- 28 Jul, 2004 40 commits
-
-
Andrea Arcangeli authored
I think I understood why some ext2 fs corruption still happens even after the last i_size fix. what happened I believe is that the writepages layer got a not a fully uptodate page (in turn with bh mapped on top of it), and then right before unlocking the page and entering the writeback mode, it freed all the bh. Without bh a not uptodate page will trigger a full readpage from disk, that overwrites the pagecache before the multi-bio gets submitted, generating fs corruption. I believe the below patch should fix it (untested) against kernel CVS. The testcases developed by Kurt showed the pagecache being overwritten with on-disk data at block offsets, and Chris as well was wondering about races between wait_on_page_writeback and readpage. the below fix just explains everything we've seen since not-fully-uptodate pages must have always bh on them and the below patch enforces just that invariant, and it should fix our pagecache-overwritten-by-disk-data problem. Signed-off-by:
Andrea Arcangeli <andrea@suse.de> Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
Linus Torvalds <torvalds@osdl.org>
-
Brian Gerst authored
Remove the orphaned PC9800 boot code. Signed-off-by:
-
Alan Cox authored
This is fairly self explanatory - ENOIOCTLCMD is an internal code outside of the -1 to -511 range. The correct return for an unknown ioctl is -ENOTTY although some Linux devices return the incorrect -EINVAL result. Patch-By:
Alan Cox <alan@redhat.com> OSDL Developer Certificate of Origin 1.0 included herein by reference Signed-off-by:
-
Martin Pool authored
It's slightly hard to see how these variables are used; this patch adds a short description. Signed-off-by:
-
Martin Pool authored
This fixes a path where an error from reloading the partition table could be lost. Signed-off-by:
-
Martin Schwidefsky authored
From: Ursula Braun-Krahl <braunu@de.ibm.com> From: Thomas Spatzier <tspat@de.ibm.com> From: Peter Tiedemann <ptiedem@de.ibm.com> network driver changes. - iucv: add missing symbolic links between /sys/bus/iucv/drivers/netiucv and /sys/class/net. - iucv: make use of the debug feature. - iucv: 0 vs. NULL cleanup, avoid build warnings. - iucv: remove unnecessary -ENODEV check after call to iucv_declare_buffer. - ctc: adjust debug feature log levels. - lcs: replace broken schedule_timeout call with msleep. - qeth: add missing link type. - qeth: treat add_hhlen attribute as decimal value. Signed-off-by:
Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by:
-
Martin Schwidefsky authored
From: Heiko Carstens <heiko.carstens@de.ibm.com> From: Maxim Shchetynin <maxim@de.ibm.com> zfcp host adapter changes: - Get rid of ZFCP_WAIT_EVENT_TIMEOUT macro. - Correct an error log mesage. Signed-off-by:
-
Martin Schwidefsky authored
s390 core changes: - Rework system call entry cleanup code to fix a potential asynchronous interrupt stack overflow if the user stack pointer happens to be in the same range as the asynchronous stack. - Replace broken schedule_timeout call with msleep. - Regenerate default configuration Signed-off-by:
-
Zwane Mwaikambo authored
We might as well recommend the user boot with the noapic kernel parameter instead of filling poor Ingo's mailbox. Signed-off-by:
Zwane Mwaikambo <zwane@fsmlabs.com> Signed-off-by:
Ingo Molnar <mingo@elte.hu> Signed-off-by:
-
Andrew Morton authored
Arjan's new warning found a bug! fs/ncpfs/inode.c: In function `ncp_notify_change': fs/ncpfs/inode.c:960: warning: ignoring return value of `inode_setattr', declared with attribute warn_unused_result Signed-off-by:
-
Jan Blunck authored
ext2_readdir() is ignoring the set error return values and always returns 0. Signed-off-by:
Jan Blunck <j.blunck@tu-harburg.de> Signed-off-by:
-
Olaf Hering authored
The alsa driver for powermacs requires i2c support. Signed-off-by:
Olaf Hering <olh@suse.de> Signed-off-by:
-
Bart Samwel authored
From: Jan Topinski <simcha@chatka.org> Add support to automatically disable laptop mode when the battery almost runs out. This prevents data loss when the battery actually runs out. Signed-off-by:
Bart Samwel <bart@samwel.tk> Signed-off-by:
-
Bart Samwel authored
From: Bart Samwel <bart@samwel.tk> A bunch of small laptop mode updates: * Mention the laptop_mode and block_dump sysctls in the relevant docs, referring to the laptop-mode.txt. Improve description of what the laptop_mode knob means (i.e. seconds between last I/O and automatic sync). * Document the most relevant config file options. * Change the installation instructions in the laptop mode doc to describe the ACPI-bound version first, because this is the easiest. Explain the non-ACPI version afterwards. Added a reference to my "packaged version" of the scripts. * Move the setting of hdparm and cpu frequency out of the ACPI script and into the core laptop mode script. Add config option DO_HD to enable hdparm. Add config setting HD to set device to do hdparm on. Rename Dax's CPU_MANAGE to DO_CPU, with 0/1 instead of yes/no, to improve consistency. * Fix a problem where the root filesystem wouln't be remounted when /etc/mtab listed its type as "unknown". Deduce the type of "unknown" filesystems from /etc/fstab. This depends on the patch that adds config files, which was submitted to Linus by Dax Kelson on July 10. Signed-off-by:
