linux/tools/perf/builtin-record.c

905 lines
21 KiB
C
Raw Normal View History

/*
* builtin-record.c
*
* Builtin record command: Record the profile of a workload
* (or a CPU, or a PID) into the perf.data output file - for
* later analysis via perf report.
*/
#define _FILE_OFFSET_BITS 64
#include "builtin.h"
#include "perf.h"
#include "util/build-id.h"
#include "util/util.h"
#include "util/parse-options.h"
#include "util/parse-events.h"
#include "util/header.h"
#include "util/event.h"
#include "util/debug.h"
#include "util/session.h"
perf symbols: Use the buildids if present With this change 'perf record' will intercept PERF_RECORD_MMAP calls, creating a linked list of DSOs, then when the session finishes, it will traverse this list and read the buildids, stashing them at the end of the file and will set up a new feature bit in the header bitmask. 'perf report' will then notice this feature and populate the 'dsos' list and set the build ids. When reading the symtabs it will refuse to load from a file that doesn't have the same build id. This improves the reliability of the profiler output, as symbols and profiling data is more guaranteed to match. Example: [root@doppio ~]# perf report | head /home/acme/bin/perf with build id b1ea544ac3746e7538972548a09aadecc5753868 not found, continuing without symbols # Samples: 2621434559 # # Overhead Command Shared Object Symbol # ........ ............... ............................. ...... # 7.91% init [kernel] [k] read_hpet 7.64% init [kernel] [k] mwait_idle_with_hints 7.60% swapper [kernel] [k] read_hpet 7.60% swapper [kernel] [k] mwait_idle_with_hints 3.65% init [kernel] [k] 0xffffffffa02339d9 [root@doppio ~]# In this case the 'perf' binary was an older one, vanished, so its symbols probably wouldn't match or would cause subtly different (and misleading) output. Next patches will support the kernel as well, reading the build id notes for it and the modules from /sys. Another patch should also introduce a new plumbing command: 'perf list-buildids' that will then be used in porcelain that is distro specific to fetch -debuginfo packages where such buildids are present. This will in turn allow for one to run 'perf record' in one machine and 'perf report' in another. Future work on having the buildid sent directly from the kernel in the PERF_RECORD_MMAP event is needed to close races, as the DSO can be changed during a 'perf record' session, but this patch at least helps with non-corner cases and current/older kernels. Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Frank Ch. Eigler <fche@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Jason Baron <jbaron@redhat.com> Cc: Jim Keniston <jkenisto@us.ibm.com> Cc: K. Prasad <prasad@linux.vnet.ibm.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Roland McGrath <roland@redhat.com> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Cc: Steven Rostedt <rostedt@goodmis.org> LKML-Reference: <1257367843-26224-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-04 20:50:43 +00:00
#include "util/symbol.h"
perf tools: Fix sparse CPU numbering related bugs 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>
2010-03-10 09:36:09 +00:00
#include "util/cpumap.h"
#include <unistd.h>
#include <sched.h>
#include <sys/mman.h>
enum write_mode_t {
WRITE_FORCE,
WRITE_APPEND
};
static int *fd[MAX_NR_CPUS][MAX_COUNTERS];
static u64 user_interval = ULLONG_MAX;
static u64 default_interval = 0;
static int nr_cpus = 0;
static unsigned int page_size;
static unsigned int mmap_pages = 128;
static unsigned int user_freq = UINT_MAX;
static int freq = 1000;
static int output;
static int pipe_output = 0;
static const char *output_name = "perf.data";
static int group = 0;
static int realtime_prio = 0;
perf: Fix endianness argument compatibility with OPT_BOOLEAN() and introduce OPT_INCR() Parsing an option from the command line with OPT_BOOLEAN on a bool data type would not work on a big-endian machine due to the manner in which the boolean was being cast into an int and incremented. For example, running 'perf probe --list' on a PowerPC machine would fail to properly set the list_events bool and would therefore print out the usage information and terminate. This patch makes OPT_BOOLEAN work as expected with a bool datatype. For cases where the original OPT_BOOLEAN was intentionally being used to increment an int each time it was passed in on the command line, this patch introduces OPT_INCR with the old behaviour of OPT_BOOLEAN (the verbose variable is currently the only such example of this). I have reviewed every use of OPT_BOOLEAN to verify that a true C99 bool was passed. Where integers were used, I verified that they were only being used for boolean logic and changed them to bools to ensure that they would not be mistakenly used as ints. The major exception was the verbose variable which now uses OPT_INCR instead of OPT_BOOLEAN. Signed-off-by: Ian Munsie <imunsie@au.ibm.com> Acked-by: David S. Miller <davem@davemloft.net> Cc: <stable@kernel.org> # NOTE: wont apply to .3[34].x cleanly, please backport Cc: Git development list <git@vger.kernel.org> Cc: Ian Munsie <imunsie@au1.ibm.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Eric B Munson <ebmunson@us.ibm.com> Cc: Valdis.Kletnieks@vt.edu Cc: WANG Cong <amwang@redhat.com> Cc: Thiago Farina <tfransosi@gmail.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Cc: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> Cc: Mike Galbraith <efault@gmx.de> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Anton Blanchard <anton@samba.org> Cc: John Kacur <jkacur@redhat.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> LKML-Reference: <1271147857-11604-1-git-send-email-imunsie@au.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-04-13 08:37:33 +00:00
