Lockdep events subsystem gathers various locking related events
such as a request, release, contention or acquisition of a lock.
The name of this event subsystem is a bit of a misnomer since
these events are not quite related to lockdep but more generally
to locking, ie: these events are not reporting lock dependencies
or possible deadlock scenario but pure locking events.
Hence this rename.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Acked-by: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Li Zefan <lizf@cn.fujitsu.com>
LKML-Reference: <1258103194-843-1-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch updates percpu related symbols under kernel/ and mm/ such
that percpu symbols are unique and don't clash with local symbols.
This serves two purposes of decreasing the possibility of global
percpu symbol collision and allowing dropping per_cpu__ prefix from
percpu symbols.
* kernel/lockdep.c: s/lock_stats/cpu_lock_stats/
* kernel/sched.c: s/init_rq_rt/init_rt_rq_var/ (any better idea?)
s/sched_group_cpus/sched_groups/
* kernel/softirq.c: s/ksoftirqd/run_ksoftirqd/a
* kernel/softlockup.c: s/(*)_timestamp/softlockup_\1_ts/
s/watchdog_task/softlockup_watchdog/
s/timestamp/ts/ for local variables
* kernel/time/timer_stats: s/lookup_lock/tstats_lookup_lock/
* mm/slab.c: s/reap_work/slab_reap_work/
s/reap_node/slab_reap_node/
* mm/vmstat.c: local variable changed to avoid collision with vmstat_work
Partly based on Rusty Russell's "alloc_percpu: rename percpu vars
which cause name clashes" patch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: (slab/vmstat) Christoph Lameter <cl@linux-foundation.org>
Reviewed-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Nick Piggin <npiggin@suse.de>
Some tracepoint magic (TRACE_EVENT(lock_acquired)) relies on
the fact that lock hold times are positive and uses div64 on
that. That triggered a build warning on MIPS, and probably
causes bad output in certain circumstances as well.
Make it truly positive.
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1254818502.21044.112.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This allows lockdep to locate symbols that are in arch-specific data
sections (such as data in Blackfin on-chip SRAM regions).
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Robin Getz <rgetz@blackfin.uclinux.org>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since lockdep has introduced BFS to avoid recursion, statistics
for recursion does not make any sense now. So remove them.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Cc: a.p.zijlstra@chello.nl
LKML-Reference: <1251542879-5211-1-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The unit is KB, so sizeof(struct circular_queue) should be
divided by 1024.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Cc: akpm@linux-foundation.org
Cc: torvalds@linux-foundation.org
Cc: davem@davemloft.net
Cc: Ming Lei <tom.leiming@gmail.com>
Cc: a.p.zijlstra@chello.nl
LKML-Reference: <1249220616-7190-1-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
We still can apply DaveM's generation count optimization to
BFS, based on the following idea:
- before doing each BFS, increase the global generation id
by 1
- if one node in the graph has been visited, mark it as
visited by storing the current global generation id into
the node's dep_gen_id field
- so we can decide if one node has been visited already, by
comparing the node's dep_gen_id with the global generation id.
By applying DaveM's generation count optimization to current
implementation of BFS, we gain the following advantages:
- we save MAX_LOCKDEP_ENTRIES/8 bytes memory;
- we remove the bitmap_zero(bfs_accessed, MAX_LOCKDEP_ENTRIES);
in each BFS, which is very time-consuming since
MAX_LOCKDEP_ENTRIES may be very large.(16384UL)
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "David S. Miller" <davem@davemloft.net>
LKML-Reference: <1248274089-6358-1-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
spin_lock_nest_lock() allows to take many instances of the same
class, this can easily lead to overflow of MAX_LOCK_DEPTH.
To avoid this overflow, we'll stop accounting instances but
start reference counting the class in the held_lock structure.
[ We could maintain a list of instances, if we'd move the hlock
stuff into __lock_acquired(), but that would require
significant modifications to the current code. ]
We restrict this mode to spin_lock_nest_lock() only, because it
degrades the lockdep quality due to lost of instance.
For lockstat this means we don't track lock statistics for any
but the first lock in the series.
