- 20 Apr, 2006 34 commits
-
-
Linus Torvalds authored
* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6: [EBTABLES]: Clean up vmalloc usage in net/bridge/netfilter/ebtables.c [NET]: Add skb->truesize assertion checking. [TCP]: Account skb overhead in tcp_fragment [SUNGEM]: Marvell PHY suspend. [LLC]: Use pskb_trim_rcsum() in llc_fixup_skb(). [NET]: sockfd_lookup_light() returns random error for -EBADFD
-
Linus Torvalds authored
* master.kernel.org:/home/rmk/linux-2.6-arm: [ARM] for_each_possible_cpu [ARM] add_memory() build fix [ARM] 3483/1: ixp23xx: update defconfig to 2.6.17-rc2 [ARM] 3482/1: ixp2000: update defconfig to 2.6.17-rc2 [ARM] 3481/1: ep93xx: update defconfig to 2.6.17-rc2 [ARM] 3480/1: ixp4xx: fix irq2gpio array type
-
KAMEZAWA Hiroyuki authored
for_each_cpu() actually iterates across all possible CPUs. We've had mistakes in the past where people were using for_each_cpu() where they should have been iterating across only online or present CPUs. This is inefficient and possibly buggy. We're renaming for_each_cpu() to for_each_possible_cpu() to avoid this in the future. This patch replaces for_each_cpu with for_each_possible_cpu. Signed-off-by:
KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
Russell King <rmk+kernel@arm.linux.org.uk>
-
Andrew Morton authored
This is back again. Offending patch is x86_64-mm-hotadd-reserve.patch arch/arm/kernel/setup.c:435: error: conflicting types for 'add_memory' include/linux/memory_hotplug.h:102: error: previous declaration of 'add_memory' was here arch/arm/kernel/setup.c:435: error: conflicting types for 'add_memory' include/linux/memory_hotplug.h:102: error: previous declaration of 'add_memory' was here Signed-off-by:
-
Lennert Buytenhek authored
Patch from Lennert Buytenhek Signed-off-by:
Lennert Buytenhek <buytenh@wantstofly.org> Signed-off-by:
-
Lennert Buytenhek authored
Patch from Lennert Buytenhek Signed-off-by:
-
Lennert Buytenhek authored
Patch from Lennert Buytenhek Signed-off-by:
-
Lennert Buytenhek authored
Patch from Lennert Buytenhek The irq2gpio array was recently converted from an array of ints to an array of chars (by patch 3368/1.) However, this array contains elements that are -1, and on ARM, the char type is unsigned by default, so this patch broke the GPIO check in ixp4xx_set_irq_type. Change the 'char' to be a 'signed char' to fix this. Signed-off-by:
Deepak Saxena <dsaxena@plexity.net> Signed-off-by:
-
git://brick.kernel.dk/data/git/linux-2.6-blockLinus Torvalds authored
* 'for-linus' of git://brick.kernel.dk/data/git/linux-2.6-block: [PATCH] block/elevator.c: remove unused exports [PATCH] splice: fix smaller sized splice reads [PATCH] Don't inherit ->splice_pipe across forks [patch] cleanup: use blk_queue_stopped [PATCH] Document online io scheduler switching
-
Linus Torvalds authored
* master.kernel.org:/pub/scm/linux/kernel/git/davem/sparc-2.6: [SPARC]: __NR_sys_splice --> __NR_splice
-
Eric Dumazet authored
It seems latest kernel has a wrong/missing __read_mostly implementation for x86_64 __read_mostly macro should be declared outside of #if CONFIG_X86_VSMP block Signed-off-by:
Eric Dumazet <dada1@cosmosbay.com> Signed-off-by:
Andi Kleen <ak@suse.de> Signed-off-by:
Linus Torvalds <torvalds@osdl.org>
-
Andi Kleen authored
AMD K7/K8 CPUs only save/restore the FOP/FIP/FDP x87 registers in FXSAVE when an exception is pending. This means the value leak through context switches and allow processes to observe some x87 instruction state of other processes. This was actually documented by AMD, but nobody recognized it as being different from Intel before. The fix first adds an optimization: instead of unconditionally calling FNCLEX after each FXSAVE test if ES is pending and skip it when not needed. Then do a x87 load from a kernel variable to clear FOP/FIP/FDP. This means other processes always will only see a constant value defined by the kernel in their FP state. I took some pain to make sure to chose a variable that's already in L1 during context switch to make the overhead of this low. Also alternative() is used to patch away the new code on CPUs who don't need it. Patch for both i386/x86-64. The problem was discovered originally by Jan Beulich. Richard Brunner provided the basic code for the workarounds, with contribution from Jan. This is CVE-2006-1056 Cc: richard.brunner@amd.com Cc: jbeulich@novell.com Signed-off-by:
-
Ingo Molnar authored
- fix mddev_lock() usage bugs in md_attr_show() and md_attr_store(). [they did not anticipate the possibility of getting a signal] - remove mddev_lock_uninterruptible() [unused] Signed-off-by:
