid); + + if (valid && len == hlen) + err = memcmp(out2, h, hlen); + +err: + mpi_free(in); + mpi_free(res); + kfree(out1); + kfree(out2); + mpi_free(pkey[0]); + mpi_free(pkey[1]); +err1: + up_read(&key->sem); + + return err; +} + +/** + * digsig_verify() - digital signature verification with public key + * @keyring: keyring to search key in + * @sig: digital signature + * @sigen: length of the signature + * @data: data + * @datalen: length of the data + * @return: 0 on success, -EINVAL otherwise + * + * Verifies data integrity against digital signature. + * Currently only RSA is supported. + * Normally hash of the content is used as a data for this function. + * + */ +int digsig_verify(struct key *keyring, const char *sig, int siglen, + const char *data, int datalen) +{ + int err = -ENOMEM; + struct signature_hdr *sh = (struct signature_hdr *)sig; + struct shash_desc *desc = NULL; + unsigned char hash[SHA1_DIGEST_SIZE]; + struct key *key; + char name[20]; + + if (siglen < sizeof(*sh) + 2) + return -EINVAL; + + if (sh->algo != PUBKEY_ALGO_RSA) + return -ENOTSUPP; + + sprintf(name, "%llX", __be64_to_cpup((uint64_t *)sh->keyid)); + + if (keyring) { + /* search in specific keyring */ + key_ref_t kref; + kref = keyring_search(make_key_ref(keyring, 1UL), + &key_type_user, name); + if (IS_ERR(kref)) + key = ERR_PTR(PTR_ERR(kref)); + else + key = key_ref_to_ptr(kref); + } else { + key = request_key(&key_type_user, name, NULL); + } + if (IS_ERR(key)) { + pr_err("key not found, id: %s\n", name); + return PTR_ERR(key); + } + + desc = kzalloc(sizeof(*desc) + crypto_shash_descsize(shash), + GFP_KERNEL); + if (!desc) + goto err; + + desc->tfm = shash; + desc->flags = CRYPTO_TFM_REQ_MAY_SLEEP; + + crypto_shash_init(desc); + crypto_shash_update(desc, data, datalen); + crypto_shash_update(desc, sig, sizeof(*sh)); + crypto_shash_final(desc, hash); + + kfree(desc); + + /* pass signature mpis address */ + err = digsig_verify_rsa(key, sig + sizeof(*sh), siglen - sizeof(*sh), + hash, sizeof(hash)); + +err: + key_put(key); + + return err ? -EINVAL : 0; +} +EXPORT_SYMBOL_GPL(digsig_verify); + +static int __init digsig_init(void) +{ + shash = crypto_alloc_shash("sha1", 0, 0); + if (IS_ERR(shash)) { + pr_err("shash allocation failed\n"); + return PTR_ERR(shash); + } + + return 0; + +} + +static void __exit digsig_cleanup(void) +{ + crypto_free_shash(shash); +} + +module_init(digsig_init); +module_exit(digsig_cleanup); + +MODULE_LICENSE("GPL"); -- cgit v1.2.3-70-g09d2 From 571806a9f70fc4a4c575ab796555dfb401c2dfc1 Mon Sep 17 00:00:00 2001 From: Wolfram Sang Date: Tue, 25 Oct 2011 15:03:42 +0200 Subject: lib: devres: add annotations for #endif While working on devres, I found those make navigating the code a tad easier. Signed-off-by: Wolfram Sang Acked-by: Grant Likely Signed-off-by: Greg Kroah-Hartman --- lib/devres.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/devres.c b/lib/devres.c index 7c0e953a748..78777aea5b3 100644 --- a/lib/devres.c +++ b/lib/devres.c @@ -348,5 +348,5 @@ void pcim_iounmap_regions(struct pci_dev *pdev, u16 mask) } } EXPORT_SYMBOL(pcim_iounmap_regions); -#endif -#endif +#endif /* CONFIG_PCI */ +#endif /* CONFIG_HAS_IOPORT */ -- cgit v1.2.3-70-g09d2 From 72f8c0bfa0de64c68ee59f40eb9b2683bffffbb0 Mon Sep 17 00:00:00 2001 From: Wolfram Sang Date: Tue, 25 Oct 2011 15:16:47 +0200 Subject: lib: devres: add convenience function to remap a resource Almost every platform_driver does the three steps get_resource, request_mem_region, ioremap. This does not only lead to a lot of code duplication, but also a huge number of similar error strings and inconsistent error codes on failure. So, introduce a helper function which simplifies remapping a resource and make it hard to do something wrong and add documentation for it. Signed-off-by: Wolfram Sang Acked-by: Grant Likely Acked-by: Tejun Heo Signed-off-by: Greg Kroah-Hartman --- Documentation/driver-model/devres.txt | 1 + include/linux/device.h | 3 +++ lib/devres.c | 51 +++++++++++++++++++++++++++++++++++ 3 files changed, 55 insertions(+) (limited to 'lib') diff --git a/Documentation/driver-model/devres.txt b/Documentation/driver-model/devres.txt index d79aead9418..10c64c8a13d 100644 --- a/Documentation/driver-model/devres.txt +++ b/Documentation/driver-model/devres.txt @@ -262,6 +262,7 @@ IOMAP devm_ioremap() devm_ioremap_nocache() devm_iounmap() + devm_request_and_ioremap() : checks resource, requests region, ioremaps pcim_iomap() pcim_iounmap() pcim_iomap_table() : array of mapped addresses indexed by BAR diff --git a/include/linux/device.h b/include/linux/device.h index ffbcf95cd97..c6335982774 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -490,6 +490,9 @@ extern int devres_release_group(struct device *dev, void *id); extern void *devm_kzalloc(struct device *dev, size_t size, gfp_t gfp); extern void devm_kfree(struct device *dev, void *p); +void __iomem *devm_request_and_ioremap(struct device *dev, + struct resource *res); + struct device_dma_parameters { /* * a low level driver may set these to teach IOMMU code about diff --git a/lib/devres.c b/lib/devres.c index 78777aea5b3..4fbc09e6e9e 100644 --- a/lib/devres.c +++ b/lib/devres.c @@ -85,6 +85,57 @@ void devm_iounmap(struct device *dev, void __iomem *addr) } EXPORT_SYMBOL(devm_iounmap); +/** + * devm_request_and_ioremap() - Check, request region, and ioremap resource + * @dev: Generic device to handle the resource for + * @res: resource to be handled + * + * Takes all necessary steps to ioremap a mem resource. Uses managed device, so + * everything is undone on driver detach. Checks arguments, so you can feed + * it the result from e.g. platform_get_resource() directly. Returns the + * remapped pointer or NULL on error. Usage example: + * + * res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + * base = devm_request_and_ioremap(&pdev->dev, res); + * if (!base) + * return -EADDRNOTAVAIL; + */ +void __iomem *devm_request_and_ioremap(struct device *dev, + struct resource *res) +{ + resource_size_t size; + const char *name; + void __iomem *dest_ptr; + + BUG_ON(!dev); + + if (!res || resource_type(res) != IORESOURCE_MEM) { + dev_err(dev, "invalid resource\n"); + return NULL; + } + + size = resource_size(res); + name = res->name ?: dev_name(dev); + + if (!devm_request_mem_region(dev, res->start, size, name)) { + dev_err(dev, "can't request region for resource %pR\n", res); + return NULL; + } + + if (res->flags & IORESOURCE_CACHEABLE) + dest_ptr = devm_ioremap(dev, res->start, size); + else + dest_ptr = devm_ioremap_nocache(dev, res->start, size); + + if (!dest_ptr) { + dev_err(dev, "ioremap failed for resource %pR\n", res); + devm_release_mem_region(dev, res->start, size); + } + + return dest_ptr; +} +EXPORT_SYMBOL(devm_request_and_ioremap); + #ifdef CONFIG_HAS_IOPORT /* * Generic iomap devres -- cgit v1.2.3-70-g09d2 From de353533753e048b5c4658f0a42365937527ac45 Mon Sep 17 00:00:00 2001 From: Dmitry Kasatkin Date: Mon, 21 Nov 2011 17:31:15 +0200 Subject: digsig: build dependency fix Fix build errors by adding Kconfig dependency on KEYS. CRYPTO dependency removed. CC security/integrity/digsig.o security/integrity/digsig.c: In function ?integrity_digsig_verify?: security/integrity/digsig.c:38:4: error: implicit declaration of function ?request_key? security/integrity/digsig.c:38:17: error: ?key_type_keyring? undeclared (first use in this function) security/integrity/digsig.c:38:17: note: each undeclared identifier is reported only once for each function it appears in make[2]: *** [security/integrity/digsig.o] Error 1 Reported-by: Randy Dunlap Signed-off-by: Dmitry Kasatkin Signed-off-by: James Morris --- lib/Kconfig | 2 +- security/integrity/Kconfig | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) (limited to 'lib') diff --git a/lib/Kconfig b/lib/Kconfig index c1a89185fe6..5634e473ff6 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -295,7 +295,7 @@ config MPILIB_EXTRA config DIGSIG tristate "In-kernel signature checker" - depends on CRYPTO + depends on KEYS select MPILIB help Digital signature verification. Currently only RSA is supported. diff --git a/security/integrity/Kconfig b/security/integrity/Kconfig index d87fa2a8fa3..d384ea92148 100644 --- a/security/integrity/Kconfig +++ b/security/integrity/Kconfig @@ -5,7 +5,7 @@ config INTEGRITY config INTEGRITY_DIGSIG boolean "Digital signature verification using multiple keyrings" - depends on INTEGRITY + depends on INTEGRITY && KEYS default n select DIGSIG help -- cgit v1.2.3-70-g09d2 From 4673ca8eb3690832e76371371955a8b02e1f59d4 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 24 Nov 2011 14:54:28 +0200 Subject: lib: move GENERIC_IOMAP to lib/Kconfig define GENERIC_IOMAP in a central location instead of all architectures. This will be helpful for the follow-up patch which makes it select other configs. Code is also a bit shorter this way. Signed-off-by: Michael S. Tsirkin --- arch/alpha/Kconfig | 4 ---- arch/cris/Kconfig | 5 +---- arch/hexagon/Kconfig | 4 +--- arch/ia64/Kconfig | 5 +---- arch/m68k/Kconfig | 4 +--- arch/openrisc/Kconfig | 3 --- arch/powerpc/platforms/Kconfig | 3 --- arch/score/Kconfig | 4 +--- arch/sh/Kconfig | 3 --- arch/unicore32/Kconfig | 4 +--- arch/x86/Kconfig | 4 +--- lib/Kconfig | 3 +++ 12 files changed, 10 insertions(+), 36 deletions(-) (limited to 'lib') diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig index 3d74801a401..3636b11ddff 100644 --- a/arch/alpha/Kconfig +++ b/arch/alpha/Kconfig @@ -70,10 +70,6 @@ config GENERIC_ISA_DMA bool default y -config GENERIC_IOMAP - bool - default n - source "init/Kconfig" source "kernel/Kconfig.freezer" diff --git a/arch/cris/Kconfig b/arch/cris/Kconfig index 408b055c585..b3abfb08aa5 100644 --- a/arch/cris/Kconfig +++ b/arch/cris/Kconfig @@ -19,10 +19,6 @@ config GENERIC_CMOS_UPDATE config ARCH_USES_GETTIMEOFFSET def_bool n -config GENERIC_IOMAP - bool - default y - config ARCH_HAS_ILOG2_U32 bool default n @@ -52,6 +48,7 @@ config CRIS select HAVE_IDE select HAVE_GENERIC_HARDIRQS select GENERIC_IRQ_SHOW + select GENERIC_IOMAP config HZ int diff --git a/arch/hexagon/Kconfig b/arch/hexagon/Kconfig index 02513c2dd5e..9059e390588 100644 --- a/arch/hexagon/Kconfig +++ b/arch/hexagon/Kconfig @@ -26,6 +26,7 @@ config HEXAGON select HAVE_ARCH_KGDB select HAVE_ARCH_TRACEHOOK select NO_IOPORT + select GENERIC_IOMAP # mostly generic routines, with some accelerated ones ---help--- Qualcomm Hexagon is a processor architecture designed for high @@ -73,9 +74,6 @@ config GENERIC_CSUM config GENERIC_IRQ_PROBE def_bool y -config GENERIC_IOMAP - def_bool y - #config ZONE_DMA # bool # default y diff --git a/arch/ia64/Kconfig b/arch/ia64/Kconfig index 27489b6dd53..2732e1b0aa3 100644 --- a/arch/ia64/Kconfig +++ b/arch/ia64/Kconfig @@ -29,6 +29,7 @@ config IA64 select GENERIC_IRQ_SHOW select ARCH_WANT_OPTIONAL_GPIOLIB select ARCH_HAVE_NMI_SAFE_CMPXCHG + select GENERIC_IOMAP default y help The Itanium Processor Family is Intel's 64-bit successor to @@ -102,10 +103,6 @@ config EFI bool default y -config GENERIC_IOMAP - bool - default y - config ARCH_CLOCKSOURCE_DATA def_bool y diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig index 361d54019bb..973e68614f2 100644 --- a/arch/m68k/Kconfig +++ b/arch/m68k/Kconfig @@ -38,9 +38,6 @@ config GENERIC_CALIBRATE_DELAY bool default y -config GENERIC_IOMAP - def_bool MMU - config TIME_LOW_RES bool default y @@ -73,6 +70,7 @@ source "kernel/Kconfig.freezer" config MMU bool "MMU-based Paged Memory Management Support" default y + select GENERIC_IOMAP help Select if you want MMU-based virtualised addressing space support by paged memory management. If unsure, say 'Y'. diff --git a/arch/openrisc/Kconfig b/arch/openrisc/Kconfig index e518a5a4cf4..081a54f1a93 100644 --- a/arch/openrisc/Kconfig +++ b/arch/openrisc/Kconfig @@ -38,9 +38,6 @@ config RWSEM_XCHGADD_ALGORITHM config GENERIC_HWEIGHT def_bool y -config GENERIC_IOMAP - def_bool y - config NO_IOPORT def_bool y diff --git a/arch/powerpc/platforms/Kconfig b/arch/powerpc/platforms/Kconfig index 3fe6d927ad7..100feed9794 100644 --- a/arch/powerpc/platforms/Kconfig +++ b/arch/powerpc/platforms/Kconfig @@ -175,9 +175,6 @@ config PPC_INDIRECT_MMIO config PPC_IO_WORKAROUNDS bool -config GENERIC_IOMAP - bool - source "drivers/cpufreq/Kconfig" menu "CPU Frequency drivers" diff --git a/arch/score/Kconfig b/arch/score/Kconfig index df169e84db4..455ce2d7682 100644 --- a/arch/score/Kconfig +++ b/arch/score/Kconfig @@ -4,6 +4,7 @@ config SCORE def_bool y select HAVE_GENERIC_HARDIRQS select GENERIC_IRQ_SHOW + select GENERIC_IOMAP choice prompt "System type" @@ -33,9 +34,6 @@ endmenu config CPU_SCORE7 bool -config GENERIC_IOMAP - def_bool y - config NO_DMA bool default y diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig index 5629e209913..5aeab58ac8b 100644 --- a/arch/sh/Kconfig +++ b/arch/sh/Kconfig @@ -84,9 +84,6 @@ config GENERIC_GPIO config GENERIC_CALIBRATE_DELAY bool -config GENERIC_IOMAP - bool - config GENERIC_CLOCKEVENTS def_bool y diff --git a/arch/unicore32/Kconfig b/arch/unicore32/Kconfig index 942ed6174f1..eeb8054c7cd 100644 --- a/arch/unicore32/Kconfig +++ b/arch/unicore32/Kconfig @@ -12,6 +12,7 @@ config UNICORE32 select GENERIC_IRQ_PROBE select GENERIC_IRQ_SHOW select ARCH_WANT_FRAME_POINTERS + select GENERIC_IOMAP help UniCore-32 is 32-bit Instruction Set Architecture, including a series of low-power-consumption RISC chip @@ -30,9 +31,6 @@ config GENERIC_CLOCKEVENTS config GENERIC_CSUM def_bool y -config GENERIC_IOMAP - def_bool y - config NO_IOPORT bool diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index cb9a1044a77..08af6457de7 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -75,6 +75,7 @@ config X86 select HAVE_BPF_JIT if (X86_64 && NET) select CLKEVT_I8253 select ARCH_HAVE_NMI_SAFE_CMPXCHG + select GENERIC_IOMAP config INSTRUCTION_DECODER def_bool (KPROBES || PERF_EVENTS) @@ -140,9 +141,6 @@ config NEED_SG_DMA_LENGTH config GENERIC_ISA_DMA def_bool ISA_DMA_API -config GENERIC_IOMAP - def_bool y - config GENERIC_BUG def_bool y depends on BUG diff --git a/lib/Kconfig b/lib/Kconfig index 32f3e5ae2be..005892723a5 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -19,6 +19,9 @@ config RATIONAL config GENERIC_FIND_FIRST_BIT bool +config GENERIC_IOMAP + bool + config CRC_CCITT tristate "CRC-CCITT functions" help -- cgit v1.2.3-70-g09d2 From 66eab4df288aaee75938ae99877c4f759fc6d56c Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 24 Nov 2011 20:45:20 +0200 Subject: lib: add GENERIC_PCI_IOMAP Many architectures want a generic pci_iomap but not the rest of iomap.c. Split that to a separate .c file and add a new config symbol. select automatically by GENERIC_IOMAP. Signed-off-by: Michael S. Tsirkin --- include/asm-generic/io.h | 4 ++-- include/asm-generic/iomap.h | 9 +++----- include/asm-generic/pci_iomap.h | 25 +++++++++++++++++++++ lib/Kconfig | 4 ++++ lib/Makefile | 1 + lib/iomap.c | 38 ++------------------------------ lib/pci_iomap.c | 48 +++++++++++++++++++++++++++++++++++++++++ 7 files changed, 85 insertions(+), 44 deletions(-) create mode 100644 include/asm-generic/pci_iomap.h create mode 100644 lib/pci_iomap.c (limited to 'lib') diff --git a/include/asm-generic/io.h b/include/asm-generic/io.h index 912088773a6..dddc61dedc2 100644 --- a/include/asm-generic/io.h +++ b/include/asm-generic/io.h @@ -19,6 +19,8 @@ #include #endif +#include + #ifndef mmiowb #define mmiowb() do {} while (0) #endif @@ -283,9 +285,7 @@ static inline void writesb(const void __iomem *addr, const void *buf, int len) #define __io_virt(x) ((void __force *) (x)) #ifndef CONFIG_GENERIC_IOMAP -/* Create a virtual mapping cookie for a PCI BAR (memory or IO) */ struct pci_dev; -extern void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max); static inline void pci_iounmap(struct pci_dev *dev, void __iomem *p) { } diff --git a/include/asm-generic/iomap.h b/include/asm-generic/iomap.h index 98dcd76ce83..8a3d4fde260 100644 --- a/include/asm-generic/iomap.h +++ b/include/asm-generic/iomap.h @@ -67,18 +67,15 @@ extern void ioport_unmap(void __iomem *); #endif #ifdef CONFIG_PCI -/* Create a virtual mapping cookie for a PCI BAR (memory or IO) */ +/* Destroy a virtual mapping cookie for a PCI BAR (memory or IO) */ struct pci_dev; -extern void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max); extern void pci_iounmap(struct pci_dev *dev, void __iomem *); #else struct pci_dev; -static inline void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max) -{ - return NULL; -} static inline void pci_iounmap(struct pci_dev *dev, void __iomem *addr) { } #endif +#include + #endif diff --git a/include/asm-generic/pci_iomap.h b/include/asm-generic/pci_iomap.h new file mode 100644 index 00000000000..8de4b73e19e --- /dev/null +++ b/include/asm-generic/pci_iomap.h @@ -0,0 +1,25 @@ +/* Generic I/O port emulation, based on MN10300 code + * + * Copyright (C) 2007 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public Licence + * as published by the Free Software Foundation; either version + * 2 of the Licence, or (at your option) any later version. + */ +#ifndef __ASM_GENERIC_PCI_IOMAP_H +#define __ASM_GENERIC_PCI_IOMAP_H + +struct pci_dev; +#ifdef CONFIG_PCI +/* Create a virtual mapping cookie for a PCI BAR (memory or IO) */ +extern void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max); +#else +static inline void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max) +{ + return NULL; +} +#endif + +#endif /* __ASM_GENERIC_IO_H */ diff --git a/lib/Kconfig b/lib/Kconfig index 005892723a5..36884b409e3 100644 --- a/lib/Kconfig +++ b/lib/Kconfig @@ -19,8 +19,12 @@ config RATIONAL config GENERIC_FIND_FIRST_BIT bool +config GENERIC_PCI_IOMAP + bool + config GENERIC_IOMAP bool + select GENERIC_PCI_IOMAP config CRC_CCITT tristate "CRC-CCITT functions" diff --git a/lib/Makefile b/lib/Makefile index a4da283f5dc..609b2adc604 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -33,6 +33,7 @@ endif lib-$(CONFIG_HOTPLUG) += kobject_uevent.o obj-$(CONFIG_GENERIC_IOMAP) += iomap.o +obj-$(CONFIG_GENERIC_PCI_IOMAP) += pci_iomap.o obj-$(CONFIG_HAS_IOMEM) += iomap_copy.o devres.o obj-$(CONFIG_CHECK_SIGNATURE) += check_signature.o obj-$(CONFIG_DEBUG_LOCKING_API_SELFTESTS) += locking-selftest.o diff --git a/lib/iomap.c b/lib/iomap.c index 5dbcb4b2d86..ada922a808e 100644 --- a/lib/iomap.c +++ b/lib/iomap.c @@ -242,45 +242,11 @@ EXPORT_SYMBOL(ioport_unmap); #endif /* CONFIG_HAS_IOPORT */ #ifdef CONFIG_PCI -/** - * pci_iomap - create a virtual mapping cookie for a PCI BAR - * @dev: PCI device that owns the BAR - * @bar: BAR number - * @maxlen: length of the memory to map - * - * Using this function you will get a __iomem address to your device BAR. - * You can access it using ioread*() and iowrite*(). These functions hide - * the details if this is a MMIO or PIO address space and will just do what - * you expect from them in the correct way. - * - * @maxlen specifies the maximum length to map. If you want to get access to - * the complete BAR without checking for its length first, pass %0 here. - * */ -void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen) -{ - resource_size_t start = pci_resource_start(dev, bar); - resource_size_t len = pci_resource_len(dev, bar); - unsigned long flags = pci_resource_flags(dev, bar); - - if (!len || !start) - return NULL; - if (maxlen && len > maxlen) - len = maxlen; - if (flags & IORESOURCE_IO) - return ioport_map(start, len); - if (flags & IORESOURCE_MEM) { - if (flags & IORESOURCE_CACHEABLE) - return ioremap(start, len); - return ioremap_nocache(start, len); - } - /* What? */ - return NULL; -} - +/* Hide the details if this is a MMIO or PIO address space and just do what + * you expect in the correct way. */ void pci_iounmap(struct pci_dev *dev, void __iomem * addr) { IO_COND(addr, /* nothing */, iounmap(addr)); } -EXPORT_SYMBOL(pci_iomap); EXPORT_SYMBOL(pci_iounmap); #endif /* CONFIG_PCI */ diff --git a/lib/pci_iomap.c b/lib/pci_iomap.c new file mode 100644 index 00000000000..4b0fdc22e68 --- /dev/null +++ b/lib/pci_iomap.c @@ -0,0 +1,48 @@ +/* + * Implement the default iomap interfaces + * + * (C) Copyright 2004 Linus Torvalds + */ +#include +#include + +#include + +#ifdef CONFIG_PCI +/** + * pci_iomap - create a virtual mapping cookie for a PCI BAR + * @dev: PCI device that owns the BAR + * @bar: BAR number + * @maxlen: length of the memory to map + * + * Using this function you will get a __iomem address to your device BAR. + * You can access it using ioread*() and iowrite*(). These functions hide + * the details if this is a MMIO or PIO address space and will just do what + * you expect from them in the correct way. + * + * @maxlen specifies the maximum length to map. If you want to get access to + * the complete BAR without checking for its length first, pass %0 here. + * */ +void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long maxlen) +{ + resource_size_t start = pci_resource_start(dev, bar); + resource_size_t len = pci_resource_len(dev, bar); + unsigned long flags = pci_resource_flags(dev, bar); + + if (!len || !start) + return NULL; + if (maxlen && len > maxlen) + len = maxlen; + if (flags & IORESOURCE_IO) + return ioport_map(start, len); + if (flags & IORESOURCE_MEM) { + if (flags & IORESOURCE_CACHEABLE) + return ioremap(start, len); + return ioremap_nocache(start, len); + } + /* What? */ + return NULL; +} + +EXPORT_SYMBOL(pci_iomap); +#endif /* CONFIG_PCI */ -- cgit v1.2.3-70-g09d2 From 90802ed9c3dbab2e067bd9fc67a30e66e6774e8f Mon Sep 17 00:00:00 2001 From: Paul Bolle Date: Mon, 5 Dec 2011 13:00:34 +0100 Subject: treewide: Fix comment and string typo 'bufer' Signed-off-by: Paul Bolle Signed-off-by: Jiri Kosina --- drivers/media/rc/ene_ir.c | 2 +- drivers/media/rc/ene_ir.h | 2 +- drivers/net/ethernet/chelsio/cxgb4vf/sge.c | 2 +- drivers/scsi/gdth.h | 2 +- drivers/usb/musb/cppi_dma.c | 2 +- drivers/video/bf54x-lq043fb.c | 2 +- drivers/video/bfin-t350mcqb-fb.c | 2 +- fs/ext4/inode.c | 2 +- fs/jbd/checkpoint.c | 2 +- fs/jbd2/checkpoint.c | 2 +- fs/xfs/xfs_buf.c | 2 +- lib/decompress_bunzip2.c | 4 ++-- lib/decompress_unlzma.c | 2 +- 13 files changed, 14 insertions(+), 14 deletions(-) (limited to 'lib') diff --git a/drivers/media/rc/ene_ir.c b/drivers/media/rc/ene_ir.c index