The event__process function is useful in processing /proc/<pid>/maps. All of
the functions that are called from event__process are defined in util/event.c.
Though its defined in builtin-top.c, it could be reused for perf probe for
uprobes. Hence moving it to util/event.c and exporting the function.
LKML-Reference: <20100802123851.GD22812@linux.vnet.ibm.com>
Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This patch adds a -C option to stat, record, top to designate a list of CPUs to
monitor. CPUs can be specified as a comma-separated list or ranges, no space
allowed.
Examples:
$ perf record -a -C0-1,4-7 sleep 1
$ perf top -C0-4
$ perf stat -a -C1,2,3,4 sleep 1
With perf record in per-thread mode with inherit mode on, samples are collected
only when the thread runs on the designated CPUs.
The -C option does not turn on system-wide mode automatically.
Cc: David S. Miller <davem@davemloft.net>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <4bff9496.d345d80a.41fe.7b00@mx.google.com>
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that if the kernel DSO has a build id because record inserted it in
the perf.data build id table in the header, or a BUILD_ID event was
inserted in the stream, we first look at the build id cache
($HOME/.debug/).
If we find it there, try to use it, allowing offline annotation in
addition to 'perf report'.
Reported-by: Stephane Eranian <eranian@google.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
OPT_SET_INT was renamed to OPT_SET_UINT since the only use in these
tools is to set something that has an enum type, that is builtin
compatible with unsigned int.
Several string constifications were done to make OPT_STRING require a
const char * type.
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
To avoid problems like the one fixed by Stephane Eranian in 3de29ca, now
we'll got this instead:
bench/sched-messaging.c:259: error: negative width in bit-field ‘<anonymous>’
bench/sched-messaging.c:261: error: negative width in bit-field ‘<anonymous>’
Which is rather cryptic, but is how BUILD_BUG_ON_ZERO works, so kernel
hackers should be already used to this.
With it in place found some problems, fixed by changing the affected
variables to sensible types or changed some OPT_INTEGER to OPT_UINTEGER.
Next csets will go thru converting each of the remaining OPT_ so that
review can be made easier by grouping changes per type per patch.
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Rename perf_event_attr::precise to perf_event_attr::precise_ip and
widen it to 2 bits. This new field describes the required precision of
the PERF_SAMPLE_IP field:
0 - SAMPLE_IP can have arbitrary skid
1 - SAMPLE_IP must have constant skid
2 - SAMPLE_IP requested to have 0 skid
3 - SAMPLE_IP must have 0 skid
And modify the Intel PEBS code accordingly. The PEBS implementation
now supports up to precise_ip == 2, where we perform the IP fixup.
Also s/PERF_RECORD_MISC_EXACT/&_IP/ to clarify its meaning, this bit
should be set for each PERF_SAMPLE_IP field known to match the actual
instruction triggering the event.
This new scheme allows for a PEBS mode that uses the buffer for more
than a single event.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Stephane Eranian <eranian@google.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Currently, perf 'live mode' writes build-ids at the end of the
session, which isn't actually useful for processing live mode events.
What would be better would be to have the build-ids sent before any of
the samples that reference them, which can be done by processing the
event stream and retrieving the build-ids on the first hit. Doing
that in perf-record itself, however, is off-limits.
This patch introduces perf-inject, which does the same job while
leaving perf-record untouched. Normal mode perf still records the
build-ids at the end of the session as it should, but for live mode,
perf-inject can be injected in between the record and report steps
e.g.:
perf record -o - ./hackbench 10 | perf inject -v -b | perf report -v -i -
perf-inject reads a perf-record event stream and repipes it to stdout.
At any point the processing code can inject other events into the
event stream - in this case build-ids (-b option) are read and
injected as needed into the event stream.
Build-ids are just the first user of perf-inject - potentially
anything that needs userspace processing to augment the trace stream
with additional information could make use of this facility.
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
LKML-Reference: <1272696080-16435-3-git-send-email-tzanussi@gmail.com>
Signed-off-by: Tom Zanussi <tzanussi@gmail.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
struct kernel_info and kerninfo__ are too vague, what they really
describe are machines, virtual ones or hosts.
There are more changes to introduce helpers to shorten function calls
and to make more clear what is really being done, but I left that for
subsequent patches.
Cc: Avi Kivity <avi@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>
Cc: Zhang, Yanmin <yanmin_zhang@linux.intel.com>
LKML-Reference: <new-submission>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Parameter --pid (or -p) of perf currently means a thread-wide
collection. For exmaple, if a process whose id is 8888 has 10
threads, 'perf top -p 8888' just collects the main thread
statistics. That's misleading. Users are used to attach a whole
process when debugging a process by gdb. To follow normal usage
style, the patch change --pid to process-wide collection and add
--tid (-t) to mean a thread-wide collection.
Usage example is:
# perf top -p 8888
# perf record -p 8888 -f sleep 10
# perf stat -p 8888 -f sleep 10
Above commands collect the statistics of all threads of process
8888.