Ingo Molnar <mingo@elte.hu> Acked-by:
Neil Brown <neilb@suse.de> Signed-off-by:
-
Adrian Bunk authored
We don't have to #if guard prototypes. This also fixes a bug observed by Randy Dunlap due to a misspelled option in the #if. Signed-off-by:
Adrian Bunk <bunk@stusta.de> Signed-off-by:
-
Adrian Bunk authored
There was a report of a regression in the ALSA driver for the same hardware. Signed-off-by:
-
Adrian Bunk authored
Update the schedule for the removal of drivers depending on OBSOLETE_OSS_DRIVER as follows: - adjust OBSOLETE_OSS_DRIVER dependencie - from the release of 2.6.16 till the release of 2.6.17: approx. two months for users to report problems with the ALSA drivers for the same hardware - after the release of 2.6.17 (and before 2.6.18): remove the subset of drivers marked at OBSOLETE_OSS_DRIVER without known regressions in the ALSA drivers for the same hardware Additionally, correct some OBSOLETE_OSS_DRIVER dependencies. A rationale of the changes is in http://lkml.org/lkml/2006/1/28/135Signed-off-by:
-
Arnaud MAZIN authored
Add a test to detect the ICH7 based Core Duo SONY laptops (such as the SZ1) as type3 models. Signed-off-by: Arnaud MAZIN < arnaud.mazin@gmail.com> Acked-by:
Stelian Pop <stelian@poppies.net> Signed-off-by:
-
Jon Masters authored
This fixes a hang in mpu401_uart.c that can occur when the mpu401 interface is non-existent or otherwise doesn't respond to commands but we issue IO anyway. snd_mpu401_uart_cmd now returns an error code that is passed up the stack so that an open() will fail immediately in such cases. Eventually discovered after wine/cxoffice would constantly cause hard lockups on my desktop immediately after loading (emulating Windows too well). Turned out that I'd recently moved my sound cards around and using /dev/sequencer now talks to a sound card with a broken MPU. This second version changes -EFAULT to -EIO and frees open resources on error too. Test booted and seems to work ok. Signed-off-by:
Jon Masters <jcm@jonmasters.org> Cc: Jaroslav Kysela <perex@suse.cz> Acked-by:
Takashi Iwai <tiwai@suse.de> Signed-off-by:
-
Jean Delvare authored
Fix for bug #6395: Fail to resume on Tecra M2 with ADM1032 and Intel 82801DBM The BIOS of the Tecra M2 doesn't like it when it has to reboot or resume after the i2c-i801 driver has left the SMBus in PEC mode. The most simple fix is to clear the PEC bit after after every transaction. That's what this driver was doing up to 2.6.15 (inclusive). Thanks to Daniele Gaffuri for the very good report. Signed-off-by:
Jean Delvare <khali@linux-fr.org> Cc: Greg KH <greg@kroah.com> Signed-off-by:
-
Johannes Goecke authored
On the MSI-K8T-NEO2 FIR ( Athlon-64, Socket 939 with VIA-K8T800- Chipset and onboard Sound,... ) the BIOS lets you choose "DISABLED" or "AUTO" for the On-Board Sound Device. If you add another PCI-Sound-Card the BIOS disables the on-board device. So far I have a Quirk, that does set the correspondent BIT in the PCI-registers to enable the soundcard. But how to ensure that the code is executed ONLY on excactly this kind of boards (not any other with similar Chipset)? Cc: Jaroslav Kysela <perex@suse.cz> Acked-by:
-
Chris Zankel authored
Remove the dependence on the async_icount structure in the TIOCGICOUNT macro for Xtensa. (Thanks Russell and Adrian for pointing this out) Signed-off-by:
Chris Zankel <chris@zankel.net> Signed-off-by:
-
Thayumanavar Sachithanantham authored
[PATCH] cs5535_gpio.c: call cdev_del() during module_exit to unmap kobject references and other cleanups During module unloading, cdev_del() must be called to unmap cdev related kobject references and other cleanups(such as inode->i_cdev being set to NULL) which prevents the OOPS upon subsequent loading, usage and unloading of modules(as seen in the mail thread http://marc.theaimsgroup.com/?l=linux-kernel&m=114533640609018&w=2). Also, remove unneeded test of gpio_base. Signed-off-by:
Thayumanavar Sachithanantham <thayumk@gmail.com> Signed-off-by:
-
Darren Jenkins authored
WARNING: drivers/video/pm2fb.o - Section mismatch: reference to .init.data: from .text after 'pm2fb_set_par' (at offset 0xd5d) WARNING: drivers/video/pm2fb.o - Section mismatch: reference to .init.data: from .text after 'pm2fb_set_par' (at offset 0xd82) They are caused because pm2fb_set_par() uses lowhsync and lowvsync which are marked __devinitdata. Signed-off-by:
Darren Jenkins <darrenrjenkins@gmail.com> Signed-off-by:
-
Christoph Lameter authored
