perf evsel: Introduce fprintf_callchain() method out of fprintf_sym()

In 'perf trace' we're just interested in printing callchains, and we
don't want to use the symbol_conf.use_callchain, so move the callchain
part to a new method.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-kcn3romzivcpxb3u75s9nz33@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf evsel: Rename print_ip() to fprintf_sym()

As it receives a FILE, and its more than just the IP, which can even be
requested not to be printed.

For consistency with other similar methods in tools/perf/, name it as
perf_evsel__fprintf_sym() and make it return the number of bytes
printed, just like 'fprintf(3)'

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-84gawlqa3lhk63nf0t9vnqnn@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Add support for printing call chains on sys_exit events.

Now, one can print the call chain for every encountered sys_exit event,
e.g.:

    $ perf trace -e nanosleep --call-graph dwarf path/to/ex_sleep
    1005.757 (1000.090 ms): ex_sleep/13167 nanosleep(...) = 0
                                             syscall_slow_exit_work ([kernel.kallsyms])
                                             syscall_return_slowpath ([kernel.kallsyms])
                                             int_ret_from_sys_call ([kernel.kallsyms])
                                             __nanosleep (/usr/lib/libc-2.23.so)
                                             [unknown] (/usr/lib/libQt5Core.so.5.6.0)
                                             QThread::sleep (/usr/lib/libQt5Core.so.5.6.0)
                                             main (path/to/ex_sleep)
                                             __libc_start_main (/usr/lib/libc-2.23.so)
                                             _start (path/to/ex_sleep)

Note that it is advised to increase the number of mmap pages to prevent
event losses when using this new feature. Often, adding `-m 10M` to the
`perf trace` invocation is enough.

This feature is also available in strace when built with libunwind via
`strace -k`. Performance wise, this solution is much better:

    $ time find path/to/linux &> /dev/null

    real    0m0.051s
    user    0m0.013s
    sys     0m0.037s

    $ time perf trace -m 800M --call-graph dwarf find path/to/linux &> /dev/null

    real    0m2.624s
    user    0m1.203s
    sys     0m1.333s

    $ time strace -k find path/to/linux  &> /dev/null

    real    0m35.398s
    user    0m10.403s
    sys     0m23.173s

Note that it is currently not possible to configure the print output.
Adding such a feature, similar to what is available in `perf script` via
its `--fields` knob can be added later on.

Signed-off-by: Milian Wolff <milian.wolff@kdab.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
LPU-Reference: 1460115255-17648-1-git-send-email-milian.wolff@kdab.com
[ Split from a larger patch, do not print the IP, left align,
  remove dup call symbol__init(), added man page entry ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf evsel: Allow passing a left alignment when printing a symbol

For callchains, etc where we want it to align just below the syscall
name, for instance, in 'perf trace'

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-uk9ekchd67651c625ltaur5y@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf evsel: Allow specifying a file to output in perf_evsel__print_ip

As this function will be used in 'perf trace'.

Cc: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/n/tip-8x297v9utnxq77onikevvlse@git.kernel.org
[ Split from a larger patch ]
Signed-off-by: Milian Wolff <milian.wolff@kdab.com>

perf bpf: Automatically create bpf-output event __bpf_stdout__

This patch removes the need to set a bpf-output event in cmdline.  By
referencing a map named '__bpf_stdout__', perf automatically creates an
event for it.

For example:

  # perf record -e ./test_bpf_trace.c usleep 100000
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.012 MB perf.data (2 samples) ]
  # perf script
           usleep  4639 [000] 261895.307826:        0            __bpf_stdout__:  ffffffff810eb9a1 ...
       BPF output: 0000: 52 61 69 73 65 20 61 20  Raise a
                   0008: 42 50 46 20 65 76 65 6e  BPF even
                   0010: 74 21 00 00              t!..
       BPF string: "Raise a BPF event!"

           usleep  4639 [000] 261895.407883:        0            __bpf_stdout__:  ffffffff8105d609 ...
       BPF output: 0000: 52 61 69 73 65 20 61 20  Raise a
                   0008: 42 50 46 20 65 76 65 6e  BPF even
                   0010: 74 21 00 00              t!..
       BPF string: "Raise a BPF event!"

  perf record -e ./test_bpf_trace.c usleep 100000

  equals to:

  perf record -e bpf-output/no-inherit=1,name=__bpf_stdout__/ \
              -e ./test_bpf_trace.c/map:__bpf_stdout__.event=__bpf_stdout__/ \
              usleep 100000

Where test_bpf_trace.c is:

  /************************ BEGIN **************************/
  #include <uapi/linux/bpf.h>
  struct bpf_map_def {
         unsigned int type;
         unsigned int key_size;
         unsigned int value_size;
         unsigned int max_entries;
  };
  #define SEC(NAME) __attribute__((section(NAME), used))
  static u64 (*ktime_get_ns)(void) =
         (void *)BPF_FUNC_ktime_get_ns;
  static int (*trace_printk)(const char *fmt, int fmt_size, ...) =
         (void *)BPF_FUNC_trace_printk;
  static int (*get_smp_processor_id)(void) =
         (void *)BPF_FUNC_get_smp_processor_id;
  static int (*perf_event_output)(void *, struct bpf_map_def *, int, void *, unsigned long) =
         (void *)BPF_FUNC_perf_event_output;

  struct bpf_map_def SEC("maps") __bpf_stdout__ = {
         .type = BPF_MAP_TYPE_PERF_EVENT_ARRAY,
         .key_size = sizeof(int),
         .value_size = sizeof(u32),
         .max_entries = __NR_CPUS__,
  };

  static inline int __attribute__((always_inline))
  func(void *ctx, int type)
  {
char output_str[] = "Raise a BPF event!";
char err_str[] = "BAD %d\n";
int err;

        err = perf_event_output(ctx, &__bpf_stdout__, get_smp_processor_id(),
        &output_str, sizeof(output_str));
if (err)
trace_printk(err_str, sizeof(err_str), err);
        return 1;
  }
  SEC("func_begin=sys_nanosleep")
  int func_begin(void *ctx) {return func(ctx, 1);}
  SEC("func_end=sys_nanosleep%return")
  int func_end(void *ctx) { return func(ctx, 2);}
  char _license[] SEC("license") = "GPL";
  int _version SEC("version") = LINUX_VERSION_CODE;
  /************************* END ***************************/

Committer note:

Testing with 'perf trace':

  # trace -e nanosleep --ev test_bpf_stdout.c usleep 1
     0.007 ( 0.007 ms): usleep/729 nanosleep(rqtp: 0x7ffc5bbc5fe0) ...
     0.007 (         ): __bpf_stdout__:Raise a BPF event!..)
     0.008 (         ): perf_bpf_probe:func_begin:(ffffffff81112460))
     0.069 (         ): __bpf_stdout__:Raise a BPF event!..)
     0.070 (         ): perf_bpf_probe:func_end:(ffffffff81112460 <- ffffffff81003d92))
     0.072 ( 0.072 ms): usleep/729  ... [continued]: nanosleep()) = 0
  #

Suggested-and-Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1460128045-97310-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf bpf: Clone bpf stdout events in multiple bpf scripts

This patch allows cloning bpf-output event configuration among multiple
bpf scripts. If there exist a map named '__bpf_output__' and not
configured using 'map:__bpf_output__.event=', this patch clones the
configuration of another '__bpf_stdout__' map. For example, following
command:

  # perf trace --ev bpf-output/no-inherit,name=evt/ \
               --ev ./test_bpf_trace.c/map:__bpf_stdout__.event=evt/ \
               --ev ./test_bpf_trace2.c usleep 100000

equals to:

  # perf trace --ev bpf-output/no-inherit,name=evt/ \
               --ev ./test_bpf_trace.c/map:__bpf_stdout__.event=evt/  \
               --ev ./test_bpf_trace2.c/map:__bpf_stdout__.event=evt/ \
               usleep 100000

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Suggested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1460128045-97310-4-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf dwarf: Guard !x86_64 definitions under #ifdef else clause

To fix the build on Fedora Rawhide (gcc 6.0.0 20160311 (Red Hat 6.0.0-0.17):

    CC       /tmp/build/perf/arch/x86/util/dwarf-regs.o
  arch/x86/util/dwarf-regs.c:66:36: error: 'x86_32_regoffset_table' defined but not used [-Werror=unused-const-variable=]
   static const struct pt_regs_offset x86_32_regoffset_table[] = {
                                      ^~~~~~~~~~~~~~~~~~~~~~
  cc1: all warnings being treated as errors

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-fghuksc1u8ln82bof4lwcj0o@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Use readdir() instead of deprecated readdir_r()

The readdir() function is thread safe as long as just one thread uses a
DIR, which is the case when parsing tracepoint event definitions, to
avoid breaking the build with glibc-2.23.90 (upcoming 2.24), use it
instead of readdir_r().

See: http://man7.org/linux/man-pages/man3/readdir.3.html

"However, in modern implementations (including the glibc implementation),
concurrent calls to readdir() that specify different directory streams
are thread-safe.  In cases where multiple threads must read from the
same directory stream, using readdir() with external synchronization is
still preferable to the use of the deprecated readdir_r(3) function."

Noticed while building on a Fedora Rawhide docker container.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-wddn49r6bz6wq4ee3dxbl7lo@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Use readdir() instead of deprecated readdir_r()

The readdir() function is thread safe as long as just one thread uses a
DIR, which is the case when synthesizing events for pre-existing threads
by traversing /proc, so, to avoid breaking the build with glibc-2.23.90
(upcoming 2.24), use it instead of readdir_r().

See: http://man7.org/linux/man-pages/man3/readdir.3.html

"However, in modern implementations (including the glibc implementation),
concurrent calls to readdir() that specify different directory streams
are thread-safe.  In cases where multiple threads must read from the
same directory stream, using readdir() with external synchronization is
still preferable to the use of the deprecated readdir_r(3) function."

Noticed while building on a Fedora Rawhide docker container.

   CC       /tmp/build/perf/util/event.o
  util/event.c: In function '__event__synthesize_thread':
  util/event.c:466:2: error: 'readdir_r' is deprecated [-Werror=deprecated-declarations]
    while (!readdir_r(tasks, &dirent, &next) && next) {
    ^~~~~
  In file included from /usr/include/features.h:368:0,
                   from /usr/include/stdint.h:25,
                   from /usr/lib/gcc/x86_64-redhat-linux/6.0.0/include/stdint.h:9,
                   from /git/linux/tools/include/linux/types.h:6,
                   from util/event.c:1:
  /usr/include/dirent.h:189:12: note: declared here

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-i1vj7nyjp2p750rirxgrfd3c@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf thread_map: Use readdir() instead of deprecated readdir_r()

The readdir() function is thread safe as long as just one thread uses a
DIR, which is the case in thread_map, so, to avoid breaking the build
with glibc-2.23.90 (upcoming 2.24), use it instead of readdir_r().

