| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | /*
 | 
					
						
							|  |  |  |  * LZO decompressor for the Linux kernel. Code borrowed from the lzo | 
					
						
							|  |  |  |  * implementation by Markus Franz Xaver Johannes Oberhumer. | 
					
						
							|  |  |  |  * | 
					
						
							|  |  |  |  * Linux kernel adaptation: | 
					
						
							|  |  |  |  * Copyright (C) 2009 | 
					
						
							|  |  |  |  * Albin Tonnerre, Free Electrons <albin.tonnerre@free-electrons.com> | 
					
						
							|  |  |  |  * | 
					
						
							|  |  |  |  * Original code: | 
					
						
							|  |  |  |  * Copyright (C) 1996-2005 Markus Franz Xaver Johannes Oberhumer | 
					
						
							|  |  |  |  * All Rights Reserved. | 
					
						
							|  |  |  |  * | 
					
						
							|  |  |  |  * lzop and the LZO library are free software; you can redistribute them | 
					
						
							|  |  |  |  * and/or modify them 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. | 
					
						
							|  |  |  |  * | 
					
						
							|  |  |  |  * This program 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; see the file COPYING. | 
					
						
							|  |  |  |  * If not, write to the Free Software Foundation, Inc., | 
					
						
							|  |  |  |  * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. | 
					
						
							|  |  |  |  * | 
					
						
							|  |  |  |  * Markus F.X.J. Oberhumer | 
					
						
							|  |  |  |  * <markus@oberhumer.com> | 
					
						
							|  |  |  |  * http://www.oberhumer.com/opensource/lzop/
 | 
					
						
							|  |  |  |  */ | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | #ifdef STATIC
 | 
					
						
							| 
									
										
										
										
											2012-08-13 17:24:24 +02:00
										 |  |  | #include "lzo/lzo1x_decompress_safe.c"
 | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | #else
 | 
					
						
							|  |  |  | #include <linux/decompress/unlzo.h>
 | 
					
						
							|  |  |  | #endif
 | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | #include <linux/types.h>
 | 
					
						
							|  |  |  | #include <linux/lzo.h>
 | 
					
						
							|  |  |  | #include <linux/decompress/mm.h>
 | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | #include <linux/compiler.h>
 | 
					
						
							|  |  |  | #include <asm/unaligned.h>
 | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | static const unsigned char lzop_magic[] = { | 
					
						
							|  |  |  | 	0x89, 0x4c, 0x5a, 0x4f, 0x00, 0x0d, 0x0a, 0x1a, 0x0a }; | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | #define LZO_BLOCK_SIZE        (256*1024l)
 | 
					
						
							|  |  |  | #define HEADER_HAS_FILTER      0x00000800L
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | #define HEADER_SIZE_MIN       (9 + 7     + 4 + 8     + 1       + 4)
 | 
					
						
							|  |  |  | #define HEADER_SIZE_MAX       (9 + 7 + 1 + 8 + 8 + 4 + 1 + 255 + 4)
 | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | STATIC inline int INIT parse_header(u8 *input, int *skip, int in_len) | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | { | 
					
						
							|  |  |  | 	int l; | 
					
						
							|  |  |  | 	u8 *parse = input; | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	u8 *end = input + in_len; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 	u8 level = 0; | 
					
						
							|  |  |  | 	u16 version; | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	/*
 | 
					
						
							|  |  |  | 	 * Check that there's enough input to possibly have a valid header. | 
					
						
							|  |  |  | 	 * Then it is possible to parse several fields until the minimum | 
					
						
							|  |  |  | 	 * size may have been used. | 
					
						
							|  |  |  | 	 */ | 
					
						
							|  |  |  | 	if (in_len < HEADER_SIZE_MIN) | 
					
						
							|  |  |  | 		return 0; | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 	/* read magic: 9 first bits */ | 
					
						
							|  |  |  | 	for (l = 0; l < 9; l++) { | 
					
						
							|  |  |  | 		if (*parse++ != lzop_magic[l]) | 
					
						
							|  |  |  | 			return 0; | 
					
						
							|  |  |  | 	} | 
					
						
							|  |  |  | 	/* get version (2bytes), skip library version (2),
 | 
					
						
							|  |  |  | 	 * 'need to be extracted' version (2) and | 
					
						
							|  |  |  | 	 * method (1) */ | 
					
						
							|  |  |  | 	version = get_unaligned_be16(parse); | 
					
						
							|  |  |  | 	parse += 7; | 
					
						
							|  |  |  | 	if (version >= 0x0940) | 
					
						
							|  |  |  | 		level = *parse++; | 
					
						
							|  |  |  | 	if (get_unaligned_be32(parse) & HEADER_HAS_FILTER) | 
					
						
							|  |  |  | 		parse += 8; /* flags + filter info */ | 
					
						
							|  |  |  | 	else | 
					
						
							|  |  |  | 		parse += 4; /* flags */ | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	/*
 | 
					
