Because the operators and parentheses have a higher precedence
than the operand characters, which is normal, then we can't
use any string containing such special characters:
()=<>!&|
To get this support and also avoid ambiguous intepretation from
the parser or the human, we can do it using double quotes so that
we keep the usual languages habits.
Then after this patch you can still declare string condition like
before:
echo name == myname
But if you want to compare against a string containing an operator
character, you can use double quotes:
echo 'name == "&myname"'
Don't forget to include the whole expression into single quotes or
the double ones will be eaten by echo.
[ Impact: support strings with special characters for tracing filters ]
Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Zhaolei <zhaolei@cn.fujitsu.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
tracing/filters: support for filters of dynamic sized arrays
Currently the filtering infrastructure supports well the
numeric types and fixed sized array types.
But the recently added __string() field uses a specific
indirect offset mechanism which requires a specific
predicate. Until now it wasn't supported.
This patch adds this support and implies very few changes,
only a new predicate is needed, the management of this specific
field can be done through the usual string helpers in the
filtering infrastructure.
[ Impact: support all kinds of strings in the tracing filters ]
Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Zhaolei <zhaolei@cn.fujitsu.com> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Steven Rostedt [Thu, 7 May 2009 02:52:15 +0000 (22:52 -0400)]
tracing: add hierarchical enabling of events
With the current event directory, you can only enable individual events.
The file debugfs/tracing/set_event is used to be able to enable or
disable several events at once. But that can still be awkward.
This patch adds hierarchical enabling of events. That is, each directory
in debugfs/tracing/events has an "enable" file. This file can enable
or disable all events within the directory and below.
will enable all events, but then disable just the irq subsystem events.
When reading one of these enable files, there are four results:
0 - all events this file affects are disabled
1 - all events this file affects are enabled
X - there is a mixture of events enabled and disabled
? - this file does not affect any event
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Thu, 7 May 2009 01:54:09 +0000 (21:54 -0400)]
tracing: reset ring buffer when removing modules with events
Li Zefan found that there's a race using the event ids of events and
modules. When a module is loaded, an event id is incremented. We only
have 16 bits for event ids (65536) and there is a possible (but highly
unlikely) race that we could load and unload a module that registers
events so many times that the event id counter overflows.
When it overflows, it then restarts and goes looking for available
ids. An id is available if it was added by a module and released.
The race is if you have one module add an id, and then is removed.
Another module loaded can use that same event id. But if the old module
still had events in the ring buffer, the new module's call back would
get bogus data. At best (and most likely) the output would just be
garbage. But if the module for some reason used pointers (not recommended)
then this could potentially crash.
The safest thing to do is just reset the ring buffer if a module that
registered events is removed.
[ Impact: prevent unpredictable results of event id overflows ]
Reported-by: Li Zefan <lizf@cn.fujitsu.com>
LKML-Reference: <49FEAFD0.30106@cn.fujitsu.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Thu, 7 May 2009 01:20:39 +0000 (21:20 -0400)]
tracing: update sample with TRACE_INCLUDE_FILE
When creating trace events for ftrace, the header file with the TRACE_EVENT
macros must also have a macro called TRACE_SYSTEM. This macro describes
the name of the system the TRACE_EVENTS are defined for. It also doubles
as a way for the define_trace.h file to include the file that included
it.
For example:
in irq.h
#define TRACE_SYSTEM irq
[...]
#include <trace/define_trace.h>
The define_trace will use TRACE_SYSTEM to include irq.h. But if the name
of the trace system does not match the name of the trace header file,
one can override it with:
Which will change define_trace.h to inclued foo_trace.h instead of foo.h
The sample comments this, but people that use the sample code will more
likely use the code and not read the comments. This patch changes the
sample code to use the TRACE_INCLUDE_FILE to better show developers how to
use it.
[ Impact: make sample less confusing to developers ]
Reported-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 6 May 2009 22:36:59 +0000 (18:36 -0400)]
ring-buffer: change test to be more latency friendly
The ring buffer benchmark/test runs a producer for 10 seconds.
This is done with preemption and interrupts enabled. But if the kernel
is not compiled with CONFIG_PREEMPT, it basically stops everything
but interrupts for 10 seconds.
Although this is just a test and is not for production, this attribute
can be quite annoying. It can also spawn badness elsewhere.
This patch solves the issues by calling "cond_resched" when the system
is not compiled with CONFIG_PREEMPT. It also keeps track of the time
spent to call cond_resched such that it does not go against the
time calculations. That is, if the task schedules away, the time scheduled
out is removed from the test data. Note, this only works for non PREEMPT
because we do not know when the task is scheduled out if we have PREEMPT
enabled.
[ Impact: prevent test from stopping the world for 10 seconds ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 6 May 2009 19:30:07 +0000 (15:30 -0400)]
ring-buffer: make moving the tail page a separate function
Ingo Molnar thought the code would be cleaner if we used a function call
instead of a goto for moving the tail page. After implementing this,
it seems that gcc still inlines the result and the output is pretty much
the same. Since this is considered a cleaner approach, might as well
implement it.
[ Impact: code clean up ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 6 May 2009 16:40:51 +0000 (12:40 -0400)]
ring-buffer: check for failed allocation in ring buffer benchmark
The result of the allocation of the ring buffer read page in the
ring buffer bench mark does not check the return to see if a page
was actually allocated. This patch fixes that.
