linux-2.6
18 years ago[PATCH] Move pidmap to pspace.h
Sukadev Bhattiprolu [Mon, 2 Oct 2006 09:17:23 +0000 (02:17 -0700)] 
[PATCH] Move pidmap to pspace.h

Move struct pidmap and PIDMAP_ENTRIES to a new file, include/linux/pspace.h
where it will be used in subsequent patches to define pid spaces.

Its a subset of Eric Biederman's patch http://lkml.org/lkml/2006/2/6/285

[akpm@osdl.org: cleanups]
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@us.ibm.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Serge Hallyn <serue@us.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] pid: simplify pid iterators
Oleg Nesterov [Mon, 2 Oct 2006 09:17:22 +0000 (02:17 -0700)] 
[PATCH] pid: simplify pid iterators

I think it is hardly possible to read the current do_each_task_pid().  The
new version is much simpler and makes the code smaller.

Only the do_each_task_pid change is tested, the do_each_pid_task isn't.

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] pids coding style use struct pidmap in next_pidmap
Eric W. Biederman [Mon, 2 Oct 2006 09:17:21 +0000 (02:17 -0700)] 
[PATCH] pids coding style use struct pidmap in next_pidmap

Use struct pidmap instead of pidmap_t.

This updates my proc: readdir race fix (take 3) patch
to account for the changes made by: Sukadev Bhattiprolu <sukadev@us.ibm.com>
to kill pidmap_t.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] pids: coding style: use struct pidmap
Sukadev Bhattiprolu [Mon, 2 Oct 2006 09:17:20 +0000 (02:17 -0700)] 
[PATCH] pids: coding style: use struct pidmap

Use struct pidmap instead of pidmap_t.

Its a subset of Eric Biederman's patch http://lkml.org/lkml/2006/2/6/271.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@us.ibm.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Serge Hallyn <serue@us.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] const struct tty_operations
Jeff Dike [Mon, 2 Oct 2006 09:17:18 +0000 (02:17 -0700)] 
[PATCH] const struct tty_operations

As part of an SMP cleanliness pass over UML, I consted a bunch of
structures in order to not have to document their locking.  One of these
structures was a struct tty_operations.  In order to const it in UML
without introducing compiler complaints, the declaration of
tty_set_operations needs to be changed, and then all of its callers need to
be fixed.

This patch declares all struct tty_operations in the tree as const.  In all
cases, they are static and used only as input to tty_set_operations.  As an
extra check, I ran an i386 allyesconfig build which produced no extra
warnings.

53 drivers are affected.  I checked the history of a bunch of them, and in
most cases, there have been only a handful of maintenance changes in the
last six months.  serial_core.c was the busiest one that I looked at.

Signed-off-by: Jeff Dike <jdike@addtoit.com>
Acked-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] fs/inode.c tweaks
Andreas Mohr [Mon, 2 Oct 2006 09:17:17 +0000 (02:17 -0700)] 
[PATCH] fs/inode.c tweaks

Only touch inode's i_mtime and i_ctime to make them equal to "now" in case
they aren't yet (don't just update timestamp unconditionally).  Uninline
the hash function to save 259 Bytes.

This tiny inode change which may improve cache behaviour also shaves off 8
Bytes from file_update_time() on i386.

Included a tiny codestyle cleanup, too.

Signed-off-by: Andreas Mohr <andi@lisas.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Remove NULL check in register_nls()
Alexey Dobriyan [Mon, 2 Oct 2006 09:17:16 +0000 (02:17 -0700)] 
[PATCH] Remove NULL check in register_nls()

Everybody passes valid pointer there.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Acked-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] file: modify struct fown_struct to use a struct pid
Eric W. Biederman [Mon, 2 Oct 2006 09:17:15 +0000 (02:17 -0700)] 
[PATCH] file: modify struct fown_struct to use a struct pid

File handles can be requested to send sigio and sigurg to processes.  By
tracking the destination processes using struct pid instead of pid_t we make
the interface safe from all potential pid wrap around problems.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] vt: Make vt_pid a struct pid (making it pid wrap around safe).
Eric W. Biederman [Mon, 2 Oct 2006 09:17:14 +0000 (02:17 -0700)] 
[PATCH] vt: Make vt_pid a struct pid (making it pid wrap around safe).

I took a good hard look at the locking and it appears the locking on vt_pid
is the console semaphore.  Every modified path is called under the console
semaphore except reset_vc when it is called from fn_SAK or do_SAK both of
which appear to be in interrupt context.  In addition I need to be careful
because in the presence of an oops the console_sem may be arbitrarily
dropped.

Which leads me to conclude the current locking is inadequate for my needs.

Given the weird cases we could hit because of oops printing instead of
introducing an extra spin lock to protect the data and keep the pid to
signal and the signal to send in sync, I have opted to use xchg on just the
struct pid * pointer instead.

Due to console_sem we will stay in sync between vt_pid and vt_mode except
for a small window during a SAK, or oops handling.  SAK handling should
kill any user space process that care, and oops handling we are broken
anyway.  Besides the worst that can happen is that I try to send the wrong
signal.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] vt: rework the console spawning variables
Eric W. Biederman [Mon, 2 Oct 2006 09:17:13 +0000 (02:17 -0700)] 
[PATCH] vt: rework the console spawning variables

This is such a rare path it took me a while to figure out how to test
this after soring out the locking.

This patch does several things.
- The variables used are moved into a structure and declared in vt_kern.h
- A spinlock is added so we don't have SMP races updating the values.
- Instead of raw pid_t value a struct_pid is used to guard against
  pid wrap around issues, if the daemon to spawn a new console dies.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] pid: implement pid_nr
Eric W. Biederman [Mon, 2 Oct 2006 09:17:12 +0000 (02:17 -0700)] 
[PATCH] pid: implement pid_nr

As we stop storing pid_t's and move to storing struct pid *.  We need a way to
get the pid_t from the struct pid to report to user space what we have stored.

Having a clean well defined way to do this is especially important as we move
to multiple pid spaces as may need to report a different value to the caller
depending on which pid space the caller is in.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] pid: export the symbols needed to use struct pid *
Eric W. Biederman [Mon, 2 Oct 2006 09:17:11 +0000 (02:17 -0700)] 
[PATCH] pid: export the symbols needed to use struct pid *

pids aren't something that drivers should care about.  However there are a lot
of helper layers in the kernel that do care, and are built as modules.  Before
I can convert them to using struct pid instead of pid_t I need to export the
appropriate symbols so they can continue to be built.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] pid: implement signal functions that take a struct pid *
Eric W. Biederman [Mon, 2 Oct 2006 09:17:10 +0000 (02:17 -0700)] 
[PATCH] pid: implement signal functions that take a struct pid *

Currently the signal functions all either take a task or a pid_t argument.
This patch implements variants that take a struct pid *.  After all of the
users have been update it is my intention to remove the variants that take a
pid_t as using pid_t can be more work (an extra hash table lookup) and
difficult to get right in the presence of multiple pid namespaces.