						
							|  |  |  | 	 * At least mode, mtime_low, filename length, and checksum must | 
					
						
							|  |  |  | 	 * be left to be parsed. If also mtime_high is present, it's OK | 
					
						
							|  |  |  | 	 * because the next input buffer check is after reading the | 
					
						
							|  |  |  | 	 * filename length. | 
					
						
							|  |  |  | 	 */ | 
					
						
							|  |  |  | 	if (end - parse < 8 + 1 + 4) | 
					
						
							|  |  |  | 		return 0; | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 	/* skip mode and mtime_low */ | 
					
						
							|  |  |  | 	parse += 8; | 
					
						
							|  |  |  | 	if (version >= 0x0940) | 
					
						
							|  |  |  | 		parse += 4;	/* skip mtime_high */ | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | 	l = *parse++; | 
					
						
							|  |  |  | 	/* don't care about the file name, and skip checksum */ | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	if (end - parse < l + 4) | 
					
						
							|  |  |  | 		return 0; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 	parse += l + 4; | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | 	*skip = parse - input; | 
					
						
							|  |  |  | 	return 1; | 
					
						
							|  |  |  | } | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | STATIC inline int INIT unlzo(u8 *input, int in_len, | 
					
						
							|  |  |  | 				int (*fill) (void *, unsigned int), | 
					
						
							|  |  |  | 				int (*flush) (void *, unsigned int), | 
					
						
							|  |  |  | 				u8 *output, int *posp, | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:14 -08:00
										 |  |  | 				void (*error) (char *x)) | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | { | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	u8 r = 0; | 
					
						
							|  |  |  | 	int skip = 0; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 	u32 src_len, dst_len; | 
					
						
							|  |  |  | 	size_t tmp; | 
					
						
							|  |  |  | 	u8 *in_buf, *in_buf_save, *out_buf; | 
					
						
							| 
									
										
										
										
											2010-04-23 13:17:58 -04:00
										 |  |  | 	int ret = -1; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 
 | 
					
						
							|  |  |  | 	if (output) { | 
					
						
							|  |  |  | 		out_buf = output; | 
					
						
							|  |  |  | 	} else if (!flush) { | 
					
						
							|  |  |  | 		error("NULL output pointer and no flush function provided"); | 
					
						
							|  |  |  | 		goto exit; | 
					
						
							|  |  |  | 	} else { | 
					
						
							|  |  |  | 		out_buf = malloc(LZO_BLOCK_SIZE); | 
					
						
							|  |  |  | 		if (!out_buf) { | 
					
						
							|  |  |  | 			error("Could not allocate output buffer"); | 
					
						
							|  |  |  | 			goto exit; | 
					
						
							|  |  |  | 		} | 
					
						
							|  |  |  | 	} | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | 	if (input && fill) { | 
					
						
							|  |  |  | 		error("Both input pointer and fill function provided, don't know what to do"); | 
					
						
							|  |  |  | 		goto exit_1; | 
					
						
							|  |  |  | 	} else if (input) { | 
					
						
							|  |  |  | 		in_buf = input; | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	} else if (!fill) { | 
					
						
							|  |  |  | 		error("NULL input pointer and missing fill function"); | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 		goto exit_1; | 
					
						
							|  |  |  | 	} else { | 
					
						
							|  |  |  | 		in_buf = malloc(lzo1x_worst_compress(LZO_BLOCK_SIZE)); | 
					
						
							|  |  |  | 		if (!in_buf) { | 
					
						
							|  |  |  | 			error("Could not allocate input buffer"); | 
					
						
							|  |  |  | 			goto exit_1; | 
					
						
							|  |  |  | 		} | 
					
						
							|  |  |  | 	} | 
					
						
							|  |  |  | 	in_buf_save = in_buf; | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | 	if (posp) | 
					
						
							|  |  |  | 		*posp = 0; | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	if (fill) { | 
					