static bool raw_samples = false;
static bool system_wide = false;
static pid_t target_pid = -1;
static pid_t target_tid = -1;
static pid_t *all_tids = NULL;
static int thread_num = 0;
static pid_t child_pid = -1;
static bool no_inherit = false;
static enum write_mode_t write_mode = WRITE_FORCE;
perf: Fix endianness argument compatibility with OPT_BOOLEAN() and introduce OPT_INCR() Parsing an option from the command line with OPT_BOOLEAN on a bool data type would not work on a big-endian machine due to the manner in which the boolean was being cast into an int and incremented. For example, running 'perf probe --list' on a PowerPC machine would fail to properly set the list_events bool and would therefore print out the usage information and terminate. This patch makes OPT_BOOLEAN work as expected with a bool datatype. For cases where the original OPT_BOOLEAN was intentionally being used to increment an int each time it was passed in on the command line, this patch introduces OPT_INCR with the old behaviour of OPT_BOOLEAN (the verbose variable is currently the only such example of this). I have reviewed every use of OPT_BOOLEAN to verify that a true C99 bool was passed. Where integers were used, I verified that they were only being used for boolean logic and changed them to bools to ensure that they would not be mistakenly used as ints. The major exception was the verbose variable which now uses OPT_INCR instead of OPT_BOOLEAN. Signed-off-by: Ian Munsie <imunsie@au.ibm.com> Acked-by: David S. Miller <davem@davemloft.net> Cc: <stable@kernel.org> # NOTE: wont apply to .3[34].x cleanly, please backport Cc: Git development list <git@vger.kernel.org> Cc: Ian Munsie <imunsie@au1.ibm.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Eric B Munson <ebmunson@us.ibm.com> Cc: Valdis.Kletnieks@vt.edu Cc: WANG Cong <amwang@redhat.com> Cc: Thiago Farina <tfransosi@gmail.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Cc: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> Cc: Mike Galbraith <efault@gmx.de> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Anton Blanchard <anton@samba.org> Cc: John Kacur <jkacur@redhat.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> LKML-Reference: <1271147857-11604-1-git-send-email-imunsie@au.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-04-13 08:37:33 +00:00
static bool call_graph = false;
static bool inherit_stat = false;
static bool no_samples = false;
static bool sample_address = false;
static long samples = 0;
static u64 bytes_written = 0;
static struct pollfd *event_array;
static int nr_poll = 0;
static int nr_cpu = 0;
static int file_new = 1;
static off_t post_processing_offset;
static struct perf_session *session;
static const char *cpu_list;
struct mmap_data {
int counter;
void *base;
unsigned int mask;
unsigned int prev;
};
static struct mmap_data mmap_array[MAX_NR_CPUS];
static unsigned long mmap_read_head(struct mmap_data *md)
{
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 10:02:48 +00:00
struct perf_event_mmap_page *pc = md->base;
long head;
head = pc->data_head;
rmb();
return head;
}
static void mmap_write_tail(struct mmap_data *md, unsigned long tail)
{
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 10:02:48 +00:00
struct perf_event_mmap_page *pc = md->base;
/*
* ensure all reads are done before we write the tail out.
*/
/* mb(); */
pc->data_tail = tail;
}
static void advance_output(size_t size)
{
bytes_written += size;
}
static void write_output(void *buf, size_t size)
{
while (size) {
int ret = write(output, buf, size);
if (ret < 0)
die("failed to write");
size -= ret;
buf += ret;
bytes_written += ret;
}
}
static int process_synthesized_event(event_t *event,
struct perf_session *self __used)
{
write_output(event, event->header.size);
return 0;
}
static void mmap_read(struct mmap_data *md)
{
unsigned int head = mmap_read_head(md);
unsigned int old = md->prev;
unsigned char *data = md->base + page_size;
unsigned long size;
void *buf;
int diff;
/*
* If we're further behind than half the buffer, there's a chance
* the writer will bite our tail and mess up the samples under us.
*
* If we somehow ended up ahead of the head, we got messed up.
*
* In either case, truncate and restart at head.
*/
diff = head - old;
if (diff < 0) {
fprintf(stderr, "WARNING: failed to keep up with mmap data\n");
/*
* head points to a known good entry, start there.
*/
old = head;
}
if (old != head)
samples++;
size = head - old;
if ((old & md->mask) + size != (head & md->mask)) {
buf = &data[old & md->mask];
size = md->mask + 1 - (old & md->mask);
old += size;
write_output(buf, size);
}
buf = &data[old & md->mask];
size = head - old;
old += size;
write_output(buf, size);
md->prev = old;
mmap_write_tail(md, old);
}
static volatile int done = 0;
static volatile int signr = -1;
static void sig_handler(int sig)
{
done = 1;
signr = sig;
}
static void sig_atexit(void)
{
if (child_pid != -1)
kill(child_pid, SIGTERM);
if (signr == -1)
return;
signal(signr, SIG_DFL);
kill(getpid(), signr);
}
static int group_fd;
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 10:02:48 +00:00
static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int nr)
{
struct perf_header_attr *h_attr;
if (nr < session->header.attrs) {
h_attr = session->header.attr[nr];
} else {
h_attr = perf_header_attr__new(a);
if (h_attr != NULL)
if (perf_header__add_attr(&session->header, h_attr) < 0) {