Currently nesting is limited to 11 bits because that was the
spare space available in held_lock. This yields a 2048
instances maximium.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Add a lockdep helper to validate that we indeed are the owner
of a lock.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
fixes a few comments and whitespaces that annoyed me.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Some cleanups of the lockdep code after the BFS series:
- Remove the last traces of the generation id
- Fixup comment style
- Move the bfs routines into lockdep.c
- Cleanup the bfs routines
[ tom.leiming@gmail.com: Fix crash ]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-11-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Add BFS statistics to the existing lockdep stats.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-10-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Also account the BFS memory usage.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
[ fix build for !PROVE_LOCKING ]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-9-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Implement lockdep_count_{for,back}ward using BFS.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-8-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Since the shortest lock dependencies' path may be obtained by BFS,
we print the shortest one by print_shortest_lock_dependencies().
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-7-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch uses BFS to implement find_usage_*wards(),which
was originally writen by DFS.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-6-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch uses BFS to implement check_noncircular() and
prints the generated shortest circle if exists.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-5-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
1,introduce match() to BFS in order to make it usable to
match different pattern;
2,also rename some functions to make them more suitable.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-4-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
1,replace %MAX_CIRCULAR_QUE_SIZE with &(MAX_CIRCULAR_QUE_SIZE-1)
since we define MAX_CIRCULAR_QUE_SIZE as power of 2;
2,use bitmap to mark if a lock is accessed in BFS in order to
clear it quickly, because we may search a graph many times.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-3-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Currently lockdep will print the 1st circle detected if it
exists when acquiring a new (next) lock.
This patch prints the shortest path from the next lock to be
acquired to the previous held lock if a circle is found.
The patch still uses the current method to check circle, and
once the circle is found, breadth-first search algorithem is
used to compute the shortest path from the next lock to the
previous lock in the forward lock dependency graph.
Printing the shortest path will shorten the dependency chain,
and make troubleshooting for possible circular locking easier.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1246201486-7308-2-git-send-email-tom.leiming@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Merge reason: tracing/core was on a .30-rc1 base and was missing out on
on a handful of tracing fixes present in .30-rc5-almost.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Steven Rostedt reported:
> OK, I think I figured this bug out. This is a lockdep issue with respect
> to tracepoints.
>
> The trace points in lockdep are called all the time. Outside the lockdep
> logic. But if lockdep were to trigger an error / warning (which this run
> did) we might be in trouble. For new locks, like the dentry->d_lock, that
> are created, they will not get a name:
>
> void lockdep_init_map(struct lockdep_map *lock, const char *name,
> struct lock_class_key *key, int subclass)
> {
> if (unlikely(!debug_locks))
> return;
>
> When a problem is found by lockdep, debug_locks becomes false. Thus we
> stop allocating names for locks. This dentry->d_lock I had, now has no
> name. Worse yet, I have CONFIG_DEBUG_VM set, that scrambles non
> initialized memory. Thus, when the trace point was hit, it had junk for
> the lock->name, and the machine crashed.
Ah, nice catch. I think we should put at least the name in regardless.
Ensure we at least initialize the trivial entries of the depmap so that
they can be relied upon, even when lockdep itself decided to pack up and
go home.
[ Impact: fix lock tracing after lockdep warnings. ]
Reported-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <1239954049.23397.4156.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: clean up
Create a sub directory in include/trace called events to keep the
trace point headers in their own separate directory. Only headers that
declare trace points should be defined in this directory.
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Neil Horman <nhorman@tuxdriver.com>
Cc: Zhao Lei <zhaolei@cn.fujitsu.com>
Cc: Eduard - Gabriel Munteanu <eduard.munteanu@linux360.ro>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
This patch lowers the number of places a developer must modify to add
new tracepoints. The current method to add a new tracepoint
into an existing system is to write the trace point macro in the
trace header with one of the macros TRACE_EVENT, TRACE_FORMAT or
DECLARE_TRACE, then they must add the same named item into the C file
with the macro DEFINE_TRACE(name) and then add the trace point.
This change cuts out the needing to add the DEFINE_TRACE(name).
Every file that uses the tracepoint must still include the trace/<type>.h
file, but the one C file must also add a define before the including
of that file.
#define CREATE_TRACE_POINTS
#include <trace/mytrace.h>
This will cause the trace/mytrace.h file to also produce the C code
necessary to implement the trace point.
