Rafael J. Wysocki reported weird, multi-seconds delays during
suspend/resume and bisected it back to:
commit 82a1fcb902
Author: Ingo Molnar <mingo@elte.hu>
Date: Fri Jan 25 21:08:02 2008 +0100
softlockup: automatically detect hung TASK_UNINTERRUPTIBLE tasks
fix it:
- restore the old wakeup mechanism
- fix break usage in do_each_thread() { } while_each_thread().
- fix the hotplug switch stmt, a fall-through case was broken.
Bisected-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Tested-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'audit.b46' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/audit-current:
[AUDIT] Add uid, gid fields to ANOM_PROMISCUOUS message
[AUDIT] ratelimit printk messages audit
[patch 2/2] audit: complement va_copy with va_end()
[patch 1/2] kernel/audit.c: warning fix
[AUDIT] create context if auditing was ever enabled
[AUDIT] clean up audit_receive_msg()
[AUDIT] make audit=0 really stop audit messages
[AUDIT] break large execve argument logging into smaller messages
[AUDIT] include audit type in audit message when using printk
[AUDIT] do not panic on exclude messages in audit_log_pid_context()
[AUDIT] Add End of Event record
[AUDIT] add session id to audit messages
[AUDIT] collect uid, loginuid, and comm in OBJ_PID records
[AUDIT] return EINTR not ERESTART*
[PATCH] get rid of loginuid races
[PATCH] switch audit_get_loginuid() to task_struct *
some printk messages from the audit system can become excessive. This
patch ratelimits those messages. It was found that messages, such as
the audit backlog lost printk message could flood the logs to the point
that a machine could take an nmi watchdog hit or otherwise become
unresponsive.
Signed-off-by: Eric Paris <eparis@redhat.com>
Complement va_copy() with va_end().
Signed-off-by: Richard Knutsson <ricknu-0@student.ltu.se>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
kernel/audit.c: In function 'audit_log_start':
kernel/audit.c:1133: warning: 'serial' may be used uninitialized in this function
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Disabling audit at runtime by auditctl doesn't mean that we can
stop allocating contexts for new processes; we don't want to miss them
when that sucker is reenabled.
(based on work from Al Viro in the RHEL kernel series)
Signed-off-by: Eric Paris <eparis@redhat.com>
generally clean up audit_receive_msg() don't free random memory if
selinux_sid_to_string fails for some reason. Move generic auditing
to a helper function
Signed-off-by: Eric Paris <eparis@redhat.com>
Some audit messages (namely configuration changes) are still emitted even if
the audit subsystem has been explicitly disabled. This patch turns those
messages off as well.
Signed-off-by: Eric Paris <eparis@redhat.com>
execve arguments can be quite large. There is no limit on the number of
arguments and a 4G limit on the size of an argument.
this patch prints those aruguments in bite sized pieces. a userspace size
limitation of 8k was discovered so this keeps messages around 7.5k
single arguments larger than 7.5k in length are split into multiple records
and can be identified as aX[Y]=
Signed-off-by: Eric Paris <eparis@redhat.com>
Currently audit drops the audit type when an audit message goes through
printk instead of the audit deamon. This is a minor annoyance in
that the audit type is no longer part of the message and the information
the audit type conveys needs to be carried in, or derived from the
message data.
The attached patch includes the type number as part of the printk.
Admittedly it isn't the type name that the audit deamon provides but I
think this is better than dropping the type completely.
Signed-pff-by: John Johansen <jjohansen@suse.de>
Signed-off-by: Eric Paris <eparis@redhat.com>
If we fail to get an ab in audit_log_pid_context this may be due to an exclude
rule rather than a memory allocation failure. If it was due to a memory
allocation failue we would have already paniced and no need to do it again.
Signed-off-by: Eric Paris <eparis@redhat.com>
This patch adds an end of event record type. It will be sent by the kernel as
the last record when a multi-record event is triggered. This will aid realtime
analysis programs since they will now reliably know they have the last record
to complete an event. The audit daemon filters this and will not write it to
disk.
