Oleg Nesterov [Sun, 27 Jan 2013 17:36:24 +0000 (18:36 +0100)]
uprobes/tracing: Fully initialize uprobe_trace_consumer before uprobe_register()
probe_event_enable() does uprobe_register() and only after that sets
utc->tu and tu->consumer/flags. This can race with uprobe_dispatcher()
which can miss these assignments or see them out of order. Nothing
really bad can happen, but this doesn't look clean/safe.
And this does not allow to use uprobe_consumer->filter() we are going
to add, it is called by uprobe_register() and it needs utc->tu.
Change this code to initialize everything before uprobe_register(), and
reset tu->consumer/flags if it fails. We can't race with event_disable(),
the caller holds event_mutex, and if we could the code would be wrong
anyway.
In fact I think uprobe_trace_consumer should die, it buys nothing but
complicates the code. We can simply add uprobe_consumer into trace_uprobe.
Josh Stone [Sun, 13 Jan 2013 18:03:34 +0000 (19:03 +0100)]
uprobes: Add exports for module use
The original pull message for uprobes (commit 654443e2) noted:
This tree includes uprobes support in 'perf probe' - but SystemTap
(and other tools) can take advantage of user probe points as well.
In order to actually be usable in module-based tools like SystemTap, the
interface needs to be exported. This patch first adds the obvious
exports for uprobe_register and uprobe_unregister. Then it also adds
one for task_user_regset_view, which is necessary to get the correct
state of userspace registers.
Signed-off-by: Josh Stone <jistone@redhat.com> Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Oleg Nesterov [Mon, 31 Dec 2012 17:20:42 +0000 (18:20 +0100)]
uprobes: Do not allocate current->utask unnecessary
handle_swbp() does get_utask() before can_skip_sstep() for no reason,
we do not need ->utask if can_skip_sstep() succeeds.
Move get_utask() to pre_ssout() who actually starts to use it. Move
the initialization of utask->active_uprobe/state as well. This way
the whole initialization is consolidated in pre_ssout().
Signed-off-by: Oleg Nesterov <oleg@redhat.com> Acked-by: Anton Arapov <anton@redhat.com>
Oleg Nesterov [Mon, 31 Dec 2012 15:39:49 +0000 (16:39 +0100)]
uprobes: Fold xol_alloc_area() into get_xol_area()
Currently only xol_get_insn_slot() does get_xol_area() + xol_alloc_area(),
but this will have more users and we do not want to copy-and-paste this
code. This patch simply moves xol_alloc_area() into get_xol_area() to
simplify the current and future code.
Oleg Nesterov [Sun, 30 Dec 2012 16:40:39 +0000 (17:40 +0100)]
uprobes: Move alloc_page() from xol_add_vma() to xol_alloc_area()
Move alloc_page() from xol_add_vma() to xol_alloc_area() to cleanup
the code. This separates the memory allocations and consolidates the
-EALREADY cleanups and the error handling.
Oleg Nesterov [Sun, 30 Dec 2012 14:54:08 +0000 (15:54 +0100)]
uprobes: Change handle_swbp() to expose bp_vaddr to handler_chain()
Change handle_swbp() to set regs->ip = bp_vaddr in advance, this is
what consumer->handler() needs but uprobe_get_swbp_addr() is not
exported.
This also simplifies the code and makes it more consistent across
the supported architectures. handle_swbp() becomes the only caller
of uprobe_get_swbp_addr().
Signed-off-by: Oleg Nesterov <oleg@redhat.com> Acked-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Oleg Nesterov [Sun, 30 Dec 2012 14:21:22 +0000 (15:21 +0100)]
uprobes/x86: Change __skip_sstep() to actually skip the whole insn
__skip_sstep() doesn't update regs->ip. Currently this is correct
but only "by accident" and it doesn't skip the whole insn. Change
it to advance ->ip by the length of the detected 0x66*0x90 sequence.
Oleg Nesterov [Sat, 29 Dec 2012 16:49:11 +0000 (17:49 +0100)]
uprobes: Teach handler_chain() to filter out the probed task
Currrently the are 2 problems with pre-filtering:
1. It is not possible to add/remove a task (mm) after uprobe_register()
2. A forked child inherits all breakpoints and uprobe_consumer can not
control this.
This patch does the first step to improve the filtering. handler_chain()
removes the breakpoints installed by this uprobe from current->mm if all
handlers return UPROBE_HANDLER_REMOVE.
Note that handler_chain() relies on ->register_rwsem to avoid the race
with uprobe_register/unregister which can add/del a consumer, or even
remove and then insert the new uprobe at the same address.
Perhaps we will add uprobe_apply_mm(uprobe, mm, is_register) and teach
copy_mm() to do filter(UPROBE_FILTER_FORK), but I think this change makes
sense anyway.
