Skip to content
  1. Feb 25, 2019
  3. Feb 23, 2019
  5. Feb 22, 2019
    • Adrian Hunter's avatar
      perf thread-stack: Improve thread_stack__no_call_return() · 1f35cd65
      Adrian Hunter authored 
      

Improve thread_stack__no_call_return() to better handle 'returns' that
do not match the stack i.e. 'no call'. See code comments for details.
The example below shows how retpolines are affected:

Example:

  $ cat simple-retpoline.c
  __attribute__((noinline)) int bar(void)
  {
          return -1;
  }

  int foo(void)
  {
          return bar() + 1;
  }

  __attribute__((indirect_branch("thunk"))) int main()
  {
          int (*volatile fn)(void) = foo;

          fn();
          return fn();
  }
  $ gcc -ggdb3 -Wall -Wextra -O2 -o simple-retpoline simple-retpoline.c
  $ objdump -d simple-retpoline
  <SNIP>
  0000000000001040 <main>:
      1040:       48 83 ec 18             sub    $0x18,%rsp
      1044:       48 8d 05 25 01 00 00    lea    0x125(%rip),%rax        # 1170 <foo>
      104b:       48 89 44 24 08          mov    %rax,0x8(%rsp)
      1050:       48 8b 44 24 08          mov    0x8(%rsp),%rax
      1055:       e8 1f 01 00 00          callq  1179 <__x86_indirect_thunk_rax>
      105a:       48 8b 44 24 08          mov    0x8(%rsp),%rax
      105f:       48 83 c4 18             add    $0x18,%rsp
      1063:       e9 11 01 00 00          jmpq   1179 <__x86_indirect_thunk_rax>
  <SNIP>
  0000000000001160 <bar>:
      1160:       b8 ff ff ff ff          mov    $0xffffffff,%eax
      1165:       c3                      retq
  <SNIP>
  0000000000001170 <foo>:
      1170:       e8 eb ff ff ff          callq  1160 <bar>
      1175:       83 c0 01                add    $0x1,%eax
      1178:       c3                      retq
  0000000000001179 <__x86_indirect_thunk_rax>:
      1179:       e8 07 00 00 00          callq  1185 <__x86_indirect_thunk_rax+0xc>
      117e:       f3 90                   pause
      1180:       0f ae e8                lfence
      1183:       eb f9                   jmp    117e <__x86_indirect_thunk_rax+0x5>
      1185:       48 89 04 24             mov    %rax,(%rsp)
      1189:       c3                      retq
  <SNIP>
  $ perf record -o simple-retpoline.perf.data -e intel_pt/cyc/u ./simple-retpoline
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0,017 MB simple-retpoline.perf.data ]
  $ perf script -i simple-retpoline.perf.data --itrace=be -s ~/libexec/perf-core/scripts/python/export-to-sqlite.py simple-retpoline.db branches calls
  2019-01-08 14:03:37.851655 Creating database...
  2019-01-08 14:03:37.863256 Writing records...
  2019-01-08 14:03:38.069750 Adding indexes
  2019-01-08 14:03:38.078799 Done
  $ ~/libexec/perf-core/scripts/python/exported-sql-viewer.py simple-retpoline.db

Before:

    main
        -> __x86_indirect_thunk_rax
            -> __x86_indirect_thunk_rax
                -> __x86_indirect_thunk_rax
                    -> bar

After:

    main
        -> __x86_indirect_thunk_rax
            -> __x86_indirect_thunk_rax
                -> foo
                    -> bar

Committer testing:

Chose "Reports", Then "Context-Sensitive Call Graph" and then go on
expanding:

Before:

simple-retpolin
   PID:PID
      _start
         _start
            __libc_start_main
               main
                   __x86_indirect_thunk_rax
                      __x86_indirect_thunk_rax
                      bar

After:

Remove the "simple.retpoline.db" file, run again the 'perf script' line
to regenerate the .db file and run the exported-sql-viewer.py again to
get the same all the way to 'main', then, from there, including 'main':

               main
                   __x86_indirect_thunk_rax
                       __x86_indirect_thunk_rax
                           foo
                               bar

Signed-off-by: default avatarAdrian Hunter <adrian.hunter@intel.com>
Tested-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: default avatarJiri Olsa <jolsa@kernel.org>
Link: http://lkml.kernel.org/r/20190109091835.5570-6-adrian.hunter@intel.com


Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      1f35cd65
    • Wei Li's avatar
      perf annotate: Fix getting source line failure · 11db1ad4
      Wei Li authored 
      The output of "perf annotate -l --stdio xxx" changed since commit 425859ff


("perf annotate: No need to calculate notes->start twice") removed notes->start
assignment in symbol__calc_lines(). It will get failed in
find_address_in_section() from symbol__tty_annotate() subroutine as the
a2l->addr is wrong. So the annotate summary doesn't report the line number of
source code correctly.