See: http://man7.org/linux/man-pages/man3/readdir.3.html

"However, in modern implementations (including the glibc implementation),
concurrent calls to readdir() that specify different directory streams
are thread-safe.  In cases where multiple threads must read from the
same directory stream, using readdir() with external synchronization is
still preferable to the use of the deprecated readdir_r(3) function."

Noticed while building on a Fedora Rawhide docker container.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-del8h2a0f40z75j4r42l96l0@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf script: Use readdir() instead of deprecated readdir_r()

The readdir() function is thread safe as long as just one thread uses a
DIR, which is the case in 'perf script', so, to avoid breaking the build
with glibc-2.23.90 (upcoming 2.24), use it instead of readdir_r().

See: http://man7.org/linux/man-pages/man3/readdir.3.html

"However, in modern implementations (including the glibc implementation),
concurrent calls to readdir() that specify different directory streams
are thread-safe.  In cases where multiple threads must read from the
same directory stream, using readdir() with external synchronization is
still preferable to the use of the deprecated readdir_r(3) function."

Noticed while building on a Fedora Rawhide docker container.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-mt3xz7n2hl49ni2vx7kuq74g@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf symbols: Adjust symbol for shared objects

He Kuang reported a problem that perf fails to get correct symbol on
Android platform in [1]. The problem can be reproduced on normal x86_64
platform. I will describe the reproducing steps in detail at the end of
commit message.

The reason of this problem is the missing of symbol adjustment for normal
shared objects. In most of the cases skipping adjustment is okay. However,
when '.text' section have different 'address' and 'offset' the result is wrong.
I checked all shared objects in my working platform, only wine dll objects and
debug objects (in .debug) have this problem. However, it is common on Android.
For example:

 $ readelf -S ./libsurfaceflinger.so | grep \.text
   [10] .text             PROGBITS         0000000000029030  00012030

This patch enables symbol adjustment for dynamic objects so the symbol
address got from elfutils would be adjusted correctly.

Now nearly all types of ELF files should adjust symbols. Makes
ss->adjust_symbols default to true.

Steps to reproduce the problem:

  $ cat ./Makefile
  PWD := $(shell pwd)
  LDFLAGS += "-Wl,-rpath=$(PWD)"
  CFLAGS += -g
  main: main.c libbuggy.so
  libbuggy.so: buggy.c
gcc -g -shared -fPIC -Wl,-Ttext-segment=0x200000 $< -o $@
  clean:
rm -rf main libbuggy.so *.o

  $ cat ./buggy.c
  int fib(int x)
  {
      return (x == 0) ? 1 : (x == 1) ? 1 : fib(x - 1) + fib(x - 2);
  }

  $ cat ./main.c
  #include <stdio.h>

  extern int fib(int x);
  int main()
  {
     int i;

     for (i = 0; i < 40; i++)
         printf("%d\n", fib(i));
     return 0;
 }

 $ make
 $ perf record ./main
 ...
 $ perf report --stdio
 # Overhead  Command  Shared Object      Symbol
 # ........  .......  .................  ...............................
 #
     14.97%  main     libbuggy.so        [.] 0x000000000000066c
      8.68%  main     libbuggy.so        [.] 0x00000000000006aa
      8.52%  main     libbuggy.so        [.] fib@plt
      7.95%  main     libbuggy.so        [.] 0x0000000000000664
      5.94%  main     libbuggy.so        [.] 0x00000000000006a9
      5.35%  main     libbuggy.so        [.] 0x0000000000000678
 ...

The correct result should be (after this patch):

  # Overhead  Command  Shared Object      Symbol
  # ........  .......  .................  ...............................
  #
      91.47%  main     libbuggy.so        [.] fib
       8.52%  main     libbuggy.so        [.] fib@plt
       0.00%  main     [kernel.kallsyms]  [k] kmem_cache_free

[1] http://lkml.kernel.org/g/1452567507-54013-1-git-send-email-hekuang@huawei.com

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1460024671-64774-3-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf symbols: Record text offset in dso to calculate objdump address

In this patch, the offset of '.text' section is stored into dso
and used here to re-calculate address to objdump.

In most of the cases, executable code is in '.text' section, so the
adjustment made to a symbol in dso__load_sym (using
sym.st_value -= shdr.sh_addr - shdr.sh_offset) should equal to
'sym.st_value -= dso->text_offset'. Therefore, adding text_offset back
get objdump address from symbol address (rip). However, it is not true
for kernel and kernel module since there could be multiple executable
sections with different offset. Exclude kernel for this reason.

After this patch, even dso->adjust_symbols is set to true for shared
objects, map__rip_2objdump() and map__objdump_2mem() would return
correct result, so perf behavior of annotate won't be changed.

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Cody P Schafer <dev@codyps.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kirill Smelkov <kirr@nexedi.com>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1460024671-64774-2-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Build syscall table .c header from kernel's syscall_64.tbl

We used libaudit to map ids to syscall names and vice-versa, but that
imposes a delay in supporting new syscalls, having to wait for libaudit
to get those new syscalls on its tables.

To remove that delay, for x86_64 initially, grab a copy of
arch/x86/entry/syscalls/syscall_64.tbl and use it to generate those
tables.

Syscalls currently not available in audit-libs:

  # trace -e copy_file_range,membarrier,mlock2,pread64,pwrite64,timerfd_create,userfaultfd
  Error: Invalid syscall copy_file_range, membarrier, mlock2, pread64, pwrite64, timerfd_create, userfaultfd
  Hint: try 'perf list syscalls:sys_enter_*'
  Hint: and: 'man syscalls'
  #

With this patch:

  # trace -e copy_file_range,membarrier,mlock2,pread64,pwrite64,timerfd_create,userfaultfd
    8505.733 ( 0.010 ms): gnome-shell/2519 timerfd_create(flags: 524288) = 36
    8506.688 ( 0.005 ms): gnome-shell/2519 timerfd_create(flags: 524288) = 40
   30023.097 ( 0.025 ms): qemu-system-x8/24629 pwrite64(fd: 18, buf: 0x7f63ae382000, count: 4096, pos: 529592320) = 4096
   31268.712 ( 0.028 ms): qemu-system-x8/24629 pwrite64(fd: 18, buf: 0x7f63afd8b000, count: 4096, pos: 2314133504) = 4096
   31268.854 ( 0.016 ms): qemu-system-x8/24629 pwrite64(fd: 18, buf: 0x7f63afda2000, count: 4096, pos: 2314137600) = 4096

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-51xfjbxevdsucmnbc4ka5r88@git.kernel.org
[ Added make dep for 'prepare' in 'LIBPERF_IN', fix by Wang Nan to fix parallell build ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Allow generating per-arch syscall table arrays

Tools should use a mechanism similar to arch/x86/entry/syscalls/ to
generate a header file with the definitions for two variables:

  static const char *syscalltbl_x86_64[] = {
[0] = "read",
[1] = "write",
  <SNIP>
[324] = "membarrier",
[325] = "mlock2",
[326] = "copy_file_range",
  };
  static const int syscalltbl_x86_64_max_id = 326;

In a per arch file that should then be included in
tools/perf/util/syscalltbl.c.

First one will be for x86_64.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-02uuamkxgccczdth8komspgp@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Move syscall table id <-> name routines to separate class

We're using libaudit for doing name to id and id to syscall name
translations, but that makes 'perf trace' to have to wait for newer
libaudit versions supporting recently added syscalls, such as
"userfaultfd" at the time of this changeset.

We have all the information right there, in the kernel sources, so move
this code to a separate place, wrapped behind functions that will
progressively use the kernel source files to extract the syscall table
for use in 'perf trace'.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-i38opd09ow25mmyrvfwnbvkj@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Beautify mode_t arguments

When reading the syscall tracepoint /format file, look for arguments of type
"mode_t" and attach a beautifier:

  [root@jouet ~]# cat ~/bin/tp_with_fields_of_type
  #!/bin/bash
  grep -w $1 /sys/kernel/tracing/events/syscalls/*/format | sed -r 's%.*sys_enter_(.*)/format.*%\1%g' | paste -d, -s
  # tp_with_fields_of_type umode_t
  chmod,creat,fchmodat,fchmod,mkdirat,mkdir,mknodat,mknod,mq_open,openat,open
  #

Testing it:

  #define S_ISUID 0004000
  #define S_ISGID 0002000
  #define S_ISVTX 0001000
  #define S_IRWXU 0000700
  #define S_IRUSR 0000400
  #define S_IWUSR 0000200
  #define S_IXUSR 0000100

  #define S_IRWXG 0000070
  #define S_IRGRP 0000040
  #define S_IWGRP 0000020
  #define S_IXGRP 0000010

  #define S_IRWXO 0000007
  #define S_IROTH 0000004
  #define S_IWOTH 0000002
  #define S_IXOTH 0000001

  # for mode in 4000 2000 1000 700 400 200 100 70 40 20 10 7 4 2 1 ; do \
      echo -n $mode '->' ; trace --no-inherit -e chmod,fchmodat,fchmod chmod $mode x; \
    done
  4000 -> 0.338 ( 0.012 ms): fchmodat(dfd: CWD, filename: x, mode: ISUID) = 0
  2000 -> 0.438 ( 0.015 ms): fchmodat(dfd: CWD, filename: x, mode: ISGID) = 0
  1000 -> 0.677 ( 0.040 ms): fchmodat(dfd: CWD, filename: x, mode: ISVTX) = 0
   700 -> 0.394 ( 0.013 ms): fchmodat(dfd: CWD, filename: x, mode: IRWXU) = 0
   400 -> 0.337 ( 0.010 ms): fchmodat(dfd: CWD, filename: x, mode: IRUSR) = 0
   200 -> 0.259 ( 0.008 ms): fchmodat(dfd: CWD, filename: x, mode: IWUSR) = 0
   100 -> 0.249 ( 0.008 ms): fchmodat(dfd: CWD, filename: x, mode: IXUSR) = 0
    70 -> 0.266 ( 0.008 ms): fchmodat(dfd: CWD, filename: x, mode: IRWXG) = 0
    40 -> 0.329 ( 0.009 ms): fchmodat(dfd: CWD, filename: x, mode: IRGRP) = 0
    20 -> 0.250 ( 0.009 ms): fchmodat(dfd: CWD, filename: x, mode: IWGRP) = 0
    10 -> 0.259 ( 0.008 ms): fchmodat(dfd: CWD, filename: x, mode: IXGRP) = 0
     7 -> 0.249 ( 0.009 ms): fchmodat(dfd: CWD, filename: x, mode: IRWXO) = 0
     4 -> 0.278 ( 0.011 ms): fchmodat(dfd: CWD, filename: x, mode: IROTH) = 0
     2 -> 0.276 ( 0.009 ms): fchmodat(dfd: CWD, filename: x, mode: IWOTH) = 0
     1 -> 0.250 ( 0.008 ms): fchmodat(dfd: CWD, filename: x, mode: IXOTH) = 0
  #
  # trace --no-inherit -e chmod,fchmodat,fchmod chmod 7777 x
     0.258 ( 0.011 ms): fchmodat(dfd: CWD, filename: x, mode: IALLUGO) = 0
  # trace --no-inherit -e chmod,fchmodat,fchmod chmod 7770 x
     0.258 ( 0.008 ms): fchmodat(dfd: CWD, filename: x, mode: ISUID|ISGID|ISVTX|IRWXU|IRWXG) = 0
  # trace --no-inherit -e chmod,fchmodat,fchmod chmod 777 x
     0.293 ( 0.012 ms): fchmodat(dfd: CWD, filename: x, mode: IRWXUGO
  #