Note: instead of checking the retcode from uc->handler, we could add
uc->filter(UPROBE_FILTER_BPHIT). But I think this is not optimal to
call 2 hooks in a row. This buys nothing, and if handler/filter do
something nontrivial they will probably do the same work twice.
Oleg Nesterov [Fri, 28 Dec 2012 16:58:38 +0000 (17:58 +0100)]
uprobes: Reintroduce uprobe_consumer->filter()
Finally add uprobe_consumer->filter() and change consumer_filter()
to actually call this method.
Note that ->filter() accepts mm_struct, not task_struct. Because:
1. We do not have for_each_mm_user(mm, task).
2. Even if we implement for_each_mm_user(), ->filter() can
use it itself.
3. It is not clear who will actually need this interface to
do the "nontrivial" filtering.
Another argument is "enum uprobe_filter_ctx", consumer->filter() can
use it to figure out why/where it was called. For example, perhaps
we can add UPROBE_FILTER_PRE_REGISTER used by build_map_info() to
quickly "nack" the unwanted mm's. In this case consumer should know
that it is called under ->i_mmap_mutex.
See the previous discussion at http://marc.info/?t=135214229700002
Perhaps we should pass more arguments, vma/vaddr?
Note: this patch obviously can't help to filter out the child created
by fork(), this will be addressed later.
Oleg Nesterov [Thu, 27 Dec 2012 17:21:11 +0000 (18:21 +0100)]
uprobes: Rationalize the usage of filter_chain()
filter_chain() was added into install_breakpoint/remove_breakpoint to
simplify the initial changes but this is sub-optimal.
This patch shifts the callsite to the callers, register_for_each_vma()
and uprobe_mmap(). This way:
- It will be easier to add the new arguments. This is the main reason,
we can do more optimizations later.
- register_for_each_vma(is_register => true) can be optimized, we only
need to consult the new consumer. The previous consumers were already
asked when they called uprobe_register().
This patch also moves the MMF_HAS_UPROBES check from remove_breakpoint(),
this allows to avoid the potentionally costly filter_chain(). Note that
register_for_each_vma(is_register => false) doesn't really need to take
->consumer_rwsem, but I don't think it makes sense to optimize this and
introduce filter_chain_lockless().
Oleg Nesterov [Sun, 25 Nov 2012 21:48:37 +0000 (22:48 +0100)]
uprobes: Kill uprobes_mutex[], separate alloc_uprobe() and __uprobe_register()
uprobe_register() and uprobe_unregister() are the only users of
mutex_lock(uprobes_hash(inode)), and the only reason why we can't
simply remove it is that we need to ensure that delete_uprobe() is
not possible after alloc_uprobe() and before consumer_add().
IOW, we need to ensure that when we take uprobe->register_rwsem
this uprobe is still valid and we didn't race with _unregister()
which called delete_uprobe() in between.
With this patch uprobe_register() simply checks uprobe_is_active()
and retries if it hits this very unlikely race. uprobes_mutex[] is
no longer needed and can be removed.
There is another reason for this change, prepare_uprobe() should be
folded into alloc_uprobe() and we do not want to hold the extra locks
around read_mapping_page/etc.
Oleg Nesterov [Sun, 25 Nov 2012 21:01:42 +0000 (22:01 +0100)]
uprobes: Introduce uprobe_is_active()
The lifetime of uprobe->rb_node and uprobe->inode is not refcounted,
delete_uprobe() is called when we detect that uprobe has no consumers,
and it would be deadly wrong to do this twice.
Change delete_uprobe() to WARN() if it was already called. We use
RB_CLEAR_NODE() to mark uprobe "inactive", then RB_EMPTY_NODE() can
be used to detect this case.
RB_EMPTY_NODE() is not used directly, we add the trivial helper for
the next change.
Oleg Nesterov [Sat, 24 Nov 2012 17:27:08 +0000 (18:27 +0100)]
uprobes: Kill UPROBE_RUN_HANDLER flag
Simply remove UPROBE_RUN_HANDLER and the corresponding code.
It can only help if uprobe has a single consumer, and in fact
it is no longer needed after handler_chain() was changed to use
->register_rwsem, we simply can not race with uprobe_register().
Oleg Nesterov [Sat, 24 Nov 2012 17:15:46 +0000 (18:15 +0100)]
uprobes: Change filter_chain() to iterate ->consumers list
Now that it safe to use ->consumer_rwsem under ->mmap_sem we can
almost finish the implementation of filter_chain(). It still lacks
the actual uc->filter(...) call but othewrwise it is ready, just
it pretends that ->filter() always returns true.