Note, if more than one trace/<type>.h is used to create the C code
it is best to list them all together.
#define CREATE_TRACE_POINTS
#include <trace/foo.h>
#include <trace/bar.h>
#include <trace/fido.h>
Thanks to Mathieu Desnoyers and Christoph Hellwig for coming up with
the cleaner solution of the define above the includes over my first
design to have the C code include a "special" header.
This patch converts sched, irq and lockdep and skb to use this new
method.
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Neil Horman <nhorman@tuxdriver.com>
Cc: Zhao Lei <zhaolei@cn.fujitsu.com>
Cc: Eduard - Gabriel Munteanu <eduard.munteanu@linux360.ro>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
While trying to optimize the new lock on reiserfs to replace
the bkl, I find the lock tracing very useful though it lacks
something important for performance (and latency) instrumentation:
the time a task waits for a lock.
That's what this patch implements:
bash-4816 [000] 202.652815: lock_contended: lock_contended: &sb->s_type->i_mutex_key
bash-4816 [000] 202.652819: lock_acquired: &rq->lock (0.000 us)
<...>-4787 [000] 202.652825: lock_acquired: &rq->lock (0.000 us)
<...>-4787 [000] 202.652829: lock_acquired: &rq->lock (0.000 us)
bash-4816 [000] 202.652833: lock_acquired: &sb->s_type->i_mutex_key (16.005 us)
As shown above, the "lock acquired" field is followed by the time
it has been waiting for the lock. Usually, a lock contended entry
is followed by a near lock_acquired entry with a non-zero time waited.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <1238975373-15739-1-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Have a better idea about exactly which loc causes a lockdep
limit overflow. Often it's a bug or inefficiency in that
subsystem.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1237376327.5069.253.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Heiko reported that we grab the graph lock with irqs enabled.
Fix this by providng the same wrapper as all other lockdep entry
functions have.
Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Nick Piggin <npiggin@suse.de>
LKML-Reference: <1237544000.24626.52.camel@twins>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup
The atomic debug modifiers are already defined in
kernel/lockdep_internals.h.
Signed-off-by: David Rientjes <rientjes@google.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
LKML-Reference: <alpine.DEB.2.00.0903050222160.30401@chino.kir.corp.google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: clarify lockdep printk text
print_irq_inversion_bug() gets handed state strings of the form
"HARDIRQ", "SOFTIRQ", "RECLAIM_FS"
and appends "-irq-{un,}safe" to them, which is either redudant for *IRQ or
confusing in the RECLAIM_FS case.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1236175192.5330.7585.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
In the recent mark_lock_irq() rework a bug snuck in that would report the
state of write locks causing irq inversion under a read lock as a read
lock.
Fix this by masking the read bit of the state when validating write
dependencies.
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1236172646.5330.7450.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The __GFP_FS annotations fail to build with CONFIG_LOCKDEP=y,
CONFIG_PROVE_LOCKING=n, ammend that.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Arnd pointed out we have the stringify macro magic already in-kernel.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
CC: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
re-add some of the comments that got lost in the refactoring.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Now that we have nice numerical relations for the states, remove the macro
magics.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Now what its only two functions, they again look rather similar.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
s/HELD_OVER/ENABLED/g
so that its similar to the hard and soft-irq names.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
s/\(LOCKF\?_ENABLED_[^ ]*\)S\(_READ\)\?\>/\1\2/g
So that the USED_IN and ENABLED have the same names.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Here is another version, with the incremental patch rolled up, and
added reclaim context annotation to kswapd, and allocation tracing
to slab allocators (which may only ever reach the page allocator
in rare cases, so it is good to put annotations here too).
Haven't tested this version as such, but it should be getting closer
to merge worthy ;)
--
After noticing some code in mm/filemap.c accidentally perform a __GFP_FS
allocation when it should not have been, I thought it might be a good idea to
try to catch this kind of thing with lockdep.
I coded up a little idea that seems to work. Unfortunately the system has to
actually be in __GFP_FS page reclaim, then take the lock, before it will mark
it. But at least that might still be some orders of magnitude more common
(and more debuggable) than an actual deadlock condition, so we have some
improvement I hope (the concept is no less complete than discovery of a lock's
interrupt contexts).