Now lets see if check by using the tracepoint for that specific syscall,
instead of raw_syscalls:sys_enter as 'trace' does for its strace fu:

  # trace --no-inherit --ev syscalls:sys_enter_fchmodat -e fchmodat chmod 666 x
     0.255 (         ): syscalls:sys_enter_fchmodat:dfd: 0xffffffffffffff9c, filename: 0x55db32a3f0f0, mode: 0x000001b6)
     0.268 ( 0.012 ms): fchmodat(dfd: CWD, filename: x, mode: IRUGO|IWUGO                     ) = 0
  #

Perfect, 0x1bc == 0666.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-18e8zfgbkj83xo87yoom43kd@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf script: Process event update events

Andreas reported following command produces no output:

  # cat test.py
  #!/usr/bin/env python

  def stat__krava(cpu, thread, time, val, ena, run):
      print "event %s cpu %d, thread %d, time %d, val %d, ena %d, run %d" % \
            ("krava", cpu, thread, time, val, ena, run)
  # perf stat -a -I 1000 -e cycles,"cpu/config=0x6530160,name=krava/" record | perf script -s test.py
  ^C
  #

The reason is that 'perf script' does not process event update events and
will never get the event name update thus the python callback is never
called.

The fix is just to add already existing callback we use in 'perf stat
report'.

Committer note:

After the patch:

  # perf stat -a -I 1000 -e cycles,"cpu/config=0x6530160,name=krava/" record | perf script -s test.py
  event krava cpu -1, thread -1, time 1000239179, val 1789051, ena 4000690920, run 4000690920
  event krava cpu -1, thread -1, time 2000479061, val 2391338, ena 4000879596, run 4000879596
  event krava cpu -1, thread -1, time 3000740802, val 1939121, ena 4000977209, run 4000977209
  event krava cpu -1, thread -1, time 4001006730, val 2356115, ena 4001000489, run 4001000489
  ^C
  #

Reported-by: Andreas Hollmann <hollmann@in.tum.de>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1460013073-18444-3-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Add dedicated unwind addr_space member into thread struct

Milian reported issue with thread::priv, which was double booked by perf
trace and DWARF unwind code. So using those together is impossible at
the moment.

Moving DWARF unwind private data into separate variable so perf trace
can keep using thread::priv.

Reported-and-Tested-by: Milian Wolff <milian.wolff@kdab.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andreas Hollmann <hollmann@in.tum.de>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1460013073-18444-2-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Introduce trim function

To be used in cases for both sides trim.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andreas Hollmann <hollmann@in.tum.de>
Cc: David Ahern <dsahern@gmail.com>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1460013073-18444-1-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Beautify pid_t arguments

When reading the syscall tracepoint /format file, look for arguments
of type "pid_t" and attach the PID beautifier, that will do a lookup
on the threads it knows, i.e. the ones that came from PERF_RECORD_COMM
events and add the COMM after the pid in such args:

Excerpt of a system wide trace for syscalls with pid_t args:

  55602.977 ( 0.006 ms): bash/12122 setpgid(pid: 24347 (bash), pgid: 24347 (bash)) = 0
  55603.024 ( 0.004 ms): bash/24347 setpgid(pid: 24347 (bash), pgid: 24347 (bash)) = 0
  55691.527 (88.397 ms): bash/12122 wait4(upid: -1, stat_addr: 0x7ffe0cee1720, options: UNTRACED|CONTINUED) ...
  55692.479 ( 0.952 ms): git/24347 wait4(upid: 24368, stat_addr: 0x7ffe030d5724) ...
  55694.549 ( 2.070 ms): pre-commit/24368 wait4(upid: -1, stat_addr: 0x7ffc94f4fc10) = 24369 (pre-commit)
  55694.575 ( 0.002 ms): pre-commit/24368 wait4(upid: -1, stat_addr: 0x7ffc94f4f650, options: NOHANG) = -1 ECHILD No child processes
  55695.934 ( 0.010 ms): pre-commit/24368 wait4(upid: -1, stat_addr: 0x7ffc94f4f2d0, options: NOHANG) = 24370 (git)
  55695.937 ( 0.001 ms): pre-commit/24368 wait4(upid: -1, stat_addr: 0x7ffc94f4f2d0, options: NOHANG) = -1 ECHILD No child processes
  55717.963 ( 0.000 ms): pre-commit/24371  ... [continued]: wait4()) = 24372
  55717.978 (21.468 ms): :24371/24371 wait4(upid: -1, stat_addr: 0x7ffc94f4f230) ...
  55718.087 ( 0.109 ms): pre-commit/24371 wait4(upid: -1, stat_addr: 0x7ffc94f4f230) = 24373 (tr)
  55718.187 ( 0.096 ms): pre-commit/24371 wait4(upid: -1, stat_addr: 0x7ffc94f4f230) = 24374 (wc)
  55718.218 ( 0.002 ms): pre-commit/24371 wait4(upid: -1, stat_addr: 0x7ffc94f4eed0, options: NOHANG) = -1 ECHILD No child processes
  55718.367 ( 0.005 ms): pre-commit/24368 wait4(upid: -1, stat_addr: 0x7ffc94f4f1d0, options: NOHANG) = 24371 (pre-commit)
  55718.369 ( 0.001 ms): pre-commit/24368 wait4(upid: -1, stat_addr: 0x7ffc94f4f1d0, options: NOHANG) = -1 ECHILD No child processes
  55741.021 (49.494 ms): git/24347  ... [continued]: wait4()) = 24368 (pre-commit)
  74146.427 (18319.601 ms): git/24347 wait4(upid: 24375 (git), stat_addr: 0x7ffe030d6824) ...
  74149.036 ( 0.891 ms): bash/24391 wait4(upid: -1, stat_addr: 0x7ffe0cee0560) = 24393 (sed)

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-75yl9hzjhb020iadc81gdj8t@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Beautify set_tid_address, getpid, getppid return values

Showing the COMM for that return, if available.

  # trace -e getpid,getppid,set_tid_address
    490.007 ( 0.005 ms): sh/8250 getpid(...) = 8250 (sh)
    490.014 ( 0.001 ms): sh/8250 getppid(...) = 7886 (make)
    491.156 ( 0.004 ms): install/8251 set_tid_address(tidptr: 0x7f204a9d4ad0) = 8251 (install)
  ^C

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-psbpplqupatom9x4uohbxid5@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Infrastructure to show COMM strings for syscalls returning PIDs

Starting with clone, waitid and wait4:

  # trace -e waitid,wait4
     1.385 ( 1.385 ms): bash/12122 wait4(upid: -1, stat_addr: 0x7ffe0cee1720, options: UNTRACED|CONTINUED) = 1210 (ls)
     1.426 ( 0.002 ms): bash/12122 wait4(upid: -1, stat_addr: 0x7ffe0cee1150, options: NOHANG|UNTRACED|CONTINUED) = 0
     3.293 ( 0.604 ms): bash/1211 wait4(upid: -1, stat_addr: 0x7ffe0cee0560                             ) = 1214 (sed)
     3.342 ( 0.002 ms): bash/1211 wait4(upid: -1, stat_addr: 0x7ffe0cee01d0, options: NOHANG            ) = -1 ECHILD No child processes
     3.576 ( 0.016 ms): bash/12122 wait4(upid: -1, stat_addr: 0x7ffe0cee0550, options: NOHANG|UNTRACED|CONTINUED) = 1211 (bash)
  ^C# trace -e clone
     0.027 ( 0.000 ms): systemd/1  ... [continued]: clone()) = 1227 (systemd)
     0.050 ( 0.000 ms): systemd/1227  ... [continued]: clone()) = 0
  ^C[root@jouet ~]#

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-lyf5d3y5j15wikjb6pe6ukoi@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Beautify wait4/waitid 'options' argument

  # trace -e waitid,wait4

   0.557 ( 0.557 ms): bash/27335 wait4(upid: -1, stat_addr: 0x7ffd02f449f0) = 27336
   1.250 ( 0.685 ms): bash/27335 wait4(upid: -1, stat_addr: 0x7ffd02f449f0) = 27337
   1.312 ( 0.002 ms): bash/27335 wait4(upid: -1, stat_addr: 0x7ffd02f44690, options: NOHANG) = -1 ECHILD No child processes
   1.550 ( 0.015 ms): bash/3856 wait4(upid: -1, stat_addr: 0x7ffd02f44990, options: NOHANG|UNTRACED|CONTINUED) = 27335
   1.552 ( 0.001 ms): bash/3856 wait4(upid: -1, stat_addr: 0x7ffd02f44990, options: NOHANG|UNTRACED|CONTINUED) = 0
  #

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-i5vlo5n5jv0amt8bkyicmdxh@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Beautify sched_setscheduler 'policy' argument

  $ trace -e sched_setscheduler chrt -f 1 usleep 1
  chrt: failed to set pid 0's policy: Operation not permitted
     0.005 ( 0.005 ms): chrt/19189 sched_setscheduler(policy: FIFO, param: 0x7ffec5273d70) = -1 EPERM Operation not permitted
  $

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-i5vlo5n5jv0amt8bkyicmdxh@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf list: Document event specifications better

Document some features for specifying events in the perf list manpage:

- Event groups
- Leader sampling
- How to specify raw PMU events in the new syntax
- Global versus per process PMUs.
- Access restrictions
- Fix Intel SDM URL

v2: Lots of new content. address review feedback.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/1459810686-15913-1-git-send-email-andi@firstfloor.org
[ Add quotes to some keywords, such as "any" ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Remove superfluous ARCH Makefile includes

Link: http://lkml.kernel.org/n/tip-yk6brsq3opuotr9by18xlkr8@git.kernel.org
Signed-off-by: Jiri Olsa <jolsa@kernel.org>

perf script perl: Do error checking on new backtrace routine

This ended up triggering these warnings when building on Ubuntu 12.04.5:

  util/scripting-engines/trace-event-perl.c: In function 'perl_process_callchain':
  util/scripting-engines/trace-event-perl.c:293:4: error: value computed is not used [-Werror=unused-value]
  util/scripting-engines/trace-event-perl.c:294:4: error: value computed is not used [-Werror=unused-value]
  util/scripting-engines/trace-event-perl.c:295:4: error: value computed is not used [-Werror=unused-value]
  util/scripting-engines/trace-event-perl.c:297:4: error: value computed is not used [-Werror=unused-value]
  util/scripting-engines/trace-event-perl.c:309:4: error: value computed is not used [-Werror=unused-value]
  cc1: all warnings being treated as errors
  mv: cannot stat `/tmp/build/perf/util/scripting-engines/.trace-event-perl.o.tmp': No such file or directory
  make[4]: *** [/tmp/build/perf/util/scripting-engines/trace-event-perl.o] Error 1

Fix it by doing error checking when building the perl data structures
related to callchains.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Dima Kogan <dima@secretsauce.net>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@gmail.com>
Fixes: f7380c12ec6c ("perf script perl: Perl scripts now get a backtrace, like the python ones")
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf probe: Check if dwarf_getlocations() is available

If not, tell the user that:

  config/Makefile:273: Old libdw.h, finding variables at given 'perf probe' point will not work, install elfutils-devel/libdw-dev >= 0.157

And return -ENOTSUPP in die_get_var_range(), failing features that
need it, like the one pointed out above.