Oleg Nesterov [Sat, 24 Nov 2012 16:29:40 +0000 (17:29 +0100)]
uprobes: Introduce uprobe->register_rwsem
Introduce uprobe->register_rwsem. It is taken for writing around
__uprobe_register/unregister.
Change handler_chain() to use this sem rather than consumer_rwsem.
The main reason for this change is that we have the nasty problem
with mmap_sem/consumer_rwsem dependency. filter_chain() needs to
protect uprobe->consumers like handler_chain(), but they can not
use the same lock. filter_chain() can be called under ->mmap_sem
(currently this is always true), but we want to allow ->handler()
to play with the probed task's memory, and this needs ->mmap_sem.
Alternatively we could use srcu, but synchronize_srcu() is very
slow and ->register_rwsem allows us to do more. In particular, we
can teach handler_chain() to do remove_breakpoint() if this bp is
"nacked" by all consumers, we know that we can't race with the
new consumer which does uprobe_register().
See also the next patches. uprobes_mutex[] is almost ready to die.
Oleg Nesterov [Fri, 23 Nov 2012 19:15:17 +0000 (20:15 +0100)]
uprobes: _register() should always do register_for_each_vma(true)
To support the filtering uprobe_register() should do
register_for_each_vma(true) every time the new consumer comes,
we need to install the previously nacked breakpoints.
Note:
- uprobes_mutex[] should die, what it actually protects is
alloc_uprobe().
- UPROBE_RUN_HANDLER should die too, obviously it can't work
unless uprobe has a single consumer. The consumer should
serialize with _register/_unregister itself. Or this flag
should live in uprobe_consumer->state.
- Perhaps we can do some optimizations later. For example, if
filter_chain() never returns false uprobe can record this
fact and avoid the unnecessary register_for_each_vma().
Oleg Nesterov [Fri, 23 Nov 2012 18:43:50 +0000 (19:43 +0100)]
uprobes: _unregister() should always do register_for_each_vma(false)
uprobe_unregister() removes the breakpoints only if the last consumer
goes away. To support the filtering it should do this every time, we
want to remove the breakpoints which nobody else want to keep.
Note: given that filter_chain() is not actually implemented, this patch
itself doesn't change the behaviour yet, register_for_each_vma(false)
is a heavy "nop" unless there are no more consumers.
Oleg Nesterov [Thu, 22 Nov 2012 17:30:15 +0000 (18:30 +0100)]
uprobes: Introduce filter_chain()
Add the new helper filter_chain(). Currently it is only placeholder,
the comment explains what is should do. We will change it later to
consult every consumer to decide whether we need to install the swbp.
Until then it works as if any consumer returns true, this matches the
current behavior.
Change install_breakpoint() to call filter_chain() instead of checking
uprobe->consumers != NULL. We obviously need this, and this equally
closes the race with _unregister().
Change remove_breakpoint() to call this helper too. Currently this is
pointless because remove_breakpoint() is only called when the last
consumer goes away, but we will change this.
Oleg Nesterov [Wed, 21 Nov 2012 16:32:30 +0000 (17:32 +0100)]
uprobes: Kill uprobe_consumer->filter()
uprobe_consumer->filter() is pointless in its current form, kill it.
We will add it back, but with the different signature/semantics. Perhaps
we will even re-introduce the callsite in handler_chain(), but not to
just skip uc->handler().
Oleg Nesterov [Wed, 21 Nov 2012 17:01:43 +0000 (18:01 +0100)]
uprobes: Kill the pointless inode/uc checks in register/unregister
register/unregister verifies that inode/uc != NULL. For what?
This really looks like "hide the potential problem", the caller
should pass the valid data.
register() also checks uc->next == NULL, probably to prevent the
double-register but the caller can do other stupid/wrong things.
If we do this check, then we should document that uc->next should
be cleared before register() and add BUG_ON().
Also add the small comment about the i_size_read() check.
Ingo Molnar [Wed, 6 Feb 2013 21:50:44 +0000 (22:50 +0100)]
Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core
Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
. Check for flex and bison before continuing building, from Borislav Petkov.
. Make event_copy local to mmaps, fixing buffer wrap around problems, from
David Ahern.
. Add option for runtime switching perf data file in perf report, just press
's' and a menu with the valid files found in the current directory will be
presented, from Feng Tang.
. Add support to display whole group data for raw columns, from Jiri Olsa.
. Fix SIGALRM and pipe read race for the rwtop perl script. from Jiri Olsa.
. Fix perf_evsel::exclude_GH handling and add a test to catch regressions, from
Jiri Olsa.
. Error checking fixes, from Namhyung Kim.
. Fix calloc argument ordering, from Paul Gortmaker.
. Fix set event list leader, from Stephane Eranian.
. Add per processor socket count aggregation in perf stat, from Stephane Eranian.
