We were always looking at the running machine /proc/modules,
even when processing a perf.data file, which only makes sense
when we're doing 'perf record' and 'perf report' on the same
machine, and in close sucession, or if we don't use modules at
all, right Peter? ;-)
Now, at 'perf record' time we read /proc/modules, find the long
path for modules, and put them as PERF_MMAP events, just like we
did to encode the reloc reference symbol for vmlinux. Talking
about that now it is encoded in .pgoff, so that we can use
.{start,len} to store the address boundaries for the kernel so
that when we reconstruct the kmaps tree we can do lookups right
away, without having to fixup the end of the kernel maps like we
did in the past (and now only in perf record).
One more step in the 'perf archive' direction when we'll finally
be able to collect data in one machine and analyse in another.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1263396139-4798-1-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Add --line option to support showing probable source-code lines.
perf probe --line SRC:LN[-LN|+NUM]
or
perf probe --line FUNC[:LN[-LN|+NUM]]
This option shows source-code with line number if the line can
be probed. Lines without line number (and blue color) means that
the line can not be probed, because debuginfo doesn't have the
information of those lines.
The argument specifies the range of lines, "source.c:100-120"
shows lines between 100th to l20th in source.c file. And
"func:10+20" shows 20 lines from 10th line of func function.
e.g.
# ./perf probe --line kernel/sched.c:1080
<kernel/sched.c:1080>
*
* called with rq->lock held and irqs disabled
*/
static void hrtick_start(struct rq *rq, u64 delay)
{
struct hrtimer *timer = &rq->hrtick_timer;
1086 ktime_t time = ktime_add_ns(timer->base->get_time(), delay);
hrtimer_set_expires(timer, time);
1090 if (rq == this_rq()) {
1091 hrtimer_restart(timer);
1092 } else if (!rq->hrtick_csd_pending) {
1093 __smp_call_function_single(cpu_of(rq), &rq->hrtick_csd,
1094 rq->hrtick_csd_pending = 1;
If you specifying function name, this shows function-relative
line number.
# ./perf probe --line schedule
<schedule:0>
asmlinkage void __sched schedule(void)
1 {
struct task_struct *prev, *next;
unsigned long *switch_count;
struct rq *rq;
int cpu;
need_resched:
preempt_disable();
9 cpu = smp_processor_id();
10 rq = cpu_rq(cpu);
11 rcu_sched_qs(cpu);
12 prev = rq->curr;
13 switch_count = &prev->nivcsw;
Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: systemtap <systemtap@sources.redhat.com>
Cc: DLE <dle-develop@lists.sourceforge.net>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Mike Galbraith <efault@gmx.de>
LKML-Reference: <20100106144534.27218.77939.stgit@dhcp-100-2-132.bos.redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
DSOs don't have this problem because the kernel emits a
PERF_MMAP for each new executable mapping it performs on
monitored threads.
To fix the kernel case we simulate the same behaviour, by having
'perf record' to synthesize a PERF_MMAP for the kernel, encoded
like this:
[root@doppio ~]# perf record -a -f sleep 1
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.344 MB perf.data (~15038 samples) ]
[root@doppio ~]# perf report -D | head -10
0xd0 [0x40]: event: 1
.
. ... raw event: size 64 bytes
. 0000: 01 00 00 00 00 00 40 00 00 00 00 00 00 00 00 00 ......@........
. 0010: 00 00 00 81 ff ff ff ff 00 00 00 00 00 00 00 00 ...............
. 0020: 00 00 00 00 00 00 00 00 5b 6b 65 72 6e 65 6c 2e ........ [kernel
. 0030: 6b 61 6c 6c 73 79 6d 73 2e 5f 74 65 78 74 5d 00 kallsyms._text]
. 0xd0
[0x40]: PERF_RECORD_MMAP 0/0: [0xffffffff81000000((nil)) @ (nil)]: [kernel.kallsyms._text]
I.e. we identify such event as having:
.pid = 0
.filename = [kernel.kallsyms.REFNAME]
.start = REFNAME addr in /proc/kallsyms at 'perf record' time
and use now a hardcoded value of '.text' for REFNAME.
Then, later, in 'perf report', if there are any kernel hits and
thus we need to resolve kernel symbols, we search for REFNAME
and if its address changed, relocation happened and we thus must
change the kernel mapping routines to one that uses .pgoff as
the relocation to apply.
This way we use the same mechanism used for the other DSOs and
don't have to do a two pass in all the kernel symbols.
Reported-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
LKML-Reference: <1262717431-1246-1-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Now a cache will be created in a ~/.debug debuginfo like
hierarchy, so that at the end of a 'perf record' session all the
binaries (with build-ids) involved get collected and indexed by
their build-ids, so that perf report can find them.