Signed-off-by: Zhang Yanmin <yanmin_zhang@linux.intel.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Sheng Yang <sheng@linux.intel.com>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jes Sorensen <Jes.Sorensen@redhat.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: zhiteng.huang@intel.com
Cc: Zachary Amsden <zamsden@redhat.com>
LKML-Reference: <1268922965-14774-3-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Before this patch this message would very briefly appear on the
screen and then the screen would get updates only on the top,
for number of interrupts received, etc, but no annotation would
be performed:
[root@doppio linux-2.6-tip]# perf top -s n_tty_write > /tmp/bla
objdump: '[kernel.kallsyms]': No such file
Now this is what the user gets:
[root@doppio linux-2.6-tip]# perf top -s n_tty_write
Can't annotate n_tty_write: No vmlinux file was found in the
path: [0] vmlinux
[1] /boot/vmlinux
[2] /boot/vmlinux-2.6.33-rc5
[3] /lib/modules/2.6.33-rc5/build/vmlinux
[4] /usr/lib/debug/lib/modules/2.6.33-rc5/vmlinux
[root@doppio linux-2.6-tip]#
This bug was introduced when we added automatic search for
vmlinux, before that time the user had to specify a vmlinux
file.
Reported-by: David S. Miller <davem@davemloft.net>
Reported-by: Ingo Molnar <mingo@elte.hu>
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>
Cc: <stable@kernel.org>
LKML-Reference: <1268664418-28328-2-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
At present, the perf subcommands that do system-wide monitoring
(perf stat, perf record and perf top) don't work properly unless
the online cpus are numbered 0, 1, ..., N-1. These tools ask
for the number of online cpus with sysconf(_SC_NPROCESSORS_ONLN)
and then try to create events for cpus 0, 1, ..., N-1.
This creates problems for systems where the online cpus are
numbered sparsely. For example, a POWER6 system in
single-threaded mode (i.e. only running 1 hardware thread per
core) will have only even-numbered cpus online.
This fixes the problem by reading the /sys/devices/system/cpu/online
file to find out which cpus are online. The code that does that is in
tools/perf/util/cpumap.[ch], and consists of a read_cpu_map()
function that sets up a cpumap[] array and returns the number of
online cpus. If /sys/devices/system/cpu/online can't be read or
can't be parsed successfully, it falls back to using sysconf to
ask how many cpus are online and sets up an identity map in cpumap[].
The perf record, perf stat and perf top code then calls
read_cpu_map() in the system-wide monitoring case (instead of
sysconf) and uses cpumap[] to get the cpu numbers to pass to
perf_event_open.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Cc: Anton Blanchard <anton@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
LKML-Reference: <20100310093609.GA3959@brick.ozlabs.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Because we may have aliases, like __GI___strcoll_l in
/lib64/libc-2.10.2.so that appears in objdump as:
$ objdump --start-address=0x0000003715a86420 \
--stop-address=0x0000003715a872dc -dS /lib64/libc-2.10.2.so
0000003715a86420 <__strcoll_l>:
3715a86420: 55 push %rbp
3715a86421: 48 89 e5 mov %rsp,%rbp
3715a86424: 41 57 push %r15
[root@doppio linux-2.6-tip]#
So look for the address exactly at the start of the line instead
so that annotation can work for in these cases.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: Kirill Smelkov <kirr@landau.phys.spbu.ru>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1265550376-12665-2-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
First, for programs and prelinked libraries, annotate code was
fooled by objdump output IPs (src->eip in the code) being
wrongly converted to absolute IPs. In such case there were no
conversion needed, but in
src->eip = strtoull(src->line, NULL, 16);
src->eip = map->unmap_ip(map, src->eip); // = eip + map->start - map->pgoff
we were reading absolute address from objdump (e.g. 8048604) and
then almost doubling it, because eip & map->start are
approximately close for small programs.
Needless to say, that later, in record_precise_ip() there was no
matching with real runtime IPs.
And second, like with `perf annotate` the problem with
non-prelinked *.so was that we were doing rip -> objdump address
conversion wrong.
Also, because unlike `perf annotate`, `perf top` code does
annotation based on absolute IPs for performance reasons(*), new
helper for mapping objdump addresse to IP is introduced.
(*) we get samples info in absolute IPs, and since we do lots of
hit-testing on absolute IPs at runtime in record_precise_ip(), it's
better to convert objdump addresses to IPs once and do no conversion
at runtime.