. Fix perf python binding breakage.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org>
I know that was the regex used, because on seeing the 1st of these
changes, I wondered "how many other instances of this are there" and I
happened to just use "calloc(sizeof" as a regex and it in turn found
these additional reversed args instances in the perf code.
In the kcalloc cases, the changes are cosmetic, since the numbers are
simply multiplied. I had no desire to go data mining in userspace to
see if the same thing held true there, however.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Joe Perches <joe@perches.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1359594349-25912-1-git-send-email-paul.gortmaker@windriver.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Jiri Olsa [Tue, 5 Feb 2013 16:05:50 +0000 (17:05 +0100)]
perf perl scripts: Fix SIGALRM and pipe read race for rwtop
Fixing rwtop script race. The issue is caused by rwtop script triggering
SIGALRM and underneath pipe reading layer reporting error when
interrupted.
Fixing this by setting SA_RESTART for rwtop SIGALRM handler, which
avoids interruption of the pipe reading layer.
The discussion for this issue & fix is here:
https://lkml.org/lkml/2012/9/18/123
Signed-off-by: Jiri Olsa <jolsa@redhat.com> Original-patch-by: Andrew Jones <drjones@redhat.com> Cc: Andrew Jones <drjones@redhat.com> Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com> Cc: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1360080351-3246-2-git-send-email-jolsa@redhat.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Jiri Olsa [Mon, 4 Feb 2013 12:32:55 +0000 (13:32 +0100)]
perf hists browser: Add support to display whole group data for raw columns
Currently we don't display group members' values for raw columns like
'Samples' and 'Period' when in group report mode.
Uniting '__hpp__percent_fmt' and '__hpp__raw_fmt' function under new
function __hpp__fmt. It's basically '__hpp__percent_fmt' code with new
'fmt_percent' bool parameter added saying whether raw number or
percentage should be printed.
This way raw columns print out all the group members when
in group report mode, like:
$ perf record -e '{cycles,cache-misses}' ls
...
$ perf report --group --show-total-period --stdio
...
# Overhead Period Command Shared Object Symbol
# ................ ........................ ....... ................. .................................
#
23.63% 11.24% 3331335 317 ls [kernel.kallsyms] [k] __lock_acquire
12.72% 0.00% 1793100 0 ls [kernel.kallsyms] [k] native_sched_clock
9.72% 0.00% 1369920 0 ls libc-2.14.90.so [.] _nl_find_locale
0.03% 0.07% 4476 2 ls [kernel.kallsyms] [k] intel_pmu_enable_all
0.00% 11.73% 0 331 ls ld-2.14.90.so [.] _dl_cache_libcmp
0.00% 11.06% 0 312 ls [kernel.kallsyms] [k] vma_interval_tree_insert
Signed-off-by: Jiri Olsa <jolsa@redhat.com> Acked-by: Namhyung Kim <namhyung@kernel.org> Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1359981185-16819-2-git-send-email-jolsa@redhat.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This patch adds:
- cpu_map__get_socket: get socked id from cpu
- cpu_map__build_socket_map: build socket map
- cpu_map__socket: gets acutal socket from logical socket
Those functions are used by uncore and processor socket-level
aggregation modes.
Signed-off-by: Stephane Eranian <eranian@google.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1360161962-9675-2-git-send-email-eranian@google.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
David Ahern [Tue, 5 Feb 2013 21:12:42 +0000 (14:12 -0700)]
perf evlist: Make event_copy local to mmaps
I am getting segfaults *after* the time sorting of perf samples where
the event type is off the charts:
(gdb) bt
\#0 0x0807b1b2 in hists__inc_nr_events (hists=0x80a99c4, type=1163281902) at util/hist.c:1225
\#1 0x08070795 in perf_session_deliver_event (session=0x80a9b90, event=0xf7a6aff8, sample=0xffffc318, tool=0xffffc520,
file_offset=0) at util/session.c:884
\#2 0x0806f9b9 in flush_sample_queue (s=0x80a9b90, tool=0xffffc520) at util/session.c:555
\#3 0x0806fc53 in process_finished_round (tool=0xffffc520, event=0x0, session=0x80a9b90) at util/session.c:645
This is bizarre because the event has already been processed once --
before it was added to the samples queue -- and the event was found to
be sane at that time.
There seem to be 2 causes:
1. perf_evlist__mmap_read updates the read location even though there
are outstanding references to events sitting in the mmap buffers via the
ordered samples queue.
2. There is a single evlist->event_copy for all evlist entries.
event_copy is used to handle an event wrapping at the mmap buffer
boundary.
This patch addresses the second problem - making event_copy local to
each perf_mmap. With this change my highly repeatable use case no longer
fails.
The first problem is much more complicated and will be the subject of a
future patch.