perf_header_attr__delete(h_attr);
h_attr = NULL;
}
}
return h_attr;
}
static void create_counter(int counter, int cpu)
{
char *filter = filters[counter];
perf: Do the big rename: Performance Counters -> Performance Events Bye-bye Performance Counters, welcome Performance Events! In the past few months the perfcounters subsystem has grown out its initial role of counting hardware events, and has become (and is becoming) a much broader generic event enumeration, reporting, logging, monitoring, analysis facility. Naming its core object 'perf_counter' and naming the subsystem 'perfcounters' has become more and more of a misnomer. With pending code like hw-breakpoints support the 'counter' name is less and less appropriate. All in one, we've decided to rename the subsystem to 'performance events' and to propagate this rename through all fields, variables and API names. (in an ABI compatible fashion) The word 'event' is also a bit shorter than 'counter' - which makes it slightly more convenient to write/handle as well. Thanks goes to Stephane Eranian who first observed this misnomer and suggested a rename. User-space tooling and ABI compatibility is not affected - this patch should be function-invariant. (Also, defconfigs were not touched to keep the size down.) This patch has been generated via the following script: FILES=$(find * -type f | grep -vE 'oprofile|[^K]config') sed -i \ -e 's/PERF_EVENT_/PERF_RECORD_/g' \ -e 's/PERF_COUNTER/PERF_EVENT/g' \ -e 's/perf_counter/perf_event/g' \ -e 's/nb_counters/nb_events/g' \ -e 's/swcounter/swevent/g' \ -e 's/tpcounter_event/tp_event/g' \ $FILES for N in $(find . -name perf_counter.[ch]); do M=$(echo $N | sed 's/perf_counter/perf_event/g') mv $N $M done FILES=$(find . -name perf_event.*) sed -i \ -e 's/COUNTER_MASK/REG_MASK/g' \ -e 's/COUNTER/EVENT/g' \ -e 's/\<event\>/event_id/g' \ -e 's/counter/event/g' \ -e 's/Counter/Event/g' \ $FILES ... to keep it as correct as possible. This script can also be used by anyone who has pending perfcounters patches - it converts a Linux kernel tree over to the new naming. We tried to time this change to the point in time where the amount of pending patches is the smallest: the end of the merge window. Namespace clashes were fixed up in a preparatory patch - and some stylistic fallout will be fixed up in a subsequent patch. ( NOTE: 'counters' are still the proper terminology when we deal with hardware registers - and these sed scripts are a bit over-eager in renaming them. I've undone some of that, but in case there's something left where 'counter' would be better than 'event' we can undo that on an individual basis instead of touching an otherwise nicely automated patch. ) Suggested-by: Stephane Eranian <eranian@google.com> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Paul Mackerras <paulus@samba.org> Reviewed-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: David Howells <dhowells@redhat.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: <linux-arch@vger.kernel.org> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 10:02:48 +00:00
struct perf_event_attr *attr = attrs + counter;
struct perf_header_attr *h_attr;
int track = !counter; /* only the first counter needs these */
int thread_index;
int ret;
struct {
u64 count;
u64 time_enabled;
u64 time_running;
u64 id;
} read_data;
attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED |
PERF_FORMAT_TOTAL_TIME_RUNNING |
PERF_FORMAT_ID;
attr->sample_type |= PERF_SAMPLE_IP | PERF_SAMPLE_TID;
if (nr_counters > 1)
attr->sample_type |= PERF_SAMPLE_ID;
/*
* We default some events to a 1 default interval. But keep
* it a weak assumption overridable by the user.
*/
if (!attr->sample_period || (user_freq != UINT_MAX &&
user_interval != ULLONG_MAX)) {
if (freq) {
attr->sample_type |= PERF_SAMPLE_PERIOD;
attr->freq = 1;
attr->sample_freq = freq;
} else {
attr->sample_period = default_interval;
}
}
if (no_samples)
attr->sample_freq = 0;
if (inherit_stat)
attr->inherit_stat = 1;
if (sample_address)
attr->sample_type |= PERF_SAMPLE_ADDR;
if (call_graph)
attr->sample_type |= PERF_SAMPLE_CALLCHAIN;
if (system_wide)
attr->sample_type |= PERF_SAMPLE_CPU;
perf trace: Sample the CPU too Sample, record, parse and print the CPU field - it had all zeroes before. Before (watch the second column, the CPU values): perf-32685 [000] 0.000000: sched_wakeup_new: task perf:32686 [120] success=1 [011] perf-32685 [000] 0.000000: sched_migrate_task: task perf:32685 [120] from: 1 to: 11 perf-32685 [000] 0.000000: sched_process_fork: parent perf:32685 child perf:32686 true-32686 [000] 0.000000: sched_wakeup: task migration/11:25 [0] success=1 [011] true-32686 [000] 0.000000: sched_wakeup: task distccd:12793 [125] success=1 [015] true-32686 [000] 0.000000: sched_wakeup: task distccd:12793 [125] success=1 [015] perf-32685 [000] 0.000000: sched_switch: task perf:32685 [120] (S) ==> swapper:0 [140] true-32686 [000] 0.000000: sched_switch: task perf:32686 [120] (R) ==> migration/11:25 [0] true-32686 [000] 0.000000: sched_switch: task perf:32686 [120] (R) ==> distccd:12793 [125] true-32686 [000] 0.000000: sched_switch: task true:32686 [120] (R) ==> distccd:12793 [125] true-32686 [000] 0.000000: sched_process_exit: task true:32686 [120] true-32686 [000] 0.000000: sched_stat_wait: task: distccd:12793 wait: 6767985949080 [ns] true-32686 [000] 0.000000: sched_stat_wait: task: distccd:12793 wait: 6767986139446 [ns] true-32686 [000] 0.000000: sched_stat_sleep: task: distccd:12793 sleep: 132844 [ns] true-32686 [000] 0.000000: sched_stat_sleep: task: distccd:12793 sleep: 