cf10ecf5ace..860c112e0fd 100644 --- a/drivers/media/rc/ene_ir.c +++ b/drivers/media/rc/ene_ir.c @@ -324,7 +324,7 @@ static int ene_rx_get_sample_reg(struct ene_device *dev) return dev->extra_buf2_address + r_pointer; } - dbg("attempt to read beyong ring bufer end"); + dbg("attempt to read beyond ring buffer end"); return 0; } diff --git a/drivers/media/rc/ene_ir.h b/drivers/media/rc/ene_ir.h index fd108d90f75..6f978e85db8 100644 --- a/drivers/media/rc/ene_ir.h +++ b/drivers/media/rc/ene_ir.h @@ -227,7 +227,7 @@ struct ene_device { /* TX buffer */ unsigned *tx_buffer; /* input samples buffer*/ - int tx_pos; /* position in that bufer */ + int tx_pos; /* position in that buffer */ int tx_len; /* current len of tx buffer */ int tx_done; /* done transmitting */ /* one more sample pending*/ diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c index 8d5d55ad102..39b9de71a20 100644 --- a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c +++ b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c @@ -684,7 +684,7 @@ out: /* * Update our accounting state to incorporate the new Free List * buffers, tell the hardware about them and return the number of - * bufers which we were able to allocate. + * buffers which we were able to allocate. */ cred = fl->avail - cred; fl->pend_cred += cred; diff --git a/drivers/scsi/gdth.h b/drivers/scsi/gdth.h index d969855ac64..d3e4d7c6f57 100644 --- a/drivers/scsi/gdth.h +++ b/drivers/scsi/gdth.h @@ -359,7 +359,7 @@ typedef struct { u32 cmd_buff_addr2; /* physical address of cmd buffer 1 */ u32 cmd_buff_u_addr2; /* reserved for 64 bit addressing */ u32 cmd_buff_indx2; /* cmd buf addr1 unique identifier */ - u32 cmd_buff_size; /* size of each cmd bufer in bytes */ + u32 cmd_buff_size; /* size of each cmd buffer in bytes */ u32 reserved1; u32 reserved2; } __attribute__((packed)) gdth_perf_modes; diff --git a/drivers/usb/musb/cppi_dma.c b/drivers/usb/musb/cppi_dma.c index 318fb4e8a88..53be7aef630 100644 --- a/drivers/usb/musb/cppi_dma.c +++ b/drivers/usb/musb/cppi_dma.c @@ -513,7 +513,7 @@ static inline int cppi_autoreq_update(struct cppi_channel *rx, if (!(val & MUSB_RXCSR_H_REQPKT)) { val |= MUSB_RXCSR_H_REQPKT | MUSB_RXCSR_H_WZC_BITS; musb_writew(regs, MUSB_RXCSR, val); - /* flush writebufer */ + /* flush writebuffer */ val = musb_readw(regs, MUSB_RXCSR); } } diff --git a/drivers/video/bf54x-lq043fb.c b/drivers/video/bf54x-lq043fb.c index 56720fb476b..46b03f53985 100644 --- a/drivers/video/bf54x-lq043fb.c +++ b/drivers/video/bf54x-lq043fb.c @@ -4,7 +4,7 @@ * Author: Michael Hennerich * * Created: - * Description: ADSP-BF54x Framebufer driver + * Description: ADSP-BF54x Framebuffer driver * * * Modified: diff --git a/drivers/video/bfin-t350mcqb-fb.c b/drivers/video/bfin-t350mcqb-fb.c index d5e12675961..7a0c05f3537 100644 --- a/drivers/video/bfin-t350mcqb-fb.c +++ b/drivers/video/bfin-t350mcqb-fb.c @@ -4,7 +4,7 @@ * Author: Michael Hennerich * * Created: - * Description: Blackfin LCD Framebufer driver + * Description: Blackfin LCD Framebuffer driver * * * Modified: diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index b1c57bf4313..d87991d71a1 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1878,7 +1878,7 @@ static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate); * a[0] = 'a'; * truncate(f, 4096); * we have in the page first buffer_head mapped via page_mkwrite call back - * but other bufer_heads would be unmapped but dirty(dirty done via the + * but other buffer_heads would be unmapped but dirty (dirty done via the * do_wp_page). So writepage should write the first block. If we modify * the mmap area beyond 1024 we will again get a page_fault and the * page_mkwrite callback will do the block allocation and mark the diff --git a/fs/jbd/checkpoint.c b/fs/jbd/checkpoint.c index f94fc48ff3a..5d1a00a5041 100644 --- a/fs/jbd/checkpoint.c +++ b/fs/jbd/checkpoint.c @@ -537,7 +537,7 @@ int cleanup_journal_tail(journal_t *journal) * them. * * Called with j_list_lock held. - * Returns number of bufers reaped (for debug) + * Returns number of buffers reaped (for debug) */ static int journal_clean_one_cp_list(struct journal_head *jh, int *released) diff --git a/fs/jbd2/checkpoint.c b/fs/jbd2/checkpoint.c index 16a698bd906..d49d202903f 100644 --- a/fs/jbd2/checkpoint.c +++ b/fs/jbd2/checkpoint.c @@ -565,7 +565,7 @@ int jbd2_cleanup_journal_tail(journal_t *journal) * * Called with the journal locked. * Called with j_list_lock held. - * Returns number of bufers reaped (for debug) + * Returns number of buffers reaped (for debug) */ static int journal_clean_one_cp_list(struct journal_head *jh, int *released) diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index cf0ac056815..33e06d22168 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -1370,7 +1370,7 @@ restart: goto restart; } /* - * clear the LRU reference count so the bufer doesn't get + * clear the LRU reference count so the buffer doesn't get * ignored in xfs_buf_rele(). */ atomic_set(&bp->b_lru_ref, 0); diff --git a/lib/decompress_bunzip2.c b/lib/decompress_bunzip2.c index a7b80c1d6a0..3380297768d 100644 --- a/lib/decompress_bunzip2.c +++ b/lib/decompress_bunzip2.c @@ -691,7 +691,7 @@ STATIC int INIT bunzip2(unsigned char *buf, int len, outbuf = malloc(BZIP2_IOBUF_SIZE); if (!outbuf) { - error("Could not allocate output bufer"); + error("Could not allocate output buffer"); return RETVAL_OUT_OF_MEMORY; } if (buf) @@ -699,7 +699,7 @@ STATIC int INIT bunzip2(unsigned char *buf, int len, else inbuf = malloc(BZIP2_IOBUF_SIZE); if (!inbuf) { - error("Could not allocate input bufer"); + error("Could not allocate input buffer"); i = RETVAL_OUT_OF_MEMORY; goto exit_0; } diff --git a/lib/decompress_unlzma.c b/lib/decompress_unlzma.c index 476c65af970..32adb73a903 100644 --- a/lib/decompress_unlzma.c +++ b/lib/decompress_unlzma.c @@ -562,7 +562,7 @@ STATIC inline int INIT unlzma(unsigned char *buf, int in_len, else inbuf = malloc(LZMA_IOBUF_SIZE); if (!inbuf) { - error("Could not allocate input bufer"); + error("Could not allocate input buffer"); goto exit_0; } -- cgit v1.2.3-70-g09d2 From 38c8dcf665a5c86818bf46f687335a25fa32eccd Mon Sep 17 00:00:00 2001 From: Paul Bolle Date: Mon, 5 Dec 2011 13:06:18 +0100 Subject: decompress_bunzip2: remove invalid vi modeline Signed-off-by: Paul Bolle Signed-off-by: Jiri Kosina --- lib/decompress_bunzip2.c | 1 - 1 file changed, 1 deletion(-) (limited to 'lib') diff --git a/lib/decompress_bunzip2.c b/lib/decompress_bunzip2.c index 3380297768d..31c5f7675fb 100644 --- a/lib/decompress_bunzip2.c +++ b/lib/decompress_bunzip2.c @@ -1,4 +1,3 @@ -/* vi: set sw = 4 ts = 4: */ /* Small bzip2 deflate implementation, by Rob Landley (rob@landley.net). Based on bzip2 decompression code by Julian R Seward (jseward@acm.org), -- cgit v1.2.3-70-g09d2 From f21ffe9f6da6d3a69c518b7345c198d48d941c34 Mon Sep 17 00:00:00 2001 From: Konrad Rzeszutek Wilk Date: Thu, 11 Aug 2011 16:50:56 -0400 Subject: swiotlb: Expose swiotlb_nr_tlb function to modules As a mechanism to detect whether SWIOTLB is enabled or not. We also fix the spelling - it was swioltb instead of swiotlb. CC: FUJITA Tomonori [v1: Ripped out swiotlb_enabled] Signed-off-by: Konrad Rzeszutek Wilk --- drivers/xen/swiotlb-xen.c | 2 +- include/linux/swiotlb.h | 2 +- lib/swiotlb.c | 5 +++-- 3 files changed, 5 insertions(+), 4 deletions(-) (limited to 'lib') diff --git a/drivers/xen/swiotlb-xen.c b/drivers/xen/swiotlb-xen.c index 8e964b91c44..4864e5d72e7 100644 --- a/drivers/xen/swiotlb-xen.c +++ b/drivers/xen/swiotlb-xen.c @@ -153,7 +153,7 @@ void __init xen_swiotlb_init(int verbose) char *m = NULL; unsigned int repeat = 3; - nr_tbl = swioltb_nr_tbl(); + nr_tbl = swiotlb_nr_tbl(); if (nr_tbl) xen_io_tlb_nslabs = nr_tbl; else { diff --git a/include/linux/swiotlb.h b/include/linux/swiotlb.h index 445702c60d0..e872526fdc5 100644 --- a/include/linux/swiotlb.h +++ b/include/linux/swiotlb.h @@ -24,7 +24,7 @@ extern int swiotlb_force; extern void swiotlb_init(int verbose); extern void swiotlb_init_with_tbl(char *tlb, unsigned long nslabs, int verbose); -extern unsigned long swioltb_nr_tbl(void); +extern unsigned long swiotlb_nr_tbl(void); /* * Enumeration for sync targets diff --git a/lib/swiotlb.c b/lib/swiotlb.c index 99093b39614..058935ef397 100644 --- a/lib/swiotlb.c +++ b/lib/swiotlb.c @@ -110,11 +110,11 @@ setup_io_tlb_npages(char *str) __setup("swiotlb=", setup_io_tlb_npages); /* make io_tlb_overflow tunable too? */ -unsigned long swioltb_nr_tbl(void) +unsigned long swiotlb_nr_tbl(void) { return io_tlb_nslabs; } - +EXPORT_SYMBOL_GPL(swiotlb_nr_tbl); /* Note that this doesn't work with highmem page */ static dma_addr_t swiotlb_virt_to_bus(struct device *hwdev, volatile void *address) @@ -321,6 +321,7 @@ void __init swiotlb_free(void) free_bootmem_late(__pa(io_tlb_start), PAGE_ALIGN(io_tlb_nslabs << IO_TLB_SHIFT)); } + io_tlb_nslabs = 0; } static int is_swiotlb_buffer(phys_addr_t paddr) -- cgit v1.2.3-70-g09d2 From fe0e94c5a7e5335ba0d200e7d3e26e9f80cda4b1 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 7 Dec 2011 13:58:26 +0200 Subject: mpi/mpi-mpow: NULL dereference on allocation failure We can't call mpi_free() on the elements if the first kzalloc() fails. Signed-off-by: Dan Carpenter Signed-off-by: Dmitry Kasatkin Signed-off-by: James Morris --- lib/mpi/mpi-mpow.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/mpi/mpi-mpow.c b/lib/mpi/mpi-mpow.c index 4cc75933c5a..7328d0d6c74 100644 --- a/lib/mpi/mpi-mpow.c +++ b/lib/mpi/mpi-mpow.c @@ -73,7 +73,7 @@ int mpi_mulpowm(MPI res, MPI *basearray, MPI *exparray, MPI m) G = kzalloc((1 << k) * sizeof *G, GFP_KERNEL); if (!G) - goto nomem; + goto err_out; /* and calculate */ tmp = mpi_alloc(mpi_get_nlimbs(m) + 1); @@ -129,5 +129,6 @@ nomem: for (i = 0; i < (1 << k); i++) mpi_free(G[i]); kfree(G); +err_out: return rc; } -- cgit v1.2.3-70-g09d2 From 74099b18b8db0670484e0caeb3425fb29ff3144e Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Mon, 5 Dec 2011 19:12:47 +0100 Subject: driver-core: skip uevent generation when nobody is listening Most network namespaces unlikely have listeners to uevents, and should benefit from skipping all the string copies. Cc: Eric Biederman Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman --- lib/kobject_uevent.