This fixes the build on older systems, such as Ubuntu 12.04.5.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Vinson Lee <vlee@freedesktop.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-9l7luqkq4gfnx7vrklkq4obs@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf config: Fix build with older toolchain.

Fix build error on Ubuntu 12.04.5 with GCC 4.6.3.

    CC       util/config.o
  util/config.c: In function ‘perf_buildid_config’:
  util/config.c:384:15: error: declaration of ‘dirname’ shadows a global declaration [-Werror=shadow]

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@plumgrid.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Taeung Song <treeze.taeung@gmail.com>
Cc: Wang Nan <wangnan0@huawei.com>
Fixes: 9cb5987c8227 ("perf config: Rework buildid_dir_command_config to perf_buildid_config")
Link: http://lkml.kernel.org/r/1459807659-9020-1-git-send-email-vlee@freedesktop.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

Merge tag 'perf-core-for-mingo-20160401' of git://git./linux/kernel/git/acme/linux into perf/core

Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:

User visible changes:

 - Do not use events that don't have timestamps when setting 'perf trace's
   base timestamp, fixing up the timestamp column for syscalls (Arnaldo Carvalho de Melo)

 - Make the 'bpf-output' sample_type be the same as tracepoint's, fixing up
   'perf trace's timestamp column for bpf events (Wang Nan)

 - Fix PMU term format max value calculation (Kan Liang)

 - Pretty print 'seccomp', 'getrandom' syscalls in 'perf trace' (Arnaldo Carvalho de Melo)

Infrastructure changes:

 - Add support for using TSC as an ARCH timestamp when synthesizing
   JIT records (Adrian Hunter)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf bpf: Add sample types for 'bpf-output' event

Before this patch we can see very large time in the events before the
'bpf-output' event. For example:

  # perf trace -vv -T --ev sched:sched_switch \
                      --ev bpf-output/no-inherit,name=evt/ \
                      --ev ./test_bpf_trace.c/map:channel.event=evt/ \
                      usleep 10
  ...
  18446744073709.551 (18446564645918.480 ms): usleep/4157 nanosleep(rqtp: 0x7ffd3f0dc4e0) ...
  18446744073709.551 (         ): evt:Raise a BPF event!..)
  179427791.076 (         ): perf_bpf_probe:func_begin:(ffffffff810eb9a0))
  179427791.081 (         ): sched:sched_switch:usleep:4157 [120] S ==> swapper/2:0 [120])
  ...

We can also see the differences between bpf-output events and
breakpoint events:

For bpf output event:
   sample_type                    IP|TID|RAW|IDENTIFIER

For tracepoint events:
   sample_type                    IP|TID|TIME|CPU|PERIOD|RAW|IDENTIFIER

This patch fix this differences by adding more sample type for
bpf-output events.

After this patch:

  # perf trace -vv -T --ev sched:sched_switch \
                      --ev bpf-output/no-inherit,name=evt/ \
                      --ev ./test_bpf_trace.c/map:channel.event=evt/ \
                      usleep 10
  ...
  179877370.878 ( 0.003 ms): usleep/5336 nanosleep(rqtp: 0x7ffff866c450) ...
  179877370.878 (         ): evt:Raise a BPF event!..)
  179877370.878 (         ): perf_bpf_probe:func_begin:(ffffffff810eb9a0))
  179877370.882 (         ): sched:sched_switch:usleep:5336 [120] S ==> swapper/4:0 [120])
  179877370.945 (         ): evt:Raise a BPF event!..)
  ...

  # ./perf trace -vv -T --ev sched:sched_switch \
                        --ev bpf-output/no-inherit,name=evt/ \
                        --ev ./test_bpf_trace.c/map:channel.event=evt/ \
                        usleep 10 2>&1 | grep sample_type
  sample_type                      IP|TID|TIME|ID|CPU|PERIOD|RAW
  sample_type                      IP|TID|TIME|ID|CPU|PERIOD|RAW
  sample_type                      IP|TID|TIME|ID|CPU|PERIOD|RAW
  sample_type                      IP|TID|TIME|ID|CPU|PERIOD|RAW
  sample_type                      IP|TID|TIME|ID|CPU|PERIOD|RAW
  sample_type                      IP|TID|TIME|ID|CPU|PERIOD|RAW

The 'IDENTIFIER' info is not required because all events have the same
sample_type.

Committer notes:

Further testing, on top of the changes making 'perf trace' avoid samples
from events without PERF_SAMPLE_TIME:

Before:

  # trace --ev bpf-output/no-inherit,name=evt/ --ev /home/acme/bpf/test_bpf_trace.c/map:channel.event=evt/ usleep 10
  <SNIP>
    0.560 ( 0.001 ms): brk(                                                   ) = 0x55e5a1df8000
    18446640227439.430 (18446640227438.859 ms): nanosleep(rqtp: 0x7ffc96643370) ...
    18446640227439.430 (         ): evt:Raise a BPF event!..)
    0.576 (         ): perf_bpf_probe:func_begin:(ffffffff81112460))
    18446640227439.430 (         ): evt:Raise a BPF event!..)
    0.645 (         ): perf_bpf_probe:func_end:(ffffffff81112460 <- ffffffff81003d92))
    0.646 ( 0.076 ms):  ... [continued]: nanosleep()) = 0
  #

After:

  # trace --ev bpf-output/no-inherit,name=evt/ --ev /home/acme/bpf/test_bpf_trace.c/map:channel.event=evt/ usleep 10
  <SNIP>
     0.292 ( 0.001 ms): brk(                          ) = 0x55c7cd6e1000
     0.302 ( 0.004 ms): nanosleep(rqtp: 0x7ffedd8bc0f0) ...
     0.302 (         ): evt:Raise a BPF event!..)
     0.303 (         ): perf_bpf_probe:func_begin:(ffffffff81112460))
     0.397 (         ): evt:Raise a BPF event!..)
     0.397 (         ): perf_bpf_probe:func_end:(ffffffff81112460 <- ffffffff81003d92))
     0.398 ( 0.100 ms):  ... [continued]: nanosleep()) = 0

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Reported-and-Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1459517202-42320-1-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Don't set the base timestamp using events without PERF_SAMPLE_TIME

This was causing bogus values to be shown at the timestamp column:

Before:

  # trace --ev bpf-output/no-inherit,name=evt/ --ev /home/acme/bpf/test_bpf_trace.c/map:channel.event=evt/ usleep 10
  94631143.385 ( 0.001 ms): brk(                                     ) = 0x555555757000
  94631143.398 ( 0.003 ms): mmap(len: 4096, prot: READ|WRITE, flags: PRIVATE|ANONYMOUS, fd: -1) = 0x7ffff7ff6000
  94631143.406 ( 0.004 ms): access(filename: 0xf7df9e10, mode: R     ) = -1 ENOENT No such file or directory
  94631143.412 ( 0.004 ms): open(filename: 0xf7df8761, flags: CLOEXEC) = 3
  94631143.415 ( 0.002 ms): fstat(fd: 3, statbuf: 0x7fffffffd6b0     ) = 0
  94631143.419 ( 0.003 ms): mmap(len: 106798, prot: READ, flags: PRIVATE, fd: 3) = 0x7ffff7fdb000
  94631143.420 ( 0.001 ms): close(fd: 3                              ) = 0
  94631143.432 ( 0.004 ms): open(filename: 0xf7ff6640, flags: CLOEXEC) = 3
  <SNIP>

After:

  # trace --ev bpf-output/no-inherit,name=evt/ --ev /home/acme/bpf/test_bpf_trace.c/map:channel.event=evt/ usleep 10
  0.022 ( 0.001 ms): brk(                                     ) = 0x55d7668a6000
  0.037 ( 0.003 ms): mmap(len: 4096, prot: READ|WRITE, flags: PRIVATE|ANONYMOUS, fd: -1) = 0x7f8fbeb97000
  0.123 ( 0.083 ms): access(filename: 0xbe995e10, mode: R     ) = -1 ENOENT No such file or directory
  0.130 ( 0.004 ms): open(filename: 0xbe994761, flags: CLOEXEC) = 3
  0.133 ( 0.002 ms): fstat(fd: 3, statbuf: 0x7fff6487a890     ) = 0
  0.138 ( 0.003 ms): mmap(len: 106798, prot: READ, flags: PRIVATE, fd: 3) = 0x7f8fbeb7c000
  0.140 ( 0.001 ms): close(fd: 3                              ) = 0
  0.151 ( 0.004 ms): open(filename: 0xbeb97640, flags: CLOEXEC) = 3
  <SNIP>

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-p7m8llv81iv55ekxexdp5n57@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Introduce function to set the base timestamp

That is used in both live runs, i.e.:

  # trace ls

As when processing events recorded in a perf.data file:

  # trace -i perf.data

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-901l6yebnzeqg7z8mbaf49xb@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Fix PMU term format max value calculation

Currently the max value of format is calculated by the bits number. It
relies on the continuity of the format.

However, uncore event format is not continuous. E.g. uncore qpi event
format can be 0-7,21.

If bit 21 is set, there is parsing issues as below.

  $ perf stat -a -e uncore_qpi_0/event=0x200002,umask=0x8/
  event syntax error: '..pi_0/event=0x200002,umask=0x8/'
                                    \___ value too big for format, maximum is 511

This patch return the real max value by setting all possible bits to 1.