131724 [ns] After: perf-32685 [001] 0.000000: sched_wakeup_new: task perf:32686 [120] success=1 [011] perf-32685 [001] 0.000000: sched_migrate_task: task perf:32685 [120] from: 1 to: 11 perf-32685 [001] 0.000000: sched_process_fork: parent perf:32685 child perf:32686 true-32686 [011] 0.000000: sched_wakeup: task migration/11:25 [0] success=1 [011] true-32686 [015] 0.000000: sched_wakeup: task distccd:12793 [125] success=1 [015] true-32686 [015] 0.000000: sched_wakeup: task distccd:12793 [125] success=1 [015] perf-32685 [001] 0.000000: sched_switch: task perf:32685 [120] (S) ==> swapper:0 [140] true-32686 [011] 0.000000: sched_switch: task perf:32686 [120] (R) ==> migration/11:25 [0] true-32686 [015] 0.000000: sched_switch: task perf:32686 [120] (R) ==> distccd:12793 [125] true-32686 [015] 0.000000: sched_switch: task true:32686 [120] (R) ==> distccd:12793 [125] true-32686 [015] 0.000000: sched_process_exit: task true:32686 [120] true-32686 [015] 0.000000: sched_stat_wait: task: distccd:12793 wait: 6767985949080 [ns] true-32686 [015] 0.000000: sched_stat_wait: task: distccd:12793 wait: 6767986139446 [ns] true-32686 [015] 0.000000: sched_stat_sleep: task: distccd:12793 sleep: 132844 [ns] true-32686 [015] 0.000000: sched_stat_sleep: task: distccd:12793 sleep: 131724 [ns] So we can now see how this workload migrated between CPUs. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-02 18:20:38 +00:00
if (raw_samples) {
perf trace: Sample timestamps as well Before: perf-21082 [013] 0.000000: sched_wakeup_new: task perf:21083 [120] success=1 [015] perf-21082 [013] 0.000000: sched_migrate_task: task perf:21082 [120] from: 13 to: 15 perf-21082 [013] 0.000000: sched_process_fork: parent perf:21082 child perf:21083 true-21083 [015] 0.000000: sched_wakeup: task migration/15:33 [0] success=1 [015] perf-21082 [013] 0.000000: sched_switch: task perf:21082 [120] (S) ==> swapper:0 [140] true-21083 [015] 0.000000: sched_switch: task perf:21083 [120] (R) ==> migration/15:33 [0] true-21083 [011] 0.000000: sched_process_exit: task true:21083 [120] After: perf-21082 [013] 14674.797613: sched_wakeup_new: task perf:21083 [120] success=1 [015] perf-21082 [013] 14674.797506: sched_migrate_task: task perf:21082 [120] from: 13 to: 15 perf-21082 [013] 14674.797610: sched_process_fork: parent perf:21082 child perf:21083 true-21083 [015] 14674.797725: sched_wakeup: task migration/15:33 [0] success=1 [015] perf-21082 [013] 14674.797722: sched_switch: task perf:21082 [120] (S) ==> swapper:0 [140] true-21083 [015] 14674.797729: sched_switch: task perf:21083 [120] (R) ==> migration/15:33 [0] true-21083 [011] 14674.798159: sched_process_exit: task true:21083 [120] Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-03 10:00:22 +00:00
attr->sample_type |= PERF_SAMPLE_TIME;
attr->sample_type |= PERF_SAMPLE_RAW;
perf trace: Sample the CPU too Sample, record, parse and print the CPU field - it had all zeroes before. Before (watch the second column, the CPU values): perf-32685 [000] 0.000000: sched_wakeup_new: task perf:32686 [120] success=1 [011] perf-32685 [000] 0.000000: sched_migrate_task: task perf:32685 [120] from: 1 to: 11 perf-32685 [000] 0.000000: sched_process_fork: parent perf:32685 child perf:32686 true-32686 [000] 0.000000: sched_wakeup: task migration/11:25 [0] success=1 [011] true-32686 [000] 0.000000: sched_wakeup: task distccd:12793 [125] success=1 [015] true-32686 [000] 0.000000: sched_wakeup: task distccd:12793 [125] success=1 [015] perf-32685 [000] 0.000000: sched_switch: task perf:32685 [120] (S) ==> swapper:0 [140] true-32686 [000] 0.000000: sched_switch: task perf:32686 [120] (R) ==> migration/11:25 [0] true-32686 [000] 0.000000: sched_switch: task perf:32686 [120] (R) ==> distccd:12793 [125] true-32686 [000] 0.000000: sched_switch: task true:32686 [120] (R) ==> distccd:12793 [125] true-32686 [000] 0.000000: sched_process_exit: task true:32686 [120] true-32686 [000] 0.000000: sched_stat_wait: task: distccd:12793 wait: 6767985949080 [ns] true-32686 [000] 0.000000: sched_stat_wait: task: distccd:12793 wait: 6767986139446 [ns] true-32686 [000] 0.000000: sched_stat_sleep: task: distccd:12793 sleep: 132844 [ns] true-32686 [000] 0.000000: sched_stat_sleep: task: distccd:12793 sleep: 131724 [ns] After: perf-32685 [001] 0.000000: sched_wakeup_new: task perf:32686 [120] success=1 [011] perf-32685 [001] 0.000000: sched_migrate_task: task perf:32685 [120] from: 1 to: 11 perf-32685 [001] 0.000000: sched_process_fork: parent perf:32685 child perf:32686 true-32686 [011] 0.000000: sched_wakeup: task migration/11:25 [0] success=1 [011] true-32686 [015] 0.000000: sched_wakeup: task distccd:12793 [125] success=1 [015] true-32686 [015] 0.000000: sched_wakeup: task distccd:12793 [125] success=1 [015] perf-32685 [001] 0.000000: sched_switch: task perf:32685 [120] (S) ==> swapper:0 [140] true-32686 [011] 0.000000: sched_switch: task perf:32686 [120] (R) ==> migration/11:25 [0] true-32686 [015] 0.000000: sched_switch: task perf:32686 [120] (R) ==> distccd:12793 [125] true-32686 [015] 0.000000: sched_switch: task true:32686 [120] (R) ==> distccd:12793 [125] true-32686 [015] 0.000000: sched_process_exit: task true:32686 [120] true-32686 [015] 0.000000: sched_stat_wait: task: distccd:12793 wait: 6767985949080 [ns] true-32686 [015] 0.000000: sched_stat_wait: task: distccd:12793 wait: 6767986139446 [ns] true-32686 [015] 0.000000: sched_stat_sleep: task: distccd:12793 sleep: 132844 [ns] true-32686 [015] 0.000000: sched_stat_sleep: task: distccd:12793 sleep: 131724 [ns] So we can now see how this workload migrated between CPUs. Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-02 18:20:38 +00:00