Signed-off-by: David Ahern <dsahern@gmail.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1360098762-61827-1-git-send-email-dsahern@gmail.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Wed, 6 Feb 2013 05:57:17 +0000 (14:57 +0900)]
perf sort: Check return value of strdup()
When setup_sorting() is called, 'str' is passed to strtok_r() but it's
not checked to have a valid pointer. As strtok_r() accepts NULL pointer
on a first argument and use the third argument in that case, it can
cause a trouble since our third argument, tmp, is not initialized.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Acked-by: Jiri Olsa <jolsa@redhat.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1360130237-9963-3-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Wed, 6 Feb 2013 05:57:16 +0000 (14:57 +0900)]
perf sort: Make setup_sorting returns an error code
Currently the setup_sorting() is called for parsing sort keys and exits
if it failed to add the sort key. As it's included in libperf it'd be
better returning an error code rather than exiting application inside of
the library.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Suggested-by: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1360130237-9963-2-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Wed, 6 Feb 2013 05:57:15 +0000 (14:57 +0900)]
perf sort: Drop ip_[lr] arguments from _sort__sym_cmp()
Current _sort__sym_cmp() function is used for comparing symbols between
two hist entries on symbol, symbol_from and symbol_to sort keys. Those
functions pass addresses of symbols but it's meaningless since it gets
over-written inside of the _sort__sym_cmp function to a start address of
the symbol. So just get rid of them.
This might cause a difference than prior output for branch stacks since
it seems not using start address of the symbol but branch address.
However AFAICS it'd be same as it gets overwritten anyway.
Also remove redundant part of code in sort__sym_cmp().
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Acked-by: Jiri Olsa <jolsa@redhat.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1360130237-9963-1-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Stephane Eranian [Thu, 31 Jan 2013 12:54:37 +0000 (13:54 +0100)]
perf evlist: Fix set event list leader
The __perf_evlist__set_leader() was setting the leader for all events in
the list except the first. Which means it assumed the first event
already had event->leader = event.
Seems like this should be the role of the function to also do this. This
is a requirement for an upcoming patch set.
Signed-off-by: Stephane Eranian <eranian@google.com> Acked-by: Jiri Olsa <jolsa@redhat.com> Tested-by: Jiri Olsa <jolsa@redhat.com> Acked-by: Namhyung Kim <namhyung@kernel.org> Cc: Ingo Molnar <mingo@elte.hu> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Namhyung Kim <namhyung.kim@lge.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/20130131125437.GA3656@quad Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Feng Tang [Sun, 3 Feb 2013 06:38:21 +0000 (14:38 +0800)]
perf report: Enable the runtime switching of perf data file
This is for tui browser only. This patch will check the returned key of
tui hists browser, if it's K_SWITH_INPUT_DATA, then recreate a session
for the new selected data file.
V2: Move the setup_brower() before the "repeat" jump point.
Signed-off-by: Feng Tang <feng.tang@intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@elte.hu> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1359873501-24541-2-git-send-email-feng.tang@intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Feng Tang [Sun, 3 Feb 2013 06:38:20 +0000 (14:38 +0800)]
perf hists browser: Add option for runtime switching perf data file
Based on perf report/top/scripts browser integration idea from acme.
This will enable user to runtime switch the data file, when this option
is selected, it will popup all the legal data files in current working
directory, and the filename selected by user is saved in the global
variable "input_name", and a new key 'K_SWITCH_INPUT_DATA' will be
passed back to the built-in command which will perform the switch.
This initial version only enables it for 'perf report'.
v2: rebase to latest 'perf/core' branch (6e1d4dd) of acme's perf tree
Signed-off-by: Feng Tang <feng.tang@intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ingo Molnar <mingo@elte.hu> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: http://lkml.kernel.org/r/1359873501-24541-1-git-send-email-feng.tang@intel.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Jacob Shin [Wed, 6 Feb 2013 17:26:27 +0000 (11:26 -0600)]
perf/x86: Move MSR address offset calculation to architecture specific files
Move counter index to MSR address offset calculation to
architecture specific files. This prepares the way for
perf_event_amd to enable counter addresses that are not
contiguous -- for example AMD Family 15h processors have 6 core
performance counters starting at 0xc0010200 and 4 northbridge
performance counters starting at 0xc0010240.
Signed-off-by: Jacob Shin <jacob.shin@amd.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Cc: Stephane Eranian <eranian@google.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1360171589-6381-5-git-send-email-jacob.shin@amd.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
Robert Richter [Wed, 6 Feb 2013 17:26:25 +0000 (11:26 -0600)]
perf/x86/amd: Generalize northbridge constraints code for family 15h
Generalize northbridge constraints code for family 10h so that
later we can reuse the same code path with other AMD processor
families that have the same northbridge event constraints.