There are two kinds of functions introduced in this patch.  The are the
general use functions kill_pgrp and kill_pid which take a priv argument that
is ultimately used to create the appropriate siginfo information, Then there
are _kill_pgrp_info, kill_pgrp_info, kill_pid_info the internal implementation
helpers that take an explicit siginfo.

The distinction is made because filling out an explcit siginfo is tricky, and
will be even more tricky when pid namespaces are introduced.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] pid: add do_each_pid_task
Eric W. Biederman [Mon, 2 Oct 2006 09:17:09 +0000 (02:17 -0700)] 
[PATCH] pid: add do_each_pid_task

To avoid pid rollover confusion the kernel needs to work with struct pid *
instead of pid_t.  Currently there is not an iterator that walks through all
of the tasks of a given pid type starting with a struct pid.  This prevents us
replacing some pid_t instances with struct pid.  So this patch adds
do_each_pid_task which walks through the set of task for a given pid type
starting with a struct pid.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] pid: implement access helpers for a tacks various process groups
Eric W. Biederman [Mon, 2 Oct 2006 09:17:09 +0000 (02:17 -0700)] 
[PATCH] pid: implement access helpers for a tacks various process groups

In the last round of cleaning up the pid hash table a more general struct pid
was introduced, that can be referenced counted.

With the more general struct pid most if not all places where we store a pid_t
we can now store a struct pid * and remove the need for a hash table lookup,
and avoid any possible problems with pid roll over.

Looking forward to the pid namespaces struct pid * gives us an absolute form a
pid so we can compare and use them without caring which pid namespace we are
in.

This patchset introduces the infrastructure needed to use struct pid instead
of pid_t, and then it goes on to convert two different kernel users that
currently store a pid_t value.

There are a lot more places to go but this is enough to get the basic idea.

Before we can merge a pid namespace patch all of the kernel pid_t users need
to be examined.  Those that deal with user space processes need to be
converted to using a struct pid *.  Those that deal with kernel processes need
to converted to using the kthread api.  A rare few that only use their current
processes pid values get to be left alone.

This patch:

task_session returns the struct pid of a tasks session.
task_pgrp    returns the struct pid of a tasks process group.
task_tgid    returns the struct pid of a tasks thread group.
task_pid     returns the struct pid of a tasks process id.

These can be used to avoid unnecessary hash table lookups, and to implement
safe pid comparisions in the face of a pid namespace.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] proc: give the root directory a task
Eric W. Biederman [Mon, 2 Oct 2006 09:17:07 +0000 (02:17 -0700)] 
[PATCH] proc: give the root directory a task

Helper functions in base.c like proc_pident_readdir and proc_pident_lookup
assume the directories have an associated task, and cannot currently be used
on the /proc root directory because it does not have such a task.

This small changes allows for base.c to be simplified and later when multiple
pid spaces are introduced it makes getting the needed context information
trivial.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] proc: modify proc_pident_lookup to be completely table driven
Eric W. Biederman [Mon, 2 Oct 2006 09:17:07 +0000 (02:17 -0700)] 
[PATCH] proc: modify proc_pident_lookup to be completely table driven

Currently proc_pident_lookup gets the names and types from a table and then
has a huge switch statement to get the inode and file operations it needs.
That is silly and is becoming increasingly hard to maintain so I just put all
of the information in the table.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] proc: reorder the functions in base.c
Eric W. Biederman [Mon, 2 Oct 2006 09:17:05 +0000 (02:17 -0700)] 
[PATCH] proc: reorder the functions in base.c

There were enough changes in my last round of cleaning up proc I had to break
up the patch series into smaller chunks, and my last chunk never got resent.

This patchset gives proc dynamic inode numbers (the static inode numbers were
a pain to maintain and prevent all kinds of things), and removes the horrible
switch statements that had to be kept in sync with everything else.  Being
fully table driver takes us 90% of the way of being able to register new
process specific attributes in proc.

This patch:

Group the functions by what they implement instead of by type of operation.
As it existed base.c was quickly approaching the point where it could not be
followed.

No functionality or code changes asside from adding/removing forward
declartions are implemented in this patch.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] proc: readdir race fix (take 3)
Eric W. Biederman [Mon, 2 Oct 2006 09:17:04 +0000 (02:17 -0700)] 
[PATCH] proc: readdir race fix (take 3)

The problem: An opendir, readdir, closedir sequence can fail to report
process ids that are continually in use throughout the sequence of system
calls.  For this race to trigger the process that proc_pid_readdir stops at
must exit before readdir is called again.

This can cause ps to fail to report processes, and it is in violation of
posix guarantees and normal application expectations with respect to
readdir.

Currently there is no way to work around this problem in user space short
of providing a gargantuan buffer to user space so the directory read all
happens in on system call.

This patch implements the normal directory semantics for proc, that
guarantee that a directory entry that is neither created nor destroyed
while reading the directory entry will be returned.  For directory that are
either created or destroyed during the readdir you may or may not see them.
 Furthermore you may seek to a directory offset you have previously seen.

These are the guarantee that ext[23] provides and that posix requires, and
more importantly that user space expects.  Plus it is a simple semantic to
implement reliable service.  It is just a matter of calling readdir a
second time if you are wondering if something new has show up.

These better semantics are implemented by scanning through the pids in
numerical order and by making the file offset a pid plus a fixed offset.

The pid scan happens on the pid bitmap, which when you look at it is
remarkably efficient for a brute force algorithm.  Given that a typical
cache line is 64 bytes and thus covers space for 64*8 == 200 pids.  There
are only 40 cache lines for the entire 32K pid space.  A typical system
will have 100 pids or more so this is actually fewer cache lines we have to
look at to scan a linked list, and the worst case of having to scan the
entire pid bitmap is pretty reasonable.

If we need something more efficient we can go to a more efficient data
structure for indexing the pids, but for now what we have should be
sufficient.

In addition this takes no additional locks and is actually less code than
what we are doing now.

Also another very subtle bug in this area has been fixed.  It is possible
to catch a task in the middle of de_thread where a thread is assuming the
thread of it's thread group leader.  This patch carefully handles that case
so if we hit it we don't fail to return the pid, that is undergoing the
de_thread dance.

Thanks to KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> for
providing the first fix, pointing this out and working on it.

[oleg@tv-sign.ru: fix it]
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] list module taint flags in Oops/panic
Randy Dunlap [Mon, 2 Oct 2006 09:17:02 +0000 (02:17 -0700)] 
[PATCH] list module taint flags in Oops/panic

When listing loaded modules during an oops or panic, also list each
module's Tainted flags if non-zero (P: Proprietary or F: Forced load only).