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'lib') diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c index ad72a03ce5e..e66e9b63261 100644 --- a/lib/kobject_uevent.c +++ b/lib/kobject_uevent.c @@ -259,6 +259,9 @@ int kobject_uevent_env(struct kobject *kobj, enum kobject_action action, struct sk_buff *skb; size_t len; + if (!netlink_has_listeners(uevent_sock, 1)) + continue; + /* allocate message with the maximum possible size */ len = strlen(action_string) + strlen(devpath) + 2; skb = alloc_skb(len + env->buflen, GFP_KERNEL); -- cgit v1.2.3-70-g09d2 From 4af679cd7cbb0a0d8774b5cdb34bffcaa4e86e52 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 13 Dec 2011 10:36:20 +0100 Subject: kref: Inline all functions These are tiny functions, there's no point in having them out-of-line. Cc: Alexey Dobriyan Cc: Eric Dumazet Cc: Ingo Molnar Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/n/tip-8eccvi2ur2fzgi00xdjlbf5z@git.kernel.org Signed-off-by: Greg Kroah-Hartman --- include/linux/kref.h | 80 ++++++++++++++++++++++++++++++++++++++++--- lib/Makefile | 2 +- lib/kref.c | 97 ---------------------------------------------------- 3 files changed, 76 insertions(+), 103 deletions(-) delete mode 100644 lib/kref.c (limited to 'lib') diff --git a/include/linux/kref.h b/include/linux/kref.h index d4a62ab2ee5..1cbae9f2ef7 100644 --- a/include/linux/kref.h +++ b/include/linux/kref.h @@ -16,15 +16,85 @@ #define _KREF_H_ #include +#include struct kref { atomic_t refcount; }; -void kref_init(struct kref *kref); -void kref_get(struct kref *kref); -int kref_put(struct kref *kref, void (*release) (struct kref *kref)); -int kref_sub(struct kref *kref, unsigned int count, - void (*release) (struct kref *kref)); +/** + * kref_init - initialize object. + * @kref: object in question. + */ +static inline void kref_init(struct kref *kref) +{ + atomic_set(&kref->refcount, 1); + smp_mb(); +} + +/** + * kref_get - increment refcount for object. + * @kref: object. + */ +static inline void kref_get(struct kref *kref) +{ + WARN_ON(!atomic_read(&kref->refcount)); + atomic_inc(&kref->refcount); + smp_mb__after_atomic_inc(); +} + +/** + * kref_put - decrement refcount for object. + * @kref: object. + * @release: pointer to the function that will clean up the object when the + * last reference to the object is released. + * This pointer is required, and it is not acceptable to pass kfree + * in as this function. + * + * Decrement the refcount, and if 0, call release(). + * Return 1 if the object was removed, otherwise return 0. Beware, if this + * function returns 0, you still can not count on the kref from remaining in + * memory. Only use the return value if you want to see if the kref is now + * gone, not present. + */ +static inline int kref_put(struct kref *kref, void (*release)(struct kref *kref)) +{ + WARN_ON(release == NULL); + WARN_ON(release == (void (*)(struct kref *))kfree); + + if (atomic_dec_and_test(&kref->refcount)) { + release(kref); + return 1; + } + return 0; +} + + +/** + * kref_sub - subtract a number of refcounts for object. + * @kref: object. + * @count: Number of recounts to subtract. + * @release: pointer to the function that will clean up the object when the + * last reference to the object is released. + * This pointer is required, and it is not acceptable to pass kfree + * in as this function. + * + * Subtract @count from the refcount, and if 0, call release(). + * Return 1 if the object was removed, otherwise return 0. Beware, if this + * function returns 0, you still can not count on the kref from remaining in + * memory. Only use the return value if you want to see if the kref is now + * gone, not present. + */ +static inline int kref_sub(struct kref *kref, unsigned int count, + void (*release)(struct kref *kref)) +{ + WARN_ON(release == NULL); + WARN_ON(release == (void (*)(struct kref *))kfree); + if (atomic_sub_and_test((int) count, &kref->refcount)) { + release(kref); + return 1; + } + return 0; +} #endif /* _KREF_H_ */ diff --git a/lib/Makefile b/lib/Makefile index a4da283f5dc..6f195ff6a1a 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -17,7 +17,7 @@ lib-y := ctype.o string.o vsprintf.o cmdline.o \ lib-$(CONFIG_MMU) += ioremap.o lib-$(CONFIG_SMP) += cpumask.o -lib-y += kobject.o kref.o klist.o +lib-y += kobject.o klist.o obj-y += bcd.o div64.o sort.o parser.o halfmd4.o debug_locks.o random32.o \ bust_spinlocks.o hexdump.o kasprintf.o bitmap.o scatterlist.o \ diff --git a/lib/kref.c b/lib/kref.c deleted file mode 100644 index 3efb882b11d..00000000000 --- a/lib/kref.c +++ /dev/null @@ -1,97 +0,0 @@ -/* - * kref.c - library routines for handling generic reference counted objects - * - * Copyright (C) 2004 Greg Kroah-Hartman - * Copyright (C) 2004 IBM Corp. - * - * based on lib/kobject.c which was: - * Copyright (C) 2002-2003 Patrick Mochel - * - * This file is released under the GPLv2. - * - */ - -#include -#include -#include - -/** - * kref_init - initialize object. - * @kref: object in question. - */ -void kref_init(struct kref *kref) -{ - atomic_set(&kref->refcount, 1); - smp_mb(); -} - -/** - * kref_get - increment refcount for object. - * @kref: object. - */ -void kref_get(struct kref *kref) -{ - WARN_ON(!atomic_read(&kref->refcount)); - atomic_inc(&kref->refcount); - smp_mb__after_atomic_inc(); -} - -/** - * kref_put - decrement refcount for object. - * @kref: object. - * @release: pointer to the function that will clean up the object when the - * last reference to the object is released. - * This pointer is required, and it is not acceptable to pass kfree - * in as this function. - * - * Decrement the refcount, and if 0, call release(). - * Return 1 if the object was removed, otherwise return 0. Beware, if this - * function returns 0, you still can not count on the kref from remaining in - * memory. Only use the return value if you want to see if the kref is now - * gone, not present. - */ -int kref_put(struct kref *kref, void (*release)(struct kref *kref)) -{ - WARN_ON(release == NULL); - WARN_ON(release == (void (*)(struct kref *))kfree); - - if (atomic_dec_and_test(&kref->refcount)) { - release(kref); - return 1; - } - return 0; -} - - -/** - * kref_sub - subtract a number of refcounts for object. - * @kref: object. - * @count: Number of recounts to subtract. - * @release: pointer to the function that will clean up the object when the - * last reference to the object is released. - * This pointer is required, and it is not acceptable to pass kfree - * in as this function. - * - * Subtract @count from the refcount, and if 0, call release(). - * Return 1 if the object was removed, otherwise return 0. Beware, if this - * function returns 0, you still can not count on the kref from remaining in - * memory. Only use the return value if you want to see if the kref is now - * gone, not present. - */ -int kref_sub(struct kref *kref, unsigned int count, - void (*release)(struct kref *kref)) -{ - WARN_ON(release == NULL); - WARN_ON(release == (void (*)(struct kref *))kfree); - - if (atomic_sub_and_test((int) count, &kref->refcount)) { - release(kref); - return 1; - } - return 0; -} - -EXPORT_SYMBOL(kref_init); -EXPORT_SYMBOL(kref_get); -EXPORT_SYMBOL(kref_put); -EXPORT_SYMBOL(kref_sub); -- cgit v1.2.3-70-g09d2 From b3e8d7b2478401b2f25f4566a90faad54f7d6d07 Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Wed, 21 Dec 2011 15:13:54 -0800 Subject: kobject: remove kset_find_obj_hinted() Now that there are no in-kernel users of this function, remove it as it is no longer needed. Signed-off-by: Kay Sievers Signed-off-by: Greg Kroah-Hartman --- include/linux/kobject.h | 2 -- lib/kobject.c | 37 ------------------------------------- 2 files changed, 39 deletions(-) (limited to 'lib') diff --git a/include/linux/kobject.h b/include/linux/kobject.h index ad81e1c5148..fc615a97e2d 100644 --- a/include/linux/kobject.h +++ b/include/linux/kobject.h @@ -191,8 +191,6 @@ static inline struct kobj_type *get_ktype(struct kobject *kobj) } extern struct kobject *kset_find_obj(struct kset *, const char *); -extern struct kobject *kset_find_obj_hinted(struct kset *, const char *, - struct kobject *); /* The global /sys/kernel/ kobject for people to chain off of */ extern struct kobject *kernel_kobj; diff --git a/lib/kobject.c b/lib/kobject.c index 640bd98a4c8..c33d7a18d63 100644 --- a/lib/kobject.c +++ b/lib/kobject.c @@ -745,44 +745,12 @@ void kset_unregister(struct kset *k) * take a reference and return the object. */ struct kobject *kset_find_obj(struct kset *kset, const char *name) -{ - return kset_find_obj_hinted(kset, name, NULL); -} - -/** - * kset_find_obj_hinted - search for object in kset given a predecessor hint. - * @kset: kset we're looking in. - * @name: object's name. - * @hint: hint to possible object's predecessor. - * - * Check the hint's next object and if it is a match return it directly, - * otherwise, fall back to the behavior of kset_find_obj(). Either way - * a reference for the returned object is held and the reference on the - * hinted object is released. - */ -struct kobject *kset_find_obj_hinted(struct kset *kset, const char *name, - struct kobject *hint) { struct kobject *k; struct kobject *ret = NULL; spin_lock(&kset->list_lock); - if (!hint) - goto slow_search; - - /* end of list detection */ - if (hint->entry.next == kset->list.next) - goto slow_search; - - k = container_of(hint->entry.next, struct kobject, entry); - if (!kobject_name(k) || strcmp(kobject_name(k), name)) - goto slow_search; - - ret = kobject_get(k); - goto unlock_exit; - -slow_search: list_for_each_entry(k, &kset->list, entry) { if (kobject_name(k) && !strcmp(kobject_name(k), name)) { ret = kobject_get(k); @@ -790,12 +758,7 @@ slow_search: } } -unlock_exit: spin_unlock(&kset->list_lock); - - if (hint) - kobject_put(hint); - return ret; } -- cgit v1.2.3-70-g09d2 From f4ae40a6a50a98ac23d4b285f739455e926a473e Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sun, 24 Jul 2011 04:33:43 -0400 Subject: switch debugfs to umode_t Signed-off-by: Al Viro --- Documentation/filesystems/debugfs.txt | 24 +++++------ arch/arm/mach-msm/smd_debug.c | 2 +- arch/s390/include/asm/debug.h | 4 +- arch/s390/kernel/debug.c | 8 ++-- arch/x86/xen/debugfs.c | 2 +- arch/x86/xen/debugfs.h | 2 +- drivers/acpi/ec_sys.c | 2 +- drivers/mmc/card/mmc_test.c | 2 +- .../net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c | 2 +- drivers/net/wireless/ath/carl9170/debug.c | 2 +- drivers/net/wireless/libertas/debugfs.c | 2 +- drivers/s390/block/dasd.c | 4 +- drivers/scsi/bfa/bfad_debugfs.c | 2 +- fs/debugfs/file.c | 22 +++++------ fs/debugfs/inode.c | 14 +++---- fs/ocfs2/cluster/netdebug.c | 2 +- include/linux/debugfs.h | 46 +++++++++++----------- include/linux/relay.h | 2 +- kernel/relay.c | 2 +- kernel/trace/blktrace.c | 2 +- kernel/trace/trace.c | 2 +- kernel/trace/trace.h | 2 +- lib/fault-inject.c | 8 ++-- mm/failslab.c | 2 +- mm/page_alloc.c | 2 +- 25 files changed, 82 insertions(+), 82 deletions(-) (limited to 'lib') diff --git a/Documentation/filesystems/debugfs.txt b/Documentation/filesystems/debugfs.txt index 742cc06e138..9281a95d689 100644 --- a/Documentation/filesystems/debugfs.txt +++ b/Documentation/filesystems/debugfs.txt @@ -35,7 +35,7 @@ described below will work. The most general way to create a file within a debugfs directory is with: - struct dentry *debugfs_create_file(const char *name, mode_t mode, + struct dentry *debugfs_create_file(const char *name, umode_t mode, struct dentry *parent, void *data, const struct file_operations *fops); @@ -53,13 +53,13 @@ actually necessary; the debugfs code provides a number of helper functions for simple situations. Files containing a single integer value can be created with any of: - struct dentry *debugfs_create_u8(const char *name, mode_t mode, + struct dentry *debugfs_create_u8(const char *name, umode_t mode, struct dentry *parent, u8 *value); - struct dentry *debugfs_create_u16(const char *name, mode_t mode, + struct dentry *debugfs_create_u16(const