-
Bart Samwel authored
From: Dax Kelson <dax@gurulabs.com> Add support for a external configuration file for /sbin/laptop_mode, and /etc/acpi/actions/battery.sh. Convert battery.sh to use CPUFreq (off by default) instead of CPU throttling (that was off by default). Cleanup some formating for 80 columns. All changes were written by me on a plane flight from Philadelphia to Salt Lake City on July 9th. :) (Note from Bart Samwel: was submitted to Linus earlier, this patch is unchanged from that. I needed to submit this into -mm for now because my other changes are dependent on Dax's changes.) Signed-off-by:
-
Samuel Thibault authored
Here is a small fix to enable muting ESS Maestro sound card thanks to the up/down buttons: when has reached the driver's minimum (! this is something like -dB value), just mute. (It was working in OSS driver, but not in ALSA) Signed-off-by:
-
Stephen Hemminger authored
Make code for hlist_for_each_safe use better code (same as hlist_for_each_entry_safe). Get rid of comment about prefetch, because that was fixed a while ago. Only current use of this is in the bridge code, that I maintain. Signed-off-by:
Stephen Hemminger <shemminger@osdl.org> Signed-off-by:
-
Roland Dreier authored
Quite a few functions in lib/string.c are not exported. I ran into this trying to use strnchr() in a module. This patch - exports every function defined in lib/string.c - adds some missing __HAVE_ARCH_xxx defines for i386 - gets rid of the exports of functions from lib/string.c in arch/s390 (BTW, why is s390 exporting things NOVERS?) Signed-off-by:
Roland Dreier <roland@topspin.com> Signed-off-by:
-
Domen Puncer authored
From: Geert Uytterhoeven <geert@linux-m68k.org> These four are called from fb_show_logo, which is exported symbol, called by fbcon_switch. From: Domen Puncer <domen@coderock.org> Signed-off-by: Maximilian Attems Signed-off-by:
-
Shane Shrybman authored
Removal of unused variable in mm/readahead.c. Signed-off-by:
-
David Eger authored
Cirrusfb updates: - Cirrusfb depends on Zorro or PCI - Revive Zorro support, using new Zorro driver framework - Merge PCI and Zorro code where possible - Use C99 initializers - Kill superfluous whitespace - #undef I/O ops before redefining them on Amiga From: Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by:
David Eger <eger@havoc.gtf.org> Signed-off-by:
-
Geert Uytterhoeven authored
Dallas 1-wire: never include <asm/delay.h> directly Signed-off-by:
Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by:
-
Geert Uytterhoeven authored
Matrox 1-wire unconditionally depends on PCI Signed-off-by:
-
Geert Uytterhoeven authored
Toshiba and VIA FIR unconditionally depend on PCI Signed-off-by:
-
Geert Uytterhoeven authored
Cisco/Aironet 34X/35X/4500/4800 unconditionally depends on the PCI DMA API, so mark it broken if !PCI Signed-off-by:
-
Geert Uytterhoeven authored
Technisat Skystar2 PCI unconditionally depends on PCI Signed-off-by:
-
Geert Uytterhoeven authored
Guillemot MAXI Radio FM 2000 unconditionally depends on PCI Signed-off-by:
-
Geert Uytterhoeven authored
Kill warnings in Hisax ISDN drivers when !PCI. Signed-off-by:
-
Geert Uytterhoeven authored
Fritz!PCI/PCIv2/PnP and HYSDN unconditionally depend on PCI Signed-off-by:
-
Geert Uytterhoeven authored
IEEE1394 core and SBP-2 unconditionally depend on the PCI DMA API, so mark them broken if !PCI Signed-off-by:
-
Geert Uytterhoeven authored
VIA686A i2c unconditionally depends on PCI Signed-off-by:
-
Geert Uytterhoeven authored
Kill warnings in Specialix serial driver when !PCI. Signed-off-by:
-
Geert Uytterhoeven authored
Kill warnings in MOXA Intellio and Smartio multiport serial drivers when !PCI. Also kill warnings about unused variables in the non-modular case. Signed-off-by:
-
Geert Uytterhoeven authored
Multi-Tech, Microgate SyncLink, and Applicom serial unconditionally depend on PCI Signed-off-by:
-
Geert Uytterhoeven authored
M68k maintainership update: - Transfer maintainership to Roman and me (ack'ed by Jes and Roman) - Update main website URL - Add Linux/m68k CVS repository website Signed-off-by:
-
Geert Uytterhoeven authored
M68k: Fix build after <asm/pgalloc.h> surgery in 2.6.8-rc1: - Add missing include on machines with a standard m68k MMU - Convert __pte_free_tlb() to a macro (like it is on most other archs) on Sun-3, to avoid include hell Signed-off-by: -
Geert Uytterhoeven authored
M68k: Add missing include, showing up due to include reshuffling in 2.6.8-rc1 Signed-off-by:
-
Geert Uytterhoeven authored
M68k: Add `volatile' to some bitops parameters, as this is required by the cpumask code in 2.6.8-rc1. Signed-off-by:
-
Geert Uytterhoeven authored
Dmasound was moved to sound/oss/ a while ago Signed-off-by:
-
Geert Uytterhoeven authored
M68k: Add missing #include <linux/config.h> Signed-off-by:
-
