Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-12-29 06:12:08 +00:00
Code Issues Packages Projects Releases Wiki Activity
235b80226b
linux/arch/x86/include/asm/xor_64.h

325 lines
7.0 KiB
C
Raw Normal View History

x86: Fix ASM_X86__ header guards Change header guards named "ASM_X86__*" to "_ASM_X86_*" since: a. the double underscore is ugly and pointless. b. no leading underscore violates namespace constraints. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
2008-10-23 05:26:29 +00:00
#ifndef _ASM_X86_XOR_64_H
#define _ASM_X86_XOR_64_H
x86: more header fixes Summary: Add missing include guards for some x86 headers. This has only had the most rudimentary testing, but is hopefully obviously correct. Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-06-10 21:45:45 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/*
* Optimized RAID-5 checksumming functions for MMX and SSE.
*
* 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; either version 2, or (at your option)
* any later version.
*
* You should have received a copy of the GNU General Public License
* (for example /usr/src/linux/COPYING); if not, write to the Free
* Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
/*
* Cache avoiding checksumming functions utilizing KNI instructions
* Copyright (C) 1999 Zach Brown (with obvious credit due Ingo)
*/
/*
* Based on
* High-speed RAID5 checksumming functions utilizing SSE instructions.
* Copyright (C) 1998 Ingo Molnar.
*/
/*
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
* x86-64 changes / gcc fixes from Andi Kleen.
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
* Copyright 2002 Andi Kleen, SuSE Labs.
*
* This hasn't been optimized for the hammer yet, but there are likely
* no advantages to be gotten from x86-64 here anyways.
*/
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
#include <asm/i387.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#define OFFS(x) "16*("#x")"
#define PF_OFFS(x) "256+16*("#x")"
#define PF0(x) " prefetchnta "PF_OFFS(x)"(%[p1]) ;\n"
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
#define LD(x, y) " movaps "OFFS(x)"(%[p1]), %%xmm"#y" ;\n"
#define ST(x, y) " movaps %%xmm"#y", "OFFS(x)"(%[p1]) ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#define PF1(x) " prefetchnta "PF_OFFS(x)"(%[p2]) ;\n"
#define PF2(x) " prefetchnta "PF_OFFS(x)"(%[p3]) ;\n"
#define PF3(x) " prefetchnta "PF_OFFS(x)"(%[p4]) ;\n"
#define PF4(x) " prefetchnta "PF_OFFS(x)"(%[p5]) ;\n"
#define PF5(x) " prefetchnta "PF_OFFS(x)"(%[p6]) ;\n"
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
#define XO1(x, y) " xorps "OFFS(x)"(%[p2]), %%xmm"#y" ;\n"
#define XO2(x, y) " xorps "OFFS(x)"(%[p3]), %%xmm"#y" ;\n"
#define XO3(x, y) " xorps "OFFS(x)"(%[p4]), %%xmm"#y" ;\n"
#define XO4(x, y) " xorps "OFFS(x)"(%[p5]), %%xmm"#y" ;\n"
#define XO5(x, y) " xorps "OFFS(x)"(%[p6]), %%xmm"#y" ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
static void
xor_sse_2(unsigned long bytes, unsigned long *p1, unsigned long *p2)
{
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
unsigned int lines = bytes >> 8;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
kernel_fpu_begin();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
asm volatile(
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#undef BLOCK
#define BLOCK(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
LD(i, 0) \
LD(i + 1, 1) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF1(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF1(i + 2) \
LD(i + 2, 2) \
LD(i + 3, 3) \
PF0(i + 4) \
PF0(i + 6) \
XO1(i, 0) \
XO1(i + 1, 1) \
XO1(i + 2, 2) \
XO1(i + 3, 3) \
ST(i, 0) \
ST(i + 1, 1) \
ST(i + 2, 2) \
ST(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF0(0)
PF0(2)
" .align 32 ;\n"
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
" 1: ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
BLOCK(0)
BLOCK(4)
BLOCK(8)
BLOCK(12)
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
" addq %[inc], %[p1] ;\n"
" addq %[inc], %[p2] ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
" decl %[cnt] ; jnz 1b"
: [p1] "+r" (p1), [p2] "+r" (p2), [cnt] "+r" (lines)
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
: [inc] "r" (256UL)
: "memory");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
kernel_fpu_end();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
static void
xor_sse_3(unsigned long bytes, unsigned long *p1, unsigned long *p2,
unsigned long *p3)
{
unsigned int lines = bytes >> 8;
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
kernel_fpu_begin();
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