Before fix:

  liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ cat common_while_1.c
  void hotspot_1(void)
  {
	volatile int i;

	for (i = 0; i < 0x10000000; i++);
	for (i = 0; i < 0x10000000; i++);
	for (i = 0; i < 0x10000000; i++);
  }

  int main(void)
  {
	hotspot_1();

	return 0;
  }
  liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ gcc common_while_1.c -g -o common_while_1

  liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ sudo ./perf record ./common_while_1
  [ perf record: Woken up 2 times to write data ]
  [ perf record: Captured and wrote 0.488 MB perf.data (12498 samples) ]
  liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ sudo ./perf annotate -l -s hotspot_1 --stdio

  Sorted summary for file /home/liwei/main_code/hulk_work/hulk/tools/perf/common_while_1
  ----------------------------------------------

   19.30 common_while_1[32]
   19.03 common_while_1[4e]
   19.01 common_while_1[16]
    5.04 common_while_1[13]
    4.99 common_while_1[4b]
    4.78 common_while_1[2c]
    4.77 common_while_1[10]
    4.66 common_while_1[2f]
    4.59 common_while_1[51]
    4.59 common_while_1[35]
    4.52 common_while_1[19]
    4.20 common_while_1[56]
    0.51 common_while_1[48]
   Percent |      Source code & Disassembly of common_while_1 for cycles:ppp (12480 samples, percent: local period)
  -----------------------------------------------------------------------------------------------------------------
         :
         :
         :
         :         Disassembly of section .text:
         :
         :         00000000000005fa <hotspot_1>:
         :         hotspot_1():
         :         void hotspot_1(void)
         :         {
    0.00 :   5fa:   push   %rbp
    0.00 :   5fb:   mov    %rsp,%rbp
         :                 volatile int i;
         :
         :                 for (i = 0; i < 0x10000000; i++);
    0.00 :   5fe:   movl   $0x0,-0x4(%rbp)
    0.00 :   605:   jmp    610 <hotspot_1+0x16>
    0.00 :   607:   mov    -0x4(%rbp),%eax
   common_while_1[10]    4.77 :   60a:   add    $0x1,%eax
   common_while_1[13]    5.04 :   60d:   mov    %eax,-0x4(%rbp)
   common_while_1[16]   19.01 :   610:   mov    -0x4(%rbp),%eax
   common_while_1[19]    4.52 :   613:   cmp    $0xfffffff,%eax
      0.00 :   618:   jle    607 <hotspot_1+0xd>
           :                 for (i = 0; i < 0x10000000; i++);
  ...

After fix:

  liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ sudo ./perf record ./common_while_1
  [ perf record: Woken up 2 times to write data ]
  [ perf record: Captured and wrote 0.488 MB perf.data (12500 samples) ]
  liwei@euler:~/main_code/hulk_work/hulk/tools/perf$ sudo ./perf annotate -l -s hotspot_1 --stdio

  Sorted summary for file /home/liwei/main_code/hulk_work/hulk/tools/perf/common_while_1
  ----------------------------------------------