[ Impact: avoid NULL dereference ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 6 May 2009 14:26:45 +0000 (10:26 -0400)]
ring-buffer: remove unneeded conditional in rb_reserve_next
The code in __rb_reserve_next checks on page overflow if it is the
original commiter and then resets the page back to the original
setting. Although this is fine, and the code is correct, it is
a bit fragil. Some experimental work I did breaks it easily.
The better and more robust solution is to have all commiters that
overflow the page, simply subtract what they added.
[ Impact: more robust ring buffer account management ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
CC kernel/trace/trace_output.o
kernel/trace/trace_output.c: In function ‘register_ftrace_event’:
kernel/trace/trace_output.c:544: warning: ‘list’ may be used uninitialized in this function
Is wrong as 'list' is always initialized - but GCC (4.3.2) does not
recognize this relationship properly.
Work around the warning by initializing the variable to NULL.
This attempts to clarify names utilized during block I/O remap
operations (partition, volume manager). It correctly matches up the
/from/ information for both device & sector. This takes in the concept
from Kosaki Motohiro and extends it to include better naming for the
"device_from" field.
[ Impact: cleanup ]
Signed-off-by: Alan D. Brunelle <alan.brunelle@hp.com> Reviewed-by: Li Zefan <lizf@cn.fujitsu.com> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Jens Axboe <jens.axboe@oracle.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
LKML-Reference: <49FF4FAE.3000301@hp.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Li Zefan [Wed, 6 May 2009 02:33:45 +0000 (10:33 +0800)]
tracing/events: fix concurrent access to ftrace_events list
A module will add/remove its trace events when it gets loaded/unloaded, so
the ftrace_events list is not "const", and concurrent access needs to be
protected.
This patch thus fixes races between loading/unloding modules and read
'available_events' or read/write 'set_event', etc.
Steven Rostedt [Wed, 6 May 2009 02:47:18 +0000 (22:47 -0400)]
ring-buffer: add benchmark and tester
This patch adds code that can benchmark the ring buffer as well as
test it. This code can be compiled into the kernel (not recommended)
or as a module.
A separate ring buffer is used to not interfer with other users, like
ftrace. It creates a producer and a consumer (option to disable creation
of the consumer) and will run for 10 seconds, then sleep for 10 seconds
and then repeat.
While running, the producer will write 10 byte loads into the ring
buffer with just putting in the current CPU number. The reader will
continually try to read the buffer. The reader will alternate from reading
the buffer via event by event, or by full pages.
The output is a pr_info, thus it will fill up the syslogs.
Starting ring buffer hammer
End ring buffer hammer
Time: 9000349 (usecs)
Overruns: 12578640
Read: 5358440 (by events)
Entries: 0
Total: 17937080
Missed: 0
Hit: 17937080
Entries per millisec: 1993
501 ns per entry
Sleeping for 10 secs
Starting ring buffer hammer
End ring buffer hammer
Time: 9936350 (usecs)
Overruns: 0
Read: 28146644 (by pages)
Entries: 74
Total: 28146718
Missed: 0
Hit: 28146718
Entries per millisec: 2832
353 ns per entry
Sleeping for 10 secs
Time: is the time the test ran
Overruns: the number of events that were overwritten and not read
Read: the number of events read (either by pages or events)
Entries: the number of entries left in the buffer
(the by pages will only read full pages)
Total: Entries + Read + Overruns
Missed: the number of entries that failed to write
Hit: the number of entries that were written
The above example shows that it takes ~353 nanosecs per entry when
there is a reader, reading by pages (and no overruns)
The event by event reader slowed the producer down to 501 nanosecs.
[ Impact: see how changes to the ring buffer affect stability and performance ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 6 May 2009 01:16:11 +0000 (21:16 -0400)]
ring-buffer: move big if statement down
In the hot path of the ring buffer "__rb_reserve_next" there's a big
if statement that does not even return back to the work flow.
code;
if (cross to next page) {
[ lots of code ]
return;
}
more code;
The condition is even the unlikely path, although we do not denote it
with an unlikely because gcc is fine with it. The condition is true when
the write crosses a page boundary, and we need to start at a new page.
Having this if statement makes it hard to read, but calling another
function to do the work is also not appropriate, because we are using a lot
of variables that were set before the if statement, and we do not want to
send them as parameters.
This patch changes it to a goto:
code;
if (cross to next page)
goto next_page;
more code;
return;
next_page:
[ lots of code]
This makes the code easier to understand, and a bit more obvious.
The output from gcc is practically identical. For some reason, gcc decided
to use different registers when I switched it to a goto. But other than that,
the logic is the same.
[ Impact: easier to read code ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Tim Abbott [Fri, 1 May 2009 00:06:11 +0000 (20:06 -0400)]
ftrace: use .sched.text, not .text.sched in recordmcount.pl
The only references in the kernel to the .text.sched section are in
recordmcount.pl. Since the code it has is intended to be example code
it should refer to real kernel sections. So change it to .sched.text
instead.
[ Impact: consistency in comments ]
Signed-off-by: Tim Abbott <tabbott@mit.edu>
LKML-Reference: <1241136371-10768-1-git-send-email-tabbott@mit.edu> Acked-by: Sam Ravnborg <sam@ravnborg.org> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Fri, 1 May 2009 22:44:45 +0000 (18:44 -0400)]
ring-buffer: record page entries in buffer page descriptor
Currently, when the ring buffer writer overflows the buffer and must
write over non consumed data, we increment the overrun counter by
reading the entries on the page we are about to overwrite. This reads
the entries one by one.