If a module is did not taint the kernel, it is just listed like
usbcore
but if it did taint the kernel, it is listed like
wizmodem(PF)

Example:
[ 3260.121718] Unable to handle kernel NULL pointer dereference at 0000000000000000 RIP:
[ 3260.121729]  [<ffffffff8804c099>] :dump_test:proc_dump_test+0x99/0xc8
[ 3260.121742] PGD fe8d067 PUD 264a6067 PMD 0
[ 3260.121748] Oops: 0002 [1] SMP
[ 3260.121753] CPU 1
[ 3260.121756] Modules linked in: dump_test(P) snd_pcm_oss snd_mixer_oss snd_seq snd_seq_device ide_cd generic ohci1394 snd_hda_intel snd_hda_codec snd_pcm snd_timer snd ieee1394 snd_page_alloc piix ide_core arcmsr aic79xx scsi_transport_spi usblp
[ 3260.121785] Pid: 5556, comm: bash Tainted: P      2.6.18-git10 #1

[Alternatively, I can look into listing tainted flags with 'lsmod',
but that won't help in oopsen/panics so much.]

[akpm@osdl.org: cleanup]
Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] make genpool allocator adhere to kernel-doc standards
Dean Nelson [Mon, 2 Oct 2006 09:17:01 +0000 (02:17 -0700)] 
[PATCH] make genpool allocator adhere to kernel-doc standards

The exported kernel interfaces of genpool allocator need to adhere to
the requirements of kernel-doc.

Signed-off-by: Dean Nelson <dcn@sgi.com>
Cc: Steve Wise <swise@opengridcomputing.com>
Acked-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] LIB: add gen_pool_destroy()
Steve Wise [Mon, 2 Oct 2006 09:17:00 +0000 (02:17 -0700)] 
[PATCH] LIB: add gen_pool_destroy()

Modules using the genpool allocator need to be able to destroy the data
structure when unloading.

Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Cc: Randy Dunlap <rdunlap@xenotime.net>
Cc: Dean Nelson <dcn@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years agoJFS: White space cleanup
Dave Kleikamp [Mon, 2 Oct 2006 14:55:27 +0000 (09:55 -0500)] 
JFS: White space cleanup

Removed trailing spaces & tabs, and spaces preceding tabs.
Also a couple very minor comment cleanups.

Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
(cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)

18 years ago[PATCH] JFS: return correct error when i-node allocation failed
Akinobu Mita [Thu, 14 Sep 2006 14:22:38 +0000 (09:22 -0500)] 
[PATCH] JFS: return correct error when i-node allocation failed

I have seen confusing behavior on JFS when I injected many intentional
slab allocation errors. The cp command failed with no disk space error
with enough disk space.

This patch makes:

- change the return value in case slab allocation failures happen
  from -ENOSPC to -ENOMEM

- ialloc() return error code so that the caller can know the reason
  of failures

Signed-off-by: Akinobu Mita <mita@miraclelinux.com>
Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
(cherry picked from 2b46f77976f798f3fe800809a1d0ed38763c71c8 commit)

18 years agoJFS: Remove shadow variable from fs/jfs/jfs_txnmgr.c:xtLog()
Tony Breeds [Mon, 11 Sep 2006 13:19:19 +0000 (08:19 -0500)] 
JFS: Remove shadow variable from fs/jfs/jfs_txnmgr.c:xtLog()

Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
Signed-off-by: Dave Kleikamp <shaggy@austin.ibm.com>
(cherry picked from bdc3d9e5af7d9c105be734dd7b5c3f1d9425a15a commit)

18 years ago[ARM] 3848/1: pxafb: Add option of fixing video modes and spitz QVGA mode support
Richard Purdie [Wed, 20 Sep 2006 21:54:21 +0000 (22:54 +0100)] 
[ARM] 3848/1: pxafb: Add option of fixing video modes and spitz QVGA mode support

Add the ability to have pxafb use only certain fixed video modes
(selected on a per platform basis). This is useful on production
hardware such as the Zaurus cxx00 models where the valid modes are
known in advance and any other modes could result in hardware damage.

Following this, add support for the cxx00 QVGA mode. Mode information
is passed to the lcd_power call to allowing the panel drivers to
configure the display hardware accordingly (corgi_lcd already contains
the functionality for the cxx00 panel).

This mirrors the setup already used by w100fb.

Signed-off-by: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[POWERPC] Fix rheap alignment problem
Li Yang [Fri, 29 Sep 2006 10:15:52 +0000 (18:15 +0800)] 
[POWERPC] Fix rheap alignment problem

Honor alignment parameter in the rheap allocator.  This is needed by
qe_lib.
Remove compile warning.

Signed-off-by: Pantelis Antoniou <pantelis@embeddedalley.com>
Signed-off-by: Li Yang <leoli@freescale.com>
Acked-by: Kumar Galak <galak@kernel.crashing.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[POWERPC] Use check_legacy_ioport() for ISAPnP
David Woodhouse [Wed, 27 Sep 2006 07:18:55 +0000 (08:18 +0100)] 
[POWERPC] Use check_legacy_ioport() for ISAPnP

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[POWERPC] Avoid NULL pointer in gpio1_interrupt
Olaf Hering [Tue, 26 Sep 2006 20:28:36 +0000 (22:28 +0200)] 
[POWERPC] Avoid NULL pointer in gpio1_interrupt

gpio1_interrupt() may dereference a NULL pointer if ioremap() fails.
But, maybe no gpio interrupt happens in the first place?

Signed-off-by: Olaf Hering <olaf@aepfle.de>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[MTD] Cleanup of 'ioremap balanced with iounmap for drivers/mtd subsystem'
Amol Lad [Mon, 2 Oct 2006 08:48:23 +0000 (09:48 +0100)] 
[MTD] Cleanup of 'ioremap balanced with iounmap for drivers/mtd subsystem'

Updated version of patch, in response to comments from Francois Romieu
<romieu@fr.zoreil.com>

Remove gratuitous casts from iounmap and initialisation of variables.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
18 years ago[MTD] fix nftl_write warning
Frederik Deweerdt [Mon, 2 Oct 2006 08:42:25 +0000 (09:42 +0100)] 
[MTD] fix nftl_write warning

Building 2.6.18-mm2 issues the following warning if CONFIG_NFTL_RW is not set:

  CC [M]  drivers/mtd/nftlcore.o
drivers/mtd/nftlcore.c:183: warning: 'nftl_write' defined but not used
The following patch only compiles nftl_write if CONFIG_NFTL_RW is set.

Signed-off-by: Frederik Deweerdt <frederik.deweerdt@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
18 years ago[POWERPC] Enable generic rtc hook for the MPC8349 mITX
Kim Phillips [Tue, 26 Sep 2006 22:46:51 +0000 (17:46 -0500)] 
[POWERPC] Enable generic rtc hook for the MPC8349 mITX

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years ago[POWERPC] Add powerpc get/set_rtc_time interface to new generic rtc class
Kim Phillips [Tue, 26 Sep 2006 22:46:37 +0000 (17:46 -0500)] 
[POWERPC] Add powerpc get/set_rtc_time interface to new generic rtc class

Add powerpc get/set_rtc_time interface to new generic rtc class. This
abstracts rtc chip specific code from the platform code for rtc-over-i2c
platforms.  Specific RTC chip support is now configured under
Device Drivers -> Real Time Clock. Setting time of day from the RTC
on startup is also configurable.

this time without the potentially platform breaking initcall.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
18 years agoInput: wistron - add support for Acer TravelMate 2424NWXCi
Ashutosh Naik [Mon, 2 Oct 2006 02:07:14 +0000 (22:07 -0400)] 
Input: wistron - add support for Acer TravelMate 2424NWXCi

The key mappings are the same as the older Acer TravelMate 240.

