Hang during boot, bisected to commit 96d60303fd
Since somewhere early in the 2.6.35 merge window I've been seeing a hang during boot. The hang last roughly 30 seconds and is followed by this: ata4.00: exception Emask 0x0 SAct 0x1 SErr 0x0 action 0x6 frozen ata4.00: failed command: READ FPDMA QUEUED ata4.00: cmd 60/08:00:00:00:00/00:00:00:00:00/40 tag 0 ncq 4... 7 Jun 2010 09:38
task_io_accounting, taskstats
* Rafael Tinoco <tinhocas(a)gmail.com> [2010-06-02 18:56:43]: Hello, I'm proposing this patch to extend taskstats capability of IO_ACCOUTING based on socket msg size. Already discussed with Balbir about it. The idea was to keep the "socket" accounting generic. Not taking in consideration witch typ... 3 Jun 2010 21:15
linux-next: build warning after merge of the rr tree
Hi Rusty, After merging the rr tree, today's linux-next build (powerpc ppc64_defconfig) produced this warning: kernel/module.c: In function 'load_module': kernel/module.c:2407: warning: 'hdr' may be used uninitialized in this function Probably introduced by commit 601b979485d01750ea17cf138e209d81acfe63b5 ("... 2 Jun 2010 22:00
[PATCH] mmc: make sdhci work with ricoh mmc controller
The current way of disabling it is not well tested by vendor and has all kinds of bugs that show up on resume from ram/disk. Old way of disabling is still supported by continuing to use CONFIG_MMC_RICOH_MMC. Based on 'http://list.drzeus.cx/pipermail/sdhci-devel/2007-December/002085.html' Signed-off-by: Maxim... 2 Jun 2010 22:00
x86, i8259: Only register sysdev for real PIC
On Wed, 2 Jun 2010 13:31:50 +0200 Adam Lackorzynski <adam(a)os.inf.tu-dresden.de> wrote: Do not register the sysdev function when the null_legacy_pic is used so that the i8259 resume, suspend and shutdown functions are not called. What are the effects of this change? Does it fix an oops or a hang, or nothing... 3 Jun 2010 13:24
BKL: autoconvert trivial users to private mutex
On Thu, Jun 03, 2010 at 02:13:15AM +0200, Arnd Bergmann wrote: All these files use the big kernel lock in a trivial way to serialize their private file operations, typically resulting from an earlier semi-automatic pushdown from VFS. None of these drivers appears to want to lock against other code... 3 Jun 2010 13:24
[RFC 5/5] BKL: introduce CONFIG_BKL.
From: Arnd Bergmann <arnd(a)relay.de.ibm.com> To make it possible to build a kernel without the big kernel lock, we need to make sure we don't build any code that still uses it, so disable the options in Kconfig. Most of the v4l, drm and oss drivers still need it too, but their respective top-level interfaces ... 2 Jun 2010 20:55
[RFC 4/5] BKL: use no BKL in llseek
We have shown that the BKL in default_llseek and other llseek operations never protects against concurrent access from another function: * Most drivers never access f_pos * Most drivers that access f_pos have their own llseek function * Those drivers that use default_llseek and still access f_pos from outs... 2 Jun 2010 20:55
[RFC 3/5] BKL: do not take BKL in do_coredump
core_pattern is not actually protected and hasn't been ever since we introduced procfs support for sysctl -- a _long_ time. Don't take it here either. Signed-off-by: Arnd Bergmann <arnd(a)arndb.de> --- fs/exec.c | 6 ------ 1 files changed, 0 insertions(+), 6 deletions(-) diff --git a/fs/exec.c b/fs/exec.c... 2 Jun 2010 20:55
[RFC 2/5] BKL: remove the BKL from kernel init code
I have shown by code review that no driver takes the BKL at init time any more, so whatever the init code was locking against is no longer there and it is now safe to remove the BKL there. Signed-off-by: Arnd Bergmann <arnd(a)arndb.de> --- init/main.c | 5 ----- kernel/trace/trace.c | 8 -------- ... 2 Jun 2010 20:55