This is not very effecient. This patch adds another entry counter
into each buffer page descriptor that keeps track of the number of
entries on the page. Now on overwrite, the overrun counter simply
needs to add the number of entries that is on the page it is about
to overwrite.
[ Impact: speed up of ring buffer in overwrite mode ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Fri, 1 May 2009 00:49:44 +0000 (20:49 -0400)]
ring-buffer: convert cpu buffer entries to local_t
The entries counter in cpu buffer is not atomic. It can be updated by
other interrupts or from another CPU (readers).
But making entries into "atomic_t" causes an atomic operation that can
hurt performance. Instead we convert it to a local_t that will increment
a counter with a local CPU atomic operation (if the arch supports it).
Instead of fighting with readers and overwrites that decrement the counter,
I added a "read" counter. Every time a reader reads an entry it is
incremented.
We already have a overrun counter and with that, the entries counter and
the read counter, we can calculate the total number of entries in the
buffer with:
(entries - overrun) - read
As long as the total number of entries in the ring buffer is less than
the word size, this will work. But since the entries counter was previously
a long, this is no different than what we had before.
Thanks to Andrew Morton for pointing out in the first version that
atomic_t does not replace unsigned long. I switched to atomic_long_t
even though it is signed. A negative count is most likely a bug.
[ Impact: keep accurate count of cpu buffer entries ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Where entries are the total number of data entries in the buffer.
overrun is the number of entries not consumed and were overwritten by
the writer.
commit overrun is the number of entries dropped due to nested writers
wrapping the buffer before the initial writer finished the commit.
nmi dropped is the number of entries dropped due to the ring buffer
lock being held when an nmi was going to write to the ring buffer.
Note, this field will be meaningless and will go away when the ring
buffer becomes lockless.
[ Impact: let userspace know what is happening in the ring buffers ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 29 Apr 2009 17:43:37 +0000 (13:43 -0400)]
ring-buffer: add counters for commit overrun and nmi dropped entries
The WARN_ON in the ring buffer when a commit is preempted and the
buffer is filled by preceding writes can happen in normal operations.
The WARN_ON makes it look like a bug, not to mention, because
it does not stop tracing and calls printk which can also recurse, this
is prone to deadlock (the WARN_ON is not in a position to recurse).
This patch removes the WARN_ON and replaces it with a counter that
can be retrieved by a tracer. This counter is called commit_overrun.
While at it, I added a nmi_dropped counter to count any time an NMI entry
is dropped because the NMI could not take the spinlock.
[ Impact: prevent deadlock by printing normal case warning ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Tue, 5 May 2009 05:15:24 +0000 (01:15 -0400)]
ring-buffer: export symbols
I'm adding a module to do a series of tests on the ring buffer as well
as benchmarks. This module needs to have more of the ring buffer API
exported. There's nothing wrong with reading the ring buffer from a
module.
[ Impact: allow modules to read pages from the ring buffer ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Jason Baron [Thu, 30 Apr 2009 17:29:42 +0000 (13:29 -0400)]
tracing: add new tracepoints docbook
Add tracepoint docbook. This will help us document and understand
what tracepoints are in the kernel. Since there are multiple
macros, and files that contain tracepoints.
"tracing: create automated trace defines" causes this compile error on s390,
as reported by Sachin Sant against linux-next:
kernel/built-in.o: In function `__do_softirq':
(.text+0x1c680): undefined reference to `__tracepoint_softirq_entry'
This happens because the definitions of the softirq tracepoints were moved
from kernel/softirq.c to kernel/irq/handle.c. Since s390 doesn't support
generic hardirqs handle.c doesn't get compiled and the definitions are
missing.
So move the tracepoints to softirq.c again.
[ Impact: fix build failure on s390 ]
Reported-by: Sachin Sant <sachinp@in.ibm.com> Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com
LKML-Reference: <20090429135139.5fac79b8@osiris.boeblingen.de.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Tom Zanussi [Tue, 28 Apr 2009 08:04:59 +0000 (03:04 -0500)]
tracing/filters: a better event parser
Replace the current event parser hack with a better one. Filters are
no longer specified predicate by predicate, but all at once and can
use parens and any of the following operators:
numeric fields:
==, !=, <, <=, >, >=
string fields:
==, !=
predicates can be combined with the logical operators:
&&, ||
examples:
"common_preempt_count > 4" > filter
"((sig >= 10 && sig < 15) || sig == 17) && comm != bash" > filter
If there was an error, the erroneous string along with an error
message can be seen by looking at the filter e.g.:
((sig >= 10 && sig < 15) || dsig == 17) && comm != bash
^
parse_error: Field not found
Currently the caret for an error always appears at the beginning of
the filter; a real position should be used, but the error message
should be useful even without it.
To clear a filter, '0' can be written to the filter file.
Filters can also be set or cleared for a complete subsystem by writing
the same filter as would be written to an individual event to the
filter file at the root of the subsytem. Note however, that if any
event in the subsystem lacks a field specified in the filter being
set, the set will fail and all filters in the subsytem are
automatically cleared. This change from the previous version was made
because using only the fields that happen to exist for a given event
would most likely result in a meaningless filter.
Because the logical operators are now implemented as predicates, the
maximum number of predicates in a filter was increased from 8 to 16.