attr->sample_type |= PERF_SAMPLE_CPU;
}
perf_counter: Fix/complete ftrace event records sampling This patch implements the kernel side support for ftrace event record sampling. A new counter sampling attribute is added: PERF_SAMPLE_TP_RECORD which requests ftrace events record sampling. In this case if a PERF_TYPE_TRACEPOINT counter is active and a tracepoint fires, we emit the tracepoint binary record to the perfcounter event buffer, as a sample. Result, after setting PERF_SAMPLE_TP_RECORD attribute from perf record: perf record -f -F 1 -a -e workqueue:workqueue_execution perf report -D 0x21e18 [0x48]: event: 9 . . ... raw event: size 72 bytes . 0000: 09 00 00 00 01 00 48 00 d0 c7 00 81 ff ff ff ff ......H........ . 0010: 0a 00 00 00 0a 00 00 00 21 00 00 00 00 00 00 00 ........!...... . 0020: 2b 00 01 02 0a 00 00 00 0a 00 00 00 65 76 65 6e +...........eve . 0030: 74 73 2f 31 00 00 00 00 00 00 00 00 0a 00 00 00 ts/1........... . 0040: e0 b1 31 81 ff ff ff ff ....... . 0x21e18 [0x48]: PERF_EVENT_SAMPLE (IP, 1): 10: 0xffffffff8100c7d0 period: 33 The raw ftrace binary record starts at offset 0020. Translation: struct trace_entry { type = 0x2b = 43; flags = 1; preempt_count = 2; pid = 0xa = 10; tgid = 0xa = 10; } thread_comm = "events/1" thread_pid = 0xa = 10; func = 0xffffffff8131b1e0 = flush_to_ldisc() What will come next? - Userspace support ('perf trace'), 'flight data recorder' mode for perf trace, etc. - The unconditional copy from the profiling callback brings some costs however if someone wants no such sampling to occur, and needs to be fixed in the future. For that we need to have an instant access to the perf counter attribute. This is a matter of a flag to add in the struct ftrace_event. - Take care of the events recursivity! Don't ever try to record a lock event for example, it seems some locking is used in the profiling fast path and lead to a tracing recursivity. That will be fixed using raw spinlock or recursivity protection. - [...] - Profit! :-) Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Pekka Enberg <penberg@cs.helsinki.fi> Cc: Gabriel Munteanu <eduard.munteanu@linux360.ro> Cc: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-06 23:25:54 +00:00
attr->mmap = track;
attr->comm = track;
attr->inherit = !no_inherit;
if (target_pid == -1 && target_tid == -1 && !system_wide) {
attr->disabled = 1;
attr->enable_on_exec = 1;
}
for (thread_index = 0; thread_index < thread_num; thread_index++) {
try_again:
fd[nr_cpu][counter][thread_index] = sys_perf_event_open(attr,
all_tids[thread_index], cpu, group_fd, 0);
if (fd[nr_cpu][counter][thread_index] < 0) {
int err = errno;
if (err == EPERM || err == EACCES)
die("Permission error - are you root?\n"
"\t Consider tweaking"
" /proc/sys/kernel/perf_event_paranoid.\n");
else if (err == ENODEV && cpu_list) {
die("No such device - did you specify"
" an out-of-range profile CPU?\n");
}
/*
* If it's cycles then fall back to hrtimer
* based cpu-clock-tick sw counter, which
* is always available even if no PMU support:
*/
if (attr->type == PERF_TYPE_HARDWARE
&& attr->config == PERF_COUNT_HW_CPU_CYCLES) {
if (verbose)
warning(" ... trying to fall back to cpu-clock-ticks\n");
attr->type = PERF_TYPE_SOFTWARE;
attr->config = PERF_COUNT_SW_CPU_CLOCK;
goto try_again;
}
printf("\n");
error("perfcounter syscall returned with %d (%s)\n",
fd[nr_cpu][counter][thread_index], strerror(err));
#if defined(__i386__) || defined(__x86_64__)
if (attr->type == PERF_TYPE_HARDWARE && err == EOPNOTSUPP)
die("No hardware sampling interrupt available."
" No APIC? If so then you can boot the kernel"
" with the \"lapic\" boot parameter to"
" force-enable it.\n");
#endif
die("No CONFIG_PERF_EVENTS=y kernel support configured?\n");
exit(-1);
}
h_attr = get_header_attr(attr, counter);
if (h_attr == NULL)
die("nomem\n");
if (!file_new) {
if (memcmp(&h_attr->attr, attr, sizeof(*attr))) {
fprintf(stderr, "incompatible append\n");
exit(-1);
}
}
if (read(fd[nr_cpu][counter][thread_index], &read_data, sizeof(read_data)) == -1) {
perror("Unable to read perf file descriptor\n");
exit(-1);
}
if (perf_header_attr__add_id(h_attr, read_data.id) < 0) {
pr_warning("Not enough memory to add id\n");
exit(-1);
}
assert(fd[nr_cpu][counter][thread_index] >= 0);
fcntl(fd[nr_cpu][counter][thread_index], F_SETFL, O_NONBLOCK);
/*
* First counter acts as the group leader:
*/
if (group && group_fd == -1)
group_fd = fd[nr_cpu][counter][thread_index];
if (counter || thread_index) {
ret = ioctl(fd[nr_cpu][counter][thread_index],
PERF_EVENT_IOC_SET_OUTPUT,
fd[nr_cpu][0][0]);
if (ret) {
error("failed to set output: %d (%s)\n", errno,
strerror(errno));
exit(-1);
}
} else {
mmap_array[nr_cpu].counter = counter;
mmap_array[nr_cpu].prev = 0;
mmap_array[nr_cpu].mask = mmap_pages*page_size - 1;
mmap_array[nr_cpu].base = mmap(NULL, (mmap_pages+1)*page_size,
PROT_READ|PROT_WRITE, MAP_SHARED, fd[nr_cpu][counter][thread_index], 0);
if (mmap_array[nr_cpu].base == MAP_FAILED) {
error("failed to mmap with %d (%s)\n", errno, strerror(errno));
exit(-1);
}
event_array[nr_poll].fd = fd[nr_cpu][counter][thread_index];
event_array[nr_poll].events = POLLIN;
nr_poll++;