Signed-off-by: Steve Grubb <sgrubb redhat com>
Signed-off-by: Eric Paris <eparis@redhat.com>
In order to correlate audit records to an individual login add a session
id. This is incremented every time a user logs in and is included in
almost all messages which currently output the auid. The field is
labeled ses= or oses=
Signed-off-by: Eric Paris <eparis@redhat.com>
Add uid, loginuid, and comm collection to OBJ_PID records. This just
gives users a little more information about the task that received a
signal. pid is rather meaningless after the fact, and even though comm
isn't great we can't collect exe reasonably on this code path for
performance reasons.
Signed-off-by: Eric Paris <eparis@redhat.com>
The syscall exit code will change ERESTART* kernel internal return codes
to EINTR if it does not restart the syscall. Since we collect the audit
info before that point we should fix those in the audit log as well.
Signed-off-by: Eric Paris <eparis@redhat.com>
Keeping loginuid in audit_context is racy and results in messier
code. Taken to task_struct, out of the way of ->audit_context
changes.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
To allow the implementation of optimized rw-locks in user space, glibc
needs a possibility to select waiters for wakeup depending on a bitset
mask.
This requires two new futex OPs: FUTEX_WAIT_BITS and FUTEX_WAKE_BITS
These OPs are basically the same as FUTEX_WAIT and FUTEX_WAKE plus an
additional argument - a bitset. Further the FUTEX_WAIT_BITS OP is
expecting an absolute timeout value instead of the relative one, which
is used for the FUTEX_WAIT OP.
FUTEX_WAIT_BITS calls into the kernel with a bitset. The bitset is
stored in the futex_q structure, which is used to enqueue the waiter
into the hashed futex waitqueue.
FUTEX_WAKE_BITS also calls into the kernel with a bitset. The wakeup
function logically ANDs the bitset with the bitset stored in each
waiters futex_q structure. If the result is zero (i.e. none of the set
bits in the bitsets is matching), then the waiter is not woken up. If
the result is not zero (i.e. one of the set bits in the bitsets is
matching), then the waiter is woken.
The bitset provided by the caller must be non zero. In case the
provided bitset is zero the kernel returns EINVAL.
Internaly the new OPs are only extensions to the existing FUTEX_WAIT
and FUTEX_WAKE functions. The existing OPs hand a bitset with all bits
set into the futex_wait() and futex_wake() functions.
Signed-off-by: Thomas Gleixner <tgxl@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The WARN_ON() in the fixup return path of futex_lock_pi() can
trigger with false positives.
The following scenario happens:
t1 holds the futex and t2 and t3 are blocked on the kernel side rt_mutex.
t1 releases the futex (and the rt_mutex) and assigned t2 to be the next
owner of the futex.
t2 is interrupted and returns w/o acquiring the rt_mutex, before t1 can
release the rtmutex.
t1 releases the rtmutex and t3 becomes the pending owner of the rtmutex.
t2 notices that it is the designated owner (user space variable) and
fails to acquire the rt_mutex via trylock, because it is not allowed to
steal the rt_mutex from t3. Now it looks at the rt_mutex pending owner (t3)
and assigns the futex and the pi_state to it.
During the fixup t4 steals the rtmutex from t3.
t2 returns from the fixup and the owner of the rt_mutex has changed from
t3 to t4.
There is no need to do another round of fixups from t2. The important
part (t2 is not returning as the user space visible owner) is
done. The further fixups are done, before either t3 or t4 return to
user space.
For the user space it is not relevant which task (t3 or t4) is the real
owner, as long as those are both in the kernel, which is guaranteed by
the serialization of the hash bucket lock. Both tasks (which ever returns
first to userspace - t4 because it locked the rt_mutex or t3 due to a signal)
are going through the lock_futex_pi() return path where the ownership is
fixed before the return to user space.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
To allow better diagnosis of tick-sched related, especially NOHZ
related problems, we need to know when the last wakeup via an irq
happened and when the CPU left the idle state.
