Roman Zippel [Sat, 25 Mar 2006 11:08:28 +0000 (03:08 -0800)]
[PATCH] remove pps support
This removes the support for pps. It's completely unused within the kernel
and is basically in the way for further cleanups. It should be easier to
readd proper support for it after the rest has been converted to NTP4
(where the pps mechanisms are quite different from NTP3 anyway).
Mark A. Greer [Sat, 25 Mar 2006 11:08:27 +0000 (03:08 -0800)]
[PATCH] serial: mpsc driver has definition of SUPPORT_SYSRQ below include of serial_core.h
The definition of SUPPORT_SYSRQ must come before #include of serial_core.h.
This patch moves the definition of SUPPORT_SYSRQ to be just after the #include
of config.h to make it consistent with 8250.c.
Mark A. Greer [Sat, 25 Mar 2006 11:08:26 +0000 (03:08 -0800)]
[PATCH] serial: mpsc driver passes bad devname to request_irq()
The devname passed to request_irq() contained a '/' which is wrong. At
a minimum, the '/' prevented the devname from showing up in
/proc/irq/<irq>/<devname>. This patch replaces the '/' with a '-' to
fixes that problem.
Chen, Kenneth W [Sat, 25 Mar 2006 11:08:16 +0000 (03:08 -0800)]
[PATCH] direct-io: bug fix in dio handling write error
There is a bug in direct-io on propagating write error up to the higher I/O
layer. When performing an async ODIRECT write to a block device, if a
device error occurred (like media error or disk is pulled), the error code
is only propagated from device driver to the DIO layer. The error code
stops at finished_one_bio(). The aysnc write, however, is supposedly have
a corresponding AIO event with appropriate return code (in this case -EIO).
Application which waits on the async write event, will hang forever since
such AIO event is lost forever (if such app did not use the timeout option
in io_getevents call. Regardless, an AIO event is lost).
The discovery of above bug leads to another discovery of potential race
window with dio->result. The fundamental problem is that dio->result is
overloaded with dual use: an indicator of fall back path for partial dio
write, and an error indicator used in the I/O completion path. In the
event of device error, the setting of -EIO to dio->result clashes with
value used to track partial write that activates the fall back path.
It was also pointed out that it is impossible to use dio->result to track
partial write and at the same time to track error returned from device
driver. Because direct_io_work can only determines whether it is a partial
write at the end of io submission and in mid stream of those io submission,
a return code could be coming back from the driver. Thus messing up all
the subsequent logic.
Proposed fix is to separating out error code returned by the IO completion
path from partial IO submit tracking. A new variable is added to dio
structure specifically to track io error returned in the completion path.
Phillip Susi [Sat, 25 Mar 2006 11:08:14 +0000 (03:08 -0800)]
[PATCH] udf: fix uid/gid options and add uid/gid=ignore and forget options
As Pekka Enberg pointed out, with the if still following the else, you can
still get a null uid written to the disk if you specify a default uid= without
uid=forget. In other words, if the desktop user is uid=1000 and the mount
option uid=1000 is given ( which is done on ubuntu automatically and probably
other distributions that use hal ), then if any other user besides uid 1000
owns a file then a 0 will be written to the media as the owning uid instead.
Jesper Juhl [Sat, 25 Mar 2006 11:08:12 +0000 (03:08 -0800)]
[PATCH] Fix memory leak in isapnp
Spotted by the Coverity checker as bug #666
akpm; there are several other `return 1;'s in there which aren't freeing
`dev'. (A fix which converts this function to single-exit would be
preferred..)
Jonathan Corbet [Sat, 25 Mar 2006 11:08:05 +0000 (03:08 -0800)]
[PATCH] radix-tree documentation cleanups
Documentation changes to help radix tree users avoid overrunning the tags
array. RADIX_TREE_TAGS moves to linux/radix-tree.h and is now known as
RADIX_TREE_MAX_TAGS (Nick Piggin's idea). Tag parameters are changed to
unsigned, and some comments are updated.
Correct some error handling on the compat version of the nfsservctl()
system. It was detecting errors while copying in the arguments from user
space, but then attempting to use the arguments anyway. This didn't seem
so good.
John Hawkes [Sat, 25 Mar 2006 11:08:02 +0000 (03:08 -0800)]
[PATCH] fix alloc_large_system_hash() roundup
The "rounded up to nearest power of 2 in size" algorithm in
alloc_large_system_hash is not correct. As coded, it takes an otherwise
acceptable power-of-2 value and doubles it. For example, we see the error
if we boot with thash_entries=2097152 which produces a hash table with 4194304 entries.