This is interesting when developing software where you want to
do a 'perf diff' with the previous build and opens avenues for
lots more interesting tools, like a 'perf diff --graph' that
takes more than two binaries into account.
Tunables for collecting just the symtabs can be added if one
doesn't want to have the full binary, but having the full binary
allows things like 'perf rerecord' or other tools that can
re-run the tests by having access to the exact binary in some
perf.data file, so it may well be interesting to keep the full
binary there.
Space consumption is minimised by trying to use hard links, a
'perf cache' tool to manage the space used, a la ccache is
required to purge older entries.
With this in place it will be possible also to introduce new
commands, 'perf archive' and 'perf restore' (or some more
suitable and future proof names) to create a cpio/tar file with
the perf data and the files in the cache that _had_ perf hits of
interest.
There are more aspects to polish, like finding the right vmlinux
file to cache, etc, but this is enough for a first step.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1261957026-15580-10-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This is really something tools need to do before asking for the
events to be processed, leaving perf_session__process_events to
do just that, process events.
Also add a msg parameter to perf_session__has_traces() so that
the right message can be printed, fixing a regression added by
me in the previous cset (right timechart message) and also
fixing 'perf kmem', that was not asking if 'perf kmem record'
was ran.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1261957026-15580-6-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Fix libdwarf include path to fit debian-like systems too.
Borislav Petkov reported:
> even after installing libdwarf-dev on my debian box here,
> make in tools/perf/ still complains that it cannot find libdwarf:
>
> Makefile:491: No libdwarf.h found or old libdwarf.h found, disables dwarf
> support. Please install libdwarf-dev/libdwarf-devel >= 20081231
>
> The problem is that the include path on debian is not
> /usr/include/libdwarf/ but simply /usr/include because the debian
> package libdwarf-dev puts the headers straight into
> /usr/include.
This patch adds -I/usr/include/libdwarf to BASIC_CFLAGS
and fix probe-finder.h to include just libdwarf.h/dwarf.h.
This patch also adds a workaround for the undefined _MIPS_SZLONG
bug in libdwarf.h.
Reported-by: Borislav Petkov <borislav.petkov@amd.com>
Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Gabor Gombas <gombasg@sztaki.hu>
Cc: systemtap <systemtap@sources.redhat.com>
Cc: DLE <dle-develop@lists.sourceforge.net>
LKML-Reference: <20091216221618.13816.83296.stgit@dhcp-100-2-132.bos.redhat.com>
[ v2: small stylistic fixlets to probe-finder.h ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
That means that almost everything you can do with 'perf report'
can be done with 'perf diff', for instance:
$ perf record -f find / > /dev/null
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.062 MB perf.data (~2699
samples) ] $ perf record -f find / > /dev/null
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.062 MB perf.data (~2687
samples) ] perf diff | head -8
9.02% +1.00% find libc-2.10.1.so [.] _IO_vfprintf_internal
2.91% -1.00% find [kernel] [k] __kmalloc
2.85% -1.00% find [kernel] [k] ext4_htree_store_dirent
1.99% -1.00% find [kernel] [k] _atomic_dec_and_lock
2.44% find [kernel] [k] half_md4_transform
$
So if you want to zoom into libc:
$ perf diff --dsos libc-2.10.1.so | head -8
37.34% find [.] _IO_vfprintf_internal
10.34% find [.] __GI_memmove
8.25% +2.00% find [.] _int_malloc
5.07% -1.00% find [.] __GI_mempcpy
7.62% +2.00% find [.] _int_free
$
And if there were multiple commands using libc, it is also
possible to aggregate them all by using --sort symbol:
$ perf diff --dsos libc-2.10.1.so --sort symbol | head -8
37.34% [.] _IO_vfprintf_internal
10.34% [.] __GI_memmove
8.25% +2.00% [.] _int_malloc
5.07% -1.00% [.] __GI_mempcpy
7.62% +2.00% [.] _int_free
$
The displacement column now is off by default, to use it:
perf diff -m --dsos libc-2.10.1.so --sort symbol | head -8
37.34% [.] _IO_vfprintf_internal
10.34% [.] __GI_memmove
8.25% +2.00% [.] _int_malloc
5.07% -1.00% +2 [.] __GI_mempcpy
7.62% +2.00% -1 [.] _int_free
$
Using -t/--field-separator can be used for scripting:
$ perf diff -t, -m --dsos libc-2.10.1.so --sort symbol | head -8
37.34, , ,[.] _IO_vfprintf_internal
10.34, , ,[.] __GI_memmove
8.25,+2.00%, ,[.] _int_malloc
5.07,-1.00%, +2,[.] __GI_mempcpy
7.62,+2.00%, -1,[.] _int_free
6.99,+1.00%, -1,[.] _IO_new_file_xsputn
1.89,-2.00%, +4,[.] __readdir64
$
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1260978567-550-1-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