Signed-off-by: Ashutosh Naik <ashutosh.naik@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
18 years agoInput: wistron - fix setting up special buttons
Reiner Herrmann [Mon, 2 Oct 2006 01:58:51 +0000 (21:58 -0400)] 
Input: wistron - fix setting up special buttons

If either wifi or bluetooth button has been detected, the code
would break off the loop. But there are laptops that have both
types of buttons, so the loop has to continue checking.

Signed-off-by: Reiner Herrmann <reiner@reiner-h.de>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
18 years ago[MIPS] Remove unused galileo-boars header files
Yoichi Yuasa [Sun, 1 Oct 2006 10:47:08 +0000 (19:47 +0900)] 
[MIPS] Remove unused galileo-boars header files

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Rename SERIAL_PORT_DEFNS for EV64120
Yoichi Yuasa [Sun, 1 Oct 2006 10:43:27 +0000 (19:43 +0900)] 
[MIPS] Rename SERIAL_PORT_DEFNS for EV64120

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Add UART IRQ number for EV64120
Yoichi Yuasa [Sun, 1 Oct 2006 10:35:28 +0000 (19:35 +0900)] 
[MIPS] Add UART IRQ number for EV64120

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Remove excite_flash.c
Yoichi Yuasa [Fri, 29 Sep 2006 09:27:07 +0000 (18:27 +0900)] 
[MIPS] Remove excite_flash.c

excite_flashtest.c is unused.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Update i8259 resources.
Yoichi Yuasa [Fri, 29 Sep 2006 09:17:51 +0000 (18:17 +0900)] 
[MIPS] Update i8259 resources.

Updated i8259 resources to same as i386.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Make unwind_stack() can dig into interrupted context
Atsushi Nemoto [Fri, 29 Sep 2006 09:02:51 +0000 (18:02 +0900)] 
[MIPS] Make unwind_stack() can dig into interrupted context

If the PC was ret_from_irq or ret_from_exception, there will be no
more normal stackframe.  Instead of stopping the unwinding, use PC and
RA saved by an exception handler to continue unwinding into the
interrupted context.  This also simplifies the CONFIG_STACKTRACE code.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Stacktrace build-fix and improvement
Atsushi Nemoto [Thu, 28 Sep 2006 10:15:33 +0000 (19:15 +0900)] 
[MIPS] Stacktrace build-fix and improvement

Fix build error due to stacktrace API change.  Now save_stack_trace()
tries to save all kernel context, including interrupts and exception.
Also some asm code are changed a bit so that we can detect the end of
current context easily.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] QEMU: Add support for little endian mips
Aurelien Jarno [Wed, 27 Sep 2006 21:07:25 +0000 (23:07 +0200)] 
[MIPS] QEMU: Add support for little endian mips

This very small patch adds support for little endian on the virtual QEMU
mips platform. The status of this platform is the same as the big endian
one, ie it is possible to boot a system with init=/bin/sh.

Signed-off-by: Aurelien Jarno <aurelien@aurel32.net>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] Remove __flush_icache_page
Atsushi Nemoto [Fri, 1 Sep 2006 15:43:07 +0000 (00:43 +0900)] 
[MIPS] Remove __flush_icache_page

__flash_icache_page is unused, so kill it.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] lockdep: update defconfigs
Atsushi Nemoto [Tue, 26 Sep 2006 14:44:16 +0000 (23:44 +0900)] 
[MIPS] lockdep: update defconfigs

Add those lines to all defconfigs.

CONFIG_LOCKDEP_SUPPORT=y
CONFIG_STACKTRACE_SUPPORT=y
CONFIG_TRACE_IRQFLAGS_SUPPORT=y

This is a patch againt linux-mips.org git tree.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] lockdep: Add STACKTRACE_SUPPORT and enable LOCKDEP_SUPPORT
Atsushi Nemoto [Tue, 26 Sep 2006 14:44:01 +0000 (23:44 +0900)] 
[MIPS] lockdep: Add STACKTRACE_SUPPORT and enable LOCKDEP_SUPPORT

Implement stacktrace interface by using unwind_stack() and enable lockdep
support in Kconfig.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[MIPS] lockdep: fix TRACE_IRQFLAGS_SUPPORT
Atsushi Nemoto [Tue, 26 Sep 2006 14:43:40 +0000 (23:43 +0900)] 
[MIPS] lockdep: fix TRACE_IRQFLAGS_SUPPORT

In handle_sys and its variants, we must reload some registers which
might be clobbered by trace_hardirqs_on().
Also we must make sure trace_hardirqs_on() called in kernel level (not
exception level).

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[SERIAL] add PNP IDs for FPI based touchscreens
Matthew Garrett [Sun, 1 Oct 2006 18:27:38 +0000 (19:27 +0100)] 
[SERIAL] add PNP IDs for FPI based touchscreens

The Compaq TC1000 and Fujitsu Stylistic range of tablet machines use
touchscreens from FPI. These are implemented as serial interfaces,
generally exposed in the ACPIPNP information on the system. This patch
adds them to the 8250_pnp driver tables, avoiding the need to mess
around with setserial to set them up.

I haven't been able to confirm what FUJ02B5, FUJ02BA and FUJ02BB are.
FUJ02B1 refers to the controller for the system hotkeys. FUJ02BC appears
to be the last in the range - after this, they moved to Wacom-based
systems.

Signed-off-by: Matthew Garrett <mjg59@srcf.ucam.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years agopccard_store_cis: fix wrong error handling
Linus Torvalds [Sun, 1 Oct 2006 20:17:44 +0000 (13:17 -0700)] 
pccard_store_cis: fix wrong error handling

The test for the error from pcmcia_replace_cis() was incorrect, and
would always trigger (because if an error didn't happen, the "ret" value
would not be zero, it would be the passed-in count).

Reported and debugged by Fabrice Bellet <fabrice@bellet.info>

Rather than just fix the single broken test, make the code in question
use an understandable code-sequence instead, fixing the whole function
to be more readable.

Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[SERIAL] Magic SysRq SAK does nothing on serial consoles
Simon Tatham [Mon, 25 Sep 2006 23:51:28 +0000 (16:51 -0700)] 
[SERIAL] Magic SysRq SAK does nothing on serial consoles

Make sysrq-K work on serial console by passing in the tty.

Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] tickle NMI watchdog on serial output.
Dave Jones [Mon, 25 Sep 2006 23:51:26 +0000 (16:51 -0700)] 
[SERIAL] tickle NMI watchdog on serial output.

Serial is _slow_ sometimes. So slow, that the NMI watchdog kicks in.