I also had to fix how objdump output is parsed (with hardcoded
8/16 characters format, which was inappropriate for ET_DYN dsos
with small addresses like '4ac')
Also note, that not all objdump output lines has associtated
IPs, e.g. look at source lines here:
000004ac <my_strlen>:
extern "C"
int my_strlen(const char *s)
4ac: 55 push %ebp
4ad: 89 e5 mov %esp,%ebp
4af: 83 ec 10 sub $0x10,%esp
{
int len = 0;
4b2: c7 45 fc 00 00 00 00 movl $0x0,-0x4(%ebp)
4b9: eb 08 jmp 4c3 <my_strlen+0x17>
while (*s) {
++len;
4bb: 83 45 fc 01 addl $0x1,-0x4(%ebp)
++s;
4bf: 83 45 08 01 addl $0x1,0x8(%ebp)
So we mark them with eip=0, and ignore such lines in annotate
lookup code.
Signed-off-by: Kirill Smelkov <kirr@landau.phys.spbu.ru>
[ Note: one hunk of this patch was applied by Mike in 57d8188 ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
LKML-Reference: <1265550376-12665-1-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
By relying on logic in dso__load_kernel_sym(), we can
automatically load vmlinux.
The only thing which needs to be adjusted, is how --sym-annotate
option is handled - now we can't rely on vmlinux been loaded
until full successful pass of dso__load_vmlinux(), but that's
not the case if we'll do sym_filter_entry setup in
symbol_filter().
So move this step right after event__process_sample() where we
know the whole dso__load_kernel_sym() pass is done.
By the way, though conceptually similar `perf top` still can't
annotate userspace - see next patches with fixes.
Signed-off-by: Kirill Smelkov <kirr@landau.phys.spbu.ru>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
LKML-Reference: <1265223128-11786-9-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
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>
There is still some more work to do to disentangle map creation
from DSO loading, but this happens only for the kernel, and for
the early adopters of perf diff, where this disentanglement
matters most, we'll be testing different kernels, so no problem
here.
Further clarification: right now we create the kernel maps for
the various modules and discontiguous kernel text maps when
loading the DSO, we should do it as a two step process, first
creating the maps, for multiple mappings with the same DSO
store, then doing the dso load just once, for the first hit on
one of the maps sharing this DSO backing store.
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: <1260741029-4430-6-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Now we have a very high level routine for simple tools to
process IP sample events:
int event__preprocess_sample(const event_t *self,
struct addr_location *al,
symbol_filter_t filter)
It receives the event itself and will insert new threads in the
global threads list and resolve the map and symbol, filling all
this info into the new addr_location struct, so that tools like
annotate and report can further process the event by creating
hist_entries in their specific way (with or without callgraphs,
etc).
It in turn uses the new next layer function:
void thread__find_addr_location(struct thread *self, u8 cpumode,
enum map_type type, u64 addr,
struct addr_location *al,
symbol_filter_t filter)
This one will, given a thread (userspace or the kernel kthread
one), will find the given type (MAP__FUNCTION now, MAP__VARIABLE
too in the near future) at the given cpumode, taking vdsos into
account (userspace hit, but kernel symbol) and will fill all
these details in the addr_location given.
Tools that need a more compact API for plain function
resolution, like 'kmem', can use this other one:
struct symbol *thread__find_function(struct thread *self, u64 addr,
symbol_filter_t filter)
So, to resolve a kernel symbol, that is all the 'kmem' tool
needs, its just a matter of calling:
sym = thread__find_function(kthread, addr, NULL);
The 'filter' parameter is needed because we do lazy
parsing/loading of ELF symtabs or /proc/kallsyms.
With this we remove more code duplication all around, which is
always good, huh? :-)
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: John Kacur <jkacur@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1259346563-12568-12-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
While implementing event__preprocess_sample, that will do all of
the symbol lookup in one convenient function, I noticed that
util/process_event.[ch] were not being used at all, then started
looking if there were other functions that could be shared
and...
All those functions really don't need to receive offset + head,
the only thing they did was common to all of them, so do it at
one place instead.
Stats about number of each type of event processed now is done
in a central place.
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frédéric Weisbecker <fweisbec@gmail.com>
Cc: John Kacur <jkacur@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
LKML-Reference: <1259346563-12568-11-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Making the routines that were so far specific to the kernel maps
useful for all threads.
This is done by making the kernel maps be contained in a kernel
"thread".
This gets the kernel specific routines closer to the userspace
counterparts, which will help in reducing the boilerplate for
resolving a symbol, as will be demonstrated in the next patches.
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: <1259346563-12568-9-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
By using an array of rb_roots in struct dso we can, from a
struct map instance to get the right symbol rb_tree more easily.
This way we can have just one symbol lookup method for struct
map instances, map__find_symbol, instead of one per symtab type
(functions, variables).
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: <1259346563-12568-6-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
And also express its configuration toggles via a struct.
Now all one has to do is to call symbol__init(NULL) if the
defaults are OK, or pass a struct symbol_conf pointer with the
desired configuration.
If a tool uses kernel_maps__find_symbol() to look at the kernel
and modules mappings for a symbol but didn't call symbol__init()
first, that will generate a one time warning too, alerting the
subcommand developer that symbol__init() must be called.
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: <1259071517-3242-2-git-send-email-acme@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