Tom Zanussi [Tue, 28 Apr 2009 08:04:53 +0000 (03:04 -0500)]
tracing/filters: distinguish between signed and unsigned fields
The new filter comparison ops need to be able to distinguish between
signed and unsigned field types, so add an is_signed flag/param to the
event field struct/trace_define_fields(). Also define a simple macro,
is_signed_type() to determine the signedness at compile time, used in the
trace macros. If the is_signed_type() macro won't work with a specific
type, a new slightly modified version of TRACE_FIELD() called
TRACE_FIELD_SIGN(), allows the signedness to be set explicitly.
[ Impact: extend trace-filter code for new feature ]
Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com>
LKML-Reference: <1240905893.6416.120.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Tom Zanussi [Tue, 28 Apr 2009 08:04:47 +0000 (03:04 -0500)]
tracing/filters: move preds into event_filter object
Create a new event_filter object, and move the pred-related members
out of the call and subsystem objects and into the filter object - the
details of the filter implementation don't need to be exposed in the
call and subsystem in any case, and it will also help make the new
parser implementation a little cleaner.
[ Impact: refactor trace-filter code to prepare for new features ]
Signed-off-by: Tom Zanussi <tzanussi@gmail.com> Acked-by: Steven Rostedt <rostedt@goodmis.org> Cc: fweisbec@gmail.com Cc: Li Zefan <lizf@cn.fujitsu.com>
LKML-Reference: <1240905887.6416.119.camel@tropicana> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Stuart Bennett [Tue, 28 Apr 2009 19:17:50 +0000 (20:17 +0100)]
tracing: x86, mmiotrace: refactor clearing/restore of page presence
* change function names to clear_* from set_*: in reality we only clear
and restore page presence, and never unconditionally set present.
Using clear_*({true, false}, ...) is therefore more honest than
set_*({false, true}, ...)
* upgrade presence storage to pteval_t: doing user-space tracing will
require saving and manipulation of the _PAGE_PROTNONE bit, in addition
to the existing _PAGE_PRESENT changes, and having multiple bools stored
and passed around does not seem optimal
[ Impact: refactor, clean up mmiotrace code ]
Signed-off-by: Stuart Bennett <stuart@freedesktop.org> Acked-by: Pekka Paalanen <pq@iki.fi> Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <1240946271-7083-4-git-send-email-stuart@freedesktop.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Steven Rostedt [Wed, 29 Apr 2009 04:26:30 +0000 (00:26 -0400)]
tracing: have splice only copy full pages
Splice works with pages, it is much more effecient to use an entire
page than to copy bits over several pages.
Using logdev to trace the internals of the splice mechanism, I was
able to see that splice can be very aggressive. When tracing is
occurring, and the reader caught up to the writer, and the writer
is on the reader page, the reader will copy what is there into the
splice page. Splice may iterate over several pages and if the
writer is still writing to the page, the reader will keep copying
bits to new pages to pass to userspace.
This patch changes it to only pass data to userspace if the page
is full (the writer has left the page). This has a small side effect
that splice can not read a partial page, and must wait for the
page to fill. This should not be an issue. If tracing has stopped,
then a use of "read" will still read all of the page.
[ Impact: better performance for ring buffer splice code ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 29 Apr 2009 04:23:13 +0000 (00:23 -0400)]
tracing: only add splice page if entries exist
The splice code allocates a page even when the ring buffer is empty.
It detects the ring buffer being empty when it it fails to copy
anything from the ring buffer into the page.
This patch adds a check to see if there is anything in the ring buffer
before allocating a page.
Thanks to logdev for letting me trace the tracer to find this.
[ Impact: speed up due to removing unnecessary allocation ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 29 Apr 2009 04:16:21 +0000 (00:16 -0400)]
tracing: fix ref count in splice pages
The pages allocated for the splice binary buffer did not initialize
the ref count correctly. This caused pages not to be freed and causes
a drastic memory leak.
Thanks to logdev I was able to trace the tracer to find where the leak
was.
[ Impact: stop memory leak when using splice ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Tue, 28 Apr 2009 15:39:34 +0000 (11:39 -0400)]
tracing: convert ftrace_dump spinlocks to raw
ftrace_dump is used for printing out the contents of the ftrace ring buffer
to the console on failure. Currently it uses a spinlock to synchronize
the output from multiple failures on different CPUs. This spin lock
currently is a normal spinlock and can cause issues with lockdep and
lock tracing.
This patch converts it to raw since it is for error handling only.
The lock is local to the ftrace_dump and is not used by any other
infrastructure.
[ Impact: prevent ftrace_dump from locking up by internal tracing ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Sat, 25 Apr 2009 03:11:22 +0000 (23:11 -0400)]
tracing/events: make modules have their own file_operations structure
For proper module reference counting, the file_operations that modules use
must have the "owner" field set to the module. Unfortunately, the trace events
use share file_operations. The same file_operations are used by all both
kernel core and all modules.
This patch makes the modules allocate their own file_operations and
copies the functions from the core kernel. This allows those file
operations to be owned by the module.
Care is taken to free this code on module unload.
Thanks to Greg KH for reminding me that file_operations must be owned
by the module to have reference counting take place.
Steven Rostedt [Fri, 24 Apr 2009 16:20:52 +0000 (12:20 -0400)]
tracing/events: reuse trace event ids after overflow
With modules being able to add trace events, and the max trace event
counter is 16 bits (65536) we can overflow the counter easily
with a simple while loop adding and removing modules that contain
trace events.
This patch links together the registered trace events and on overflow
searches for available trace event ids. It will still fail if
over 65536 events are registered, but considering that a typical
kernel only has 22000 functions, 65000 events should be sufficient.