Rob Landley [Sat, 25 Mar 2006 11:07:58 +0000 (03:07 -0800)]
[PATCH] Ext2 flags shouldn't report "nogrpid"
If I mount ext2 "rw", I want it to say "rw", not "rw,nogrpid".
I caught this writing an automated regression test script for the busybox
mount command. The symptom is
/dev/loop0 on /images/ext2.dir type ext2 (rw,nogrpid)
instead of:
/dev/loop0 on /images/ext2.dir type ext2 (rw)
Dave Jones [Sat, 25 Mar 2006 11:07:53 +0000 (03:07 -0800)]
[PATCH] remove dead address from maintainers list.
Mailing this address gives.. Sorry your message to [email protected] cannot
be delivered. This account has been disabled or discontinued [#102].
mta129.mail.re4.yahoo.com)
Zdenek Pavlas [Sat, 25 Mar 2006 11:07:49 +0000 (03:07 -0800)]
[PATCH] BLK_DEV_INITRD: do not require BLK_DEV_RAM=y
Initramfs initrd images do not need a ramdisk device, so remove this
restriction in Kconfig. BLK_DEV_RAM=n saves about 13k on i386. Also
without ramdisk device there's no need for "dry run", so initramfs unpacks
much faster.
People using cramfs, squashfs, or gzipped ext2/minix initrd images are
probably smart enough not to turn off ramdisk support by accident.
Andrew Morton [Sat, 25 Mar 2006 11:07:48 +0000 (03:07 -0800)]
[PATCH] find_task_by_pid() needs tasklist_lock
A couple of places are forgetting to take it.
The kswapd case is probably unimportant. keventd_create_kthread() was racy.
The whole thing is a bit flakey: you start a kernel thread, get its pid from
kernel_thread() then look up its task_struct.
a) It assumes that pid recycling takes a "long" time.
b) We get a task_struct but no reference was taken on it. The owner of the
kswapd and kthread task_struct*'s must assume that the new thread won't
exit unexpectedly. Because if it does, they're left holding dead memory
and any attempt to control or stop that task will crash.
Kirill Korotaev [Sat, 25 Mar 2006 11:07:45 +0000 (03:07 -0800)]
[PATCH] Reduce sched latency in shrink_dcache_sb()
This patch reduces scheduling latency in shrink_dcache_sb() noticed during
remounting of big partitions with many cached dentries. The same latency
fix was applied to select_parent() long ago.
Chris Wright [Sat, 25 Mar 2006 11:07:41 +0000 (03:07 -0800)]
[PATCH] make cap_ptrace enforce PTRACE_TRACME checks
PTRACE_TRACEME doesn't have proper capabilities validation when parent is
less privileged than child. Issue pointed out by Ram Gupta
<[email protected]>.
Note: I haven't identified a strong security issue, and it's a small ABI
change that could break apps that rely on existing behaviour (which allows
parent that is less privileged than child to ptrace when child does
PTRACE_TRACEME).
Chris Wright [Sat, 25 Mar 2006 11:07:41 +0000 (03:07 -0800)]
[PATCH] refactor capable() to one implementation, add __capable() helper
Move capable() to kernel/capability.c and eliminate duplicate
implementations. Add __capable() function which can be used to check for
capabiilty of any process.
Davide Libenzi [Sat, 25 Mar 2006 11:07:39 +0000 (03:07 -0800)]
[PATCH] POLLRDHUP/EPOLLRDHUP handling for half-closed devices notifications
Implement the half-closed devices notifiation, by adding a new POLLRDHUP
(and its alias EPOLLRDHUP) bit to the existing poll/select sets. Since the
existing POLLHUP handling, that does not report correctly half-closed
devices, was feared to be changed, this implementation leaves the current
POLLHUP reporting unchanged and simply add a new bit that is set in the few
places where it makes sense. The same thing was discussed and conceptually
agreed quite some time ago:
http://lkml.org/lkml/2003/7/12/116
Since this new event bit is added to the existing Linux poll infrastruture,
even the existing poll/select system calls will be able to use it. As far
as the existing POLLHUP handling, the patch leaves it as is. The
pollrdhup-2.6.16.rc5-0.10.diff defines the POLLRDHUP for all the existing
archs and sets the bit in the six relevant files. The other attached diff
is the simple change required to sys/epoll.h to add the EPOLLRDHUP
definition.
There is "a stupid program" to test POLLRDHUP delivery here:
http://www.xmailserver.org/pollrdhup-test.c
It tests poll(2), but since the delivery is same epoll(2) will work equally.