asm volatile(
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#undef BLOCK
#define BLOCK(i) \
PF1(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF1(i + 2) \
LD(i, 0) \
LD(i + 1, 1) \
LD(i + 2, 2) \
LD(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF2(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF2(i + 2) \
PF0(i + 4) \
PF0(i + 6) \
XO1(i, 0) \
XO1(i + 1, 1) \
XO1(i + 2, 2) \
XO1(i + 3, 3) \
XO2(i, 0) \
XO2(i + 1, 1) \
XO2(i + 2, 2) \
XO2(i + 3, 3) \
ST(i, 0) \
ST(i + 1, 1) \
ST(i + 2, 2) \
ST(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF0(0)
PF0(2)
" .align 32 ;\n"
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
" 1: ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
BLOCK(0)
BLOCK(4)
BLOCK(8)
BLOCK(12)
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
" addq %[inc], %[p1] ;\n"
" addq %[inc], %[p2] ;\n"
" addq %[inc], %[p3] ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
" decl %[cnt] ; jnz 1b"
: [cnt] "+r" (lines),
[p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3)
: [inc] "r" (256UL)
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
: "memory");
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
kernel_fpu_end();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
static void
xor_sse_4(unsigned long bytes, unsigned long *p1, unsigned long *p2,
unsigned long *p3, unsigned long *p4)
{
unsigned int lines = bytes >> 8;
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
kernel_fpu_begin();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
asm volatile(
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#undef BLOCK
#define BLOCK(i) \
PF1(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF1(i + 2) \
LD(i, 0) \
LD(i + 1, 1) \
LD(i + 2, 2) \
LD(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF2(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF2(i + 2) \
XO1(i, 0) \
XO1(i + 1, 1) \
XO1(i + 2, 2) \
XO1(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF3(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF3(i + 2) \
PF0(i + 4) \
PF0(i + 6) \
XO2(i, 0) \
XO2(i + 1, 1) \
XO2(i + 2, 2) \
XO2(i + 3, 3) \
XO3(i, 0) \
XO3(i + 1, 1) \
XO3(i + 2, 2) \
XO3(i + 3, 3) \
ST(i, 0) \
ST(i + 1, 1) \
ST(i + 2, 2) \
ST(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF0(0)
PF0(2)
" .align 32 ;\n"
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
" 1: ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
BLOCK(0)
BLOCK(4)
BLOCK(8)
BLOCK(12)
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
" addq %[inc], %[p1] ;\n"
" addq %[inc], %[p2] ;\n"
" addq %[inc], %[p3] ;\n"
" addq %[inc], %[p4] ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
" decl %[cnt] ; jnz 1b"
: [cnt] "+c" (lines),
[p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4)
: [inc] "r" (256UL)
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
: "memory" );
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
kernel_fpu_end();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
static void
xor_sse_5(unsigned long bytes, unsigned long *p1, unsigned long *p2,
unsigned long *p3, unsigned long *p4, unsigned long *p5)
{
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
unsigned int lines = bytes >> 8;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
kernel_fpu_begin();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
asm volatile(
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#undef BLOCK
#define BLOCK(i) \
PF1(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF1(i + 2) \
LD(i, 0) \
LD(i + 1, 1) \
LD(i + 2, 2) \
LD(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF2(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF2(i + 2) \
XO1(i, 0) \
XO1(i + 1, 1) \
XO1(i + 2, 2) \
XO1(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF3(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF3(i + 2) \
XO2(i, 0) \
XO2(i + 1, 1) \
XO2(i + 2, 2) \
XO2(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF4(i) \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
PF4(i + 2) \
PF0(i + 4) \
PF0(i + 6) \
XO3(i, 0) \
XO3(i + 1, 1) \
XO3(i + 2, 2) \
XO3(i + 3, 3) \
XO4(i, 0) \
XO4(i + 1, 1) \
XO4(i + 2, 2) \
XO4(i + 3, 3) \
ST(i, 0) \
ST(i + 1, 1) \
ST(i + 2, 2) \
ST(i + 3, 3) \
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
PF0(0)
PF0(2)
" .align 32 ;\n"
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
" 1: ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
BLOCK(0)
BLOCK(4)
BLOCK(8)
BLOCK(12)
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
" addq %[inc], %[p1] ;\n"
" addq %[inc], %[p2] ;\n"
" addq %[inc], %[p3] ;\n"
" addq %[inc], %[p4] ;\n"