Reported-by: Li Zefan <lizf@cn.fujitsu.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Lai Jiangshan [Fri, 24 Apr 2009 03:27:05 +0000 (11:27 +0800)]
ring_buffer: compressed event header
RB_MAX_SMALL_DATA = 28bytes is too small for most tracers, it wastes
an 'u32' to save the actually length for events which data size > 28.
This fix uses compressed event header and enlarges RB_MAX_SMALL_DATA.
[ Impact: saves about 0%-12.5%(depends on tracer) memory in ring_buffer ]
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
LKML-Reference: <49F13189.3090000@cn.fujitsu.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Fri, 24 Apr 2009 03:26:18 +0000 (23:26 -0400)]
tracing: fix cut and paste macro error
In case a module uses the TRACE_EVENT macro for creating automated
events in ftrace, it may choose to use a different file name
than the defined system name, or choose to use a different path than
the default "include/trace/events" include path.
If this is done, then before including trace/define_trace.h the
header would define either "TRACE_INCLUDE_FILE" for the file
name or "TRACE_INCLUDE_PATH" for the include path.
If it does not define these, then the define_trace.h defines them
instead. If define trace defines them, then define_trace.h should
also undefine them before exiting. To do this a macro is used
to note this:
Chris Wright [Thu, 23 Apr 2009 17:21:38 +0000 (10:21 -0700)]
x86: use native register access for native tlb flushing
currently these are paravirtulaized, doesn't appear any callers rely on
this (no pv_ops backends are using native_tlb and overriding cr3/4
access).
[ Impact: fix lockdep warning with paravirt and function tracer ]
Signed-off-by: Chris Wright <chrisw@sous-sol.org>
LKML-Reference: <20090423172138.GR3036@sequoia.sous-sol.org> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Thu, 26 Mar 2009 15:43:36 +0000 (11:43 -0400)]
tracing: add size checks for exported ftrace internal structures
The events exported by TRACE_EVENT are automated and are guaranteed
to be correct when used.
The internal ftrace structures on the other hand are more manually
exported. These require the ftrace maintainer to make sure they
are up to date.
This patch adds a size check to help flag when a type changes in
an internal ftrace data structure, and the update needs to be reflected
in the export.
If a export is incorrect, then the only harm is that the user space
tools will not know how to correctly read the internal structures of
ftrace.
[ Impact: help prevent inconsistent ftrace format print outs ]
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Steven Rostedt [Thu, 26 Mar 2009 15:03:29 +0000 (11:03 -0400)]
tracing: increase size of number of possible events
With the new event tracing registration, we must increase the number
of events that can be registered. Currently the type field is only
one byte, which leaves us only 256 possible events.
Since we do not save the CPU number in the tracer anymore (it is determined
by the per cpu ring buffer that is used) we have an extra byte to use.
This patch increases the size of type from 1 byte (256 events) to
2 bytes (65,536 events).
It also adds a WARN_ON_ONCE if we exceed that limit.
[ Impact: allow more than 255 events ]
Signed-off-by: Steven Rostedt <srostedt@redhat.com>
Steven Rostedt [Thu, 26 Mar 2009 14:25:24 +0000 (10:25 -0400)]
tracing/wakeup: move access to wakeup_cpu into spinlock
The code had the following outside the lock:
if (next != wakeup_task)
return;
pc = preempt_count();
/* The task we are waiting for is waking up */
data = wakeup_trace->data[wakeup_cpu];
On initialization, wakeup_task is NULL and wakeup_cpu -1. This code
is not under a lock. If wakeup_task is set on another CPU as that
task is waking up, we can see the wakeup_task before wakeup_cpu is
set. If we read wakeup_cpu while it is still -1 then we will have
a bad data pointer.
This patch moves the reading of wakeup_cpu within the protection of
the spinlock used to protect the writing of wakeup_cpu and wakeup_task.
[ Impact: remove possible race causing invalid pointer dereference ]
Reported-by: Maneesh Soni <maneesh@in.ibm.com> Signed-off-by: Steven Rostedt <srostedt@redhat.com>
tracing/lock: provide lock_acquired event support for dynamic size string
Now that we can support the dynamic sized string, make the lock tracing
able to use it, making it safe against modules removal and consuming
the right amount of memory needed for each lock name
Changes in v2:
adapt to the __ending_string() updates and the opening_string() removal.
[ Impact: protect lock tracer against module removal ]
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Steven Rostedt <rostedt@goodmis.org>
tracing/events: provide string with undefined size support
This patch provides the support for dynamic size strings on
event tracing.
The key concept is to use a structure with an ending char array field of
undefined size and use such ability to allocate the minimal size on the
ring buffer to make one or more string entries fit inside, as opposite
to a fixed length strings with upper bound.
The strings themselves are represented using fields which have an offset
value from the beginning of the entry.
This patch provides three new macros:
__string(item, src)
This one declares a string to the structure inside TP_STRUCT__entry.
You need to provide the name of the string field and the source that will
be copied inside.
This will also add the dynamic size of the string needed for the ring
buffer entry allocation.
A stack allocated structure is used to temporarily store the offset
of each strings, avoiding double calls to strlen() on each event
insertion.
__get_str(field)
This one will give you a pointer to the string you have created. This
is an abstract helper to resolve the absolute address given the field
name which is a relative address from the beginning of the trace_structure.
__assign_str(dst, src)
Use this macro to automatically perform the string copy from src to
dst. src must be a variable to assign and dst is the name of a __string
field.