						
							|  |  |  | 		/*
 | 
					
						
							|  |  |  | 		 * Start from in_buf + HEADER_SIZE_MAX to make it possible | 
					
						
							|  |  |  | 		 * to use memcpy() to copy the unused data to the beginning | 
					
						
							|  |  |  | 		 * of the buffer. This way memmove() isn't needed which | 
					
						
							|  |  |  | 		 * is missing from pre-boot environments of most archs. | 
					
						
							|  |  |  | 		 */ | 
					
						
							|  |  |  | 		in_buf += HEADER_SIZE_MAX; | 
					
						
							|  |  |  | 		in_len = fill(in_buf, HEADER_SIZE_MAX); | 
					
						
							|  |  |  | 	} | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	if (!parse_header(in_buf, &skip, in_len)) { | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 		error("invalid header"); | 
					
						
							|  |  |  | 		goto exit_2; | 
					
						
							|  |  |  | 	} | 
					
						
							|  |  |  | 	in_buf += skip; | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	in_len -= skip; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 	if (fill) { | 
					
						
							|  |  |  | 		/* Move the unused data to the beginning of the buffer. */ | 
					
						
							|  |  |  | 		memcpy(in_buf_save, in_buf, in_len); | 
					
						
							|  |  |  | 		in_buf = in_buf_save; | 
					
						
							|  |  |  | 	} | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 	if (posp) | 
					
						
							|  |  |  | 		*posp = skip; | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | 	for (;;) { | 
					
						
							|  |  |  | 		/* read uncompressed block size */ | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		if (fill && in_len < 4) { | 
					
						
							|  |  |  | 			skip = fill(in_buf + in_len, 4 - in_len); | 
					
						
							|  |  |  | 			if (skip > 0) | 
					
						
							|  |  |  | 				in_len += skip; | 
					
						
							|  |  |  | 		} | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		if (in_len < 4) { | 
					
						
							|  |  |  | 			error("file corrupted"); | 
					
						
							|  |  |  | 			goto exit_2; | 
					
						
							|  |  |  | 		} | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 		dst_len = get_unaligned_be32(in_buf); | 
					
						
							|  |  |  | 		in_buf += 4; | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		in_len -= 4; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 
 | 
					
						
							|  |  |  | 		/* exit if last block */ | 
					
						
							|  |  |  | 		if (dst_len == 0) { | 
					
						
							|  |  |  | 			if (posp) | 
					
						
							|  |  |  | 				*posp += 4; | 
					
						
							|  |  |  | 			break; | 
					
						
							|  |  |  | 		} | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | 		if (dst_len > LZO_BLOCK_SIZE) { | 
					
						
							|  |  |  | 			error("dest len longer than block size"); | 
					
						
							|  |  |  | 			goto exit_2; | 
					
						
							|  |  |  | 		} | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | 		/* read compressed block size, and skip block checksum info */ | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		if (fill && in_len < 8) { | 
					
						
							|  |  |  | 			skip = fill(in_buf + in_len, 8 - in_len); | 
					
						
							|  |  |  | 			if (skip > 0) | 
					
						
							|  |  |  | 				in_len += skip; | 
					
						
							|  |  |  | 		} | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		if (in_len < 8) { | 
					
						
							|  |  |  | 			error("file corrupted"); | 
					
						
							|  |  |  | 			goto exit_2; | 
					
						
							|  |  |  | 		} | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 		src_len = get_unaligned_be32(in_buf); | 
					
						
							|  |  |  | 		in_buf += 8; | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		in_len -= 8; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		if (src_len <= 0 || src_len > dst_len) { | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 			error("file corrupted"); | 
					
						
							|  |  |  | 			goto exit_2; | 
					
						
							|  |  |  | 		} | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | 		/* decompress */ | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		if (fill && in_len < src_len) { | 
					
						
							|  |  |  | 			skip = fill(in_buf + in_len, src_len - in_len); | 
					
						
							|  |  |  | 			if (skip > 0) | 
					
						
							|  |  |  | 				in_len += skip; | 
					
						
							|  |  |  | 		} | 
					
						
							|  |  |  | 		if (in_len < src_len) { | 
					
						
							|  |  |  | 			error("file corrupted"); | 
					
						
							|  |  |  | 			goto exit_2; | 
					
						
							|  |  |  | 		} | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 		tmp = dst_len; | 
					
						
							| 
									
										
										
										
											2010-04-23 13:17:58 -04:00
										 |  |  | 
 | 
					
						
							|  |  |  | 		/* When the input data is not compressed at all,
 | 
					
						
							|  |  |  | 		 * lzo1x_decompress_safe will fail, so call memcpy() | 
					
						
							|  |  |  | 		 * instead */ | 
					
						
							|  |  |  | 		if (unlikely(dst_len == src_len)) | 
					
						
							|  |  |  | 			memcpy(out_buf, in_buf, src_len); | 
					
						
							|  |  |  | 		else { | 
					
						
							|  |  |  | 			r = lzo1x_decompress_safe((u8 *) in_buf, src_len, | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 						out_buf, &tmp); | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
										