NMI Watchdog detected LOCKUP on CPU2CPU 2
Modules linked in: loop usb_storage md5 ipv6 parport_pc lp parport autofs4 i2c_dev i2c_core rfcomm l2cap bluetooth sunrpc pcdPid: 3138, comm: gpm Not tainted 2.6.11-1.1290_FC4smp
RIP: 0010:[<ffffffff80273b8a>] <ffffffff80273b8a>{serial_in+106}
RSP: 0018:ffff81003afc3d50  EFLAGS: 00000002
RAX: 0000000000000020 RBX: 0000000000000020 RCX: 0000000000000000
RDX: 00000000000003fd RSI: 0000000000000005 RDI: ffffffff804dcd60
RBP: 00000000000024fc R08: 000000000000000a R09: 0000000000000033
R10: ffff81001beb7c20 R11: 0000000000000020 R12: ffffffff804dcd60
R13: ffffffff804ade76 R14: 000000000000002b R15: 000000000000002c
FS:  00002aaaaaac4920(0000) GS:ffffffff804fca00(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00002aaaaabcb000 CR3: 000000003c0d0000 CR4: 00000000000006e0
Process gpm (pid: 3138, threadinfo ffff81003afc2000, task ffff81003eb63780)
Stack: ffffffff80275f2e 0000000000000000 ffffffff80448380 0000000000007d6b
       000000000000002c fffffffffffffbbf 0000000000000292 0000000000008000
       ffffffff80138e8c 0000000000007d97
Call Trace:<ffffffff80275f2e>{serial8250_console_write+270} <ffffffff80138e8c>{__call_console_drivers+76}
       <ffffffff8013914b>{release_console_sem+315} <ffffffff80260325>{con_open+149}
       <ffffffff80254e99>{tty_open+537} <ffffffff80192713>{chrdev_open+387}
       <ffffffff80188824>{dentry_open+260} <ffffffff80188994>{filp_open+68}
       <ffffffff80187b73>{get_unused_fd+227} <ffffffff80188a6c>{sys_open+76}
       <ffffffff8010ebc6>{tracesys+209}

Code: 0f b6 c0 c3 66 90 41 57 49 89 f7 41 56 41 be 00 01 00 00 41
console shuts up ...

I initially did the patch below a year ago for the Fedora kernel, and have
been keeping it up to date since.  I recently got the same thing happening
on a vanilla kernel, so figured it was time to repost this.

Signed-off-by: Dave Jones <davej@redhat.com>
Acked-by: Alan Cox <alan@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[PATCH] rtc-sysfs fix
Andrew Morton [Sun, 1 Oct 2006 09:22:41 +0000 (02:22 -0700)] 
[PATCH] rtc-sysfs fix

It's not clear how this thinko got through..

Cc: Olaf Hering <olaf@aepfle.de>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <alessandro.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[MTD] fix printk warning
Jeff Garzik [Sun, 1 Oct 2006 16:16:00 +0000 (12:16 -0400)] 
[MTD] fix printk warning

gcc spits out this warning:

drivers/mtd/mtd_blkdevs.c: In function â€˜do_blktrans_request’:
drivers/mtd/mtd_blkdevs.c:72: warning: format â€˜%ld’ expects type â€˜long int’, but argument 2 has type â€˜unsigned int’

This could be fixed any number of ways, including use of BUG().
rq_data_dir() only returns 0 or 1, so this entire case is superfluous.
I did the most simple fix.

Signed-off-by: Jeff Garzik <jeff@garzik.org>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
18 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6
David Woodhouse [Sun, 1 Oct 2006 16:55:53 +0000 (17:55 +0100)] 
Merge branch 'master' of git://git./linux/kernel/git/torvalds/linux-2.6

Manually resolve conflict in include/mtd/Kbuild

Signed-off-by: David Woodhouse <dwmw2@infradead.org>
18 years ago[SERIAL] Fix oops when removing suspended serial port
Russell King [Sun, 1 Oct 2006 16:17:40 +0000 (17:17 +0100)] 
[SERIAL] Fix oops when removing suspended serial port

A serial card might have been removed when the system is resumed.
This results in a suspended port being shut down, which results in
the ports shutdown method being called twice in a row.  This causes
BUGs.  Avoid this by tracking the suspended state separately from
the initialised state.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] Fix resume handling bug
Russell King [Sun, 1 Oct 2006 16:14:07 +0000 (17:14 +0100)] 
[SERIAL] Fix resume handling bug

Unfortunately, pcmcia_dev_present() returns false when a device is
suspended, so checking this on resume does not work too well.  Omit
this test.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] Remove wrong asm/serial.h inclusions
Russell King [Wed, 30 Aug 2006 09:27:15 +0000 (10:27 +0100)] 
[SERIAL] Remove wrong asm/serial.h inclusions

asm/serial.h is supposed to contain the definitions for the architecture
specific 8250 ports for the 8250 driver.  It may also define BASE_BAUD,
but this is the base baud for the architecture specific ports _only_.

Therefore, nothing other than the 8250 driver should be including this
header file.  In order to move towards this goal, here is a patch which
removes some of the more obvious incorrect includes of the file.

Acked-by: Paul Fulghum <paulkf@microgate.com>
Acked-by: Tony Luck <tony.luck@intel.com>
Acked-by: Ralf Baechle <ralf@linux-mips.org>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] CONFIG_PM=n slim: drivers/serial/8250_pci.c
Alexey Dobriyan [Mon, 25 Sep 2006 23:51:27 +0000 (16:51 -0700)] 
[SERIAL] CONFIG_PM=n slim: drivers/serial/8250_pci.c

Remove some code which is unneeded if CONFIG_PM=n.

Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] OMAP1510 serial fix for 115200 baud
Jonathan McDowell [Tue, 15 Aug 2006 06:05:32 +0000 (23:05 -0700)] 
[SERIAL] OMAP1510 serial fix for 115200 baud

The patch below is necessary for 115200 baud on an OMAP1510 internal UART.
It's been in the linux-omap tree for some time and with it applied to a
vanilla Linus git tree the serial console on the Amstrad Delta (which is
OMAP1510 based and whose initial bootloader runs at 115200) works fine (it
doesn't without it).

Signed-off-by: Jonathan McDowell <noodles@earth.li>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] returning proper error from serial core driver
Ram Gupta [Tue, 15 Aug 2006 06:05:29 +0000 (23:05 -0700)] 
[SERIAL] returning proper error from serial core driver

Fix the issue of returning 0 even in case of error from uart_set_info
function.  Now it returns the error EBUSY when it can not set new port.

Signed-off-by: Ram Gupta <r.gupta@astronautics.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] Make uart_line_info() correctly tell MMIO from I/O port
Sergei Shtylyov [Sun, 3 Sep 2006 20:04:20 +0000 (00:04 +0400)] 
[SERIAL] Make uart_line_info() correctly tell MMIO from I/O port

/proc/tty/driver/serial incorrectly claims that UARTs having iotype of
UPIO_MEM32, UPIO_AU, or UPIO_TSI are I/O mapped.

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] suspend/resume handlers don't have level arg anymore
Sergei Shtylyov [Mon, 11 Sep 2006 16:32:58 +0000 (20:32 +0400)] 
[SERIAL] suspend/resume handlers don't have level arg anymore

8250.c and serial_txx9.c port suspend/resume handler still have this obsolete
argument documented...