Signed-off-by: Robert Richter <rric@kernel.org> Signed-off-by: Jacob Shin <jacob.shin@amd.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Cc: Stephane Eranian <eranian@google.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Link: http://lkml.kernel.org/r/1360171589-6381-3-git-send-email-jacob.shin@amd.com Signed-off-by: Ingo Molnar <mingo@kernel.org>
tracing: Init current_trace to nop_trace and remove NULL checks
On early boot up, when the ftrace ring buffer is initialized, the
static variable current_trace is initialized to &nop_trace.
Before this initialization, current_trace is NULL and will never
become NULL again. It is always reassigned to a ftrace tracer.
Several places check if current_trace is NULL before it uses
it, and this check is frivolous, because at the point in time
when the checks are made the only way current_trace could be
NULL is if ftrace failed its allocations at boot up, and the
paths to these locations would probably not be possible.
By initializing current_trace to &nop_trace where it is declared,
current_trace will never be NULL, and we can remove all these
checks of current_trace being NULL which never needed to be
checked in the first place.
Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Hiraku Toyooka <hiraku.toyooka.gu@hitachi.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Ingo Molnar [Fri, 1 Feb 2013 10:17:09 +0000 (11:17 +0100)]
Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/core
Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:
. Make some POWER7 events available in sysfs, equivalent to
what was done on x86, from Sukadev Bhattiprolu.
. Add event group view, from Namyung Kim:
To use it, 'perf record' should group events when recording. And then perf
report parses the saved group relation from file header and prints them
together if --group option is provided. You can use 'perf evlist' command to
see event group information:
$ perf record -e '{ref-cycles,cycles}' noploop 1
[ perf record: Woken up 2 times to write data ]
[ perf record: Captured and wrote 0.385 MB perf.data (~16807 samples) ]
$ perf evlist --group
{ref-cycles,cycles}
With this example, default perf report will show you each event
separately like this:
As you can see the Overhead column now contains both of ref-cycles and
cycles and header line shows group information also - 'anon group {
ref-cycles, cycles }'. The output is sorted by period of group leader
first.
If perf.data file doesn't contain group information, this --group
option does nothing. So if you want enable event group view by
default you can set it in ~/.perfconfig file:
$ cat ~/.perfconfig
[report]
group = true
It can be overridden with command line if you want:
$ perf report --no-group
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Ingo Molnar <mingo@kernel.org>
This patchset addes two new sets of files to sysfs for POWER architecture.
- perf event config format in /sys/devices/cpu/format/event
- generic and POWER-specific perf events in /sys/devices/cpu/events/
The format of the first file is already documented in:
sysfs-bus-event_source-devices-format
Document the format of the second set of files '/sys/devices/cpu/events/*'
which would also become part of the ABI.
Changelog[v4]:
[Jiri Olsa]: Mention that multiple event= like terms can be specified
in the 'events' file.
[Jiri Olsa]: Remove the documentation for the 'config format' file
as it is already documented in 'Documentation/ABI/testing/'.
[Jiri Olsa]: Move ABI documentation from 'stable/' to 'testing/'
Changelog[v3]:
[Greg KH] Include ABI documentation.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Acked-by: Jiri Olsa <jolsa@redhat.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Anton Blanchard <anton@au1.ibm.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Robert Richter <robert.richter@amd.com> Cc: Stephane Eranian <eranian@google.com> Cc: linuxppc-dev@ozlabs.org Link: http://lkml.kernel.org/r/20130123062645.GG13720@us.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This patch is based on commits that implement this functionality on x86.
Eg:
commit a47473939db20e3961b200eb00acf5fcf084d755
Author: Jiri Olsa <jolsa@redhat.com>
Date: Wed Oct 10 14:53:11 2012 +0200
perf/x86: Make hardware event translations available in sysfs
Changelog:[v2]
[Jiri Osla] Drop EVENT_ID() macro since it is only used once.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Anton Blanchard <anton@au1.ibm.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Robert Richter <robert.richter@amd.com> Cc: Stephane Eranian <eranian@google.com> Cc: linuxppc-dev@ozlabs.org Link: http://lkml.kernel.org/r/20130123062454.GD13720@us.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Define and use macros to identify perf events codes This would make it
easier and more readable when these event codes need to be used in more
than one place.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Acked-by: Jiri Olsa <jolsa@redhat.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Anton Blanchard <anton@au1.ibm.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Robert Richter <robert.richter@amd.com> Cc: Stephane Eranian <eranian@google.com> Cc: linuxppc-dev@ozlabs.org Link: http://lkml.kernel.org/r/20130123062353.GB13720@us.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 22 Jan 2013 09:09:46 +0000 (18:09 +0900)]
perf report: Add report.group config option
Add report.group config option for setting default value of event
group view. It affects the report output only if perf.data contains
event group info.