I guess we could even do the same thing with __GFP_IO (normal reclaim), and
even GFP_NOIO locks too... but filesystems will have the most locks and fiddly
code paths, so let's start there and see how it goes.
It *seems* to work. I did a quick test.
=================================
[ INFO: inconsistent lock state ]
2.6.28-rc6-00007-ged31348-dirty #26
---------------------------------
inconsistent {in-reclaim-W} -> {ov-reclaim-W} usage.
modprobe/8526 [HC0[0]:SC0[0]:HE1:SE1] takes:
(testlock){--..}, at: [<ffffffffa0020055>] brd_init+0x55/0x216 [brd]
{in-reclaim-W} state was registered at:
[<ffffffff80267bdb>] __lock_acquire+0x75b/0x1a60
[<ffffffff80268f71>] lock_acquire+0x91/0xc0
[<ffffffff8070f0e1>] mutex_lock_nested+0xb1/0x310
[<ffffffffa002002b>] brd_init+0x2b/0x216 [brd]
[<ffffffff8020903b>] _stext+0x3b/0x170
[<ffffffff80272ebf>] sys_init_module+0xaf/0x1e0
[<ffffffff8020c3fb>] system_call_fastpath+0x16/0x1b
[<ffffffffffffffff>] 0xffffffffffffffff
irq event stamp: 3929
hardirqs last enabled at (3929): [<ffffffff8070f2b5>] mutex_lock_nested+0x285/0x310
hardirqs last disabled at (3928): [<ffffffff8070f089>] mutex_lock_nested+0x59/0x310
softirqs last enabled at (3732): [<ffffffff8061f623>] sk_filter+0x83/0xe0
softirqs last disabled at (3730): [<ffffffff8061f5b6>] sk_filter+0x16/0xe0
other info that might help us debug this:
1 lock held by modprobe/8526:
#0: (testlock){--..}, at: [<ffffffffa0020055>] brd_init+0x55/0x216 [brd]
stack backtrace:
Pid: 8526, comm: modprobe Not tainted 2.6.28-rc6-00007-ged31348-dirty #26
Call Trace:
[<ffffffff80265483>] print_usage_bug+0x193/0x1d0
[<ffffffff80266530>] mark_lock+0xaf0/0xca0
[<ffffffff80266735>] mark_held_locks+0x55/0xc0
[<ffffffffa0020000>] ? brd_init+0x0/0x216 [brd]
[<ffffffff802667ca>] trace_reclaim_fs+0x2a/0x60
[<ffffffff80285005>] __alloc_pages_internal+0x475/0x580
[<ffffffff8070f29e>] ? mutex_lock_nested+0x26e/0x310
[<ffffffffa0020000>] ? brd_init+0x0/0x216 [brd]
[<ffffffffa002006a>] brd_init+0x6a/0x216 [brd]
[<ffffffffa0020000>] ? brd_init+0x0/0x216 [brd]
[<ffffffff8020903b>] _stext+0x3b/0x170
[<ffffffff8070f8b9>] ? mutex_unlock+0x9/0x10
[<ffffffff8070f83d>] ? __mutex_unlock_slowpath+0x10d/0x180
[<ffffffff802669ec>] ? trace_hardirqs_on_caller+0x12c/0x190
[<ffffffff80272ebf>] sys_init_module+0xaf/0x1e0
[<ffffffff8020c3fb>] system_call_fastpath+0x16/0x1b
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (63 commits)
stacktrace: provide save_stack_trace_tsk() weak alias
rcu: provide RCU options on non-preempt architectures too
printk: fix discarding message when recursion_bug
futex: clean up futex_(un)lock_pi fault handling
"Tree RCU": scalable classic RCU implementation
futex: rename field in futex_q to clarify single waiter semantics
x86/swiotlb: add default swiotlb_arch_range_needs_mapping
x86/swiotlb: add default phys<->bus conversion
x86: unify pci iommu setup and allow swiotlb to compile for 32 bit
x86: add swiotlb allocation functions
swiotlb: consolidate swiotlb info message printing
swiotlb: support bouncing of HighMem pages
swiotlb: factor out copy to/from device
swiotlb: add arch hook to force mapping
swiotlb: allow architectures to override phys<->bus<->phys conversions
swiotlb: add comment where we handle the overflow of a dma mask on 32 bit
rcu: fix rcutorture behavior during reboot
resources: skip sanity check of busy resources
swiotlb: move some definitions to header
swiotlb: allow architectures to override swiotlb pool allocation
...