Signed-off-by: Kan Liang <kan.liang@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Link: http://lkml.kernel.org/r/1459365375-14285-1-git-send-email-kan.liang@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf intel-pt/bts: Define JITDUMP_USE_ARCH_TIMESTAMP

For Intel PT / BTS, define the environment variable that selects TSC
timestamps in the jitdump file.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1457426333-30260-1-git-send-email-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf jit: Add support for using TSC as a timestamp

Intel PT uses TSC as a timestamp, so add support for using TSC instead
of the monotonic clock.  Use of TSC is selected by an environment
variable "JITDUMP_USE_ARCH_TIMESTAMP" and flagged in the jitdump file
with flag JITDUMP_FLAGS_ARCH_TIMESTAMP.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1457426330-30226-1-git-send-email-adrian.hunter@intel.com
[ Added the fixup from He Kuang to make it build on other arches, ]
[ such as aarch64, to avoid inserting this bisectiong breakage upstream ]
Link: http://lkml.kernel.org/r/1459482572-129494-1-git-send-email-hekuang@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Add time conversion event

Intel PT uses the time members from the perf_event_mmap_page to convert
between TSC and perf time.

Due to a lack of foresight when Intel PT was implemented, those time
members were recorded in the (implementation dependent) AUXTRACE_INFO
event, the structure of which is generally inaccessible outside of the
Intel PT decoder.  However now the conversion between TSC and perf time
is needed when processing a jitdump file when Intel PT has been used for
tracing.

So add a user event to record the time members.  'perf record' will
synthesize the event if the information is available.  And session
processing will put a copy of the event on the session so that tools
like 'perf inject' can easily access it.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1457426324-30158-1-git-send-email-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Pretty print getrandom() args

  # trace -e getrandom
  35622.560 ( 0.023 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  35622.585 ( 0.006 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  35622.594 ( 0.004 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  35627.395 ( 0.010 ms): libvirtd/1353 getrandom(buf: 0x7f7a1bfa35c0, count: 16, flags: NONBLOCK    ) = 16
  35630.940 ( 0.013 ms): fwupd/16120 getrandom(buf: 0x7f63243aa5c0, count: 16, flags: NONBLOCK      ) = 16
  35718.613 ( 0.015 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  35718.629 ( 0.005 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  35718.637 ( 0.004 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  35719.355 ( 0.010 ms): libvirtd/1353 getrandom(buf: 0x7f7a1bfa35c0, count: 16, flags: NONBLOCK    ) = 16
  35721.042 ( 0.030 ms): fwupd/16120 getrandom(buf: 0x7f63243aa5c0, count: 16, flags: NONBLOCK      ) = 16
  41090.830 ( 0.012 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  41090.845 ( 0.004 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  41090.851 ( 0.004 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  41091.750 ( 0.010 ms): libvirtd/1353 getrandom(buf: 0x7f7a1bfa35c0, count: 16, flags: NONBLOCK    ) = 16
  41091.823 ( 0.006 ms): fwupd/16120 getrandom(buf: 0x7f63243aa5c0, count: 16, flags: NONBLOCK      ) = 16
  41122.078 ( 0.053 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  41122.129 ( 0.009 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  41122.139 ( 0.004 ms): systemd-udevd/631 getrandom(buf: 0x55621e3c18f0, count: 16, flags: NONBLOCK) = 16
  41124.492 ( 0.007 ms): libvirtd/1353 getrandom(buf: 0x7f7a1bfa35c0, count: 16, flags: NONBLOCK    ) = 16
  41124.470 ( 0.013 ms): fwupd/16120 getrandom(buf: 0x7f63243aa5c0, count: 16, flags: NONBLOCK      ) = 16
  41590.832 ( 0.014 ms): chrome/5957 getrandom(buf: 0x7fabac7b15b0, count: 16, flags: NONBLOCK      ) = 16
  41590.884 ( 0.004 ms): chrome/5957 getrandom(buf: 0x7fabac7b15c0, count: 16, flags: NONBLOCK      ) = 16

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-gca0n1p3aca3depey703ph2q@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Pretty print seccomp() args

E.g:

  # trace -e seccomp
   200.061 (0.009 ms): :2441/2441 seccomp(op: FILTER, flags: TSYNC                       ) = -1 EFAULT Bad address
   200.910 (0.121 ms): :2441/2441 seccomp(op: FILTER, flags: TSYNC, uargs: 0x7fff57479fe0) = 0

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Milian Wolff <milian.wolff@kdab.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-t369uckshlwp4evkks4bcoo7@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf trace: Do not process PERF_RECORD_LOST twice

We catch this record to provide a visual indication that events are
getting lost, then call the default method to allow extra logging shared
with the other tools to take place.

This extra logging was done twice because we were continuing to the
"default" clause where machine__process_event() will end up calling
machine__process_lost_event() again, fix it.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-wus2zlhw3qo24ye84ewu4aqw@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf/ring_buffer: Prepare writing into the ring-buffer from the end

Convert perf_output_begin() to __perf_output_begin() and make the later
function able to write records from the end of the ring-buffer.

Following commits will utilize the 'backward' flag.

This is the core patch to support writing to the ring-buffer backwards,
which will be introduced by upcoming patches to support reading from
overwritable ring-buffers.

In theory, this patch should not introduce any extra performance
overhead since we use always_inline, but it does not hurt to double
check that assumption:

When CONFIG_OPTIMIZE_INLINING is disabled, the output object is nearly
identical to original one. See:

   http://lkml.kernel.org/g/56F52E83.70409@huawei.com

When CONFIG_OPTIMIZE_INLINING is enabled, the resuling object file becomes
smaller:

 $ size kernel/events/ring_buffer.o*
   text       data        bss        dec        hex    filename
   4641          4          8       4653       122d kernel/events/ring_buffer.o.old
   4545          4          8       4557       11cd kernel/events/ring_buffer.o.new

Performance testing results:

Calling 3000000 times of 'close(-1)', use gettimeofday() to check
duration.  Use 'perf record -o /dev/null -e raw_syscalls:*' to capture
system calls. In ns.

Testing environment:

 CPU    : Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz
 Kernel : v4.5.0

                     MEAN         STDVAR
  BASE            800214.950    2853.083
  PRE            2253846.700    9997.014
  POST           2257495.540    8516.293

Where 'BASE' is pure performance without capturing. 'PRE' is test
result of pure 'v4.5.0' kernel. 'POST' is test result after this
patch.

Considering the stdvar, this patch doesn't hurt performance, within
noise margin.

For testing details, see:

  http://lkml.kernel.org/g/56F89DCD.1040202@huawei.com

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <pi3orama@163.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: Zefan Li <lizefan@huawei.com>
Link: http://lkml.kernel.org/r/1459147292-239310-4-git-send-email-wangnan0@huawei.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/core: Set event's default ::overflow_handler()

Set a default event->overflow_handler in perf_event_alloc() so don't
need to check event->overflow_handler in __perf_event_overflow().
Following commits can give a different default overflow_handler.

Initial idea comes from Peter:

  http://lkml.kernel.org/r/20130708121557.GA17211@twins.programming.kicks-ass.net

Since the default value of event->overflow_handler is not NULL, existing
'if (!overflow_handler)' checks need to be changed.

is_default_overflow_handler() is introduced for this.

No extra performance overhead is introduced into the hot path because in the
original code we still need to read this handler from memory. A conditional
branch is avoided so actually we remove some instructions.

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <pi3orama@163.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: Zefan Li <lizefan@huawei.com>
Link: http://lkml.kernel.org/r/1459147292-239310-3-git-send-email-wangnan0@huawei.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/ring_buffer: Introduce new ioctl options to pause and resume the ring-buffer

Add new ioctl() to pause/resume ring-buffer output.

In some situations we want to read from the ring-buffer only when we
ensure nothing can write to the ring-buffer during reading. Without
this patch we have to turn off all events attached to this ring-buffer
to achieve this.

This patch is a prerequisite to enable overwrite support for the
perf ring-buffer support. Following commits will introduce new methods
support reading from overwrite ring buffer. Before reading, caller
must ensure the ring buffer is frozen, or the reading is unreliable.

Signed-off-by: Wang Nan <wangnan0@huawei.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <pi3orama@163.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: Zefan Li <lizefan@huawei.com>
Link: http://lkml.kernel.org/r/1459147292-239310-2-git-send-email-wangnan0@huawei.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

ftrace/perf: Check sample types only for sampling events

Currently we check sample type for ftrace:function events
even if it's not created as a sampling event. That prevents
creating ftrace_function event in counting mode.

Make sure we check sample types only for sampling events.

Before:
  $ sudo perf stat -e ftrace:function ls
  ...

   Performance counter stats for 'ls':

     <not supported>      ftrace:function

         0.001983662 seconds time elapsed

After:
  $ sudo perf stat -e ftrace:function ls
  ...

   Performance counter stats for 'ls':

              44,498      ftrace:function

         0.037534722 seconds time elapsed

Suggested-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1458138873-1553-2-git-send-email-jolsa@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/x86/intel/bts: Move transaction start/stop to start/stop callbacks

As per AUX buffer management requirement, AUX output has to happen between
pmu::start and pmu::stop calls so that perf_event_stop() actually stops it
and therefore perf can free the AUX data after it has called pmu::stop.

This patch moves perf_aux_output_{begin,end} from bts_event_{add,del} to
bts_event_{start,stop}. As a bonus, we get rid of bts_buffer_is_full(),
which is already taken care of by perf_aux_output_begin() anyway.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: vince@deater.net
Link: http://lkml.kernel.org/r/1457098969-21595-6-git-send-email-alexander.shishkin@linux.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/x86/intel/pt: Move transaction start/stop to PMU start/stop callbacks

As per AUX buffer management requirement, AUX output has to happen between
pmu::start and pmu::stop calls so that perf_event_stop() actually stops it
and therefore perf can free the AUX data after it has called pmu::stop.

This patch moves perf_aux_output_{begin,end} from pt_event_{add,del} to
pt_event_{start,stop}. As a bonus, we get rid of pt_buffer_is_full(),
which is already taken care of by perf_aux_output_begin() anyway.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: vince@deater.net
Link: http://lkml.kernel.org/r/1457098969-21595-5-git-send-email-alexander.shishkin@linux.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/ring_buffer: Document AUX API usage

In order to ensure safe AUX buffer management, we rely on the assumption
that pmu::stop() stops its ongoing AUX transaction and not just the hw.

This patch documents this requirement for the perf_aux_output_{begin,end}()
APIs.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: vince@deater.net
Link: http://lkml.kernel.org/r/1457098969-21595-4-git-send-email-alexander.shishkin@linux.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/core: Free AUX pages in unmap path

Now that we can ensure that when ring buffer's AUX area is on the way
to getting unmapped new transactions won't start, we only need to stop
all events that can potentially be writing aux data to our ring buffer.