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] 8250 resourse management fixes
Sergei Shtylyov [Sat, 9 Sep 2006 18:23:56 +0000 (22:23 +0400)] 
[SERIAL] 8250 resourse management fixes

I think register ranges obviously need to be claimed/released for all UARTs
including those with UPIO_MEM32 and UPIO_TSI iotype.

Also, serial8250_request_rsa_resources() returns false positives with
UPIO_MEM32, UPIO_AU, and UPIO_TSI iotype -- I don't think this makes any sense.

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] serial_cs: Add quirk for brainboxes 2-port RS232 card
Russell King [Sat, 16 Sep 2006 22:07:46 +0000 (23:07 +0100)] 
[SERIAL] serial_cs: Add quirk for brainboxes 2-port RS232 card

Mauro Ziliani reports that this card has a higher clock rate.
Rather than tweak the 8250 driver to handle this, add a quirk to
pass the correct clock rate to the driver.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] serial_cs: handle Nokia multi->single port bodge via config quirk
Russell King [Sat, 16 Sep 2006 22:04:37 +0000 (23:04 +0100)] 
[SERIAL] serial_cs: handle Nokia multi->single port bodge via config quirk

According to the existing code, Nokia only make single-port cards,
but are detected as multi-port cards.  Handle this in roughly the
same way via the config quirk - forcing it to be a real single port
card (info->multi=0) changes the way we allocate the IO memory,
which might stop the card working.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] serial_cs: add configuration quirk
Russell King [Sat, 16 Sep 2006 22:00:54 +0000 (23:00 +0100)] 
[SERIAL] serial_cs: add configuration quirk

Add a quirk primerily to handle tweaks to the link->conf structure,
eg as required for Socket cards.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] serial_cs: Convert Oxford 950 / Possio GCC wakeup quirk
Russell King [Sat, 16 Sep 2006 21:45:41 +0000 (22:45 +0100)] 
[SERIAL] serial_cs: Convert Oxford 950 / Possio GCC wakeup quirk

Move the Oxford Semi OX950 / Possio GCC wakeup handling to a quirk
wakeup handler.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] serial_cs: convert IBM post-init handling to a quirk
Russell King [Sat, 16 Sep 2006 20:34:11 +0000 (21:34 +0100)] 
[SERIAL] serial_cs: convert IBM post-init handling to a quirk

Move IBM quirk handling into its own quirk entry.  Note that doing
quirk handling after we've registered the ports is racy, but since
I don't know if moving this will have an undesired effect, it's
probably better to leave where it is.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] serial_cs: allow wildcarded quirks
Russell King [Sat, 16 Sep 2006 20:26:16 +0000 (21:26 +0100)] 
[SERIAL] serial_cs: allow wildcarded quirks

Some quirks we will introduce next apply to (eg) all cards of one
manufacturer.  Therefore, we need a way to list these in the quirk
table - use ~0 - this is not a possible device ID value.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] serial_cs: convert multi-port table to quirk table
Russell King [Sat, 16 Sep 2006 20:09:41 +0000 (21:09 +0100)] 
[SERIAL] serial_cs: convert multi-port table to quirk table

- rename multi_id table to serial_quirk / quirks[]
- use named initialisers
- store a pointer to the quirk table in the serial_info structure
  so we can use the quirk table entry later.
- apply multi-port quirk after the multi-port guessing code,
  but only if it's != -1.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] serial_cs: Use clean up multiport card detection
Russell King [Sat, 16 Sep 2006 19:53:05 +0000 (20:53 +0100)] 
[SERIAL] serial_cs: Use clean up multiport card detection

- Use ARRAY_SIZE() instead of home grown based version.
- use parse->manfid.card rather than le16_to_cpu(buf[1]) -
  manfid.card is already converted to this format.
- use info->prodid in subsequent tests rather than
  parse->manfid.card.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[SERIAL] Remove m32r_sio dependency on asm/serial.h
Russell King [Wed, 30 Aug 2006 14:30:39 +0000 (15:30 +0100)] 
[SERIAL] Remove m32r_sio dependency on asm/serial.h

m32r_sio re-uses a custom defined BASE_BAUD from asm/serial.h,
and replaces SERIAL_PORT_DFNS with its own driver private copy.
Since asm/serial.h is supposed to define 8250-based ports using
these symbols, this isn't a sane idea.

Hence, eliminate asm/serial.h from m32r_sio.c.

Acked-by: Hirokazu Takata <takata@linux-m32r.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[netdrvr] hp100: encapsulate all non-module code
Jeff Garzik [Sun, 1 Oct 2006 11:32:20 +0000 (07:32 -0400)] 
[netdrvr] hp100: encapsulate all non-module code

The previous '#ifndef MODULE' block did not cover all the
static-build-only code.  Now it does.

Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years agodrivers/net/wireless/{airo,ipw2100}: fix error handling bugs
Jeff Garzik [Sun, 1 Oct 2006 11:31:09 +0000 (07:31 -0400)] 
drivers/net/wireless/{airo,ipw2100}: fix error handling bugs

airo:
* fix oops, if !CONFIG_PROC_FS (create_proc_entry always returns NULL)

* handle pci_register_driver() failure.  if it fails, we really do
  want to exit, rather than (as a comment indicates) return success
  because-we-are-a-library.

* #if 0 have_isa_dev variable, which is assigned a value but never used

ipw2100:
* handle sysfs_create_group() failure

* handle driver_create_file() failure

Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[netdrvr] phy: Fix bugs in error handling
Jeff Garzik [Sun, 1 Oct 2006 11:27:46 +0000 (07:27 -0400)] 
[netdrvr] phy: Fix bugs in error handling

The recent __must_check stuff flagged some error handling bugs.

phy/fixed.c:
* handle device_bind_driver() failure

phy/phy_device.c:
* handle device_bind_driver() failure
* release rwsem upon failure

Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years agokbuild: trivial documentation fixes
Robert P. J. Day [Mon, 25 Sep 2006 19:55:51 +0000 (15:55 -0400)] 
kbuild: trivial documentation fixes

Signed-off-by: "Robert P. J. Day" <rpjday@mindspring.com>
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
18 years agokconfig: fix saving alternate kconfig file in parent dir
Sam Ravnborg [Sun, 1 Oct 2006 09:48:53 +0000 (11:48 +0200)] 
kconfig: fix saving alternate kconfig file in parent dir

This fixes bugzilla entry: 7182
http://bugzilla.kernel.org/show_bug.cgi?id=7182

With this patch we no longer append the directory part twice
before saving the config file.
This patch has been sent to Roman Zippel for review with no feedback.
It is so obviously simple that this should be OK to apply it anyway.

Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
18 years agokbuild: make modpost processing configurable
Sam Ravnborg [Sun, 1 Oct 2006 09:35:24 +0000 (11:35 +0200)] 
kbuild: make modpost processing configurable

On request from Al Viro make modpost processing configurable.