Of course you can mix-up any __field or __array inside this
TRACE_EVENT. The position of the __string or __assign_str
doesn't matter.
Changes in v2:
Address the suggestion of Steven Rostedt: drop the opening_string() macro
and redefine __ending_string() to get the size of the string to be copied
instead of overwritting the whole ring buffer allocation.
Changes in v3:
Address other suggestions of Steven Rostedt and Peter Zijlstra with
some changes: drop the __ending_string and the need to have only one
string field.
Use offsets instead of absolute addresses.
[ Impact: allow more compact memory usage for string tracing ]
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Li Zefan [Wed, 22 Apr 2009 08:53:34 +0000 (16:53 +0800)]
tracing/events: make struct trace_entry->type to be int type
struct trace_entry->type is unsigned char, while trace event's id is
int type, thus for a event with id >= 256, it's entry->type is cast
to (id % 256), and then we can't see the trace output of this event.
# insmod trace-events-sample.ko
# echo foo_bar > /mnt/tracing/set_event
# cat /debug/tracing/events/trace-events-sample/foo_bar/id
256
# cat /mnt/tracing/trace_pipe
<...>-3548 [001] 215.091142: Unknown type 0
<...>-3548 [001] 216.089207: Unknown type 0
<...>-3548 [001] 217.087271: Unknown type 0
<...>-3548 [001] 218.085332: Unknown type 0
[ Impact: fix output for trace events with id >= 256 ]
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <49EEDB0E.5070207@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
Steven Rostedt [Tue, 21 Apr 2009 13:41:26 +0000 (09:41 -0400)]
ring-buffer: only warn on wrap if buffer is bigger than two pages
On boot up, to save memory, ftrace allocates the minimum buffer
which is two pages. Ftrace also goes through a series of tests
(when configured) on boot up. These tests can fill up a page within
a single interrupt.
The ring buffer also has a WARN_ON when it detects that the buffer was
completely filled within a single commit (other commits are allowed to
be nested).
Combine the small buffer on start up, with the tests that can fill more
than a single page within an interrupt, this can trigger the WARN_ON.
This patch makes the WARN_ON only happen when the ring buffer consists
of more than two pages.
[ Impact: prevent false WARN_ON in ftrace startup tests ]
Steven Rostedt [Mon, 20 Apr 2009 22:16:44 +0000 (18:16 -0400)]
tracing: use nowakeup version of commit for function event trace tests
The startup tests for the event tracer also runs with the function
tracer enabled. The "wakeup" version of the trace commit was used
which can grab spinlocks. If a task was preempted by an NMI
that called a function being traced, it could deadlock due to the
function tracer trying to grab the same lock.
Thanks to Frederic Weisbecker for pointing out where the bug was.
Steven Rostedt [Mon, 20 Apr 2009 20:16:11 +0000 (16:16 -0400)]
tracing: use recursive counter over irq level
Althought using the irq level (hardirq_count, softirq_count and in_nmi)
was nice to detect bad recursion right away, but since the counters are
not atomically updated with respect to the interrupts, the function tracer
might trigger the test from an interrupt handler before the hardirq_count
is updated. This will trigger a false warning.
This patch converts the recursive detection to a simple counter.
If the depth is greater than 16 then the recursive detection will trigger.
16 is more than enough for any nested interrupts.
Steven Rostedt [Mon, 20 Apr 2009 17:32:44 +0000 (13:32 -0400)]
tracing: remove recursive test from ring_buffer_event_discard
The ring_buffer_event_discard is not tied to ring_buffer_lock_reserve.
It can be called inside or outside the reserve/commit. Even if it
is called inside the reserve/commit the commit part must also be called.
Only ring_buffer_discard_commit can be used as a replacement for
ring_buffer_unlock_commit.
This patch removes the trace_recursive_unlock from ring_buffer_event_discard
since it would be the wrong place to do so.
[Impact: prevent breakage in trace recursive testing ]
Cc: Frederic Weisbecker <fweisbec@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Mon, 20 Apr 2009 17:24:21 +0000 (13:24 -0400)]
tracing: fix recursive test level calculation
The recursive tests to detect same level recursion in the ring buffers
did not account for the hard/softirq_counts to be shifted. Thus the
numbers could be larger than then mask to be tested.
This patch includes the shift for the calculation of the irq depth.
[ Impact: stop false positives in trace recursion detection ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Mon, 20 Apr 2009 14:59:34 +0000 (10:59 -0400)]
tracing: rename EVENT_TRACER config to ENABLE_EVENT_TRACING
Currently we have two configs: EVENT_TRACING and EVENT_TRACER.
All tracers enable EVENT_TRACING. The EVENT_TRACER is only a
convenience to enable the EVENT_TRACING when no other tracers
are enabled.
The names EVENT_TRACER and EVENT_TRACING are too similar and confusing.
This patch renames EVENT_TRACER to ENABLE_EVENT_TRACING to be more
appropriate to what it actually does, as well as add a comment in
the help menu to explain the option's purpose.
[ Impact: rename config option to reduce confusion ]
Reported-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Mon, 20 Apr 2009 14:47:36 +0000 (10:47 -0400)]
tracing: create menuconfig for tracing infrastructure
During testing we often use randconfig to test various kernels.
The current configuration set up does not give an easy way to disable
all tracing with a single config. The case where randconfig would
test all tracing disabled is very unlikely.
This patch adds a config option to enable or disable all tracing.
It is hooked into the tracing menu just like other submenus are done.