perf tools: Implement counter output multiplexing Finish the -M/--multiplex option implementation: - separate it out from group_fd - correctly set it via the ioctl and dont mmap counters that are multiplexed - modify the perf record event loop to deal with buffer-less counters. - remove the -g option from perf sched record - account for unordered events in perf sched latency - (add -f to perf sched record to ease measurements) - skip idle threads (pid==0) in latency output The result is better latency output by 'perf sched latency': ----------------------------------------------------------------------------------- Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | ----------------------------------------------------------------------------------- ksoftirqd/8 | 0.071 ms | 2 | avg: 0.458 ms | max: 0.913 ms | at-spi-registry | 0.609 ms | 19 | avg: 0.013 ms | max: 0.023 ms | perf | 3.316 ms | 16 | avg: 0.013 ms | max: 0.054 ms | Xorg | 0.392 ms | 19 | avg: 0.011 ms | max: 0.018 ms | sleep | 0.537 ms | 2 | avg: 0.009 ms | max: 0.009 ms | ----------------------------------------------------------------------------------- TOTAL: | 4.925 ms | 58 | --------------------------------------------- Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Mike Galbraith <efault@gmx.de> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frederic Weisbecker <fweisbec@gmail.com> LKML-Reference: <new-submission> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 16:15:54 +00:00
}
if (filter != NULL) {
ret = ioctl(fd[nr_cpu][counter][thread_index],
PERF_EVENT_IOC_SET_FILTER, filter);
if (ret) {
error("failed to set filter with %d (%s)\n", errno,
strerror(errno));
exit(-1);
}
}
}
}
static void open_counters(int cpu)
{
int counter;
group_fd = -1;
for (counter = 0; counter < nr_counters; counter++)
create_counter(counter, cpu);
nr_cpu++;
}
static int process_buildids(void)
{
u64 size = lseek(output, 0, SEEK_CUR);
if (size == 0)
return 0;
session->fd = output;
return __perf_session__process_events(session, post_processing_offset,
size - post_processing_offset,
size, &build_id__mark_dso_hit_ops);
}
static void atexit_header(void)
{
if (!pipe_output) {
session->header.data_size += bytes_written;
process_buildids();
perf_header__write(&session->header, output, true);
}
}
static void event__synthesize_guest_os(struct machine *machine, void *data)
{
int err;
char *guest_kallsyms;
char path[PATH_MAX];
struct perf_session *psession = data;
if (machine__is_host(machine))
return;
/*
*As for guest kernel when processing subcommand record&report,
*we arrange module mmap prior to guest kernel mmap and trigger
*a preload dso because default guest module symbols are loaded
*from guest kallsyms instead of /lib/modules/XXX/XXX. This
*method is used to avoid symbol missing when the first addr is
*in module instead of in guest kernel.
*/
err = event__synthesize_modules(process_synthesized_event,
psession, machine);
if (err < 0)
pr_err("Couldn't record guest kernel [%d]'s reference"
" relocation symbol.\n", machine->pid);
if (machine__is_default_guest(machine))
guest_kallsyms = (char *) symbol_conf.default_guest_kallsyms;
else {
sprintf(path, "%s/proc/kallsyms", machine->root_dir);
guest_kallsyms = path;
}
/*
* We use _stext for guest kernel because guest kernel's /proc/kallsyms
* have no _text sometimes.
*/
err = event__synthesize_kernel_mmap(process_synthesized_event,
psession, machine, "_text");
if (err < 0)
err = event__synthesize_kernel_mmap(process_synthesized_event,
psession, machine, "_stext");
if (err < 0)
pr_err("Couldn't record guest kernel [%d]'s reference"
" relocation symbol.\n", machine->pid);
}
static struct perf_event_header finished_round_event = {
.size = sizeof(struct perf_event_header),
.type = PERF_RECORD_FINISHED_ROUND,
};
static void mmap_read_all(void)
{
int i;
for (i = 0; i < nr_cpu; i++) {
if (mmap_array[i].base)
mmap_read(&mmap_array[i]);
}
if (perf_header__has_feat(&session->header, HEADER_TRACE_INFO))
write_output(&finished_round_event, sizeof(finished_round_event));
}
static int __cmd_record(int argc, const char **argv)
{
int i, counter;
struct stat st;
int flags;
int err;
unsigned long waking = 0;
int child_ready_pipe[2], go_pipe[2];
const bool forks = argc > 0;
char buf;
struct machine *machine;
page_size = sysconf(_SC_PAGE_SIZE);
atexit(sig_atexit);
signal(SIGCHLD, sig_handler);
signal(SIGINT, sig_handler);
if (forks && (pipe(child_ready_pipe) < 0 || pipe(go_pipe) < 0)) {
perror("failed to create pipes");
exit(-1);
}
if (!strcmp(output_name, "-"))
pipe_output = 1;
else if (!stat(output_name, &st) && st.st_size) {
if (write_mode == WRITE_FORCE) {
char oldname[PATH_MAX];
snprintf(oldname, sizeof(oldname), "%s.old",
output_name);
unlink(oldname);
rename(output_name, oldname);
}
} else if (write_mode == WRITE_APPEND) {
write_mode = WRITE_FORCE;
}
flags = O_CREAT|O_RDWR;
if (write_mode == WRITE_APPEND)
file_new = 0;
else
flags |= O_TRUNC;
if (pipe_output)
output = STDOUT_FILENO;
else
output = open(output_name, flags, S_IRUSR | S_IWUSR);
if (output < 0) {
perror("failed to create output file");
exit(-1);
}
session = perf_session__new(output_name, O_WRONLY,
perf: add perf-inject builtin 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>
2010-05-01 06:41:20 +00:00
write_mode == WRITE_FORCE, false);
if (session == NULL) {
pr_err("Not enough memory for reading perf file header\n");
return -1;
}
if (!file_new) {
err = perf_header__read(session, output);
if (err < 0)
return err;
}
if (have_tracepoints(attrs, nr_counters))
perf_header__set_feat(&session->header, HEADER_TRACE_INFO);
atexit(atexit_header);
if (forks) {
child_pid = fork();
if (child_pid < 0) {
perror("failed to fork");
exit(-1);
}
if (!child_pid) {
if (pipe_output)
dup2(2, 1);
close(child_ready_pipe[0]);
close(go_pipe[1]);
fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC);
/*
* Do a dummy execvp to get the PLT entry resolved,
* so we avoid the resolver overhead on the real
* execvp call.