" addq %[inc], %[p5] ;\n"
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
" decl %[cnt] ; jnz 1b"
: [cnt] "+c" (lines),
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
[p1] "+r" (p1), [p2] "+r" (p2), [p3] "+r" (p3), [p4] "+r" (p4),
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
[p5] "+r" (p5)
: [inc] "r" (256UL)
: "memory");
x86, fpu: always use kernel_fpu_begin/end() for in-kernel FPU usage use kernel_fpu_begin/end() instead of unconditionally accessing cr0 and saving/restoring just the few used xmm/ymm registers. This has some advantages like: * If the task's FPU state is already active, then kernel_fpu_begin() will just save the user-state and avoiding the read/write of cr0. In general, cr0 accesses are much slower. * Manual save/restore of xmm/ymm registers will affect the 'modified' and the 'init' optimizations brought in the by xsaveopt/xrstor infrastructure. * Foward compatibility with future vector register extensions will be a problem if the xmm/ymm registers are manually saved and restored (corrupting the extended state of those vector registers). With this patch, there was no significant difference in the xor throughput using AVX, measured during boot. Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com> Link: http://lkml.kernel.org/r/1345842782-24175-5-git-send-email-suresh.b.siddha@intel.com Cc: Jim Kukunas <james.t.kukunas@linux.intel.com> Cc: NeilBrown <neilb@suse.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2012-08-24 21:13:00 +00:00
kernel_fpu_end();
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
}
static struct xor_block_template xor_block_sse = {
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
.name = "generic_sse",
.do_2 = xor_sse_2,
.do_3 = xor_sse_3,
.do_4 = xor_sse_4,
.do_5 = xor_sse_5,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
};
raid5: add AVX optimized RAID5 checksumming Optimize RAID5 xor checksumming by taking advantage of 256-bit YMM registers introduced in AVX. Signed-off-by: Jim Kukunas <james.t.kukunas@linux.intel.com> Signed-off-by: NeilBrown <neilb@suse.de>
2012-05-22 03:54:04 +00:00
/* Also try the AVX routines */
UAPI: (Scripted) Convert #include "..." to #include <path/...> in kernel system headers Convert #include "..." to #include <path/...> in kernel system headers. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Thomas Gleixner <tglx@linutronix.de> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Acked-by: Dave Jones <davej@redhat.com>
2012-10-02 17:01:25 +00:00
#include <asm/xor_avx.h>
raid5: add AVX optimized RAID5 checksumming Optimize RAID5 xor checksumming by taking advantage of 256-bit YMM registers introduced in AVX. Signed-off-by: Jim Kukunas <james.t.kukunas@linux.intel.com> Signed-off-by: NeilBrown <neilb@suse.de>
2012-05-22 03:54:04 +00:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
#undef XOR_TRY_TEMPLATES
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
#define XOR_TRY_TEMPLATES \
do { \
raid5: add AVX optimized RAID5 checksumming Optimize RAID5 xor checksumming by taking advantage of 256-bit YMM registers introduced in AVX. Signed-off-by: Jim Kukunas <james.t.kukunas@linux.intel.com> Signed-off-by: NeilBrown <neilb@suse.de>
2012-05-22 03:54:04 +00:00
AVX_XOR_SPEED; \
include/asm-x86/xor_64.h: checkpatch cleanups - formatting only Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-03-23 08:04:03 +00:00
xor_speed(&xor_block_sse); \
} while (0)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-16 22:20:36 +00:00
/* We force the use of the SSE xor block because it can write around L2.
We may also be able to load into the L1 only depending on how the cpu
deals with a load to a line that is being prefetched. */
raid5: add AVX optimized RAID5 checksumming Optimize RAID5 xor checksumming by taking advantage of 256-bit YMM registers introduced in AVX. Signed-off-by: Jim Kukunas <james.t.kukunas@linux.intel.com> Signed-off-by: NeilBrown <neilb@suse.de>
2012-05-22 03:54:04 +00:00
#define XOR_SELECT_TEMPLATE(FASTEST) \
AVX_SELECT(&xor_block_sse)
x86: more header fixes Summary: Add missing include guards for some x86 headers. This has only had the most rudimentary testing, but is hopefully obviously correct. Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-06-10 21:45:45 +00:00
x86: Fix ASM_X86__ header guards Change header guards named "ASM_X86__*" to "_ASM_X86_*" since: a. the double underscore is ugly and pointless. b. no leading underscore violates namespace constraints. Signed-off-by: H. Peter Anvin <hpa@zytor.com>
2008-10-23 05:26:29 +00:00
#endif /* _ASM_X86_XOR_64_H */
Reference in New Issue Copy Permalink