[ Impact: allow randconfig to easily produce all traces disabled ]
Reported-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Mon, 20 Apr 2009 14:27:58 +0000 (10:27 -0400)]
tracing: change branch profiling to a choice selection
This patch makes the branch profiling into a choice selection:
None - no branch profiling
likely/unlikely - only profile likely/unlikely branches
all - profile all branches
The all profiler will also enable the likely/unlikely branches.
This does not change the way the profiler works or the dependencies
between the profilers.
What this patch does, is keep the branch profiling from being selected
by an allyesconfig make. The branch profiler is very intrusive and
it is known to break various architecture builds when selected as an
allyesconfig.
[ Impact: prevent branch profiler from being selected in allyesconfig ]
Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Reported-by: Al Viro <viro@zeniv.linux.org.uk> Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Reported-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
tracing/ring-buffer: Add unlock recursion protection on discard
The pair of helpers trace_recursive_lock() and trace_recursive_unlock()
have been introduced recently to provide generic tracing recursion
protection.
They are used in a symetric way:
- trace_recursive_lock() on buffer reserve
- trace_recursive_unlock() on buffer commit
However sometimes, we don't commit but discard on entry
to the buffer, ie: in case of filter checking.
Then we must also unlock the recursion protection on discard time,
otherwise the tracing gets definitely deactivated and a warning
is raised spuriously, such as:
tracing/core: Add current context on tracing recursion warning
In case of tracing recursion detection, we only get the stacktrace.
But the current context may be very useful to debug the issue.
This patch adds the softirq/hardirq/nmi context with the warning
using lockdep context display to have a familiar output.
v2: Use printk_once()
v3: drop {hardirq,softirq}_context which depend on lockdep,
only keep what is part of current->trace_recursion,
sufficient to debug the warning source.
[ Impact: print context necessary to debug recursion ]
Steven Rostedt [Fri, 17 Apr 2009 01:41:52 +0000 (21:41 -0400)]
tracing: add same level recursion detection
The tracing infrastructure allows for recursion. That is, an interrupt
may interrupt the act of tracing an event, and that interrupt may very well
perform its own trace. This is a recursive trace, and is fine to do.
The problem arises when there is a bug, and the utility doing the trace
calls something that recurses back into the tracer. This recursion is not
caused by an external event like an interrupt, but by code that is not
expected to recurse. The result could be a lockup.
This patch adds a bitmask to the task structure that keeps track
of the trace recursion. To find the interrupt depth, the following
algorithm is used:
Here, level will be the depth of interrutps and softirqs, and even handles
the nmi. Then the corresponding bit is set in the recursion bitmask.
If the bit was already set, we know we had a recursion at the same level
and we warn about it and fail the writing to the buffer.
After the data has been committed to the buffer, we clear the bit.
No atomics are needed. The only races are with interrupts and they reset
the bitmask before returning anywy.
[ Impact: detect same irq level trace recursion ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Fri, 17 Apr 2009 17:02:22 +0000 (13:02 -0400)]
tracing/events: enable code with EVENT_TRACING not EVENT_TRACER
The CONFIG_EVENT_TRACER is the way to turn on event tracing when no
other tracing has been configured. All code to get enabled should
depend on CONFIG_EVENT_TRACING. That is what is enabled when TRACING
(or CONFIG_EVENT_TRACER) is selected.
This patch enables the include/trace/ftrace.h file when
CONFIG_EVENT_TRACING is enabled.
[ Impact: fix warning in event tracer selftest ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
tracing: add saved_cmdlines file to show cached task comms
Export the cached task comms to userspace. This allows user apps to translate
the pids from a trace into their respective task command lines.
[ Impact: let userspace apps reading binary buffer know comm's of pids ]
Signed-off-by: Avadh Patel <avadh4all@gmail.com>
[ added error checking and use of buf pointer to index file_buf ] Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 15 Apr 2009 20:53:47 +0000 (16:53 -0400)]
tracing/events/ring-buffer: expose format of ring buffer headers to users
Currently, every thing needed to read the binary output from the
ring buffers is available, with the exception of the way the ring
buffers handles itself internally.
This patch creates two special files in the debugfs/tracing/events
directory:
Steven Rostedt [Wed, 15 Apr 2009 17:36:40 +0000 (13:36 -0400)]
tracing/events: add startup tests for events
As events start to become popular, and the new way to add tracing
infrastructure into ftrace, it is important to catch any problems
that might happen with a mistake in the TRACE_EVENT macro.
This patch introduces a startup self test on the registered trace
events. Note, it can only do a generic test, any type of testing that
needs more involement is needed to be implemented by the tracepoint
creators.
The test goes down one by one enabling a trace point and running
some random tasks (random in the sense that I just made them up).
Those tasks are creating threads, grabbing mutexes and spinlocks
and using workqueues.
After testing each event individually, it does the same test after
enabling each system of trace points. Like sched, irq, lockdep.
Then finally it enables all tracepoints and performs the tasks again.