char *name, umode_t mode, struct dentry *parent, u16 *value); - struct dentry *debugfs_create_u32(const char *name, mode_t mode, + struct dentry *debugfs_create_u32(const char *name, umode_t mode, struct dentry *parent, u32 *value); - struct dentry *debugfs_create_u64(const char *name, mode_t mode, + struct dentry *debugfs_create_u64(const char *name, umode_t mode, struct dentry *parent, u64 *value); These files support both reading and writing the given value; if a specific @@ -67,13 +67,13 @@ file should not be written to, simply set the mode bits accordingly. The values in these files are in decimal; if hexadecimal is more appropriate, the following functions can be used instead: - struct dentry *debugfs_create_x8(const char *name, mode_t mode, + struct dentry *debugfs_create_x8(const char *name, umode_t mode, struct dentry *parent, u8 *value); - struct dentry *debugfs_create_x16(const char *name, mode_t mode, + struct dentry *debugfs_create_x16(const char *name, umode_t mode, struct dentry *parent, u16 *value); - struct dentry *debugfs_create_x32(const char *name, mode_t mode, + struct dentry *debugfs_create_x32(const char *name, umode_t mode, struct dentry *parent, u32 *value); - struct dentry *debugfs_create_x64(const char *name, mode_t mode, + struct dentry *debugfs_create_x64(const char *name, umode_t mode, struct dentry *parent, u64 *value); These functions are useful as long as the developer knows the size of the @@ -81,7 +81,7 @@ value to be exported. Some types can have different widths on different architectures, though, complicating the situation somewhat. There is a function meant to help out in one special case: - struct dentry *debugfs_create_size_t(const char *name, mode_t mode, + struct dentry *debugfs_create_size_t(const char *name, umode_t mode, struct dentry *parent, size_t *value); @@ -90,7 +90,7 @@ a variable of type size_t. Boolean values can be placed in debugfs with: - struct dentry *debugfs_create_bool(const char *name, mode_t mode, + struct dentry *debugfs_create_bool(const char *name, umode_t mode, struct dentry *parent, u32 *value); A read on the resulting file will yield either Y (for non-zero values) or @@ -104,7 +104,7 @@ Finally, a block of arbitrary binary data can be exported with: unsigned long size; }; - struct dentry *debugfs_create_blob(const char *name, mode_t mode, + struct dentry *debugfs_create_blob(const char *name, umode_t mode, struct dentry *parent, struct debugfs_blob_wrapper *blob); diff --git a/arch/arm/mach-msm/smd_debug.c b/arch/arm/mach-msm/smd_debug.c index 8736afff82f..0c56a5aaf58 100644 --- a/arch/arm/mach-msm/smd_debug.c +++ b/arch/arm/mach-msm/smd_debug.c @@ -215,7 +215,7 @@ static const struct file_operations debug_ops = { .llseek = default_llseek, }; -static void debug_create(const char *name, mode_t mode, +static void debug_create(const char *name, umode_t mode, struct dentry *dent, int (*fill)(char *buf, int max)) { diff --git a/arch/s390/include/asm/debug.h b/arch/s390/include/asm/debug.h index 18124b75a7a..9d88db1f55d 100644 --- a/arch/s390/include/asm/debug.h +++ b/arch/s390/include/asm/debug.h @@ -73,7 +73,7 @@ typedef struct debug_info { struct dentry* debugfs_entries[DEBUG_MAX_VIEWS]; struct debug_view* views[DEBUG_MAX_VIEWS]; char name[DEBUG_MAX_NAME_LEN]; - mode_t mode; + umode_t mode; } debug_info_t; typedef int (debug_header_proc_t) (debug_info_t* id, @@ -124,7 +124,7 @@ debug_info_t *debug_register(const char *name, int pages, int nr_areas, int buf_size); debug_info_t *debug_register_mode(const char *name, int pages, int nr_areas, - int buf_size, mode_t mode, uid_t uid, + int buf_size, umode_t mode, uid_t uid, gid_t gid); void debug_unregister(debug_info_t* id); diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c index 5ad6bc078bf..6848828b962 100644 --- a/arch/s390/kernel/debug.c +++ b/arch/s390/kernel/debug.c @@ -74,7 +74,7 @@ static ssize_t debug_input(struct file *file, const char __user *user_buf, static int debug_open(struct inode *inode, struct file *file); static int debug_close(struct inode *inode, struct file *file); static debug_info_t *debug_info_create(const char *name, int pages_per_area, - int nr_areas, int buf_size, mode_t mode); + int nr_areas, int buf_size, umode_t mode); static void debug_info_get(debug_info_t *); static void debug_info_put(debug_info_t *); static int debug_prolog_level_fn(debug_info_t * id, @@ -330,7 +330,7 @@ debug_info_free(debug_info_t* db_info){ static debug_info_t* debug_info_create(const char *name, int pages_per_area, int nr_areas, - int buf_size, mode_t mode) + int buf_size, umode_t mode) { debug_info_t* rc; @@ -688,7 +688,7 @@ debug_close(struct inode *inode, struct file *file) */ debug_info_t *debug_register_mode(const char *name, int pages_per_area, - int nr_areas, int buf_size, mode_t mode, + int nr_areas, int buf_size, umode_t mode, uid_t uid, gid_t gid) { debug_info_t *rc = NULL; @@ -1090,7 +1090,7 @@ debug_register_view(debug_info_t * id, struct debug_view *view) int rc = 0; int i; unsigned long flags; - mode_t mode; + umode_t mode; struct dentry *pde; if (!id) diff --git a/arch/x86/xen/debugfs.c b/arch/x86/xen/debugfs.c index 7c0fedd98ea..ef1db1900d8 100644 --- a/arch/x86/xen/debugfs.c +++ b/arch/x86/xen/debugfs.c @@ -109,7 +109,7 @@ static const struct file_operations u32_array_fops = { .llseek = no_llseek, }; -struct dentry *xen_debugfs_create_u32_array(const char *name, mode_t mode, +struct dentry *xen_debugfs_create_u32_array(const char *name, umode_t mode, struct dentry *parent, u32 *array, unsigned elements) { diff --git a/arch/x86/xen/debugfs.h b/arch/x86/xen/debugfs.h index e2813208483..78d25499be5 100644 --- a/arch/x86/xen/debugfs.h +++ b/arch/x86/xen/debugfs.h @@ -3,7 +3,7 @@ struct dentry * __init xen_init_debugfs(void); -struct dentry *xen_debugfs_create_u32_array(const char *name, mode_t mode, +struct dentry *xen_debugfs_create_u32_array(const char *name, umode_t mode, struct dentry *parent, u32 *array, unsigned elements); diff --git a/drivers/acpi/ec_sys.c b/drivers/acpi/ec_sys.c index 6c47ae9793a..b258cab9061 100644 --- a/drivers/acpi/ec_sys.c +++ b/drivers/acpi/ec_sys.c @@ -105,7 +105,7 @@ int acpi_ec_add_debugfs(struct acpi_ec *ec, unsigned int ec_device_count) { struct dentry *dev_dir; char name[64]; - mode_t mode = 0400; + umode_t mode = 0400; if (ec_device_count == 0) { acpi_ec_debugfs_dir = debugfs_create_dir("ec", NULL); diff --git a/drivers/mmc/card/mmc_test.c b/drivers/mmc/card/mmc_test.c index b038c4a9468..e99bdc18002 100644 --- a/drivers/mmc/card/mmc_test.c +++ b/drivers/mmc/card/mmc_test.c @@ -2949,7 +2949,7 @@ static void mmc_test_free_dbgfs_file(struct mmc_card *card) } static int __mmc_test_register_dbgfs_file(struct mmc_card *card, - const char *name, mode_t mode, const struct file_operations *fops) + const char *name, umode_t mode, const struct file_operations *fops) { struct dentry *file = NULL; struct mmc_test_dbgfs_file *df; diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c b/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c index da9072bfca8..f5a24d99ef4 100644 --- a/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c +++ b/drivers/net/ethernet/chelsio/cxgb4vf/cxgb4vf_main.c @@ -2000,7 +2000,7 @@ static const struct file_operations interfaces_proc_fops = { */ struct cxgb4vf_debugfs_entry { const char *name; /* name of debugfs node */ - mode_t mode; /* file system mode */ + umode_t mode; /* file system mode */ const struct file_operations *fops; }; diff --git a/drivers/net/wireless/ath/carl9170/debug.c b/drivers/net/wireless/ath/carl9170/debug.c index de57f90e1d5..3c164226687 100644 --- a/drivers/net/wireless/ath/carl9170/debug.c +++ b/drivers/net/wireless/ath/carl9170/debug.c @@ -56,7 +56,7 @@ static int carl9170_debugfs_open(struct inode *inode, struct file *file) struct carl9170_debugfs_fops { unsigned int read_bufsize; - mode_t attr; + umode_t attr; char *(*read)(struct ar9170 *ar, char *buf, size_t bufsize, ssize_t *len); ssize_t (*write)(struct ar9170 *aru, const char *buf, size_t size); diff --git a/drivers/net/wireless/libertas/debugfs.c b/drivers/net/wireless/libertas/debugfs.c index d8d8f0d0899..c192671610f 100644 --- a/drivers/net/wireless/libertas/debugfs.c +++ b/drivers/net/wireless/libertas/debugfs.c @@ -704,7 +704,7 @@ out_unlock: struct lbs_debugfs_files { const char *name; - int perm; + umode_t perm; struct file_operations fops; }; diff --git a/drivers/s390/block/dasd.c b/drivers/s390/block/dasd.c index 65894f05a80..42986d7bcf9 100644 --- a/drivers/s390/block/dasd.c +++ b/drivers/s390/block/dasd.c @@ -1073,7 +1073,7 @@ static const struct file_operations dasd_stats_global_fops = { static void dasd_profile_init(struct dasd_profile *profile, struct dentry *base_dentry) { - mode_t mode; + umode_t mode; struct dentry *pde; if (!base_dentry) @@ -1112,7 +1112,7 @@ static void dasd_statistics_removeroot(void) static void dasd_statistics_createroot(void) { - mode_t mode; + umode_t mode; struct dentry *pde; dasd_debugfs_root_entry = NULL; diff --git a/drivers/scsi/bfa/bfad_debugfs.c b/drivers/scsi/bfa/bfad_debugfs.c index dee1a094c2c..caca9b7c830 100644 --- a/drivers/scsi/bfa/bfad_debugfs.c +++ b/drivers/scsi/bfa/bfad_debugfs.c @@ -472,7 +472,7 @@ static const struct file_operations bfad_debugfs_op_regwr = { struct bfad_debugfs_entry { const char *name; - mode_t mode; + umode_t mode; const struct file_operations *fops; }; diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index 90f76575c05..d5016606fb2 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -95,7 +95,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u8_wo, NULL, debugfs_u8_set, "%llu\n"); * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ -struct dentry *debugfs_create_u8(const char *name, mode_t mode, +struct dentry *debugfs_create_u8(const char *name, umode_t mode, struct dentry *parent, u8 *value) { /* if there are no write bits set, make read only */ @@ -147,7 +147,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u16_wo, NULL, debugfs_u16_set, "%llu\n"); * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ -struct dentry *debugfs_create_u16(const char *name, mode_t mode, +struct dentry *debugfs_create_u16(const char *name, umode_t mode, struct dentry *parent, u16 *value) { /* if there are no write bits set, make read only */ @@ -199,7 +199,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u32_wo, NULL, debugfs_u32_set, "%llu\n"); * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ -struct dentry *debugfs_create_u32(const char *name, mode_t mode, +struct dentry *debugfs_create_u32(const char *name, umode_t mode, struct dentry *parent, u32 *value) { /* if there are no write bits set, make read only */ @@ -252,7 +252,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n"); * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ -struct dentry *debugfs_create_u64(const char *name, mode_t mode, +struct dentry *debugfs_create_u64(const char *name, umode_t mode, struct dentry *parent, u64 *value) { /* if there are no write bits set, make read only */ @@ -298,7 +298,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_x64, debugfs_u64_get, debugfs_u64_set, "0x%016llx\n * @value: a pointer to the variable that the file should read to and write * from. */ -struct dentry *debugfs_create_x8(const char *name, mode_t mode, +struct dentry *debugfs_create_x8(const char *name, umode_t mode, struct dentry *parent, u8 *value) { /* if there are no write bits set, make read only */ @@ -322,7 +322,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x8); * @value: a pointer to the variable that the file should read to and write * from. */ -struct dentry *debugfs_create_x16(const char *name, mode_t mode, +struct dentry *debugfs_create_x16(const char *name, umode_t mode, struct dentry *parent, u16 *value) { /* if there are no write bits set, make read only */ @@ -346,7 +346,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x16); * @value: a pointer to the variable that the file should read to and write * from. */ -struct dentry *debugfs_create_x32(const char *name, mode_t mode, +struct dentry *debugfs_create_x32(const char *name, umode_t mode, struct dentry *parent, u32 *value) { /* if there are no write bits set, make read only */ @@ -370,7 +370,7 @@ EXPORT_SYMBOL_GPL(debugfs_create_x32); * @value: a pointer to the variable that the file should read to and write * from. */ -struct dentry *debugfs_create_x64(const char *name, mode_t mode, +struct dentry *debugfs_create_x64(const char *name, umode_t mode, struct dentry *parent, u64 *value) { return debugfs_create_file(name, mode, parent, value, &fops_x64); @@ -401,7 +401,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_size_t, debugfs_size_t_get, debugfs_size_t_set, * @value: a pointer to the variable that the file should read to and write * from. */ -struct dentry *debugfs_create_size_t(const char *name, mode_t mode, +struct dentry *debugfs_create_size_t(const char *name, umode_t mode, struct dentry *parent, size_t *value) { return debugfs_create_file(name, mode, parent, value, &fops_size_t); @@ -473,7 +473,7 @@ static const struct file_operations fops_bool = { * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ -struct dentry *debugfs_create_bool(const char *name, mode_t mode, +struct dentry *debugfs_create_bool(const char *name, umode_t mode, struct dentry *parent, u32 *value) { return debugfs_create_file(name, mode, parent, value, &fops_bool); @@ -518,7 +518,7 @@ static const struct file_operations fops_blob = { * %NULL or !%NULL instead as to eliminate the need for #ifdef in the calling * code. */ -struct dentry *debugfs_create_blob(const char *name, mode_t mode, +struct dentry *debugfs_create_blob(const char *name, umode_t mode, struct dentry *parent, struct debugfs_blob_wrapper *blob) { diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index c9dc08d0c10..956d5ddddf6 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -30,7 +30,7 @@ static struct vfsmount *debugfs_mount; static int debugfs_mount_count; static bool debugfs_registered; -static struct inode *debugfs_get_inode(struct super_block *sb, int mode, dev_t dev, +static struct inode *debugfs_get_inode(struct super_block *sb, umode_t mode, dev_t dev, void *data, const struct file_operations *fops) { @@ -69,7 +69,7 @@ static struct inode *debugfs_get_inode(struct super_block *sb, int mode, dev_t d /* SMP-safe */ static int debugfs_mknod(struct inode *dir, struct dentry *dentry, - int mode, dev_t dev, void *data, + umode_t mode, dev_t dev, void *data, const struct file_operations *fops) { struct inode *inode; @@ -87,7 +87,7 @@ static int debugfs_mknod(struct inode *dir, struct dentry *dentry, return error; } -static int debugfs_mkdir(struct inode *dir, struct dentry *dentry, int mode, +static int debugfs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode, void *data, const struct file_operations *fops) { int res; @@ -101,14 +101,14 @@ static int debugfs_mkdir(struct inode *dir, struct dentry *dentry, int mode, return res; } -static int debugfs_link(struct inode *dir, struct dentry *dentry, int mode, +static int debugfs_link(struct inode *dir, struct dentry *dentry, umode_t mode, void *data, const struct file_operations *fops) { mode = (mode & S_IALLUGO) | S_IFLNK; return debugfs_mknod(dir, dentry, mode, 0, data, fops); } -static int debugfs_create(struct inode *dir, struct dentry *dentry, int mode, +static int debugfs_create(struct inode *dir, struct dentry *dentry, umode_t mode, void *data, const struct file_operations *fops) { int res; @@ -146,7 +146,7 @@ static struct file_system_type debug_fs_type = { .kill_sb = kill_litter_super, }; -static int debugfs_create_by_name(const char *name, mode_t mode, +static int debugfs_create_by_name(const char *name, umode_t mode, struct dentry *parent, struct dentry **dentry, void *data, @@ -214,7 +214,7 @@ static int debugfs_create_by_name(const char *name, mode_t mode, * If debugfs is not enabled in the kernel, the value -%ENODEV will be * returned. */ -struct dentry *debugfs_create_file(const char *name, mode_t mode, +struct dentry *debugfs_create_file(const char *name, umode_t mode, struct dentry *parent, void *data, const struct file_operations *fops) { diff --git a/fs/ocfs2/cluster/netdebug.c b/fs/ocfs2/cluster/netdebug.c index dc45deb19e6..73ba81928bc 100644 --- a/fs/ocfs2/cluster/netdebug.c +++ b/fs/ocfs2/cluster/netdebug.c @@ -553,7 +553,7 @@ void o2net_debugfs_exit(void) int o2net_debugfs_init(void) { - mode_t mode = S_IFREG|S_IRUSR; + umode_t mode = S_IFREG|S_IRUSR; o2net_dentry = debugfs_create_dir(O2NET_DEBUG_DIR, NULL); if (o2net_dentry) diff --git a/include/linux/debugfs.h b/include/linux/debugfs.h index e7d9b20ddc5..d1ac841e8dc 100644 --- a/include/linux/debugfs.h +++ b/include/linux/debugfs.h @@ -34,7 +34,7 @@ extern struct dentry *arch_debugfs_dir; extern const struct file_operations debugfs_file_operations; extern const struct inode_operations debugfs_link_operations; -struct dentry *debugfs_create_file(const char *name, mode_t mode, +struct dentry *debugfs_create_file(const char *name, umode_t mode, struct dentry *parent, void *data, const struct file_operations *fops); @@ -49,28 +49,28 @@ void debugfs_remove_recursive(struct dentry *dentry); struct dentry *debugfs_rename(struct dentry *old_dir, struct dentry *old_dentry, struct dentry *new_dir, const char *new_name); -struct dentry *debugfs_create_u8(const char *name, mode_t mode, +struct dentry *debugfs_create_u8(const char *name, umode_t mode, struct dentry *parent, u8 *value); -struct dentry *debugfs_create_u16(const char *name, mode_t mode, +struct dentry *debugfs_create_u16(const char *name, umode_t mode, struct dentry *parent, u16 *value); -struct dentry *debugfs_create_u32(const char *name, mode_t mode, +struct dentry *debugfs_create_u32(const char *name, umode_t mode, struct dentry *parent, u32 *value); -struct dentry *debugfs_create_u64(const char *name, mode_t mode, +struct dentry *debugfs_create_u64(const char *name, umode_t mode, struct dentry *parent, u64 *value); -struct dentry *debugfs_create_x8(const char *name, mode_t mode, +struct dentry *debugfs_create_x8(const char *name, umode_t mode, struct dentry *parent, u8 *value); -struct dentry *debugfs_create_x16(const char *name, mode_t mode, +struct dentry *debugfs_create_x16(const char *name, umode_t mode, struct dentry *parent, u16 *value); -struct dentry *debugfs_create_x32(const char *name, mode_t mode, +struct dentry *debugfs_create_x32(const char *name, umode_t mode, struct dentry *parent, u32 *value); -struct dentry *debugfs_create_x64(const char *name, mode_t mode, +struct dentry *debugfs_create_x64(const char *name, umode_t mode, struct dentry *parent, u64 *value); -struct dentry *debugfs_create_size_t(const char *name, mode_t mode, +struct dentry *debugfs_create_size_t(const char *name, umode_t mode, struct dentry *parent, size_t *value); -struct dentry *debugfs_create_bool(const char *name, mode_t mode, +struct dentry *debugfs_create_bool(const char *name, umode_t mode, struct dentry *parent, u32 *value); -struct dentry *debugfs_create_blob(const char *name, mode_t mode, +struct dentry *debugfs_create_blob(const char *name, umode_t mode, struct dentry *parent, struct debugfs_blob_wrapper *blob); @@ -86,7 +86,7 @@ bool debugfs_initialized(void); * want to duplicate the design decision mistakes of procfs and devfs again. */ -static inline struct dentry *debugfs_create_file(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_file(const char *name, umode_t mode, struct dentry *parent, void *data, const struct file_operations *fops) { @@ -118,70 +118,70 @@ static inline struct dentry *debugfs_rename(struct dentry *old_dir, struct dentr return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_u8(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_u8(const char *name, umode_t mode, struct dentry *parent, u8 *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_u16(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_u16(const char *name, umode_t mode, struct dentry *parent, u16 *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_u32(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_u32(const char *name, umode_t mode, struct dentry *parent, u32 *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_u64(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_u64(const char *name, umode_t mode, struct dentry *parent, u64 *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_x8(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_x8(const char *name, umode_t mode, struct dentry *parent, u8 *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_x16(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_x16(const char *name, umode_t mode, struct dentry *parent, u16 *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_x32(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_x32(const char *name, umode_t mode, struct dentry *parent, u32 *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_size_t(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_size_t(const char *name, umode_t mode, struct dentry *parent, size_t *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_bool(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_bool(const char *name, umode_t mode, struct dentry *parent, u32 *value) { return ERR_PTR(-ENODEV); } -static inline struct dentry *debugfs_create_blob(const char *name, mode_t mode, +static inline struct dentry *debugfs_create_blob(const char *name, umode_t mode, struct dentry *parent, struct debugfs_blob_wrapper *blob) { diff --git a/include/linux/relay.h b/include/linux/relay.h index 14a86bc7102..a822fd71fd6 100644 --- a/include/linux/relay.h +++ b/include/linux/relay.h @@ -144,7 +144,7 @@ struct rchan_callbacks */ struct dentry *(*create_buf_file)(const char *filename, struct dentry *parent, - int mode, + umode_t mode, struct rchan_buf *buf, int *is_global); diff --git a/kernel/relay.c b/kernel/relay.c index 226fade4d72..4335e1d7ee2 100644 --- a/kernel/relay.c +++ b/kernel/relay.c @@ -302,7 +302,7 @@ static void buf_unmapped_default_callback(struct rchan_buf *buf, */ static struct dentry *create_buf_file_default_callback(const char *filename, struct dentry *parent, - int mode, + umode_t mode, struct