gather_stats() is called with a spinlock held from check_pte_range. We cannot reschedule with a lock held. Signed-off-by:
Christoph Lameter <clameter@sgi.com> Signed-off-by:
-
Corey Minyard authored
Add myself as the IPMI maintainer. Signed-off-by:
Corey Minyard <minyard@acm.org> Signed-off-by:
-
Ananth N Mavinakayanahalli authored
In cases where a struct kretprobe's *_handler fields are non-NULL, it is possible to cause a system crash, due to the possibility of calls ending up in zombie functions. Documentation clearly states that unused *_handlers should be set to NULL, but kprobe users sometimes fail to do so. Fix it by setting the non-relevant fields of the struct kretprobe to NULL. Signed-off-by:
Ananth N Mavinakayanahalli <ananth@in.ibm.com> Acked-by:
Jim Keniston <jkenisto@us.ibm.com> Signed-off-by:
-
Adrian Bunk authored
This patch removes the following unused EXPORT_SYMBOL's: - elv_requeue_request - elv_completed_request They are only used by the block core, hence they need not be exported. Signed-off-by:
Jens Axboe <axboe@suse.de>
-
Jens Axboe authored
Signed-off-by: -
Jens Axboe authored
It's really task private, so clear that field on fork after copying task structure. Signed-off-by: -
Coywolf Qi Hunt authored
This cleanup the source to use blk_queue_stopped. Signed-off-by:
Coywolf Qi Hunt <qiyong@freeforge.net> Signed-off-by:
-
Valdis Kletnieks authored
We added the ability to change a block device's IO elevator scheduler both at kernel boot and on-the-fly, but we only documented the elevator= boot parameter. Add a quick how-to on doing it on the fly. Signed-off-by:
Valdis Kletnieks <valdis.kletnieks@vt.edu> Signed-off-by:
-
Jayachandran C authored
Make all the vmalloc calls in net/bridge/netfilter/ebtables.c follow the standard convention. Remove unnecessary casts, and use '*object' instead of 'type'. Signed-off-by:
Jayachandran C. <c.jayachandran@gmail.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Add some sanity checking. truesize should be at least sizeof(struct sk_buff) plus the current packet length. If not, then truesize is seriously mangled and deserves a kernel log message. Currently we'll do the check for release of stream socket buffers. But we can add checks to more spots over time. Incorporating ideas from Herbert Xu. Signed-off-by: -
Herbert Xu authored
Make sure that we get the full sizeof(struct sk_buff) plus the data size accounted for in skb->truesize. This will create invariants that will allow adding assertion checks on skb->truesize. Signed-off-by:
Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by:
-
- 19 Apr, 2006 6 commits
-
-
OGAWA Hirofumi authored
Those also break userland regs like following. 00000000 <sys_chown16>: 0: 0f b7 44 24 0c movzwl 0xc(%esp),%eax 5: 83 ca ff or $0xffffffff,%edx 8: 0f b7 4c 24 08 movzwl 0x8(%esp),%ecx d: 66 83 f8 ff cmp $0xffffffff,%ax 11: 0f 44 c2 cmove %edx,%eax 14: 66 83 f9 ff cmp $0xffffffff,%cx 18: 0f 45 d1 cmovne %ecx,%edx 1b: 89 44 24 0c mov %eax,0xc(%esp) 1f: 89 54 24 08 mov %edx,0x8(%esp) 23: e9 fc ff ff ff jmp 24 <sys_chown16+0x24> where the tailcall at the end overwrites the incoming stack-frame. Signed-off-by:OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> [ I would _really_ like to have a way to tell gcc about calling conventions. The "prevent_tail_call()" macro is pretty ugly ] Signed-off-by:
-
Johannes Berg authored
In a short discussion with Benjamin Herrenschmidt he mentioned that Marvell PHYs are powered down the same way as the other ones we currently handle. Thus actually do that, hopefully saving some power during suspend. Signed-off-by:
Johannes Berg <johannes@sipsolutions.net> Acked-by:
Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by:
-
David S. Miller authored
Kernel Bugzilla #6409 If we use plain skb_trim(), that's wrong, because if the SKB is cloned, and it can be because we unshared it in the caller, we have to allow reallocation. The pskb_trim*() family of routines is therefore the most appropriate here. Signed-off-by: -
Hua Zhong authored
This applies to 2.6.17-rc2. There is a missing initialization of err in sockfd_lookup_light() that could return random error for an invalid file handle. Signed-off-by:
Hua Zhong <hzhong@gmail.com> Signed-off-by:
-
David S. Miller authored
Signed-off-by: -
Linus Torvalds authored
* 'for-linus' of master.kernel.org:/pub/scm/linux/kernel/git/roland/infiniband: IB/mthca: make a function static IB/ipath: Fix whitespace IB/ipath: Make more names static IB/mad: Fix RMPP version check during agent registration IB/srp: Remove request from list when SCSI abort succeeds
-