A user can write .perfconfig file like below to enable group view by
default:
$ cat ~/.perfconfig
[report]
group = true
And it can be disabled through command line:
$ perf report --no-group
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1358845787-1350-19-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 22 Jan 2013 09:09:43 +0000 (18:09 +0900)]
perf report: Bypass non-leader events when event group is enabled
Since we have all necessary information in the leader events and other
members don't, bypass members. Member events will be shown along with
the leaders if event group is enabled.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Acked-by: Jiri Olsa <jolsa@redhat.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1358845787-1350-16-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 22 Jan 2013 09:09:42 +0000 (18:09 +0900)]
perf gtk/browser: Trim column header string when event group enabled
When event group feature is enabled, each column header is expanded to
match with the whole group column width. But this is not needed for
GTK+ browser since ti usually use variable-width fonts. So trim it.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1358845787-1350-15-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 22 Jan 2013 09:09:36 +0000 (18:09 +0900)]
perf gtk/browser: Convert hpp helpers to a function
The hpp helpers do same job for each field so it was implemented as
macro in order to access those fields easily. But it gets cumbersome to
maintain a large function in a macro as the function grows. Factor it
out to a function with a little helper macro to access field.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1358845787-1350-9-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 22 Jan 2013 09:09:35 +0000 (18:09 +0900)]
perf hists browser: Convert hpp helpers to a function
The hpp helpers do same job for each field so it was implemented as
macro in order to access those fields easily. But it gets cumbersome
to maintain a large function in a macro as the function grows. Factor
it out to a function with a little helper macro to access field.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1358845787-1350-8-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 22 Jan 2013 09:09:33 +0000 (18:09 +0900)]
perf hists: Resort hist entries using group members for output
When event group is enabled, sorting hist entries on periods for output
should consider groups members' period also. To do that, build period
table using link/pair information and compare the table.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Acked-by: Jiri Olsa <jolsa@redhat.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1358845787-1350-6-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 22 Jan 2013 09:09:32 +0000 (18:09 +0900)]
perf report: Make another loop for linking group hists
Now the event grouping viewing requires linking all member hists in a
group to the leader's. Thus hists__output_resort should be called after
linking all events in evlist.
Introduce symbol_conf.event_group flag to determine whether the feature
is enabled or not.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Acked-by: Jiri Olsa <jolsa@redhat.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1358845787-1350-5-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Namhyung Kim [Tue, 22 Jan 2013 09:09:29 +0000 (18:09 +0900)]
perf tools: Keep group information
Add a few of group-related field in struct perf_{evlist,evsel} so that
the group information in a evlist can be known easily. It only counts
groups which have more than 1 members since leader-only groups are
treated as non-group events.
Signed-off-by: Namhyung Kim <namhyung@kernel.org> Acked-by: Jiri Olsa <jolsa@redhat.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Paul Mackerras <paulus@samba.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Stephane Eranian <eranian@google.com> Link: http://lkml.kernel.org/r/1358845787-1350-2-git-send-email-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Hiraku Toyooka [Wed, 26 Dec 2012 02:53:00 +0000 (11:53 +0900)]
tracing: Make a snapshot feature available from userspace
Ftrace has a snapshot feature available from kernel space and
latency tracers (e.g. irqsoff) are using it. This patch enables
user applictions to take a snapshot via debugfs.
Add "snapshot" debugfs file in "tracing" directory.
snapshot:
This is used to take a snapshot and to read the output of the
snapshot.
# echo 1 > snapshot
This will allocate the spare buffer for snapshot (if it is
not allocated), and take a snapshot.
# cat snapshot
This will show contents of the snapshot.
# echo 0 > snapshot
This will free the snapshot if it is allocated.
Any other positive values will clear the snapshot contents if
the snapshot is allocated, or return EINVAL if it is not allocated.
Link: http://lkml.kernel.org/r/20121226025300.3252.86850.stgit@liselsia Cc: Jiri Olsa <jolsa@redhat.com> Cc: David Sharp <dhsharp@google.com> Signed-off-by: Hiraku Toyooka <hiraku.toyooka.gu@hitachi.com>
[
Fixed irqsoff selftest and also a conflict with a change
that fixes the update_max_tr.
] Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Hiraku Toyooka [Wed, 26 Dec 2012 02:52:52 +0000 (11:52 +0900)]
tracing: Replace static old_tracer check of tracer name
Currently the trace buffer read functions use a static variable
"old_tracer" for detecting if the current tracer changes. This
was suitable for a single trace file ("trace"), but to add a
snapshot feature that will use the same function for its file,
a check against a static variable is not sufficient.
To use the output functions for two different files, instead of
storing the current tracer in a static variable, as the trace
iterator descriptor contains a pointer to the original current
tracer's name, that pointer can now be used to check if the
current tracer has changed between different reads of the trace
file.