Fix up trivial conflicts in
arch/x86/kernel/Makefile
arch/x86/mm/init_32.c
include/linux/hardirq.h
as per Ingo's suggestions.
Impact: introduce new lockdep API
Allow to change a held lock's class. Basically the same as the existing
code to change a subclass therefore reuse all that.
The XFS code will be able to use this to annotate their inode locking.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: fix for lockdep and ftrace
The raw_local_irq_save/restore confuses lockdep. This patch
converts them to the local_irq_save/restore variants.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: fix build warning
this warning:
kernel/lockdep.c:584: warning: ‘print_lock_dependencies’ defined but not used
triggers because print_lock_dependencies() is only used if both
CONFIG_TRACE_IRQFLAGS and CONFIG_PROVE_LOCKING are enabled.
But adding #ifdefs is not an option here - it would spread out to 4-5
other helper functions and uglify the file. So mark this function
as __used - it's static and the compiler can eliminate it just fine.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: prettify /proc/lockdep_info
Just feel odd that not all lines of lockdep info are aligned.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: fix lockdep lock-api-caller output when irqsoff tracing is enabled
81d68a96 "ftrace: trace irq disabled critical timings" added wrappers around
trace_hardirqs_on/off_caller. However these functions use
__builtin_return_address(0) to figure out which function actually disabled
or enabled irqs. The result is that we save the ips of trace_hardirqs_on/off
instead of the real caller. Not very helpful.
However since the patch from Steven the ip already gets passed. So use that
and get rid of __builtin_return_address(0) in these two functions.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When we failed to get tasklist_lock eventually (count equals 0),
we should only print " ignoring it.\n", and not print
" locked it.\n" needlessly.
Signed-off-by: Qinghuang Feng <qhfeng.kernel@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
We currently only provide points that have to wait on contention, also
lists the points we have to wait for.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Fix bad contention counting in /proc/lock_stat.
/proc/lockstat tries to gather per-ip contention
statistics per-lock. This was failing due to
a garbage per-ip index selector being used.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Since f82b217e35 lockdep can output spurious
warnings related to hwirqs due to hardirq_off shrinkage from int to bit-sized
flag. Guard it with double negation to fix the warning.
Signed-off-by: Dmitry Baryshkov <dbaryshkov@gmail.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Use WARN() instead of a printk+WARN_ON() pair; this way the message
becomes part of the warning section for better reporting/collection.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
When we enable DEBUG_LOCK_ALLOC but do not enable PROVE_LOCKING and or
LOCK_STAT, lock_alloc() and lock_release() turn into nops, even though
we should be doing hlock checking (check=1).
This causes a false warning and a lockdep self-disable.
Rectify this.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Solve this by marking the classes as unused and not printing information
about the unused classes.
Reported-by: Eric Sesterhenn <snakebyte@gmx.de>
Signed-off-by: Rabin Vincent <rabin@rab.in>
Acked-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
On Fri, 2008-08-01 at 16:26 -0700, Linus Torvalds wrote:
> On Fri, 1 Aug 2008, David Miller wrote:
> >
> > Taking more than a few locks of the same class at once is bad
> > news and it's better to find an alternative method.
>
> It's not always wrong.
>
> If you can guarantee that anybody that takes more than one lock of a
> particular class will always take a single top-level lock _first_, then
> that's all good. You can obviously screw up and take the same lock _twice_
> (which will deadlock), but at least you cannot get into ABBA situations.
>
> So maybe the right thing to do is to just teach lockdep about "lock
> protection locks". That would have solved the multi-queue issues for
> networking too - all the actual network drivers would still have taken
> just their single queue lock, but the one case that needs to take all of
> them would have taken a separate top-level lock first.
>
> Never mind that the multi-queue locks were always taken in the same order:
> it's never wrong to just have some top-level serialization, and anybody
> who needs to take <n> locks might as well do <n+1>, because they sure as
> hell aren't going to be on _any_ fastpaths.