Add two fields (idle_waketime, idle_exittime) to the tick_sched
structure and add them to the timer_list output.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
xtime_cache needs to be updated whenever xtime and or wall_to_monotic
are changed. Otherwise users of xtime_cache might see a stale (and in
the case of timezone changes utterly wrong) value until the next
update happens.
Fixup the obvious places, which miss this update.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: John Stultz <johnstul@us.ibm.com>
Tested-by: Dhaval Giani <dhaval@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
this patch:
commit 37bb6cb409
Author: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Fri Jan 25 21:08:32 2008 +0100
hrtimer: unlock hrtimer_wakeup
Broke hrtimer_init_sleeper() users. It forgot to fix up the futex
caller of this function to detect the failed queueing and messed up
the do_nanosleep() caller in that it could leak a TASK_INTERRUPTIBLE
state.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The recent UDP patch exposed this bug in the audit code. It
was calling pskb_expand_head without increasing skb->truesize.
The caller of pskb_expand_head needs to do so because that function
is designed to be called in places where truesize is already fixed
and therefore it doesn't update its value.
Because the audit system is using it in a place where the truesize
has not yet been fixed, it needs to update its value manually.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
It may be used by the modules nfs.ko and sunrpc.ko
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
[ Made it a regular export rather than GPL-only - Linus ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'task_killable' of git://git.kernel.org/pub/scm/linux/kernel/git/willy/misc: (22 commits)
Remove commented-out code copied from NFS
NFS: Switch from intr mount option to TASK_KILLABLE
Add wait_for_completion_killable
Add wait_event_killable
Add schedule_timeout_killable
Use mutex_lock_killable in vfs_readdir
Add mutex_lock_killable
Use lock_page_killable
Add lock_page_killable
Add fatal_signal_pending
Add TASK_WAKEKILL
exit: Use task_is_*
signal: Use task_is_*
sched: Use task_contributes_to_load, TASK_ALL and TASK_NORMAL
ptrace: Use task_is_*
power: Use task_is_*
wait: Use TASK_NORMAL
proc/base.c: Use task_is_*
proc/array.c: Use TASK_REPORT
perfmon: Use task_is_*
...
Fixed up conflicts in NFS/sunrpc manually..
i was debugging early crashes and wondered where all the printks
went. The reason: ignore_loglevel_setup() was not called yet ...
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This removes the extra struct task_struct *p parameter in inc_nr_running
and dec_nr_running functions.
Signed-off by: Jerry Stralko <gerb.stralko@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Michel Dänzr has bisected an interactivity problem with
plus-reniced tasks back to this commit:
810e95ccd5 is first bad commit
commit 810e95ccd5
Author: Peter Zijlstra <a.p.zijlstra@chello.nl>
Date: Mon Oct 15 17:00:14 2007 +0200
sched: another wakeup_granularity fix
unit mis-match: wakeup_gran was used against a vruntime
fix this by assymetrically scaling the vtime of positive reniced
tasks.
Bisected-by: Michel Dänzer <michel@tungstengraphics.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The reason why we are getting better wakeup latencies for
!FAIR_USER_SCHED is because of this snippet of code in place_entity():
if (!initial) {
/* sleeps upto a single latency don't count. */
if (sched_feat(NEW_FAIR_SLEEPERS) && entity_is_task(se))
^^^^^^^^^^^^^^^^^^
vruntime -= sysctl_sched_latency;
/* ensure we never gain time by being placed backwards. */
vruntime = max_vruntime(se->vruntime, vruntime);
}
NEW_FAIR_SLEEPERS feature gives credit for sleeping only to tasks and
not group-level entities. With the patch attached, I could see that
wakeup latencies with FAIR_USER_SCHED are restored to the same level as
!FAIR_USER_SCHED.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
these bugs are harder to find than they seem, a stackdump helps.
make it dependent on CONFIG_DEBUG_SHIRQ so that people can turn it off
if it annoys them.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
During the work on the x86 32 and 64 bit backtrace code I found it useful
to have a simple test module to test a process and irq context backtrace.