Namhyung Kim [Thu, 27 Dec 2012 02:49:45 +0000 (11:49 +0900)]
tracing: Use sched_clock_cpu for trace_clock_global
For systems with an unstable sched_clock, all cpu_clock() does is enable/
disable local irq during the call to sched_clock_cpu(). And for stable
systems they are same.
trace_clock_global() already disables interrupts, so it can call
sched_clock_cpu() directly.
perf tools, powerpc: Fix compile warnings in tests/attr.c
We print several '__u64' quantities using '%llu'. On powerpc, we by
default include '<asm-generic/int-l64.h> which results in __u64 being an
unsigned long. This causes compile warnings which are treated as errors
due to '-Werror'.
By defining __SANE_USERSPACE_TYPES__ we include <asm-generic/int-ll64.h>
and define __u64 as unsigned long long.
Changelog[v2]:
[Michael Ellerman] Use __SANE_USERSPACE_TYPES__ and avoid PRIu64
format specifier - which as Jiri Olsa pointed out, breaks on x86-64.
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Cc: Anton Blanchard <anton@au1.ibm.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Michael Ellerman <ellerman@au1.ibm.com> Cc: linuxppc-dev@ozlabs.org Link: http://lkml.kernel.org/r/20130124054439.GA31588@us.ibm.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Stephane Eranian [Tue, 29 Jan 2013 11:47:44 +0000 (12:47 +0100)]
perf stat: Add interval printing
This patch adds a new printing mode for perf stat. It allows interval
printing. That means perf stat can now print event deltas at regular
time interval. This is useful to detect phases in programs.
The -I option enables interval printing. It expects an interval duration
in milliseconds. Minimum is 100ms. Once, activated perf stat prints
events deltas since last printout. All modes are supported.
Ingo Molnar [Thu, 6 Dec 2012 12:51:59 +0000 (13:51 +0100)]
perf: Add 'perf bench numa mem' NUMA performance measurement suite
Add a suite of NUMA performance benchmarks.
The goal was simulate the behavior and access patterns of real NUMA
workloads, via a wide range of parameters, so this tool goes well
beyond simple bzero() measurements that most NUMA micro-benchmarks use:
- It processes the data and creates a chain of data dependencies,
like a real workload would. Neither the compiler, nor the
kernel (via KSM and other optimizations) nor the CPU can
eliminate parts of the workload.
- It randomizes the initial state and also randomizes the target
addresses of the processing - it's not a simple forward scan
of addresses.
- It provides flexible options to set process, thread and memory
relationship information: -G sets "global" memory shared between
all test processes, -P sets "process" memory shared by all
threads of a process and -T sets "thread" private memory.
- There's a NUMA convergence monitoring and convergence latency
measurement option via -c and -m.
- Micro-sleeps and synchronization can be injected to provoke lock
contention and scheduling, via the -u and -S options. This simulates
IO and contention.
- The -x option instructs the workload to 'perturb' itself artificially
every N seconds, by moving to the first and last CPU of the system
periodically. This way the stability of convergence equilibrium and
the number of steps taken for the scheduler to reach equilibrium again
can be measured.
- The amount of work can be specified via the -l loop count, and/or
via a -s seconds-timeout value.
- CPU and node memory binding options, to test hard binding scenarios.
THP can be turned on and off via madvise() calls.
- Live reporting of convergence progress in an 'at glance' output format.
Printing of convergence and deconvergence events.
The 'perf bench numa mem -a' option will start an array of about 30
individual tests that will each output such measurements:
tracing/fgraph: Adjust fgraph depth before calling trace return callback
While debugging the virtual cputime with the function graph tracer
with a max_depth of 1 (most common use of the max_depth so far),
I found that I was missing kernel execution because of a race condition.
The code for the return side of the function has a slight race:
The ftrace_pop_return_trace() initializes the trace structure for
the callback. The ftrace_graph_return() uses the trace structure
for its own use as that structure is on the stack and is local
to this function. Then the curr_ret_stack is decremented which
is what the trace.depth is set to.
If an interrupt comes in after the ftrace_graph_return() but
before the curr_ret_stack, then the called function will get
a depth of 2. If max_depth is set to 1 this function will be
ignored.
The problem is that the trace has already been called, and the
timestamp for that trace will not reflect the time the function
was about to re-enter userspace. Calls to the interrupt will not
be traced because the max_depth has prevented this.
To solve this issue, the ftrace_graph_return() can safely be
moved after the current->curr_ret_stack has been updated.
This way the timestamp for the return callback will reflect
the actual time.
If an interrupt comes in after the curr_ret_stack update and
ftrace_graph_return(), it will be traced. It may look a little
confusing to see it within the other function, but at least
it will not be lost.
Cc: Frederic Weisbecker <fweisbec@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
projects / karo-tx-linux.git / log