The output to the console on bootup will look like this when everything
works:
Running tests on trace events:
Testing event kfree_skb: OK
Testing event kmalloc: OK
Testing event kmem_cache_alloc: OK
Testing event kmalloc_node: OK
Testing event kmem_cache_alloc_node: OK
Testing event kfree: OK
Testing event kmem_cache_free: OK
Testing event irq_handler_exit: OK
Testing event irq_handler_entry: OK
Testing event softirq_entry: OK
Testing event softirq_exit: OK
Testing event lock_acquire: OK
Testing event lock_release: OK
Testing event sched_kthread_stop: OK
Testing event sched_kthread_stop_ret: OK
Testing event sched_wait_task: OK
Testing event sched_wakeup: OK
Testing event sched_wakeup_new: OK
Testing event sched_switch: OK
Testing event sched_migrate_task: OK
Testing event sched_process_free: OK
Testing event sched_process_exit: OK
Testing event sched_process_wait: OK
Testing event sched_process_fork: OK
Testing event sched_signal_send: OK
Running tests on trace event systems:
Testing event system skb: OK
Testing event system kmem: OK
Testing event system irq: OK
Testing event system lockdep: OK
Testing event system sched: OK
Running tests on all trace events:
Testing all events: OK
[ folded in:
tracing: add #include <linux/delay.h> to fix build failure in test_work()
This build failure occured on a few rare configs:
kernel/trace/trace_events.c: In function ‘test_work’:
kernel/trace/trace_events.c:975: error: implicit declaration of function ‘udelay’
kernel/trace/trace_events.c:980: error: implicit declaration of function ‘msleep’
delay.h is included in way too many other headers, hiding cases
where new usage is added without header inclusion.
[ Impact: build fix ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
]
[ Impact: add event tracer self-tests ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Wed, 15 Apr 2009 17:24:06 +0000 (13:24 -0400)]
ftrace: use module notifier for function tracer
The hooks in the module code for the function tracer must be called
before any of that module code runs. The function tracer hooks
modify the module (replacing calls to mcount to nops). If the code
is executed while the change occurs, then the CPU can take a GPF.
To handle the above with a bit of paranoia, I originally implemented
the hooks as calls directly from the module code.
After examining the notifier calls, it looks as though the start up
notify is called before any of the module's code is executed. This makes
the use of the notify safe with ftrace.
Only the startup notify is required to be "safe". The shutdown simply
removes the entries from the ftrace function list, and does not modify
any code.
This change has another benefit. It removes a issue with a reverse dependency
in the mutexes of ftrace_lock and module_mutex.
[ Impact: fix lock dependency bug, cleanup ]
Cc: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Li Zefan [Wed, 15 Apr 2009 03:02:56 +0000 (11:02 +0800)]
blktrace: fix context-info when mixed-using blk tracer and trace events
When current tracer is set to blk tracer, TRACE_ITER_CONTEXT_INFO is
unset, but actually context-info is printed:
pdflush-431 [000] 821.181576: 8,0 P N [pdflush]
And then if we enable TRACE_ITER_CONTEXT_INFO:
# echo context-info > trace_options
We'll see context-info printed twice. What's worse, when we use blk
tracer and trace events at the same time, we'll see no context-info
for trace events at all:
jbd2_commit_logging: dev dm-0:8 transaction 333227
jbd2_end_commit: dev dm-0:8 transaction 333227 head 332814
rm-25433 [001] 9578.307485: 8,18 m N cfq25433 slice expired t=0
rm-25433 [001] 9578.307486: 8,18 m N cfq25433 put_queue
This patch adds blk_tracer->set_flags(), and context-info flag is unset
only when we set the output to classic mode.
Note after this patch, one should unset context-info explicitly if he
wants to get binary output that can be parsed by blkparse:
Steven Rostedt [Tue, 14 Apr 2009 23:39:12 +0000 (19:39 -0400)]
tracing/events: move trace point headers into include/trace/events
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>
Steven Rostedt [Tue, 14 Apr 2009 22:49:38 +0000 (18:49 -0400)]
tracing/events: fix lockdep system name
Impact: fix compile error of lockdep event tracer
Ingo Molnar pointed out that the system name for the lockdep tracer was "lock"
which is used to include the event trace file name. It should be "lockdep"
Reported-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Fri, 10 Apr 2009 18:53:50 +0000 (14:53 -0400)]
tracing/events: add support for modules to TRACE_EVENT
Impact: allow modules to add TRACE_EVENTS on load
This patch adds the final hooks to allow modules to use the TRACE_EVENT
macro. A notifier and a data structure are used to link the TRACE_EVENTs
defined in the module to connect them with the ftrace event tracing system.
It also adds the necessary automated clean ups to the trace events when a
module is removed.
Cc: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Fri, 10 Apr 2009 22:12:50 +0000 (18:12 -0400)]
tracing/events: add export symbols for trace events in modules
Impact: let modules add trace events
The trace event code requires some functions to be exported to allow
modules to use TRACE_EVENT. This patch adds EXPORT_SYMBOL_GPL to the
necessary functions.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Fri, 10 Apr 2009 17:52:20 +0000 (13:52 -0400)]
tracing/events: convert event call sites to use a link list
Impact: makes it possible to define events in modules
The events are created by reading down the section that they are linked
in by the macros. But this is not scalable to modules. This patch converts
the manipulations to use a global link list, and on boot up it adds
the items in the section to the list.
This change will allow modules to add their tracing events to the list as
well.
Note, this change alone does not permit modules to use the TRACE_EVENT macros,
but the change is needed for them to eventually do so.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Steven Rostedt [Mon, 13 Apr 2009 16:25:37 +0000 (12:25 -0400)]
tracing/events: move the ftrace event tracing code to core
This patch moves the ftrace creation into include/trace/ftrace.h and
simplifies the work of developers in adding new tracepoints.
Just the act of creating the trace points in include/trace and including
define_trace.h will create the events in the debugfs/tracing/events
directory.
This patch removes the need of include/trace/trace_events.h
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
projects / linux-beck.git / log