Since the existing backtrace code was buggy, I figure it might be useful
to have such a test module in the kernel so that maybe we can even
detect such bugs earlier..
[ mingo@elte.hu: build fix ]
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Here is a quick and naive smoke test for kprobes. This is intended to
just verify if some unrelated change broke the *probes subsystem. It is
self contained, architecture agnostic and isn't of any great use by itself.
This needs to be built in the kernel and runs a basic set of tests to
verify if kprobes, jprobes and kretprobes run fine on the kernel. In case
of an error, it'll print out a message with a "BUG" prefix.
This is a start; we intend to add more tests to this bucket over time.
Thanks to Jim Keniston and Masami Hiramatsu for comments and suggestions.
Tested on x86 (32/64) and powerpc.
Signed-off-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Acked-by: Masami Hiramatsu <mhiramat@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Unlike oopses, WARN_ON() currently does't print the loaded modules list.
This makes it harder to take action on certain bug reports. For example,
recently there were a set of WARN_ON()s reported in the mac80211 stack,
which were just signalling a driver bug. It takes then anther round trip
to the bug reporter (if he responds at all) to find out which driver
is at fault.
Another issue is that, unlike oopses, WARN_ON() doesn't currently printk
the helpful "cut here" line, nor the "end of trace" marker.
Now that WARN_ON() is out of line, the size increase due to this is
minimal and it's worth adding.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
A quick grep shows that there are currently 1145 instances of WARN_ON
in the kernel. Currently, WARN_ON is pretty much entirely inlined,
which makes it hard to enhance it without growing the size of the kernel
(and getting Andrew unhappy).
This patch build on top of Olof's patch that introduces __WARN,
and places the slowpath out of line. It also uses Ingo's suggestion
to not use __FUNCTION__ but to use kallsyms to do the lookup;
this saves a ton of extra space since gcc doesn't need to store the function
string twice now:
3936367 833603 624736 5394706 525112 vmlinux.before
3917508 833603 624736 5375847 520767 vmlinux-slowpath
15Kb savings...
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
CC: Andrew Morton <akpm@linux-foundation.org>
CC: Olof Johansson <olof@lixom.net>
Acked-by: Matt Meckall <mpm@selenic.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This is useful to debug problems with interrupt handlers that return
sometimes IRQ_NONE.
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This allows to change them at runtime using sysfs. No need to
reboot to set them.
I only added aliases (kernel.noirqdebug etc.) so the old options
still work.
Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This adds a generic definition of compat_sys_ptrace that calls
compat_arch_ptrace, parallel to sys_ptrace/arch_ptrace. Some
machines needing this already define a function by that name.
The new generic function is defined only on machines that
put #define __ARCH_WANT_COMPAT_SYS_PTRACE into asm/ptrace.h.
Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This adds a compat_ptrace_request that is the analogue of ptrace_request
for the things that 32-on-64 ptrace implementations can share in common.
So far there are just a couple of requests handled generically.
Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This makes ptrace_request handle {PEEK,POKE}{TEXT,DATA} directly.
Every arch_ptrace that could call generic_ptrace_peekdata already
has a default case calling ptrace_request, so this keeps things
simpler for the arch code.
Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The break_lock data structure and code for spinlocks is quite nasty.
Not only does it double the size of a spinlock but it changes locking to
a potentially less optimal trylock.
Put all of that under CONFIG_GENERIC_LOCKBREAK, and introduce a
__raw_spin_is_contended that uses the lock data itself to determine whether
there are waiters on the lock, to be used if CONFIG_GENERIC_LOCKBREAK is
not set.
Rename need_lockbreak to spin_needbreak, make it use spin_is_contended to
decouple it from the spinlock implementation, and make it typesafe (rwlocks
do not have any need_lockbreak sites -- why do they even get bloated up
with that break_lock then?).