Bryan Holty [Sat, 25 Mar 2006 11:07:37 +0000 (03:07 -0800)]
[PATCH] IRQ: prevent enabling of previously disabled interrupt
This fix prevents re-disabling and enabling of a previously disabled
interrupt. On an SMP system with irq balancing enabled; If an interrupt is
disabled from within its own interrupt context with disable_irq_nosync and is
also earmarked for processor migration, the interrupt is blindly moved to the
other processor and enabled without regard for its current "enabled" state.
If there is an interrupt pending, it will unexpectedly invoke the irq handler
on the new irq owning processor (even though the irq was previously disabled)
The more intuitive fix would be to invoke disable_irq_nosync and
enable_irq, but since we already have the desc->lock from __do_IRQ, we
cannot call them directly. Instead we can use the same logic to disable
and enable found in disable_irq_nosync and enable_irq, with regards to the
desc->depth.
This now prevents a disabled interrupt from being re-disabled, and more
importantly prevents a disabled interrupt from being incorrectly enabled on
a different processor.
Andrew Morton [Sat, 25 Mar 2006 11:07:35 +0000 (03:07 -0800)]
[PATCH] ads7846 build fix
sparc32:
drivers/input/touchscreen/ads7846.c: In function `ads7846_read12_ser':
drivers/input/touchscreen/ads7846.c:206: warning: implicit declaration of function `disable_irq'
drivers/input/touchscreen/ads7846.c:208: warning: implicit declaration of function `enable_irq'
Andrew Morton [Sat, 25 Mar 2006 11:07:35 +0000 (03:07 -0800)]
[PATCH] tpm: sparc32 build fix
In file included from drivers/char/tpm/tpm_nsc.c:23:
drivers/char/tpm/tpm.h: In function `tpm_read_index':
drivers/char/tpm/tpm.h:92: warning: implicit declaration of function `outb'
drivers/char/tpm/tpm.h:93: warning: implicit declaration of function `inb'
The /dev/mem and /dev/kmem write handlers weren't fully POSIX compliant in
that they wouldn't always force the file pointer to be updated when
returning success status.
The /dev/port write handler was inconsistent with the /dev/mem and
/dev/kmem handlers in that when encountering a -EFAULT condition after
already having written a number of items it would return -EFAULT rather
than the number of bytes written.
There were a number of conflicting naming schemes used in the v9fs project.
The directory was fs/9p, but MAINTAINERS and Documentation referred to
v9fs. The module name itself was 9p2000, and the file system type was 9P.
This patch attempts to clean that up, changing all references to 9p in
order to match the directory name. We'll also start using 9p instead of
v9fs as our patch prefix.
There is also a minor consistency cleanup in the options changing the name
option to uname in order to more closely match the Plan 9 options.
Update license boilerplate to specify GPLv2 and remove the (at your option
clause). This change was agreed to by all the copyright holders (approvals
can be found on v9fs-developer mailing list).
Remove the assumption that driver_register() returns the number of devices
bound to the driver. In fact, it returns zero for success or a negative
error value.
zorro_module_init() used the device count to automatically unregister and
unload drivers that found no devices. That might have worked at one time,
but has been broken for some time because zorro_register_driver() returned
either a negative error or a positive count (never zero). So it could only
unregister on failure, when it's not needed anyway.
This functionality could be resurrected in individual drivers by counting
devices in their .probe() methods.
Bjorn Helgaas [Sat, 25 Mar 2006 11:07:19 +0000 (03:07 -0800)]
[PATCH] EISA: tidy-up driver_register() return value
Remove the assumption that driver_register() returns the number of devices
bound to the driver. In fact, it returns zero for success or a negative
error value.
Remove the assumption that driver_register() returns the number of devices
bound to the driver. In fact, it returns zero for success or a negative
error value.
dio_module_init() used the device count to automatically unregister and
unload drivers that found no devices. That might have worked at one time,
but has been broken for some time because dio_register_driver() returned
either a negative error or a positive count (never zero). So it could only
unregister on failure, when it's not needed anyway.
This functionality could be resurrected in individual drivers by counting
devices in their .probe() methods.
Andrew Morton [Sat, 25 Mar 2006 11:07:15 +0000 (03:07 -0800)]
[PATCH] initcall failure reporting
We presently ignore the return values from initcalls. But that can carry
useful debugging information. So print it out if it's non-zero.
It turns out the -ENODEV happens quite a lot, due to built-in drivers which
have no hardware to drive. So suppress that unless initcall_debug was
specified.
Also make the warning message more friendly by printing the name of the
initcall function.
Also drop the KERN_DEBUG from the initcall_debug message. If we specified
inticall_debug then we obviously want to see the messages.