>
> So the simplest solution really sounds like just teaching lockdep about
> that one special case. It's not "nesting" exactly, although it's obviously
> related to it.
Do as Linus suggested. The lock protection lock is called nest_lock.
Note that we still have the MAX_LOCK_DEPTH (48) limit to consider, so anything
that spills that it still up shit creek.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
this can be used to reset a held lock's subclass, for arbitrary-depth
iterated data structures such as trees or lists which have per-node
locks.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When we traverse the graph, either forwards or backwards, we
are interested in whether a certain property exists somewhere
in a node reachable in the graph.
Therefore it is never necessary to traverse through a node more
than once to get a correct answer to the given query.
Take advantage of this property using a global ID counter so that we
need not clear all the markers in all the lock_class entries before
doing a traversal. A new ID is choosen when we start to traverse, and
we continue through a lock_class only if it's ID hasn't been marked
with the new value yet.
This short-circuiting is essential especially for high CPU count
systems. The scheduler has a runqueue per cpu, and needs to take
two runqueue locks at a time, which leads to long chains of
backwards and forwards subgraphs from these runqueue lock nodes.
Without the short-circuit implemented here, a graph traversal on
a runqueue lock can take up to (1 << (N - 1)) checks on a system
with N cpus.
For anything more than 16 cpus or so, lockdep will eventually bring
the machine to a complete standstill.
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'core/locking' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
lockdep: fix kernel/fork.c warning
lockdep: fix ftrace irq tracing false positive
lockdep: remove duplicate definition of STATIC_LOCKDEP_MAP_INIT
lockdep: add lock_class information to lock_chain and output it
lockdep: add lock_class information to lock_chain and output it
lockdep: output lock_class key instead of address for forward dependency output
__mutex_lock_common: use signal_pending_state()
mutex-debug: check mutex magic before owner
Fixed up conflict in kernel/fork.c manually
It is based on x86/master branch of git-x86 tree, and has been tested
on x86_64 platform.
ChangeLog:
v2:
- Enclosing proc file system related code into CONFIG_PROVE_LOCKING.
- Fix nr_chain_hlocks update code.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch records array of lock_class into lock_chain, and export
lock_chain information via /proc/lockdep_chains.
It is based on x86/master branch of git-x86 tree, and has been tested
on x86_64 platform.
Signed-off-by: Huang Ying <ying.huang@intel.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
With the introduction of ftrace, it is possible to recurse into
the lockdep functions via the mcount call. To prevent possible
lockups, updating the lockdep_recursion counter on grabbing the internal
lockdep_lock should prevent deadlocks.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch removes the "notrace" annotation from lockdep and adds the debugging
files in the kernel director to those that should not be compiled with
"-pg" mcount tracing.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Add notrace annotations to lockdep to keep ftrace from causing
recursive problems with lock tracing and debugging.
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Add each lock class to the all_lock_classes list when it is
first registered.
Previously, lock classes were added to all_lock_classes when
the lock class was first used. Since one of the uses of the
list is to find unused locks, this didn't work well.
Signed-off-by: Dale Farnsworth <dale@farnsworth.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
this patch extends the soft-lockup detector to automatically
detect hung TASK_UNINTERRUPTIBLE tasks. Such hung tasks are
printed the following way:
------------------>
INFO: task prctl:3042 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message
prctl D fd5e3793 0 3042 2997
f6050f38 00000046 00000001 fd5e3793 00000009 c06d8264 c06dae80 00000286
f6050f40 f6050f00 f7d34d90 f7d34fc8 c1e1be80 00000001 f6050000 00000000
f7e92d00 00000286 f6050f18 c0489d1a f6050f40 00006605 00000000 c0133a5b
Call Trace:
[<c04883a5>] schedule_timeout+0x6d/0x8b
[<c04883d8>] schedule_timeout_uninterruptible+0x15/0x17
[<c0133a76>] msleep+0x10/0x16
[<c0138974>] sys_prctl+0x30/0x1e2
[<c0104c52>] sysenter_past_esp+0x5f/0xa5
=======================
2 locks held by prctl/3042:
#0: (&sb->s_type->i_mutex_key#5){--..}, at: [<c0197d11>] do_fsync+0x38/0x7a
#1: (jbd_handle){--..}, at: [<c01ca3d2>] journal_start+0xc7/0xe9
<------------------
the current default timeout is 120 seconds. Such messages are printed
up to 10 times per bootup. If the system has crashed already then the
messages are not printed.
if lockdep is enabled then all held locks are printed as well.
this feature is a natural extension to the softlockup-detector (kernel
locked up without scheduling) and to the NMI watchdog (kernel locked up
with IRQs disabled).