Having done that, we can safely free the AUX pages and corresponding
PMU data, as this time it is guaranteed to be the last aux reference
holder.

This partially reverts:

  57ffc5ca679 ("perf: Fix AUX buffer refcounting")

... which was made to defer deallocation that was otherwise possible
from an NMI context. Now it is no longer the case; the last call to
rb_free_aux() that drops the last AUX reference has to happen in
perf_mmap_close() on that AUX area.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: vince@deater.net
Link: http://lkml.kernel.org/r/87d1qtz23d.fsf@ashishki-desk.ger.corp.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/ring_buffer: Refuse to begin AUX transaction after rb->aux_mmap_count drops

When ring buffer's AUX area is unmapped and rb->aux_mmap_count drops to
zero, new AUX transactions into this buffer can still be started,
even though the buffer in en route to deallocation.

This patch adds a check to perf_aux_output_begin() for rb->aux_mmap_count
being zero, in which case there is no point starting new transactions,
in other words, the ring buffers that pass a certain point in
perf_mmap_close will not have their events sending new data, which
clears path for freeing those buffers' pages right there and then,
provided that no active transactions are holding the AUX reference.

Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: vince@deater.net
Link: http://lkml.kernel.org/r/1457098969-21595-2-git-send-email-alexander.shishkin@linux.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/core: Verify we have a single perf_hw_context PMU

There should (and can) only be a single PMU for perf_hw_context
events.

This is because of how we schedule events: once a hardware event fails to
schedule (the PMU is 'full') we stop trying to add more. The trivial
'fix' would break the Round-Robin scheduling we do.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/x86: Move Kconfig.perf and other perf configuration bits to events/Kconfig

Ingo says:

 "If we do a separate file we should have it in arch/x86/events/Kconfig
  (not in arch/x86/Kconfig.perf), and also move some of the other bits,
  such as PERF_EVENTS_AMD_POWER?"

Suggested-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/x86/msr: Add AMD IRPERF (Instructions Retired) performance counter

AMD Zeppelin (Family 17h, Model 00h) introduces an instructions
retired performance counter which is indicated by
CPUID.8000_0008H:EBX[1]. A dedicated Instructions Retired MSR register
(MSR 0xC000_000E9) increments once for every instruction retired.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Borislav Petkov <bp@suse.de>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Jacob Shin <jacob.w.shin@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Robert Richter <rric@kernel.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1454056197-5893-3-git-send-email-ray.huang@amd.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/x86/msr: Add AMD PTSC (Performance Time-Stamp Counter) support

AMD Carrizo (Family 15h, Model 60h) introduces a time-stamp counter
which is indicated by CPUID.8000_0001H:ECX[27]. It increments at a 100
MHz rate in all P-states, and C states, S0, or S1. The frequency is
about 100MHz. This counter will be used to calculate processor power
and other parts. So add an interface into the MSR PMU to get the PTSC
counter value.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Borislav Petkov <bp@suse.de>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Jacob Shin <jacob.w.shin@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Robert Richter <rric@kernel.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1454056197-5893-2-git-send-email-ray.huang@amd.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

x86/perf/intel/cstate: Modularize driver

Add the exit function and allow the driver to be built as a module.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/20160320185623.658869675@linutronix.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>

x86/perf/intel/cstate: Sanitize error handling

There is no point in WARN_ON() inside of a well known init function. We
already know the call stack and it's really not of critical importance whether
the registration of a PMU fails.

Aside of that for consistency reasons it's just pointless to try to register
another PMU if the first register attempt failed. There is also no value in
keeping one PMU if the second one can not be registered.

Make it consistent so we can finaly modularize the driver.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/20160320185623.579794064@linutronix.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>

x86/perf/intel/cstate: Sanitize probing

The whole probing functionality can simply be expressed with model matching
and a bunch of structures describing the variants. This is a first step to
make that driver modular.

While at it, get rid of completely pointless comments and name the enums so
they are self explaining.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
[ Reworked probing to clear msr[].attr for all !present msrs. ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/20160320185623.500381872@linutronix.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>

x86/perf/intel/cstate: Make cstate hotplug handling actually work

The current implementation aside of being an incomprehensible mess is broken.

  # cat /sys/bus/event_source/devices/cstate_core/cpumask
  0-17

That's on a quad socket machine with 72 physical cores! Qualitee stuff.

So it's not a surprise that event migration in case of CPU hotplug does not
work either.

  # perf stat -e cstate_core/c6-residency/ -C 1 sleep 60 &
  # echo 0 >/sys/devices/system/cpu/cpu1/online

Tracing cstate_pmu_event_update gives me:

 [001] cstate_pmu_event_update <-event_sched_out

After the fix it properly moves the event:

 [001] cstate_pmu_event_update <-event_sched_out
 [073] cstate_pmu_event_update <-__perf_event_read
 [073] cstate_pmu_event_update <-event_sched_out

The migration of pkg events does not work either. Not that I'm surprised.

I really could not be bothered to decode that loop mess and simply replaced it
by querying the proper cpumasks which give us the answer in a comprehensible
way.

This also requires to direct the event to the current active reader CPU in
cstate_pmu_event_init() otherwise the hotplug logic can't work.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
[ Added event->cpu < 0 test to not explode]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/20160320185623.422519970@linutronix.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>

x86/perf/intel/rapl: Make the Intel RAPL PMU driver modular

By default, the RAPL driver will be built into the kernel. If it is
configured as a module, the supported CPU model can be auto loaded.

Also clean up the code of rapl_pmu_init().

Based-on-a-patch-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Kan Liang <kan.liang@intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1458372050-2420-2-git-send-email-kan.liang@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

x86/perf/intel/uncore: Make the Intel uncore PMU driver modular

By default, the uncore driver will be built into the kernel. If it is
configured as a module, the supported CPU model can be auto loaded.

This patch also cleans up the code of uncore_cpu_init() and
uncore_pci_init().

Based-on-a-patch-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Kan Liang <kan.liang@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/1458462817-2475-1-git-send-email-kan.liang@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

Merge branch 'perf/urgent' into perf/core, to fix up fixes before queueing up new changes

Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/x86/amd/ibs: Fix pmu::stop() nesting

Patch 5a50f5291701 ("perf/x86/ibs: Fix race with IBS_STARTING state")
closed a big hole while opening another, smaller hole.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Fixes: 5a50f5291701 ("perf/x86/ibs: Fix race with IBS_STARTING state")
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/core: Don't leak event in the syscall error path

In the error path, event_file not being NULL is used to determine
whether the event itself still needs to be free'd, so fix it up to
avoid leaking.

Reported-by: Leon Yu <chianglungyu@gmail.com>
Signed-off-by: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Fixes: 130056275ade ("perf: Do not double free")
Link: http://lkml.kernel.org/r/87twk06yxp.fsf@ashishki-desk.ger.corp.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf/core: Fix time tracking bug with multiplexing

Stephane reported that commit:

  3cbaa5906967 ("perf: Fix ctx time tracking by introducing EVENT_TIME")

introduced a regression wrt. time tracking, as easily observed by:

> This patch introduce a bug in the time tracking of events when
> multiplexing is used.
>
> The issue is easily reproducible with the following perf run:
>
>  $ perf stat -a -C 0 -e branches,branches,branches,branches,branches,branches -I 1000
>      1.000730239            652,394      branches   (66.41%)
>      1.000730239            597,809      branches   (66.41%)
>      1.000730239            593,870      branches   (66.63%)
>      1.000730239            651,440      branches   (67.03%)
>      1.000730239            656,725      branches   (66.96%)
>      1.000730239      <not counted>      branches
>
> One branches event is shown as not having run. Yet, with
> multiplexing, all events should run especially with a 1s (-I 1000)
> interval. The delta for time_running comes out to 0. Yet, the event
> has run because the kernel is actually multiplexing the events. The
> problem is that the time tracking is the kernel and especially in
> ctx_sched_out() is wrong now.
>
> The problem is that in case that the kernel enters ctx_sched_out() with the
> following state:
>    ctx->is_active=0x7 event_type=0x1
>    Call Trace:
>     [<ffffffff813ddd41>] dump_stack+0x63/0x82
>     [<ffffffff81182bdc>] ctx_sched_out+0x2bc/0x2d0
>     [<ffffffff81183896>] perf_mux_hrtimer_handler+0xf6/0x2c0
>     [<ffffffff811837a0>] ? __perf_install_in_context+0x130/0x130
>     [<ffffffff810f5818>] __hrtimer_run_queues+0xf8/0x2f0
>     [<ffffffff810f6097>] hrtimer_interrupt+0xb7/0x1d0
>     [<ffffffff810509a8>] local_apic_timer_interrupt+0x38/0x60
>     [<ffffffff8175ca9d>] smp_apic_timer_interrupt+0x3d/0x50
>     [<ffffffff8175ac7c>] apic_timer_interrupt+0x8c/0xa0
>
> In that case, the test:
>       if (is_active & EVENT_TIME)
>
> will be false and the time will not be updated. Time must always be updated on
> sched out.

Fix this by always updating time if EVENT_TIME was set, as opposed to
only updating time when EVENT_TIME changed.

Reported-by: Stephane Eranian <eranian@google.com>
Tested-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Cc: kan.liang@intel.com
Cc: namhyung@kernel.org
Fixes: 3cbaa5906967 ("perf: Fix ctx time tracking by introducing EVENT_TIME")
Link: http://lkml.kernel.org/r/20160329072644.GB3408@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>

Merge tag 'perf-core-for-mingo-20160330' of git://git./linux/kernel/git/acme/linux into perf/core

Pull perf/core improvements and fixes:

User visible changes:

  - Add support for skipping itrace instructions, useful to fast forward
    processor trace (Intel PT, BTS) to right after initialization code at the start
    of a workload (Andi Kleen)

  - Add support for backtraces in perl 'perf script's (Dima Kogan)

  - Add -U/-K (--all-user/--all-kernel) options to 'perf mem' (Jiri Olsa)

  - Make -f/--force option documentation consistent across tools (Jiri Olsa)

Infrastructure changes:

  - Add 'perf test' to check for event times (Jiri Olsa)

  - 'perf config' cleanups (Taeung Song)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

Merge tag 'perf-urgent-for-mingo-20160330' of git://git./linux/kernel/git/acme/linux into perf/urgent

Pull perf/urgent fixes from Arnaldo Carvalho de Melo:

 - Fix determination of a callchain node's childlessness in
   the top/report TUI, which was preventing navigating some
   callchains, --stdio unnaffected (Andres Freund)

 - Fix jitdump's genelf assumption that PowerPC is big endian
   only (Anton Blanchard)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf jit: genelf makes assumptions about endian

Commit 9b07e27f88b9 ("perf inject: Add jitdump mmap injection support")
incorrectly assumed that PowerPC is big endian only.