Rene Herman [Sat, 25 Mar 2006 11:07:13 +0000 (03:07 -0800)]
[PATCH] parport: move PP_MAJOR from ppdev.h to major.h
Today I wondered about /dev/parport<n> after not seeing anything in
drivers/parport register char-major-99. Having PP_MAJOR in
include/linux/major.h would've allowed me to more quickly determine that it
was the ppdev driver driving these.
In latest -mm a number of section mismatch warnings are generated for
floppy.o like the following:
WARNING: drivers/block/floppy.o - Section mismatch: reference to \
.init.data: from .text between 'init_module' (at offset 0x6976) and \
'cleanup_module'
The warning are caused by a reference to floppy_init() which is __init from
init_module() which is not declared __init. Declaring init_module() _init
fixes this.
Sam Ravnborg [Sat, 25 Mar 2006 11:07:11 +0000 (03:07 -0800)]
[PATCH] ide: fix section mismatch warning
In latest -mm ide-code.o gave a number of warnings like the following:
WARNING: drivers/ide/ide-core.o - Section mismatch: reference to \
.init.text: from .text between 'init_module' (at offset 0x1f97) and \
'cleanup_module'
The warning was caused by init_module() calling parse_option() and
ide_init() both declared __init.
Declaring init_module() __init fixes the warnings.
Nick Piggin [Sat, 25 Mar 2006 11:07:09 +0000 (03:07 -0800)]
[PATCH] inotify: lock avoidance with parent watch status in dentry
Previous inotify work avoidance is good when inotify is completely unused,
but it breaks down if even a single watch is in place anywhere in the
system. Robin Holt notices that udev is one such culprit - it slows down a
512-thread application on a 512 CPU system from 6 seconds to 22 minutes.
Solve this by adding a flag in the dentry that tells inotify whether or not
its parent inode has a watch on it. Event queueing to parent will skip
taking locks if this flag is cleared. Setting and clearing of this flag on
all child dentries versus event delivery: this is no in terms of race
cases, and that was shown to be equivalent to always performing the check.
The essential behaviour is that activity occuring _after_ a watch has been
added and _before_ it has been removed, will generate events.
Rusty Russell [Sat, 25 Mar 2006 11:07:05 +0000 (03:07 -0800)]
[PATCH] Remove MODULE_PARM
MODULE_PARM was actually breaking: recent gcc version optimize them out as
unused. It's time to replace the last users, which are generally in the
most unloved drivers anyway.
Andrew Morton [Sat, 25 Mar 2006 11:07:04 +0000 (03:07 -0800)]
[PATCH] hysdn: remove custom types
hysdn defines its own types: ulong, uint, uchar and word.
Problem is, the module_param macros rely upon some of those identifiers having
special meanings too. The net effect is that module_param() and friends
cannot be used in ISDN because of this namespace clash.
So remove the hysdn-private defines and open-code them all.
Oleg Drokin [Sat, 25 Mar 2006 11:07:02 +0000 (03:07 -0800)]
[PATCH] Add lookup_instantiate_filp usage warning
I think it would be nice to put an usage warning in header of
lookup_instantiate_filp() to indicate it is unsafe to use it on anything
but regular files (even that is potentially unsafe, but there your ->open()
is usually in your hands anyway), so that others won't fall into the same
trap I did.
Oleg Drokin [Sat, 25 Mar 2006 11:07:01 +0000 (03:07 -0800)]
[PATCH] Introduce FMODE_EXEC file flag
Introduce FMODE_EXEC file flag, to indicate that file is being opened for
execution. This is useful for distributed filesystems to maintain
consistent behavior for returning ETXTBUSY when opening for write and
execution happens on different nodes.
akpm:
Needed by Lustre at present. I assume their objective to to work towards
being able to install Lustre on an unmodified distro kernel, which seems
sane. It should have zero runtime cost.
Trond and Chuck indicate that NFS4 can probably use this too, for the same
thing.
Jeff Mahoney [Sat, 25 Mar 2006 11:07:00 +0000 (03:07 -0800)]
[PATCH] reiserfs: reiserfs_file_write() will lose error code when a 0-length write occurs w/ O_SYNC
When an error occurs in reiserfs_file_write before any data is written, and
O_SYNC is set, the return code of generic_osync_write will overwrite the
error code, losing it.
This patch ensures that generic_osync_inode() doesn't run under an error
condition, losing the error. This duplicates the logic from
generic_file_buffered_write().
Reiserfs does not handle transaction ID overflow correctly. Transaction ID
== 0 causes reiserfs to crash. The patch fixes all places where the
transaction ID is incremented.
This patch fixes a bug in reiserfs truncate. A transaction might overflow
when truncating long highly fragmented file. The fix is to split
truncation into several transactions to avoid overflowing.
projects / linux.git / log