   33.34 common_while_1.c:5
   33.34 common_while_1.c:6
   33.32 common_while_1.c:7
   Percent |      Source code & Disassembly of common_while_1 for cycles:ppp (12482 samples, percent: local period)
  -----------------------------------------------------------------------------------------------------------------
         :
         :
         :
         :         Disassembly of section .text:
         :
         :         00000000000005fa <hotspot_1>:
         :         hotspot_1():
         :         void hotspot_1(void)
         :         {
    0.00 :   5fa:   push   %rbp
    0.00 :   5fb:   mov    %rsp,%rbp
         :                 volatile int i;
         :
         :                 for (i = 0; i < 0x10000000; i++);
    0.00 :   5fe:   movl   $0x0,-0x4(%rbp)
    0.00 :   605:   jmp    610 <hotspot_1+0x16>
    0.00 :   607:   mov    -0x4(%rbp),%eax
   common_while_1.c:5    4.70 :   60a:   add    $0x1,%eax
    4.89 :   60d:   mov    %eax,-0x4(%rbp)
   common_while_1.c:5   19.03 :   610:   mov    -0x4(%rbp),%eax
   common_while_1.c:5    4.72 :   613:   cmp    $0xfffffff,%eax
    0.00 :   618:   jle    607 <hotspot_1+0xd>
         :                 for (i = 0; i < 0x10000000; i++);
    0.00 :   61a:   movl   $0x0,-0x4(%rbp)
    0.00 :   621:   jmp    62c <hotspot_1+0x32>
    0.00 :   623:   mov    -0x4(%rbp),%eax
   common_while_1.c:6    4.54 :   626:   add    $0x1,%eax
    4.73 :   629:   mov    %eax,-0x4(%rbp)
   common_while_1.c:6   19.54 :   62c:   mov    -0x4(%rbp),%eax
   common_while_1.c:6    4.54 :   62f:   cmp    $0xfffffff,%eax
  ...

Signed-off-by: default avatarWei Li <liwei391@huawei.com>
Acked-by: default avatarJiri Olsa <jolsa@kernel.org>
Tested-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Jin Yao <yao.jin@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Fixes: 425859ff ("perf annotate: No need to calculate notes->start twice")
Link: http://lkml.kernel.org/r/20190221095716.39529-1-liwei391@huawei.com


Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      11db1ad4
  7. Feb 21, 2019
  9. Feb 20, 2019
    • Arnaldo Carvalho de Melo's avatar
      perf trace: Allow dumping a BPF map after setting up BPF events · ff7a4f98
      Arnaldo Carvalho de Melo authored 
      Initial use case:

Dumping the maps setup by tools/perf/examples/bpf/augmented_raw_syscalls.c,
which so far are just booleans, showing just non-zeroed entries:

  # cat ~/.perfconfig
  [llvm]
	dump-obj = true
	clang-opt = -g
  [trace]
	#add_events = /home/acme/git/perf/tools/perf/examples/bpf/augmented_raw_syscalls.o
	add_events = /wb/augmented_raw_syscalls.o
  $ date
  Tue Feb 19 16:29:33 -03 2019
  $ ls -la /wb/augmented_raw_syscalls.o
  -rwxr-xr-x. 1 root root 14048 Jan 24 12:09 /wb/augmented_raw_syscalls.o
  $ file /wb/augmented_raw_syscalls.o
  /wb/augmented_raw_syscalls.o: ELF 64-bit LSB relocatable, eBPF, version 1 (SYSV), with debug_info, not stripped
  $
  # trace -e recvmmsg,sendmmsg --map-dump foobar
  ERROR: BPF map "foobar" not found
  # trace -e recvmmsg,sendmmsg --map-dump filtered_pids
  ERROR: BPF map "filtered_pids" not found
  # trace -e recvmmsg,sendmmsg --map-dump pids_filtered
  [2583] = 1,
  [2267] = 1,
  ^Z
  [1]+  Stopped                 trace -e recvmmsg,sendmmsg --map-dump pids_filtered
  # pidof trace
  2267
  # ps ax|grep gnome-terminal|grep -v grep
  2583 ?        Ssl   58:33 /usr/libexec/gnome-terminal-server
  ^C
  # trace -e recvmmsg,sendmmsg --map-dump syscalls
  [299] = 1,
  [307] = 1,
  ^C
  # grep x64_recvmmsg arch/x86/entry/syscalls/syscall_64.tbl
  299	64	recvmmsg		__x64_sys_recvmmsg
  # grep x64_sendmmsg arch/x86/entry/syscalls/syscall_64.tbl
  307	64	sendmmsg		__x64_sys_sendmmsg
  #

Next step probably will be something like 'perf stat's --interval-print and
--interval-clear.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Martin KaFai Lau <kafai@fb.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Yonghong Song <yhs@fb.com>
Link: https://lkml.kernel.org/n/tip-ztxj25rtx37ixo9cfajt8ocy@git.kernel.org


Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      ff7a4f98
    • Arnaldo Carvalho de Melo's avatar
      perf bpf: Add bpf_map dumper · d19f8564
      Arnaldo Carvalho de Melo authored 
      At some point I'll suggest moving this to libbpf, for now I'll
experiment with ways to dump BPF maps set by events in 'perf trace',
starting with a very basic dumper for the current very limited needs
of the augmented_raw_syscalls code: dumping booleans.