*/
execvp("", (char **)argv);
/*
* Tell the parent we're ready to go
*/
close(child_ready_pipe[1]);
/*
* Wait until the parent tells us to go.
*/
if (read(go_pipe[0], &buf, 1) == -1)
perror("unable to read pipe");
execvp(argv[0], (char **)argv);
perror(argv[0]);
exit(-1);
}
if (!system_wide && target_tid == -1 && target_pid == -1)
all_tids[0] = child_pid;
close(child_ready_pipe[1]);
close(go_pipe[0]);
/*
* wait for child to settle
*/
if (read(child_ready_pipe[0], &buf, 1) == -1) {
perror("unable to read pipe");
exit(-1);
}
close(child_ready_pipe[0]);
}
nr_cpus = read_cpu_map(cpu_list);
if (nr_cpus < 1) {
perror("failed to collect number of CPUs\n");
return -1;
}
if (!system_wide && no_inherit && !cpu_list) {
open_counters(-1);
} else {
for (i = 0; i < nr_cpus; i++)
open_counters(cpumap[i]);
}
if (pipe_output) {
err = perf_header__write_pipe(output);
if (err < 0)
return err;
} else if (file_new) {
err = perf_header__write(&session->header, output, false);
if (err < 0)
return err;
perf tools: Handle relocatable kernels 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>
2010-01-05 18:50:31 +00:00
}
post_processing_offset = lseek(output, 0, SEEK_CUR);
if (pipe_output) {
err = event__synthesize_attrs(&session->header,
process_synthesized_event,
session);
if (err < 0) {
pr_err("Couldn't synthesize attrs.\n");
return err;
}
err = event__synthesize_event_types(process_synthesized_event,
session);
if (err < 0) {
pr_err("Couldn't synthesize event_types.\n");
return err;
}
if (have_tracepoints(attrs, nr_counters)) {
/*
* FIXME err <= 0 here actually means that
* there were no tracepoints so its not really
* an error, just that we don't need to
* synthesize anything. We really have to
* return this more properly and also
* propagate errors that now are calling die()
*/
err = event__synthesize_tracing_data(output, attrs,
nr_counters,
process_synthesized_event,
session);
if (err <= 0) {
pr_err("Couldn't record tracing data.\n");
return err;
}
advance_output(err);
}
}
machine = perf_session__find_host_machine(session);
if (!machine) {
pr_err("Couldn't find native kernel information.\n");
return -1;
}
perf tools: Handle relocatable kernels 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>
2010-01-05 18:50:31 +00:00
err = event__synthesize_kernel_mmap(process_synthesized_event,
session, machine, "_text");
if (err < 0)
err = event__synthesize_kernel_mmap(process_synthesized_event,
session, machine, "_stext");
perf tools: Handle relocatable kernels 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>
2010-01-05 18:50:31 +00:00
if (err < 0) {
pr_err("Couldn't record kernel reference relocation symbol.\n");
return err;
}
err = event__synthesize_modules(process_synthesized_event,
session, machine);
if (err < 0) {
pr_err("Couldn't record kernel reference relocation symbol.\n");
return err;
}
if (perf_guest)
perf_session__process_machines(session, event__synthesize_guest_os);
if (!system_wide && cpu_list)
event__synthesize_thread(target_tid, process_synthesized_event,
session);
else
event__synthesize_threads(process_synthesized_event, session);
if (realtime_prio) {
struct sched_param param;
param.sched_priority = realtime_prio;
if (sched_setscheduler(0, SCHED_FIFO, &param)) {
pr_err("Could not set realtime priority.\n");
exit(-1);
}
}
/*
* Let the child rip
*/
if (forks)
close(go_pipe[1]);
for (;;) {
int hits = samples;
int thread;
mmap_read_all();
if (hits == samples) {
if (done)
break;
err = poll(event_array, nr_poll, -1);
waking++;
}
if (done) {
for (i = 0; i < nr_cpu; i++) {
for (counter = 0;
counter < nr_counters;
counter++) {
for (thread = 0;
thread < thread_num;
thread++)
ioctl(fd[i][counter][thread],
PERF_EVENT_IOC_DISABLE);
}
}
}
}
fprintf(stderr, "[ perf record: Woken up %ld times to write data ]\n", waking);
/*
* Approximate RIP event size: 24 bytes.