										
											2010-04-23 13:17:58 -04:00
										 |  |  | 			if (r != LZO_E_OK || dst_len != tmp) { | 
					
						
							|  |  |  | 				error("Compressed data violation"); | 
					
						
							|  |  |  | 				goto exit_2; | 
					
						
							|  |  |  | 			} | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 		} | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:20 -08:00
										 |  |  | 		if (flush && flush(out_buf, dst_len) != dst_len) | 
					
						
							|  |  |  | 			goto exit_2; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 		if (output) | 
					
						
							|  |  |  | 			out_buf += dst_len; | 
					
						
							|  |  |  | 		if (posp) | 
					
						
							|  |  |  | 			*posp += src_len + 12; | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 
 | 
					
						
							|  |  |  | 		in_buf += src_len; | 
					
						
							|  |  |  | 		in_len -= src_len; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 		if (fill) { | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 			/*
 | 
					
						
							|  |  |  | 			 * If there happens to still be unused data left in | 
					
						
							|  |  |  | 			 * in_buf, move it to the beginning of the buffer. | 
					
						
							|  |  |  | 			 * Use a loop to avoid memmove() dependency. | 
					
						
							|  |  |  | 			 */ | 
					
						
							|  |  |  | 			if (in_len > 0) | 
					
						
							|  |  |  | 				for (skip = 0; skip < in_len; ++skip) | 
					
						
							|  |  |  | 					in_buf_save[skip] = in_buf[skip]; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 			in_buf = in_buf_save; | 
					
						
							| 
									
										
										
										
											2011-01-12 17:01:21 -08:00
										 |  |  | 		} | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | 	} | 
					
						
							|  |  |  | 
 | 
					
						
							| 
									
										
										
										
											2010-04-23 13:17:58 -04:00
										 |  |  | 	ret = 0; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | exit_2: | 
					
						
							|  |  |  | 	if (!input) | 
					
						
							| 
									
										
										
										
											2012-01-12 17:21:01 -08:00
										 |  |  | 		free(in_buf_save); | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | exit_1: | 
					
						
							|  |  |  | 	if (!output) | 
					
						
							|  |  |  | 		free(out_buf); | 
					
						
							|  |  |  | exit: | 
					
						
							| 
									
										
										
										
											2010-04-23 13:17:58 -04:00
										 |  |  | 	return ret; | 
					
						
							| 
									
										
											  
											
												lib: add support for LZO-compressed kernels
This patch series adds generic support for creating and extracting
LZO-compressed kernel images, as well as support for using such images on
the x86 and ARM architectures, and support for creating and using
LZO-compressed initrd and initramfs images.
Russell King said:
: Testing on a Cortex A9 model:
: - lzo decompressor is 65% of the time gzip takes to decompress a kernel
: - lzo kernel is 9% larger than a gzip kernel
:
: which I'm happy to say confirms your figures when comparing the two.
:
: However, when comparing your new gzip code to the old gzip code:
: - new is 99% of the size of the old code
: - new takes 42% of the time to decompress than the old code
:
: What this means is that for a proper comparison, the results get even better:
: - lzo is 7.5% larger than the old gzip'd kernel image
: - lzo takes 28% of the time that the old gzip code took
:
: So the expense seems definitely worth the effort.  The only reason I
: can think of ever using gzip would be if you needed the additional
: compression (eg, because you have limited flash to store the image.)
:
: I would argue that the default for ARM should therefore be LZO.
This patch:
The lzo compressor is worse than gzip at compression, but faster at
extraction.  Here are some figures for an ARM board I'm working on:
Uncompressed size: 3.24Mo
gzip  1.61Mo 0.72s
lzo   1.75Mo 0.48s
So for a compression ratio that is still relatively close to gzip, it's
much faster to extract, at least in that case.
This part contains:
 - Makefile routine to support lzo compression
 - Fixes to the existing lzo compressor so that it can be used in
   compressed kernels
 - wrapper around the existing lzo1x_decompress, as it only extracts one
   block at a time, while we need to extract a whole file here
 - config dialog for kernel compression
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: cleanup]
Signed-off-by: Albin Tonnerre <albin.tonnerre@free-electrons.com>
Tested-by: Wu Zhangjin <wuzhangjin@gmail.com>
Acked-by: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Russell King <rmk@arm.linux.org.uk>
Acked-by: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
											
										 
											2010-01-08 14:42:42 -08:00
										 |  |  | } | 
					
						
							|  |  |  | 
 | 
					
						
							|  |  |  | #define decompress unlzo
 |