KBUILD_MODPOST_WARN can be set to make modpost warn instead of
error out in case on unresolved symbols in final module link.

KBUILD_MODPOST_NOFINAL can be set to avoid the final and timeconsuming
.c file generation and link of .ko files. This is solely useful for
speeding up when doing compile checks with for example allmodconfig

Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davej/agpgart
Linus Torvalds [Sun, 1 Oct 2006 07:40:55 +0000 (00:40 -0700)] 
Merge /pub/scm/linux/kernel/git/davej/agpgart

* master.kernel.org:/pub/scm/linux/kernel/git/davej/agpgart:
  [AGPGART] printk fixups.
  [AGPGART] Use pci_get_slot not pci_find_slot

18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davej/cpufreq
Linus Torvalds [Sun, 1 Oct 2006 07:40:35 +0000 (00:40 -0700)] 
Merge /pub/scm/linux/kernel/git/davej/cpufreq

* master.kernel.org:/pub/scm/linux/kernel/git/davej/cpufreq:
  [CPUFREQ] Make acpi-cpufreq unsticky again.
  [CPUFREQ] longhaul: remove duplicated code.
  [CPUFREQ] Longhaul - Disable arbiter CLE266
  [CPUFREQ] Fix section mismatch warning
  [CPUFREQ] Fix cut-n-paste bug in suspend printk

18 years ago[PATCH] Some config.h removals
Zachary Amsden [Sun, 1 Oct 2006 06:29:38 +0000 (23:29 -0700)] 
[PATCH] Some config.h removals

During tracking down a PAE compile failure, I found that config.h was being
included in a bunch of places in i386 code.  It is no longer necessary, so
drop it.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] paravirt: update pte hook
Zachary Amsden [Sun, 1 Oct 2006 06:29:38 +0000 (23:29 -0700)] 
[PATCH] paravirt: update pte hook

Add a pte_update_hook which notifies about pte changes that have been made
without using the set_pte / clear_pte interfaces.  This allows shadow mode
hypervisors which do not trap on page table access to maintain synchronized
shadows.

It also turns out, there was one pte update in PAE mode that wasn't using any
accessor interface at all for setting NX protection.  Considering it is PAE
specific, and the accessor is i386 specific, I didn't want to add a generic
encapsulation of this behavior yet.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] paravirt: remove set pte atomic
Zachary Amsden [Sun, 1 Oct 2006 06:29:37 +0000 (23:29 -0700)] 
[PATCH] paravirt: remove set pte atomic

Now that ptep_establish has a definition in PAE i386 3-level paging code, the
only paging model which is insane enough to have multi-word hardware PTEs
which are not efficient to set atomically, we can remove the ghost of
set_pte_atomic from other architectures which falesly duplicated it, and
remove all knowledge of it from the generic pgtable code.

set_pte_atomic is now a private pte operator which is specific to i386

Signed-off-by: Zachary Amsden <zach@vmware.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] paravirt: optimize ptep establish for pae
Zachary Amsden [Sun, 1 Oct 2006 06:29:36 +0000 (23:29 -0700)] 
[PATCH] paravirt: optimize ptep establish for pae

The ptep_establish macro is only used on user-level PTEs, for P->P mapping
changes.  Since these always happen under protection of the pagetable lock,
the strong synchronization of a 64-bit cmpxchg is not needed, in fact, not
even a lock prefix needs to be used.  We can simply instead clear the P-bit,
followed by a normal set.  The write ordering is still important to avoid the
possibility of the TLB snooping a partially written PTE and getting a bad
mapping installed.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] paravirt: kpte flush
Zachary Amsden [Sun, 1 Oct 2006 06:29:35 +0000 (23:29 -0700)] 
[PATCH] paravirt: kpte flush

Create a new PTE function which combines clearing a kernel PTE with the
subsequent flush.  This allows the two to be easily combined into a single
hypercall or paravirt-op.  More subtly, reverse the order of the flush for
kmap_atomic.  Instead of flushing on establishing a mapping, flush on clearing
a mapping.  This eliminates the possibility of leaving stale kmap entries
which may still have valid TLB mappings.  This is required for direct mode
hypervisors, which need to reprotect all mappings of a given page when
changing the page type from a normal page to a protected page (such as a page
table or descriptor table page).  But it also provides some nicer semantics
for real hardware, by providing extra debug-proofing against using stale
mappings, as well as ensuring that no stale mappings exist when changing the
cacheability attributes of a page, which could lead to cache conflicts when
two different types of mappings exist for the same page.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] paravirt: combine flush accessed dirty.patch
Zachary Amsden [Sun, 1 Oct 2006 06:29:34 +0000 (23:29 -0700)] 
[PATCH] paravirt: combine flush accessed dirty.patch

Remove ptep_test_and_clear_{dirty|young} from i386, and instead use the
dominating functions, ptep_clear_flush_{dirty|young}.  This allows the TLB
page flush to be contained in the same macro, and allows for an eager
optimization - if reading the PTE initially returned dirty/accessed, we can
assume the fact that no subsequent update to the PTE which cleared accessed /
dirty has occurred, as the only way A/D bits can change without holding the
page table lock is if a remote processor clears them.  This eliminates an
extra branch which came from the generic version of the code, as we know that
no other CPU could have cleared the A/D bit, so the flush will always be
needed.

We still export these two defines, even though we do not actually define
the macros in the i386 code:

 #define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
 #define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_DIRTY

The reason for this is that the only use of these functions is within the
generic clear_flush functions, and we want a strong guarantee that there
are no other users of these functions, so we want to prevent the generic
code from defining them for us.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] paravirt: lazy mmu mode hooks.patch
Zachary Amsden [Sun, 1 Oct 2006 06:29:33 +0000 (23:29 -0700)] 
[PATCH] paravirt: lazy mmu mode hooks.patch

Implement lazy MMU update hooks which are SMP safe for both direct and shadow
page tables.  The idea is that PTE updates and page invalidations while in
lazy mode can be batched into a single hypercall.  We use this in VMI for
shadow page table synchronization, and it is a win.  It also can be used by
PPC and for direct page tables on Xen.

For SMP, the enter / leave must happen under protection of the page table
locks for page tables which are being modified.  This is because otherwise,
you end up with stale state in the batched hypercall, which other CPUs can
race ahead of.  Doing this under the protection of the locks guarantees the
synchronization is correct, and also means that spurious faults which are
generated during this window by remote CPUs are properly handled, as the page
fault handler must re-check the PTE under protection of the same lock.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] paravirt: pte clear not present
Zachary Amsden [Sun, 1 Oct 2006 06:29:31 +0000 (23:29 -0700)] 
[PATCH] paravirt: pte clear not present

Change pte_clear_full to a more appropriately named pte_clear_not_present,
allowing optimizations when not-present mapping changes need not be reflected
in the hardware TLB for protected page table modes.  There is also another
case that can use it in the fremap code.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] paravirt: remove read hazard from cow
Zachary Amsden [Sun, 1 Oct 2006 06:29:30 +0000 (23:29 -0700)] 
[PATCH] paravirt: remove read hazard from cow

We don't want to read PTEs directly like this after they have been modified,
as a lazy MMU implementation of direct page tables may not have written the
updated PTE back to memory yet.