rchan_buf *buf, int *is_global) { diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 16fc34a0806..cdea7b56b0c 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -402,7 +402,7 @@ static int blk_remove_buf_file_callback(struct dentry *dentry) static struct dentry *blk_create_buf_file_callback(const char *filename, struct dentry *parent, - int mode, + umode_t mode, struct rchan_buf *buf, int *is_global) { diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index f2bd275bb60..660b069a0f9 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -4385,7 +4385,7 @@ static const struct file_operations trace_options_core_fops = { }; struct dentry *trace_create_file(const char *name, - mode_t mode, + umode_t mode, struct dentry *parent, void *data, const struct file_operations *fops) diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 092e1f8d18d..0154c0b850d 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -312,7 +312,7 @@ void tracing_reset_current(int cpu); void tracing_reset_current_online_cpus(void); int tracing_open_generic(struct inode *inode, struct file *filp); struct dentry *trace_create_file(const char *name, - mode_t mode, + umode_t mode, struct dentry *parent, void *data, const struct file_operations *fops); diff --git a/lib/fault-inject.c b/lib/fault-inject.c index 4f7554025e3..b4801f51b60 100644 --- a/lib/fault-inject.c +++ b/lib/fault-inject.c @@ -149,7 +149,7 @@ static int debugfs_ul_get(void *data, u64 *val) DEFINE_SIMPLE_ATTRIBUTE(fops_ul, debugfs_ul_get, debugfs_ul_set, "%llu\n"); -static struct dentry *debugfs_create_ul(const char *name, mode_t mode, +static struct dentry *debugfs_create_ul(const char *name, umode_t mode, struct dentry *parent, unsigned long *value) { return debugfs_create_file(name, mode, parent, value, &fops_ul); @@ -169,7 +169,7 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_stacktrace_depth, debugfs_ul_get, debugfs_stacktrace_depth_set, "%llu\n"); static struct dentry *debugfs_create_stacktrace_depth( - const char *name, mode_t mode, + const char *name, umode_t mode, struct dentry *parent, unsigned long *value) { return debugfs_create_file(name, mode, parent, value, @@ -193,7 +193,7 @@ static int debugfs_atomic_t_get(void *data, u64 *val) DEFINE_SIMPLE_ATTRIBUTE(fops_atomic_t, debugfs_atomic_t_get, debugfs_atomic_t_set, "%lld\n"); -static struct dentry *debugfs_create_atomic_t(const char *name, mode_t mode, +static struct dentry *debugfs_create_atomic_t(const char *name, umode_t mode, struct dentry *parent, atomic_t *value) { return debugfs_create_file(name, mode, parent, value, &fops_atomic_t); @@ -202,7 +202,7 @@ static struct dentry *debugfs_create_atomic_t(const char *name, mode_t mode, struct dentry *fault_create_debugfs_attr(const char *name, struct dentry *parent, struct fault_attr *attr) { - mode_t mode = S_IFREG | S_IRUSR | S_IWUSR; + umode_t mode = S_IFREG | S_IRUSR | S_IWUSR; struct dentry *dir; dir = debugfs_create_dir(name, parent); diff --git a/mm/failslab.c b/mm/failslab.c index 0dd7b8fec71..fefaabaab76 100644 --- a/mm/failslab.c +++ b/mm/failslab.c @@ -35,7 +35,7 @@ __setup("failslab=", setup_failslab); static int __init failslab_debugfs_init(void) { struct dentry *dir; - mode_t mode = S_IFREG | S_IRUSR | S_IWUSR; + umode_t mode = S_IFREG | S_IRUSR | S_IWUSR; dir = fault_create_debugfs_attr("failslab", NULL, &failslab.attr); if (IS_ERR(dir)) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 2b8ba3aebf6..99930ec7d14 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1408,7 +1408,7 @@ static int should_fail_alloc_page(gfp_t gfp_mask, unsigned int order) static int __init fail_page_alloc_debugfs(void) { - mode_t mode = S_IFREG | S_IRUSR | S_IWUSR; + umode_t mode = S_IFREG | S_IRUSR | S_IWUSR; struct dentry *dir; dir = fault_create_debugfs_attr("fail_page_alloc", NULL, -- cgit v1.2.3-70-g09d2 From fb7ebfe4108e2cdfa2bb88e57148087717463dfa Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 4 Jan 2012 15:50:02 -0800 Subject: PCI: Increase resource array mask bit size in pcim_iomap_regions() DEVICE_COUNT_RESOURCE will be bigger than 16 when SRIOV supported is enabled. Let them pass with int just like pci_enable_resources(). Signed-off-by: Yinghai Lu Signed-off-by: Jesse Barnes --- include/linux/pci.h | 6 +++--- lib/devres.c | 6 +++--- 2 files changed, 6 insertions(+), 6 deletions(-) (limited to 'lib') diff --git a/include/linux/pci.h b/include/linux/pci.h index 5d06e340da9..a16b1df3def 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1432,10 +1432,10 @@ static inline void pci_fixup_device(enum pci_fixup_pass pass, void __iomem *pcim_iomap(struct pci_dev *pdev, int bar, unsigned long maxlen); void pcim_iounmap(struct pci_dev *pdev, void __iomem *addr); void __iomem * const *pcim_iomap_table(struct pci_dev *pdev); -int pcim_iomap_regions(struct pci_dev *pdev, u16 mask, const char *name); -int pcim_iomap_regions_request_all(struct pci_dev *pdev, u16 mask, +int pcim_iomap_regions(struct pci_dev *pdev, int mask, const char *name); +int pcim_iomap_regions_request_all(struct pci_dev *pdev, int mask, const char *name); -void pcim_iounmap_regions(struct pci_dev *pdev, u16 mask); +void pcim_iounmap_regions(struct pci_dev *pdev, int mask); extern int pci_pci_problems; #define PCIPCI_FAIL 1 /* No PCI PCI DMA */ diff --git a/lib/devres.c b/lib/devres.c index 7c0e953a748..2c851bd903b 100644 --- a/lib/devres.c +++ b/lib/devres.c @@ -253,7 +253,7 @@ EXPORT_SYMBOL(pcim_iounmap); * * Request and iomap regions specified by @mask. */ -int pcim_iomap_regions(struct pci_dev *pdev, u16 mask, const char *name) +int pcim_iomap_regions(struct pci_dev *pdev, int mask, const char *name) { void __iomem * const *iomap; int i, rc; @@ -306,7 +306,7 @@ EXPORT_SYMBOL(pcim_iomap_regions); * * Request all PCI BARs and iomap regions specified by @mask. */ -int pcim_iomap_regions_request_all(struct pci_dev *pdev, u16 mask, +int pcim_iomap_regions_request_all(struct pci_dev *pdev, int mask, const char *name) { int request_mask = ((1 << 6) - 1) & ~mask; @@ -330,7 +330,7 @@ EXPORT_SYMBOL(pcim_iomap_regions_request_all); * * Unmap and release regions specified by @mask. */ -void pcim_iounmap_regions(struct pci_dev *pdev, u16 mask) +void pcim_iounmap_regions(struct pci_dev *pdev, int mask) { void __iomem * const *iomap; int i; -- cgit v1.2.3-70-g09d2 From 96b62067f970ff529c98913311d33f4b57b453dc Mon Sep 17 00:00:00 2001 From: Steve Hodgson Date: Tue, 10 Jan 2012 15:09:47 -0800 Subject: btree: export btree_get_prev() so modules can use btree_for_each The btree_for_each API is implemented with macros that internally call btree_get_prev(), so if btree_get_prev() isn't exported then modules fail to link if they try to use one of the btree_for_each macros. Since the rest of the btree API is exported, we should keep things orthogonal and make this work too. Signed-off-by: Roland Dreier Signed-off-by: Steve Hodgson Acked-by: Joern Engel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/btree.c | 1 + 1 file changed, 1 insertion(+) (limited to 'lib') diff --git a/lib/btree.c b/lib/btree.c index 2a34392bcec..e5ec1e9c1aa 100644 --- a/lib/btree.c +++ b/lib/btree.c @@ -357,6 +357,7 @@ miss: } return NULL; } +EXPORT_SYMBOL_GPL(btree_get_prev); static int getpos(struct btree_geo *geo, unsigned long *node, unsigned long *key) -- cgit v1.2.3-70-g09d2 From 5742332dea5560d6c449b007d9539dbdc8ee531b Mon Sep 17 00:00:00 2001 From: Joakim Tjernlund Date: Tue, 10 Jan 2012 15:10:18 -0800 Subject: crc32: optimize inner loop Taking a pointer reference to each row in the crc table matrix, one can reduce the inner loop with a few insn's Signed-off-by: Joakim Tjernlund Cc: Bob Pearson Cc: Frank Zago Cc: Eric Dumazet Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/crc32.c | 21 +++++++++++---------- 1 file changed, 11 insertions(+), 10 deletions(-) (limited to 'lib') diff --git a/lib/crc32.c b/lib/crc32.c index a6e633a48ce..4b35d2b4437 100644 --- a/lib/crc32.c +++ b/lib/crc32.c @@ -51,20 +51,21 @@ static inline u32 crc32_body(u32 crc, unsigned char const *buf, size_t len, const u32 (*tab)[256]) { # ifdef __LITTLE_ENDIAN -# define DO_CRC(x) crc = tab[0][(crc ^ (x)) & 255] ^ (crc >> 8) -# define DO_CRC4 crc = tab[3][(crc) & 255] ^ \ - tab[2][(crc >> 8) & 255] ^ \ - tab[1][(crc >> 16) & 255] ^ \ - tab[0][(crc >> 24) & 255] +# define DO_CRC(x) crc = t0[(crc ^ (x)) & 255] ^ (crc >> 8) +# define DO_CRC4 crc = t3[(crc) & 255] ^ \ + t2[(crc >> 8) & 255] ^ \ + t1[(crc >> 16) & 255] ^ \ + t0[(crc >> 24) & 255] # else -# define DO_CRC(x) crc = tab[0][((crc >> 24) ^ (x)) & 255] ^ (crc << 8) -# define DO_CRC4 crc = tab[0][(crc) & 255] ^ \ - tab[1][(crc >> 8) & 255] ^ \ - tab[2][(crc >> 16) & 255] ^ \ - tab[3][(crc >> 24) & 255] +# define DO_CRC(x) crc = t0[((crc >> 24) ^ (x)) & 255] ^ (crc << 8) +# define DO_CRC4 crc = t0[(crc) & 255] ^ \ + t1[(crc >> 8) & 255] ^ \ + t2[(crc >> 16) & 255] ^ \ + t3[(crc >> 24) & 255] # endif const u32 *b; size_t rem_len; + const u32 *t0=tab[0], *t1=tab[1], *t2=tab[2], *t3=tab[3]; /* Align it */ if (unlikely((long)buf & 3 && len)) { -- cgit v1.2.3-70-g09d2 From e2bdb933ab8b7db71c318a4ddcf78a9fffd61ecb Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 12 Jan 2012 17:20:41 -0800 Subject: radix_tree: take radix_tree_path off stack Down, down in the deepest depths of GFP_NOIO page reclaim, we have shrink_page_list() calling __remove_mapping() calling __delete_from_ swap_cache() or __delete_from_page_cache(). You would not expect those to need much stack, but in fact they call radix_tree_delete(): which declares a 192-byte radix_tree_path array on its stack (to record the node,offsets it visits when descending, in case it needs to ascend to update them). And if any tag is still set [1], that calls radix_tree_tag_clear(), which declares a further such 192-byte radix_tree_path array on the stack. (At least we have interrupts disabled here, so won't then be pushing registers too.) That was probably a good choice when most users were 32-bit (array of half the size), and adding fields to radix_tree_node would have bloated it unnecessarily. But nowadays many are 64-bit, and each radix_tree_node contains a struct rcu_head, which is only used when freeing; whereas the radix_tree_path info is only used for updating the tree (deleting, clearing tags or setting tags if tagged) when a lock must be held, of no interest when accessing the tree locklessly. So add a parent pointer to the radix_tree_node, in union with the rcu_head, and remove all uses of the radix_tree_path. There would be space in that union to save the offset when descending as before (we can argue that a lock must already be held to exclude other users), but recalculating it when ascending is both easy (a constant shift and a constant mask) and uncommon, so it seems better just to do that. Two little optimizations: no need to decrement height when descending, adjusting shift is enough; and once radix_tree_tag_if_tagged() has set tag on a node and its ancestors, it need not ascend from that node again. perf on the radix tree test harness reports radix_tree_insert() as 2% slower (now having to set parent), but radix_tree_delete() 