Simplify things by consolidating the define of GEN_ELF_ENDIAN and checking
for __BYTE_ORDER == __BIG_ENDIAN.

The PowerPC checks were also incorrect, they do not match what gcc
emits. We should first look for __powerpc64__, then __powerpc__.

Signed-off-by: Anton Blanchard <anton@samba.org>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>
Cc: Carl Love <cel@us.ibm.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org
Fixes: 9b07e27f88b9 ("perf inject: Add jitdump mmap injection support")
Link: http://lkml.kernel.org/r/20160329175944.33a211cc@kryten
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf hists: Fix determination of a callchain node's childlessness

The 4b3a3212233a ("perf hists browser: Support flat callchains") commit
over-aggressively tried to optimize callchain_node__init_have_children().

That lead to --tui mode not allowing to expand call chain elements if a
call chain element had only one parent. That's why --inverted callgraphs
looked halfway sane, but plain ones didn't.

Revert that individual optimization, it wasn't really related to the
rest of the commit.

Signed-off-by: Andres Freund <andres@anarazel.de>
Acked-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Fixes: 4b3a3212233a ("perf hists browser: Support flat callchains")
Link: http://lkml.kernel.org/r/20160330190245.GB13305@awork2.anarazel.de
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Add support for skipping itrace instructions

When using 'perf script' to look at PT traces it is often useful to
ignore the initialization code at the beginning.

On larger traces which may have many millions of instructions in
initialization code doing that in a pipeline can be very slow, with perf
script spending a lot of CPU time calling printf and writing data.

This patch adds an extension to the --itrace argument that skips 'n'
events (instructions, branches or transactions) at the beginning. This
is much more efficient.

v2:
Add support for BTS (Adrian Hunter)
Document in itrace.txt
Fix branch check
Check transactions and instructions too

Committer note:

To test intel_pt one needs to make sure VT-x isn't active, i.e.
stopping KVM guests on the test machine, as described by Andi Kleen
at http://lkml.kernel.org/r/20160301234953.GD23621@tassilo.jf.intel.com

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1459187142-20035-1-git-send-email-andi@firstfloor.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf script perl: Perl scripts now get a backtrace, like the python ones

We have some infrastructure to use perl or python to analyze logs
generated by perf.  Prior to this patch, only the python tools had
access to backtrace information.  This patch makes this information
available to perl scripts as well.  Example:

  Let's look at malloc() calls made by the seq utility.  First we
  create a probe point:

      $ perf probe -x /lib/x86_64-linux-gnu/libc.so.6 malloc
      Added new events:
      ...

  Now we run seq, while monitoring malloc() calls with perf

      $ perf record --call-graph=dwarf -e probe_libc:malloc seq 5
      1
      2
      3
      4
      5
      [ perf record: Woken up 1 times to write data ]
      [ perf record: Captured and wrote 0.064 MB perf.data (6 samples) ]

  We can use perf to look at its log to see the malloc calls and the backtrace

      $ perf script
      seq 14195 [000] 1927993.748254: probe_libc:malloc: (7f9ff8edd320) bytes=0x22
                  7f9ff8edd320 malloc (/lib/x86_64-linux-gnu/libc-2.22.so)
                  7f9ff8e8eab0 set_binding_values.part.0 (/lib/x86_64-linux-gnu/libc-2.22.so)
                  7f9ff8e8eda1 __bindtextdomain (/lib/x86_64-linux-gnu/libc-2.22.so)
                        401b22 main (/usr/bin/seq)
                  7f9ff8e82610 __libc_start_main (/lib/x86_64-linux-gnu/libc-2.22.so)
                        402799 _start (/usr/bin/seq)
      ...

  We can also use the scripting facilities.  We create a skeleton perl
  script that simply prints out the events

      $ perf script -g perl
      generated Perl script: perf-script.pl

  We can then use this script to see the malloc() calls with a
  backtrace.  Prior to this patch, the backtrace was not available to
  the perl scripts.

      $ perf script -s perf-script.pl
      probe_libc::malloc  0 1927993.748254260  14195 seq   __probe_ip=140325052863264, bytes=34
              [7f9ff8edd320] malloc
              [7f9ff8e8eab0] set_binding_values.part.0
              [7f9ff8e8eda1] __bindtextdomain
              [401b22] main
              [7f9ff8e82610] __libc_start_main
              [402799] _start
      ...

Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Link: http://lkml.kernel.org/r/87mvphzld0.fsf@secretsauce.net
Signed-off-by: Dima Kogan <dima@secretsauce.net>

perf config: Rename 'v' to 'home' in set_buildid_dir()

Change the variable name 'v' to 'home' to make it more readable.

Signed-off-by: Taeung Song <treeze.taeung@gmail.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1459099340-16911-3-git-send-email-treeze.taeung@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf config: Rework buildid_dir_command_config to perf_buildid_config

To avoid repeated calling perf_config() remove
buildid_dir_command_config() and add new perf_buildid_config into
perf_default_config.

Because perf_config() is already called with perf_default_config at
main().

Signed-off-by: Taeung Song <treeze.taeung@gmail.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1459099340-16911-2-git-send-email-treeze.taeung@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf config: Remove duplicated set_buildid_dir calls

Signed-off-by: Taeung Song <treeze.taeung@gmail.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1459099340-16911-1-git-send-email-treeze.taeung@gmail.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tests: Add test to check for event times

This test creates software event 'cpu-clock' attaches it in several ways
and checks that enabled and running times match.

Committer notes:

Testing it:

  [acme@jouet linux]$ perf test -v times
  44: Test events times                                        :
  --- start ---
  test child forked, pid 27170
  attaching to spawned child, enable on exec
    OK    : ena 307328, run 307328
  attaching to current thread as enabled
    OK    : ena 7826, run 7826
  attaching to current thread as disabled
    OK    : ena 738, run 738
  attaching to CPU 0 as enabled
    SKIP  : not enough rights
  attaching to CPU 0 as enabled
    SKIP  : not enough rights
  test child finished with -2
  ---- end ----
  Test events times: Skip
  [acme@jouet linux]$

  [root@jouet ~]# perf test times
  44: Test events times                                        : Ok
  [root@jouet ~]# perf test -v times
  44: Test events times                                        :
  --- start ---
  test child forked, pid 27306
  attaching to spawned child, enable on exec
    OK    : ena 479290, run 479290
  attaching to current thread as enabled
    OK    : ena 11356, run 11356
  attaching to current thread as disabled
    OK    : ena 987, run 987
  attaching to CPU 0 as enabled
    OK    : ena 3717, run 3717
  attaching to CPU 0 as enabled
    OK    : ena 2323, run 2323
  test child finished with 0
  ---- end ----
  Test events times: Ok
  [root@jouet ~]#

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1458823940-24583-7-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Make -f/--force option documentation consistent across tools

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1458823940-24583-6-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Make hists__collapse_insert_entry static

No need to export hists__collapse_insert_entry function.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1458823940-24583-4-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf mem: Add -U/-K (--all-user/--all-kernel) options

Add -U/-K (--all-user/--all-kernel) options to use the perf record
--all-user/--all-kernel options.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1458823940-24583-3-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

Merge tag 'perf-urgent-for-mingo-20160329' of git://git./linux/kernel/git/acme/linux into perf/urgent

Pull perf/urgent fix from Arnaldo Carvalho de Melo:

 - Add missing initialization of perf_sample.cpumode in synthesized samples,
   affects jitdump, records for pre-existing threads and records synthesized
   from processor trace data, noticed while testing intel_pt events with
   'perf script' (Arnaldo Carvalho de Melo)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf tools: Add missing initialization of perf_sample.cpumode in synthesized samples

In 473398a21d28 ("perf tools: Add cpumode to struct perf_sample"), I
missed some places where perf_sample fields are directly initialized in
addition to what is done in perf_evsel__parse_sample(), namely when
synthesizing PERF_RECORD_{MMAP*,COMM,FORK,EXIT} for pre-existing threads
and also in intel_pt and intel_bts when synthesizing events from
processor trace, the jitdump code also was affected, fix it.

The problem was noticed with running:

  # perf record -e intel_pt//u true
  # perf script

Where the samples wouldn't get resolved because perf_sample.cpumode
would be left as zero, i.e. PERF_RECORD_MISC_CPUMODE_UNKNOWN, not
resolving as kernel, hypervisor or user cpu modes.

Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: 473398a21d28 ("perf tools: Add cpumode to struct perf_sample")
Link: http://lkml.kernel.org/n/tip-n5sdauxgk24d5nun8kuuu2mh@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

Merge tag 'perf-urgent-for-mingo-20160328' of git://git./linux/kernel/git/acme/linux into perf/urgent

Pull perf/urgent fix from Arnaldo Carvalho de Melo:

  - Fix build break on PowerPC due to missing headers with prototypes for
    functions defined in tools/perf/arch/powerpc/util/header.c (Sukadev Bhattiprolu)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf tools: Fix build break on powerpc

Commit 531d2410635c ("perf tools: Do not include stringify.h from the
kernel sources") seems to have accidentially removed the inclusion of
"util/header.h" from "arch/powerpc/util/header.c".

"util/header.h" provides the prototype for get_cpuid() and is needed to
build perf on Powerpc:

arch/powerpc/util/header.c:17:1: error: no previous prototype for 'get_cpuid' [-Werror=missing-prototypes]

Reported-by: Michael Ellerman <mpe@ellerman.id.au>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Fixes: 531d2410635c ("perf tools: Do not include stringify.h from the kernel sources")
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
[ Included "util.h" too, to get the scnprintf() prototype ]
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf/x86: Move events_sysfs_show() outside CPU_SUP_INTEL

randconfig builds can sometimes disable CONFIG_CPU_SUP_INTEL while
enabling the AMD power reporting PMU driver, resulting in this
build failure:

  arch/x86/kernel/cpu/perf_event.h:663:31: error: 'events_sysfs_show' undeclared here (not in a function)

To fix it, move events_sysfs_show() outside of #ifdef CONFIG_CPU_SUP_INTEL.