[ Gautham R Shenoy <ego@in.ibm.com>: CPU hotplug fixes. ]
[ Andrew Morton <akpm@linux-foundation.org>: build warning fix. ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Michael Wu noticed in his lkml post at
http://marc.info/?l=linux-kernel&m=119396182726091&w=2
that certain wireless drivers ended up having their name in module
memory, which would then crash the kernel on module unload.
The patch he proposed was a bit clumsy in that it increased the size of
a lockdep entry significantly; the patch below tries another approach,
it checks, on module teardown, if the name of a class is in module space
and then zaps the class. This is very similar to what we already do
with keys that are in module space.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Lockdep, during self-test (when it was simulating double unlocks) was
sometimes unconditionally unlocking a spinlock when it had not been
locked. This won't work for ticket locks.
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
make cli/sti annotation warnings easier to interpret.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Torsten Kaiser wrote:
| static inline int in_range(const void *start, const void *addr, const void *end)
| {
| return addr >= start && addr <= end;
| }
| This will return true, if addr is in the range of start (including)
| to end (including).
|
| But debug_check_no_locks_freed() seems does:
| const void *mem_to = mem_from + mem_len
| -> mem_to is the last byte of the freed range, that fits in_range
| lock_from = (void *)hlock->instance;
| -> first byte of the lock
| lock_to = (void *)(hlock->instance + 1);
| -> first byte of the next lock, not last byte of the lock that is being checked!
|
| The test is:
| if (!in_range(mem_from, lock_from, mem_to) &&
| !in_range(mem_from, lock_to, mem_to))
| continue;
| So it tests, if the first byte of the lock is in the range that is freed ->OK
| And if the first byte of the *next* lock is in the range that is freed
| -> Not OK.
We can also simplify in_range checks, we need only 2 comparisons, not 4.
If the lock is not in memory range, it should be either at the left of range
or at the right.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
fix the oops that can be seen in:
http://bugzilla.kernel.org/attachment.cgi?id=13828&action=view
it is not safe to print the locks of running tasks.
(even with this fix we have a small race - but this is a debug
function after all.)
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Fix a typo in the __lock_acquire comment.
Signed-off-by: Gautham R Shenoy <ego@in.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The task_struct->pid member is going to be deprecated, so start
using the helpers (task_pid_nr/task_pid_vnr/task_pid_nr_ns) in
the kernel.
The first thing to start with is the pid, printed to dmesg - in
this case we may safely use task_pid_nr(). Besides, printks produce
more (much more) than a half of all the explicit pid usage.
[akpm@linux-foundation.org: git-drm went and changed lots of stuff]
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: Dave Airlie <airlied@linux.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In the following scenario:
code path 1:
my_function() -> lock(L1); ...; flush_workqueue(); ...
code path 2:
run_workqueue() -> my_work() -> ...; lock(L1); ...
you can get a deadlock when my_work() is queued or running
but my_function() has acquired L1 already.
This patch adds a pseudo-lock to each workqueue to make lockdep
warn about this scenario.
[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Acked-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Provide a check to validate that we do not hold any locks when switching
back to user-space.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
It is possible for the current->curr_chain_key to become inconsistent with the
current index if the chain fails to validate. The end result is that future
lock_acquire() operations may inadvertently fail to find a hit in the cache
resulting in a new node being added to the graph for every acquire.
Signed-off-by: Gregory Haskins <ghaskins@novell.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When I started adding support for lockdep to 64-bit powerpc, I got a
lockdep_init_error and with this patch was able to pinpoint why and where
to put lockdep_init(). Let's support this generally for others adding
lockdep support to their architecture.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
__acquire
|
lock _____
| \
| __contended
| |
| wait
| _______/
|/
|
__acquired
|
__release
|
unlock
We measure acquisition and contention bouncing.