Signed-off-by: Zachary Amsden <zach@vmware.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] invalidate_inode_pages2(): ignore page refcounts
Andrew Morton [Sun, 1 Oct 2006 06:29:29 +0000 (23:29 -0700)] 
[PATCH] invalidate_inode_pages2(): ignore page refcounts

The recent fix to invalidate_inode_pages() (git commit 016eb4a) managed to
unfix invalidate_inode_pages2().

The problem is that various bits of code in the kernel can take transient refs
on pages: the page scanner will do this when inspecting a batch of pages, and
the lru_cache_add() batching pagevecs also hold a ref.

Net result is transient failures in invalidate_inode_pages2().  This affects
NFS directory invalidation (observed) and presumably also block-backed
direct-io (not yet reported).

Fix it by reverting invalidate_inode_pages2() back to the old version which
ignores the page refcounts.

We may come up with something more clever later, but for now we need a 2.6.18
fix for NFS.

Cc: Chuck Lever <cel@citi.umich.edu>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Support piping into commands in /proc/sys/kernel/core_pattern
Andi Kleen [Sun, 1 Oct 2006 06:29:28 +0000 (23:29 -0700)] 
[PATCH] Support piping into commands in /proc/sys/kernel/core_pattern

Using the infrastructure created in previous patches implement support to
pipe core dumps into programs.

This is done by overloading the existing core_pattern sysctl
with a new syntax:

|program

When the first character of the pattern is a '|' the kernel will instead
threat the rest of the pattern as a command to run.  The core dump will be
written to the standard input of that program instead of to a file.

This is useful for having automatic core dump analysis without filling up
disks.  The program can do some simple analysis and save only a summary of
the core dump.

The core dump proces will run with the privileges and in the name space of
the process that caused the core dump.

I also increased the core pattern size to 128 bytes so that longer command
lines fit.

Most of the changes comes from allowing core dumps without seeks.  They are
fairly straight forward though.

One small incompatibility is that if someone had a core pattern previously
that started with '|' they will get suddenly new behaviour.  I think that's
unlikely to be a real problem though.

Additional background:

> Very nice, do you happen to have a program that can accept this kind of
> input for crash dumps?  I'm guessing that the embedded people will
> really want this functionality.

I had a cheesy demo/prototype.  Basically it wrote the dump to a file again,
ran gdb on it to get a backtrace and wrote the summary to a shared directory.
Then there was a simple CGI script to generate a "top 10" crashes HTML
listing.

Unfortunately this still had the disadvantage to needing full disk space for a
dump except for deleting it afterwards (in fact it was worse because over the
pipe holes didn't work so if you have a holey address map it would require
more space).

Fortunately gdb seems to be happy to handle /proc/pid/fd/xxx input pipes as
cores (at least it worked with zsh's =(cat core) syntax), so it would be
likely possible to do it without temporary space with a simple wrapper that
calls it in the right way.  I ran out of time before doing that though.

The demo prototype scripts weren't very good.  If there is really interest I
can dig them out (they are currently on a laptop disk on the desk with the
laptop itself being in service), but I would recommend to rewrite them for any
serious application of this and fix the disk space problem.

Also to be really useful it should probably find a way to automatically fetch
the debuginfos (I cheated and just installed them in advance).  If nobody else
does it I can probably do the rewrite myself again at some point.

My hope at some point was that desktops would support it in their builtin
crash reporters, but at least the KDE people I talked too seemed to be happy
with their user space only solution.

Alan sayeth:

  I don't believe that piping as such as neccessarily the right model, but
  the ability to intercept and processes core dumps from user space is asked
  for by many enterprise users as well.  They want to know about, capture,
  analyse and process core dumps, often centrally and in automated form.

[akpm@osdl.org: loff_t != unsigned long]
Signed-off-by: Andi Kleen <ak@suse.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Create call_usermodehelper_pipe()
Andi Kleen [Sun, 1 Oct 2006 06:29:27 +0000 (23:29 -0700)] 
[PATCH] Create call_usermodehelper_pipe()

A new member in the ever growing family of call_usermode* functions is
born.  The new call_usermodehelper_pipe() function allows to pipe data to
the stdin of the called user mode progam and behaves otherwise like the
normal call_usermodehelp() (except that it always waits for the child to
finish)

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Some cleanup in the pipe code
Andi Kleen [Sun, 1 Oct 2006 06:29:26 +0000 (23:29 -0700)] 
[PATCH] Some cleanup in the pipe code

Split the big and hard to read do_pipe function into smaller pieces.

This creates new create_write_pipe/free_write_pipe/create_read_pipe
functions.  These functions are made global so that they can be used by
other parts of the kernel.

The resulting code is more generic and easier to read and has cleaner error
handling and less gotos.

[akpm@osdl.org: cleanup]
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ioremap balanced with iounmap for drivers/serial/sunsu.c
Amol Lad [Sun, 1 Oct 2006 06:29:25 +0000 (23:29 -0700)] 
[PATCH] ioremap balanced with iounmap for drivers/serial/sunsu.c

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: David S. Miller <davem@sunset.davemloft.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ioremap balanced with iounmap for drivers/serial/mux.c
Amol Lad [Sun, 1 Oct 2006 06:29:25 +0000 (23:29 -0700)] 
[PATCH] ioremap balanced with iounmap for drivers/serial/mux.c

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ioremap balanced with iounmap for drivers/serial/mpsc.c
Amol Lad [Sun, 1 Oct 2006 06:29:24 +0000 (23:29 -0700)] 
[PATCH] ioremap balanced with iounmap for drivers/serial/mpsc.c

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ioremap balanced with iounmap for drivers/serial/mpc52xx_uart.c
Amol Lad [Sun, 1 Oct 2006 06:29:23 +0000 (23:29 -0700)] 
[PATCH] ioremap balanced with iounmap for drivers/serial/mpc52xx_uart.c

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ioremap balanced with iounmap for drivers/serial/ip22zilog.c
Amol Lad [Sun, 1 Oct 2006 06:29:22 +0000 (23:29 -0700)] 
[PATCH] ioremap balanced with iounmap for drivers/serial/ip22zilog.c

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ioremap balanced with iounmap for drivers/serial/ioc4_serial.c
Amol Lad [Sun, 1 Oct 2006 06:29:21 +0000 (23:29 -0700)] 
[PATCH] ioremap balanced with iounmap for drivers/serial/ioc4_serial.c

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Brent Casavant <bcasavan@sgi.com>
Cc: Pat Gefre <pfg@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ioremap balanced with iounmap for drivers/serial/8250_gsc.c
Amol Lad [Sun, 1 Oct 2006 06:29:21 +0000 (23:29 -0700)] 
[PATCH] ioremap balanced with iounmap for drivers/serial/8250_gsc.c

ioremap must be balanced by an iounmap and failing to do so can result
in a memory leak.

Signed-off-by: Amol Lad <amol@verismonetworks.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>