Reported-by: Randy Dunlap <rdunlap@infradead.org>
Reported-by: build test robot <lkp@intel.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Borislav Petkov <bp@suse.de>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Sherry Hurwitz <sherry.hurwitz@amd.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: acme@kernel.org
Cc: kbuild-all@01.org
Cc: linux-next@vger.kernel.org
Cc: spg_linux_kernel@amd.com
Link: http://lkml.kernel.org/r/1458875905-4278-1-git-send-email-ray.huang@amd.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>

Merge tag 'perf-urgent-for-mingo-20160324' of git://git./linux/kernel/git/acme/linux into perf/urgent

Pull perf/urgent fixes from Arnaldo Carvalho de Melo:

 - Fix detached tarball building due to missing 'perf bench memcpy' headers
   (Arnaldo Carvalho de Melo)

 - Fix tarpkg build test error output redirection (Arnaldo Carvalho de Melo)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf bench: Fix detached tarball building due to missing 'perf bench memcpy' headers

A change on kernel files included by the 'perf bench memcpy' code grew some new
include deps, breaking the detached tarball build:

  $ make -C tools/perf build-test
  make: Entering directory '/home/acme/git/linux/tools/perf'
  - tarpkg: ./tests/perf-targz-src-pkg .
  tests/make:302: recipe for target 'tarpkg' failed
  make[1]: *** [tarpkg] Error 2
  Makefile:102: recipe for target 'build-test' failed
  make: *** [build-test] Error 2
  make: Leaving directory '/home/acme/git/linux/tools/perf'
  $ cat tools/perf/tarpkg
  ./tests/perf-targz-src-pkg .
    PERF_VERSION = 4.5.g05f5ec
    PERF_VERSION = 4.5.g05f5ec
  In file included from bench/mem-memcpy-x86-64-asm.S:9:0:
  bench/../../../arch/x86/lib/memcpy_64.S:5:29: fatal error: asm/cpufeatures.h: No such file or directory
  compilation terminated.
  mv: cannot stat ‘bench/.mem-memcpy-x86-64-asm.o.tmp’: No such file or directory
  make[5]: *** [bench/mem-memcpy-x86-64-asm.o] Error 1
  make[5]: *** Waiting for unfinished jobs....
  make[4]: *** [bench] Error 2
  make[4]: *** Waiting for unfinished jobs....
  make[3]: *** [perf-in.o] Error 2
  make[3]: *** Waiting for unfinished jobs....
  make[2]: *** [all] Error 2
  $

Add arch/*/include/asm/*features.h to tools/perf/MANIFEST so that we can
continue to use detached tarballs to build perf.

Now it builds ok, doing it manually:

  $ make help | grep perf
    perf-tar-src-pkg    - Build perf-4.5.0.tar source tarball
    perf-targz-src-pkg  - Build perf-4.5.0.tar.gz source tarball
    perf-tarbz2-src-pkg - Build perf-4.5.0.tar.bz2 source tarball
    perf-tarxz-src-pkg  - Build perf-4.5.0.tar.xz source tarball
  $ ls -la perf-4.5.0.tar
  ls: cannot access perf-4.5.0.tar: No such file or directory
  $ make perf-tar-src-pkg
    TAR
    PERF_VERSION = 4.5.g32c25b
  $ ls -la perf-4.5.0.tar
  -rw-rw-r--. 1 acme acme 6318080 Mar 24 11:52 perf-4.5.0.tar
  $ mv perf-4.5.0.tar /tmp
  $ cd /tmp
  $ tar xf perf-4.5.0.tar
  $ cd perf-4.5.0/tools/perf
  $ make > /dev/null
  PERF_VERSION = 4.5.g32c25b
  $ ls -la perf
  -rwxrwxr-x. 1 acme acme 14046416 Mar 24 11:53 perf
  $ ./perf --version
  perf version 4.5.g32c25b
  $ perf bench
  Usage:
perf bench [<common options>] <collection> <benchmark> [<options>]

        # List of all available benchmark collections:

         sched: Scheduler and IPC benchmarks
           mem: Memory access benchmarks
          numa: NUMA scheduling and MM benchmarks
         futex: Futex stressing benchmarks
           all: All benchmarks

  $ perf bench mem

        # List of available benchmarks for collection 'mem':

        memcpy: Benchmark for memcpy() functions
        memset: Benchmark for memset() functions
           all: Run all memory access benchmarks

  $ perf bench mem memcpy
  # Running 'mem/memcpy' benchmark:
  # function 'default' (Default memcpy() provided by glibc)
  # Copying 1MB bytes ...

        15.024038 GB/sec
  # function 'x86-64-unrolled' (unrolled memcpy() in arch/x86/lib/memcpy_64.S)
  # Copying 1MB bytes ...

        17.438616 GB/sec
  # function 'x86-64-movsq' (movsq-based memcpy() in arch/x86/lib/memcpy_64.S)
  # Copying 1MB bytes ...

        25.040064 GB/sec
  # function 'x86-64-movsb' (movsb-based memcpy() in arch/x86/lib/memcpy_64.S)
  # Copying 1MB bytes ...

        25.040064 GB/sec
  $

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: David Ahern <dsahern@gmail.com>
Cc: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-2c2sncwffuabw58fj1pw86gu@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tests: Fix tarpkg build test error output redirection

So we need to trow away just stdout, leaving stderr to be caught by
the build tests infrastructure, so that we can see what went wrong
when the tarpkg build test fails:

  $ make -C tools/perf build-test
  make: Entering directory '/home/acme/git/linux/tools/perf'
  - tarpkg: ./tests/perf-targz-src-pkg .
  tests/make:302: recipe for target 'tarpkg' failed
  make[1]: *** [tarpkg] Error 2
  Makefile:102: recipe for target 'build-test' failed
  make: *** [build-test] Error 2
  make: Leaving directory '/home/acme/git/linux/tools/perf'
  $ cat tools/perf/tarpkg
  ./tests/perf-targz-src-pkg .
    PERF_VERSION = 4.5.g05f5ec
    PERF_VERSION = 4.5.g05f5ec
  In file included from bench/mem-memcpy-x86-64-asm.S:9:0:
  bench/../../../arch/x86/lib/memcpy_64.S:5:29: fatal error: asm/cpufeatures.h: No such file or directory
  compilation terminated.
  mv: cannot stat ‘bench/.mem-memcpy-x86-64-asm.o.tmp’: No such file or directory
  make[5]: *** [bench/mem-memcpy-x86-64-asm.o] Error 1
  make[5]: *** Waiting for unfinished jobs....
  make[4]: *** [bench] Error 2
  make[4]: *** Waiting for unfinished jobs....
  make[3]: *** [perf-in.o] Error 2
  make[3]: *** Waiting for unfinished jobs....
  make[2]: *** [all] Error 2
  $

So the test flow is:

1. Run: 'make -C tools/perf build-test'

2. One of its tests failed, in this case, the 'tarpkg' one

3. Look at what went wrong, by looking at the output of that test, in
   tools/perf/tarpkg

Admittedly, this should be shortcircuited to showing what went wrong directly
from the 'make build-test' step, but lets first fix this tarpkg one and the
problem it spotted, which should be fixed by adding some extra file to the
tools/perf/MANIFEST so that detached tarballs continue being self contained and
build successfully.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-ynld6egoxolmftcddpnd7oh6@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

Merge tag 'perf-core-for-mingo-20160323' of git://git./linux/kernel/git/acme/linux into perf/urgent

Pull perf/core improvements and fixes:

User visible fixes:

 - Fix documentation of :ppp modifier in 'perf list' (Andi Kleen)

 - Fix silly nodes bitfield bits/bytes length assertion in 'perf bench numa' (Jakub Jelen)

 - Remove redundant CPU output in libtraceevent (Steven Rostedt)

 - Remove 'core_id' check in topology 'perf test' (Sukadev Bhattiprolu)

Infrastructure changes/fixes:

 - Record text offset in dso to calculate objdump address, to use with
   modules in addition to vDSO symbol address calculations (Wang Nan)

 - Move utilities.mak from perf to tools/scripts/ (Arnaldo Carvalho de Melo)

 - Add cpumode to the perf_sample struct, this way we don't need to pass
   the union event to the machine and thread resolving routines, shortening
   function signatures and allowing the future introduction of a way
   to use tracepoint events instead of the unavailable HW cycles counter on
   powerpc guests in perf kvm by just hooking on perf_evsel__parse_sample,
   at the end (Arnaldo Carvalho de Melo)

 - Remove/unexport die() related infrastructure, that at some point will
   finally be removed (Arnaldo Carvalho de Melo)

 - Adopt linux/stringify.h from the kernel sources, not to touch this
   kernel header from tools/ (Arnaldo Carvalho de Melo)

 - Stop using strbuf for things we can instead trivially use libc's asprintf()
   (Arnaldo Carvalho de Melo)

 - Ditch tools/lib/util/abspath.c, its only exported function was used at just
   one place and can be replaced by libc's realpath() (Arnaldo Carvalho de Melo)

 - Use strerror_r() in the llvm infrastructure, tread safe, its what is used
   elsewhere in tools/perf/ (Arnaldo Carvalho de Melo)

Cleanups:

 - Removed misplaced or needless __maybe_unused/export (Arnaldo Carvalho de Melo)

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

perf llvm: Use strerror_r instead of the thread unsafe strerror one

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-5njrq9dltckgm624omw9ljgu@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf llvm: Use realpath to canonicalize paths

To kill the last user of make_nonrelative_path(), that gets ditched,
one more panicking function killed.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-3hu56rvyh4q5gxogovb6ko8a@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Unexport some methods unused outside strbuf.c

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-nq1wvtky4mpu0nupjyar7sbw@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf probe: No need to use formatting strbuf method

We have addch() for chars, add() for fixed size data, and addstr() for
variable length strings, use them.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-0ap02fn2xtvpduj2j6b2o1j4@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf help: Use asprintf instead of adhoc equivalents

That doesn't chekcs malloc return and that, when using strbuf, if it
can't grow, just explodes away via die().

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-vr8qsjbwub7e892hpa9msz95@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Remove unused perf_pathdup, xstrdup functions

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-s87zi5d03m6rz622y1z6rlsa@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Do not include stringify.h from the kernel sources

Use instead the copy just made to tools/include/linux/.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-q736w12nwy98x5ox2hamp5ow@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools include: Copy linux/stringify.h from the kernel

There is code in tools/ that is directly including this file from the
kernel, and this is verboten for a while, copy it so that the next csets
can fix this situation.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-e0r3nks2uai020ndghvxv5qw@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

tools lib traceevent: Remove redundant CPU output

Commit a6745330789f ("tools lib traceevent: Split pevent_print_event()
into specific functionality functions") broke apart the function
pevent_print_event() into three functions.

The first function prints the comm, pid and CPU, the second prints the
timestamp.

But that commit added the printing of the CPU in the timestamp function,
which now causes pevent_print_event() to duplicate the CPU output.

Remove the redundant printing of the record's CPU from the timestamp
function.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Fixes: a6745330789f ("tools lib traceevent: Split pevent_print_event() into specific functionality functions")
Link: http://lkml.kernel.org/r/20160323101628.459375d2@gandalf.local.home
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Remove needless 'extern' from function prototypes

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-w246stf7ponfamclsai6b9zo@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Simplify die() mechanism

This should die altogether, but for now lets remove a bit of this stuff,
as it is not used at all.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-ade3n99xscldhg5mx2vzd8p3@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf tools: Remove unused DIE_IF macro

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-elxg25jd4dhwod4wqbko87qh@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

perf script: Remove lots of unused arguments

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>