This is done by recording a cpu stamp in each lock instance.
Contention bouncing requires the cpu stamp to be set on acquisition. Hence we
move __acquired into the generic path.
__acquired is then used to measure acquisition bouncing by comparing the
current cpu with the old stamp before replacing it.
__contended is used to measure contention bouncing (only useful for preemptable
locks)
[akpm@linux-foundation.org: cleanups]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
- update the copyright notices
- use the default hash function
- fix a thinko in a BUILD_BUG_ON
- add a WARN_ON to spot inconsitent naming
- fix a termination issue in /proc/lock_stat
[akpm@linux-foundation.org: cleanups]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Present all this fancy new lock statistics information:
*warning, _wide_ output ahead*
(output edited for purpose of brevity)
# cat /proc/lock_stat
lock_stat version 0.1
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
class name contentions waittime-min waittime-max waittime-total acquisitions holdtime-min holdtime-max holdtime-total
-----------------------------------------------------------------------------------------------------------------------------------------------------------------
&inode->i_mutex: 14458 6.57 398832.75 2469412.23 6768876 0.34 11398383.65 339410830.89
---------------
&inode->i_mutex 4486 [<ffffffff802a08f9>] pipe_wait+0x86/0x8d
&inode->i_mutex 0 [<ffffffff802a01e8>] pipe_write_fasync+0x29/0x5d
&inode->i_mutex 0 [<ffffffff802a0e18>] pipe_read+0x74/0x3a5
&inode->i_mutex 0 [<ffffffff802a1a6a>] do_lookup+0x81/0x1ae
.................................................................................................................................................................
&inode->i_data.tree_lock-W: 491 0.27 62.47 493.89 2477833 0.39 468.89 1146584.25
&inode->i_data.tree_lock-R: 65 0.44 4.27 48.78 26288792 0.36 184.62 10197458.24
--------------------------
&inode->i_data.tree_lock 46 [<ffffffff80277095>] __do_page_cache_readahead+0x69/0x24f
&inode->i_data.tree_lock 31 [<ffffffff8026f9fb>] add_to_page_cache+0x31/0xba
&inode->i_data.tree_lock 0 [<ffffffff802770ee>] __do_page_cache_readahead+0xc2/0x24f
&inode->i_data.tree_lock 0 [<ffffffff8026f6e4>] find_get_page+0x1a/0x58
.................................................................................................................................................................
proc_inum_idr.lock: 0 0.00 0.00 0.00 36 0.00 65.60 148.26
proc_subdir_lock: 0 0.00 0.00 0.00 3049859 0.00 106.81 1563212.42
shrinker_rwsem-W: 0 0.00 0.00 0.00 5 0.00 1.73 3.68
shrinker_rwsem-R: 0 0.00 0.00 0.00 633 2.57 246.57 10909.76
'contentions' and 'acquisitions' are the number of such events measured (since
the last reset). The waittime- and holdtime- (min, max, total) numbers are
presented in microseconds.
If there are any contention points, the lock class is presented in the block
format (as i_mutex and tree_lock above), otherwise a single line of output is
presented.
The output is sorted on absolute number of contentions (read + write), this
should get the worst offenders presented first, so that:
# grep : /proc/lock_stat | head
will quickly show who's bad.
The stats can be reset using:
# echo 0 > /proc/lock_stat
[bunk@stusta.de: make 2 functions static]
[akpm@linux-foundation.org: fix printk warning]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Jason Baron <jbaron@redhat.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Frederic Weisbecker
Tejun Heo
Peter Zijlstra
Mike Frysinger
Ming Lei
Ingo Molnar
Peter Zijlstra
Steven Rostedt
Linus Torvalds
David Rientjes
Nick Piggin
Li Zefan
Heiko Carstens
qinghuang feng
Zhu Yi
Joe Korty
Dmitry Baryshkov
Arjan van de Ven
Rabin Vincent
Dave Jones
David Miller
Huang, Ying
Steven Rostedt
Dale Farnsworth
Oleg Nesterov
Gautham R Shenoy
Pavel Emelyanov
Johannes Berg
Gregory Haskins