*/
fprintf(stderr,
"[ perf record: Captured and wrote %.3f MB %s (~%lld samples) ]\n",
(double)bytes_written / 1024.0 / 1024.0,
output_name,
bytes_written / 24);
return 0;
}
static const char * const record_usage[] = {
"perf record [<options>] [<command>]",
"perf record [<options>] -- <command> [<options>]",
NULL
};
static bool force, append_file;
static const struct option options[] = {
OPT_CALLBACK('e', "event", NULL, "event",
"event selector. use 'perf list' to list available events",
parse_events),
OPT_CALLBACK(0, "filter", NULL, "filter",
"event filter", parse_filter),
OPT_INTEGER('p', "pid", &target_pid,
"record events on existing process id"),
OPT_INTEGER('t', "tid", &target_tid,
"record events on existing thread id"),
OPT_INTEGER('r', "realtime", &realtime_prio,
"collect data with this RT SCHED_FIFO priority"),
OPT_BOOLEAN('R', "raw-samples", &raw_samples,
"collect raw sample records from all opened counters"),
OPT_BOOLEAN('a', "all-cpus", &system_wide,
"system-wide collection from all CPUs"),
OPT_BOOLEAN('A', "append", &append_file,
"append to the output file to do incremental profiling"),
OPT_STRING('C', "cpu", &cpu_list, "cpu",
"list of cpus to monitor"),
OPT_BOOLEAN('f', "force", &force,
"overwrite existing data file (deprecated)"),
OPT_U64('c', "count", &user_interval, "event period to sample"),
OPT_STRING('o', "output", &output_name, "file",
"output file name"),
OPT_BOOLEAN('i', "no-inherit", &no_inherit,
"child tasks do not inherit counters"),
OPT_UINTEGER('F', "freq", &user_freq, "profile at this frequency"),
OPT_UINTEGER('m', "mmap-pages", &mmap_pages, "number of mmap data pages"),
OPT_BOOLEAN('g', "call-graph", &call_graph,
"do call-graph (stack chain/backtrace) recording"),
perf: Fix endianness argument compatibility with OPT_BOOLEAN() and introduce OPT_INCR() Parsing an option from the command line with OPT_BOOLEAN on a bool data type would not work on a big-endian machine due to the manner in which the boolean was being cast into an int and incremented. For example, running 'perf probe --list' on a PowerPC machine would fail to properly set the list_events bool and would therefore print out the usage information and terminate. This patch makes OPT_BOOLEAN work as expected with a bool datatype. For cases where the original OPT_BOOLEAN was intentionally being used to increment an int each time it was passed in on the command line, this patch introduces OPT_INCR with the old behaviour of OPT_BOOLEAN (the verbose variable is currently the only such example of this). I have reviewed every use of OPT_BOOLEAN to verify that a true C99 bool was passed. Where integers were used, I verified that they were only being used for boolean logic and changed them to bools to ensure that they would not be mistakenly used as ints. The major exception was the verbose variable which now uses OPT_INCR instead of OPT_BOOLEAN. Signed-off-by: Ian Munsie <imunsie@au.ibm.com> Acked-by: David S. Miller <davem@davemloft.net> Cc: <stable@kernel.org> # NOTE: wont apply to .3[34].x cleanly, please backport Cc: Git development list <git@vger.kernel.org> Cc: Ian Munsie <imunsie@au1.ibm.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Eric B Munson <ebmunson@us.ibm.com> Cc: Valdis.Kletnieks@vt.edu Cc: WANG Cong <amwang@redhat.com> Cc: Thiago Farina <tfransosi@gmail.com> Cc: Masami Hiramatsu <mhiramat@redhat.com> Cc: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> Cc: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Cc: Arjan van de Ven <arjan@linux.intel.com> Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> Cc: Mike Galbraith <efault@gmx.de> Cc: Tom Zanussi <tzanussi@gmail.com> Cc: Anton Blanchard <anton@samba.org> Cc: John Kacur <jkacur@redhat.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Steven Rostedt <rostedt@goodmis.org> LKML-Reference: <1271147857-11604-1-git-send-email-imunsie@au.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-04-13 08:37:33 +00:00
OPT_INCR('v', "verbose", &verbose,
"be more verbose (show counter open errors, etc)"),
OPT_BOOLEAN('s', "stat", &inherit_stat,
"per thread counts"),
OPT_BOOLEAN('d', "data", &sample_address,
"Sample addresses"),
OPT_BOOLEAN('n', "no-samples", &no_samples,
"don't sample"),
OPT_END()
};
int cmd_record(int argc, const char **argv, const char *prefix __used)
{
int i,j;
argc = parse_options(argc, argv, options, record_usage,
PARSE_OPT_STOP_AT_NON_OPTION);
if (!argc && target_pid == -1 && target_tid == -1 &&
!system_wide && !cpu_list)
usage_with_options(record_usage, options);
if (force && append_file) {
fprintf(stderr, "Can't overwrite and append at the same time."
" You need to choose between -f and -A");
usage_with_options(record_usage, options);
} else if (append_file) {
write_mode = WRITE_APPEND;
} else {
write_mode = WRITE_FORCE;
}
symbol__init();
if (!nr_counters) {
nr_counters = 1;
attrs[0].type = PERF_TYPE_HARDWARE;
attrs[0].config = PERF_COUNT_HW_CPU_CYCLES;
}
if (target_pid != -1) {
target_tid = target_pid;
thread_num = find_all_tid(target_pid, &all_tids);
if (thread_num <= 0) {
fprintf(stderr, "Can't find all threads of pid %d\n",
target_pid);
usage_with_options(record_usage, options);
}
} else {
all_tids=malloc(sizeof(pid_t));
if (!all_tids)
return -ENOMEM;
all_tids[0] = target_tid;
thread_num = 1;
}
for (i = 0; i < MAX_NR_CPUS; i++) {
for (j = 0; j < MAX_COUNTERS; j++) {
fd[i][j] = malloc(sizeof(int)*thread_num);
if (!fd[i][j])
return -ENOMEM;
}
}
event_array = malloc(
sizeof(struct pollfd)*MAX_NR_CPUS*MAX_COUNTERS*thread_num);
if (!event_array)
return -ENOMEM;
if (user_interval != ULLONG_MAX)
default_interval = user_interval;
if (user_freq != UINT_MAX)
freq = user_freq;
/*
* User specified count overrides default frequency.
*/
if (default_interval)
freq = 0;
else if (freq) {
default_interval = freq;
} else {
fprintf(stderr, "frequency and count are zero, aborting\n");
exit(EXIT_FAILURE);
}
return __cmd_record(argc, argv);
}