24% faster. Surely that's an exaggeration from rtth's artificially low map shift 3, but forcing it back to 6 still rates radix_tree_delete() 8% faster. [1] Can a pagecache tag (dirty, writeback or towrite) actually still be set at the time of radix_tree_delete()? Perhaps not if the filesystem is well-behaved. But although I've not tracked any stack overflow down to this cause, I have observed a curious case in which a dirty tag is set and left set on tmpfs: page migration's migrate_page_copy() happens to use __set_page_dirty_nobuffers() to set PageDirty on the newpage, and that sets PAGECACHE_TAG_DIRTY as a side-effect - harmless to a filesystem which doesn't use tags, except for this stack depth issue. Signed-off-by: Hugh Dickins Cc: Jan Kara Cc: Dave Chinner Cc: Mel Gorman Cc: Nai Xia Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/radix-tree.c | 154 +++++++++++++++++++++++++++---------------------------- 1 file changed, 76 insertions(+), 78 deletions(-) (limited to 'lib') diff --git a/lib/radix-tree.c b/lib/radix-tree.c index d9df7454519..dc63d081839 100644 --- a/lib/radix-tree.c +++ b/lib/radix-tree.c @@ -48,16 +48,14 @@ struct radix_tree_node { unsigned int height; /* Height from the bottom */ unsigned int count; - struct rcu_head rcu_head; + union { + struct radix_tree_node *parent; /* Used when ascending tree */ + struct rcu_head rcu_head; /* Used when freeing node */ + }; void __rcu *slots[RADIX_TREE_MAP_SIZE]; unsigned long tags[RADIX_TREE_MAX_TAGS][RADIX_TREE_TAG_LONGS]; }; -struct radix_tree_path { - struct radix_tree_node *node; - int offset; -}; - #define RADIX_TREE_INDEX_BITS (8 /* CHAR_BIT */ * sizeof(unsigned long)) #define RADIX_TREE_MAX_PATH (DIV_ROUND_UP(RADIX_TREE_INDEX_BITS, \ RADIX_TREE_MAP_SHIFT)) @@ -256,6 +254,7 @@ static inline unsigned long radix_tree_maxindex(unsigned int height) static int radix_tree_extend(struct radix_tree_root *root, unsigned long index) { struct radix_tree_node *node; + struct radix_tree_node *slot; unsigned int height; int tag; @@ -274,18 +273,23 @@ static int radix_tree_extend(struct radix_tree_root *root, unsigned long index) if (!(node = radix_tree_node_alloc(root))) return -ENOMEM; - /* Increase the height. */ - node->slots[0] = indirect_to_ptr(root->rnode); - /* Propagate the aggregated tag info into the new root */ for (tag = 0; tag < RADIX_TREE_MAX_TAGS; tag++) { if (root_tag_get(root, tag)) tag_set(node, tag, 0); } + /* Increase the height. */ newheight = root->height+1; node->height = newheight; node->count = 1; + node->parent = NULL; + slot = root->rnode; + if (newheight > 1) { + slot = indirect_to_ptr(slot); + slot->parent = node; + } + node->slots[0] = slot; node = ptr_to_indirect(node); rcu_assign_pointer(root->rnode, node); root->height = newheight; @@ -331,6 +335,7 @@ int radix_tree_insert(struct radix_tree_root *root, if (!(slot = radix_tree_node_alloc(root))) return -ENOMEM; slot->height = height; + slot->parent = node; if (node) { rcu_assign_pointer(node->slots[offset], slot); node->count++; @@ -504,47 +509,41 @@ EXPORT_SYMBOL(radix_tree_tag_set); void *radix_tree_tag_clear(struct radix_tree_root *root, unsigned long index, unsigned int tag) { - /* - * The radix tree path needs to be one longer than the maximum path - * since the "list" is null terminated. - */ - struct radix_tree_path path[RADIX_TREE_MAX_PATH + 1], *pathp = path; + struct radix_tree_node *node = NULL; struct radix_tree_node *slot = NULL; unsigned int height, shift; + int uninitialized_var(offset); height = root->height; if (index > radix_tree_maxindex(height)) goto out; - shift = (height - 1) * RADIX_TREE_MAP_SHIFT; - pathp->node = NULL; + shift = height * RADIX_TREE_MAP_SHIFT; slot = indirect_to_ptr(root->rnode); - while (height > 0) { - int offset; - + while (shift) { if (slot == NULL) goto out; + shift -= RADIX_TREE_MAP_SHIFT; offset = (index >> shift) & RADIX_TREE_MAP_MASK; - pathp[1].offset = offset; - pathp[1].node = slot; + node = slot; slot = slot->slots[offset]; - pathp++; - shift -= RADIX_TREE_MAP_SHIFT; - height--; } if (slot == NULL) goto out; - while (pathp->node) { - if (!tag_get(pathp->node, tag, pathp->offset)) + while (node) { + if (!tag_get(node, tag, offset)) goto out; - tag_clear(pathp->node, tag, pathp->offset); - if (any_tag_set(pathp->node, tag)) + tag_clear(node, tag, offset); + if (any_tag_set(node, tag)) goto out; - pathp--; + + index >>= RADIX_TREE_MAP_SHIFT; + offset = index & RADIX_TREE_MAP_MASK; + node = node->parent; } /* clear the root's tag bit */ @@ -646,8 +645,7 @@ unsigned long radix_tree_range_tag_if_tagged(struct radix_tree_root *root, unsigned int iftag, unsigned int settag) { unsigned int height = root->height; - struct radix_tree_path path[height]; - struct radix_tree_path *pathp = path; + struct radix_tree_node *node = NULL; struct radix_tree_node *slot; unsigned int shift; unsigned long tagged = 0; @@ -671,14 +669,8 @@ unsigned long radix_tree_range_tag_if_tagged(struct radix_tree_root *root, shift = (height - 1) * RADIX_TREE_MAP_SHIFT; slot = indirect_to_ptr(root->rnode); - /* - * we fill the path from (root->height - 2) to 0, leaving the index at - * (root->height - 1) as a terminator. Zero the node in the terminator - * so that we can use this to end walk loops back up the path. - */ - path[height - 1].node = NULL; - for (;;) { + unsigned long upindex; int offset; offset = (index >> shift) & RADIX_TREE_MAP_MASK; @@ -686,12 +678,10 @@ unsigned long radix_tree_range_tag_if_tagged(struct radix_tree_root *root, goto next; if (!tag_get(slot, iftag, offset)) goto next; - if (height > 1) { + if (shift) { /* Go down one level */ - height--; shift -= RADIX_TREE_MAP_SHIFT; - path[height - 1].node = slot; - path[height - 1].offset = offset; + node = slot; slot = slot->slots[offset]; continue; } @@ -701,15 +691,27 @@ unsigned long radix_tree_range_tag_if_tagged(struct radix_tree_root *root, tag_set(slot, settag, offset); /* walk back up the path tagging interior nodes */ - pathp = &path[0]; - while (pathp->node) { + upindex = index; + while (node) { + upindex >>= RADIX_TREE_MAP_SHIFT; + offset = upindex & RADIX_TREE_MAP_MASK; + /* stop if we find a node with the tag already set */ - if (tag_get(pathp->node, settag, pathp->offset)) + if (tag_get(node, settag, offset)) break; - tag_set(pathp->node, settag, pathp->offset); - pathp++; + tag_set(node, settag, offset); + node = node->parent; } + /* + * Small optimization: now clear that node pointer. + * Since all of this slot's ancestors now have the tag set + * from setting it above, we have no further need to walk + * back up the tree setting tags, until we update slot to + * point to another radix_tree_node. + */ + node = NULL; + next: /* Go to next item at level determined by 'shift' */ index = ((index >> shift) + 1) << shift; @@ -724,8 +726,7 @@ next: * last_index is guaranteed to be in the tree, what * we do below cannot wander astray. */ - slot = path[height - 1].node; - height++; + slot = slot->parent; shift += RADIX_TREE_MAP_SHIFT; } } @@ -1299,7 +1300,7 @@ static inline void radix_tree_shrink(struct radix_tree_root *root) /* try to shrink tree height */ while (root->height > 0) { struct radix_tree_node *to_free = root->rnode; - void *newptr; + struct radix_tree_node *slot; BUG_ON(!radix_tree_is_indirect_ptr(to_free)); to_free = indirect_to_ptr(to_free); @@ -1320,10 +1321,12 @@ static inline void radix_tree_shrink(struct radix_tree_root *root) * (to_free->slots[0]), it will be safe to dereference the new * one (root->rnode) as far as dependent read barriers go. */ - newptr = to_free->slots[0]; - if (root->height > 1) - newptr = ptr_to_indirect(newptr); - root->rnode = newptr; + slot = to_free->slots[0]; + if (root->height > 1) { + slot->parent = NULL; + slot = ptr_to_indirect(slot); + } + root->rnode = slot; root->height--; /* @@ -1363,16 +1366,12 @@ static inline void radix_tree_shrink(struct radix_tree_root *root) */ void *radix_tree_delete(struct radix_tree_root *root, unsigned long index) { - /* - * The radix tree path needs to be one longer than the maximum path - * since the "list" is null terminated. - */ - struct radix_tree_path path[RADIX_TREE_MAX_PATH + 1], *pathp = path; + struct radix_tree_node *node = NULL; struct radix_tree_node *slot = NULL; struct radix_tree_node *to_free; unsigned int height, shift; int tag; - int offset; + int uninitialized_var(offset); height = root->height; if (index > radix_tree_maxindex(height)) @@ -1385,39 +1384,35 @@ void *radix_tree_delete(struct radix_tree_root *root, unsigned long index) goto out; } slot = indirect_to_ptr(slot); - - shift = (height - 1) * RADIX_TREE_MAP_SHIFT; - pathp->node = NULL; + shift = height * RADIX_TREE_MAP_SHIFT; do { if (slot == NULL) goto out; - pathp++; + shift -= RADIX_TREE_MAP_SHIFT; offset = (index >> shift) & RADIX_TREE_MAP_MASK; - pathp->offset = offset; - pathp->node = slot; + node = slot; slot = slot->slots[offset]; - shift -= RADIX_TREE_MAP_SHIFT; - height--; - } while (height > 0); + } while (shift); if (slot == NULL) goto out; /* - * Clear all tags associated with the just-deleted item + * Clear all tags associated with the item to be deleted. + * This way of doing it would be inefficient, but seldom is any set. */ for (tag = 0; tag < RADIX_TREE_MAX_TAGS; tag++) { - if (tag_get(pathp->node, tag, pathp->offset)) + if (tag_get(node, tag, offset)) radix_tree_tag_clear(root, index, tag); } to_free = NULL; /* Now free the nodes we do not need anymore */ - while (pathp->node) { - pathp->node->slots[pathp->offset] = NULL; - pathp->node->count--; + while (node) { + node->slots[offset] = NULL; + node->count--; /* * Queue the node for deferred freeing after the * last reference to it disappears (set NULL, above). @@ -1425,17 +1420,20 @@ void *radix_tree_delete(struct radix_tree_root *root, unsigned long index) if (to_free) radix_tree_node_free(to_free); - if (pathp->node->count) { - if (pathp->node == indirect_to_ptr(root->rnode)) + if (node->count) { + if (node == indirect_to_ptr(root->rnode)) radix_tree_shrink(root); goto out; } /* Node with zero slots in use so free it */ - to_free = pathp->node; - pathp--; + to_free = node; + index >>= RADIX_TREE_MAP_SHIFT; + offset = index & RADIX_TREE_MAP_MASK; + node = node->parent; } + root_tag_clear_all(root); root->height = 0; root->rnode = NULL; -- cgit v1.2.3-70-g09d2 From 35f1526845a9d804206883e19bd257d3dcef758f Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 12 Jan 2012 17:21:01 -0800 Subject: unlzo: fix input buffer free unlzo modifies the pointer to in_buf, so we have to free the original buffer, not the modified pointer. Signed-off-by: Sascha Hauer Cc: Lasse Collin Cc: Namhyung Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- lib/decompress_unlzo.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'lib') diff --git a/lib/decompress_unlzo.c b/lib/decompress_unlzo.c index 5a7a2adf4c4..4531294fa62 100644 --- a/lib/decompress_unlzo.c +++ b/lib/decompress_unlzo.c @@ -279,7 +279,7 @@ STATIC inline int INIT unlzo(u8 *input, int in_len, ret = 0; exit_2: if (!input) - free(in_buf); + free(in_buf_save); exit_1: if (!output) free(out_buf); -- cgit v1.2.3-70-g09d2