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
We have a lot of code which differs only by the naming of specific
members of structures that contain registers. In order to enable
additional unifications, this patch drops the e- or r- size prefix
from the register names in struct pt_regs, and drops the x- prefixes
for segment registers on the 32-bit side.
This patch also performs the equivalent renames in some additional
places that might be candidates for unification in the future.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This makes ptrace_request handle PTRACE_SINGLEBLOCK along with
PTRACE_CONT et al. The new generic code makes use of the
arch_has_block_step macro and generic entry points on machines
that define them.
[ mingo@elte.hu: bugfix ]
Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This makes ptrace_request handle all the ptrace requests that wake
up the traced task. These do low-level ptrace implementation magic
that is not arch-specific and should be kept out of arch code. The
implementations on each arch usually do the same thing. The new
generic code makes use of the arch_has_single_step macro and generic
entry points to handle PTRACE_SINGLESTEP.
Signed-off-by: Roland McGrath <roland@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
various changes to the in_p/out_p delay details:
- add the io_delay=none method
- make each method selectable from the kernel config
- simplify the delay code a bit by getting rid of an indirect function call
- add the /proc/sys/kernel/io_delay_type sysctl
- change 'io_delay=standard|alternate' to io_delay=0x80 and io_delay=0xed
- make the io delay config not depend on CONFIG_DEBUG_KERNEL
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: "David P. Reed" <dpreed@reed.com>
Current idle time in kstat is based on jiffies and is coarse grained.
tick_sched.idle_sleeptime is making some attempt to keep track of idle time
in a fine grained manner. But, it is not handling the time spent in
interrupts fully.
Make tick_sched.idle_sleeptime accurate with respect to time spent on
handling interrupts and also add tick_sched.idle_lastupdate, which keeps
track of last time when idle_sleeptime was updated.
This statistics will be crucial for cpufreq-ondemand governor, which can
shed some conservative gaurd band that is uses today while setting the
frequency. The ondemand changes that uses the exact idle time is coming
soon.
Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
I recently noticed on one of my boxes that when synched with an NTP
server, the drift value reported for the system was ~283ppm. While in
some cases, clock hardware can be that bad, it struck me as unusual as
the system was using the acpi_pm clocksource, which is one of the more
trustworthy and accurate clocksources on x86 hardware.
I brought up another system and let it sync to the same NTP server, and
I noticed a similar 280some ppm drift.
In looking at the code, I found that the acpi_pm's constant frequency
was being computed correctly at boot-up, however once the system was up,
even without the ntp daemon running, the clocksource's frequency was
being modified by the clocksource_adjust() function.
Digging deeper, I realized that in the code that keeps track of how much
the clocksource is skewing from the ntp desired time, we were using
different lengths to establish how long an time interval was.
The clocksource was being setup with the following interval:
NTP_INTERVAL_LENGTH = NSEC_PER_SEC/NTP_INTERVAL_FREQ
While the ntp code was using the tick_length_base value:
tick_length_base ~= (tick_usec * NSEC_PER_USEC * USER_HZ)
/NTP_INTERVAL_FREQ
The subtle difference is:
(tick_usec * NSEC_PER_USEC * USER_HZ) != NSEC_PER_SEC
This difference in calculation was causing the clocksource correction
code to apply a correction factor to the clocksource so the two
intervals were the same, however this results in the actual frequency of
the clocksource to be made incorrect. I believe this difference would
affect all clocksources, although to differing degrees depending on the
clocksource resolution.
The issue was introduced when my HZ free ntp patch landed in 2.6.21-rc1,
so my apologies for the mistake, and for not noticing it until now.
The following patch, corrects the clocksource's initialization code so
it uses the same interval length as the code in ntp.c. After applying
this patch, the drift value for the same system went from ~283ppm to
only 2.635ppm.
I believe this patch to be good, however it does affect all arches and
I've only tested on x86, so some caution is advised. I do think it would
be a likely candidate for a stable 2.6.24.x release.
Any thoughts or feedback would be appreciated.
Signed-off-by: John Stultz <johnstul@us.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