Having functions that apply to the map keys and values and do table
lookup in things like syscall id to string tables should come next.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Martin KaFai Lau <kafai@fb.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Yonghong Song <yhs@fb.com>
Link: https://lkml.kernel.org/n/tip-lz14w0esqyt1333aon05jpwc@git.kernel.org


Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      d19f8564
    • Thomas Richter's avatar
      perf test: Fix failure of 'evsel-tp-sched' test on s390 · 03d30971
      Thomas Richter authored 
      Commit 489338a7 ("perf tests evsel-tp-sched: Fix bitwise operator")
causes test case 14 "Parse sched tracepoints fields" to fail on s390.

This test succeeds on x86.

In fact this test now fails on all architectures with type char treated
as type unsigned char.

The root cause is the signed-ness of character arrays in the tracepoints
sched_switch for structure members prev_comm and next_comm.

On s390 the output of:

 [root@m35lp76 perf]# cat /sys/kernel/debug/tracing/events/sched/sched_switch/format
 name: sched_switch
 ID: 287
 format:
   field:unsigned short common_type; offset:0; size:2;	signed:0;
   ...
   field:char prev_comm[16]; offset:8; size:16;	signed:0;
   ...
   field:char next_comm[16]; offset:40; size:16; signed:0;

reveals the character arrays prev_comm and next_comm are per
default unsigned char and have values in the range of 0..255.

On x86 both fields are signed as this output shows:
 [root@f29]# cat /sys/kernel/debug/tracing/events/sched/sched_switch/format
 name: sched_switch
 ID: 287
 format:
   field:unsigned short common_type; offset:0; size:2;	signed:0;
   ...
   field:char prev_comm[16]; offset:8; size:16;	signed:1;
   ...
   field:char next_comm[16]; offset:40; size:16; signed:1;

and the character arrays prev_comm and next_comm are per default signed
char and have values in the range of -1..127.  The implementation of
type char is architecture specific.

Since the character arrays in both tracepoints sched_switch and
sched_wakeup should contain ascii characters, simply omit the check for
signedness in the test case.

Output before:

  [root@m35lp76 perf]# ./perf test -F 14
  14: Parse sched tracepoints fields                        :
  --- start ---
  sched:sched_switch: "prev_comm" signedness(0) is wrong, should be 1
  sched:sched_switch: "next_comm" signedness(0) is wrong, should be 1
  sched:sched_wakeup: "comm" signedness(0) is wrong, should be 1
  ---- end ----
  14: Parse sched tracepoints fields                        : FAILED!
  [root@m35lp76 perf]#

Output after:

  [root@m35lp76 perf]# ./perf test -Fv 14
  14: Parse sched tracepoints fields                        :
  --- start ---
  ---- end ----
  Parse sched tracepoints fields: Ok
  [root@m35lp76 perf]#

Fixes: 489338a7

 ("perf tests evsel-tp-sched: Fix bitwise operator")

Signed-off-by: default avatarThomas Richter <tmricht@linux.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Link: http://lkml.kernel.org/r/20190219153639.31267-1-tmricht@linux.ibm.com


Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      03d30971
    • Jonas Rabenstein's avatar
      perf doc: Fix documentation of the Flags section in perf.data · 8c23a522
      Jonas Rabenstein authored 
      

According to the current documentation the flags section is placed after
the file header itself but the code assumes to find the flags section
after the data section. This change updates the documentation to that
assumption.

Signed-off-by: default avatarJonas Rabenstein <jonas.rabenstein@studium.uni-erlangen.de>
Acked-by: default avatarJiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Link: http://lkml.kernel.org/r/20190219154515.3954-2-jonas.rabenstein@studium.uni-erlangen.de


Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      8c23a522
    • Jonas Rabenstein's avatar
      perf doc: Fix HEADER_CMDLINE description in perf.data documentation · 7a663c0f
      Jonas Rabenstein authored 
      

The content of the HEADER_CMDLINE feature header is a perf_header_string_list
of the argument vector and not a perf_header_string of the commandline.

Signed-off-by: default avatarJonas Rabenstein <jonas.rabenstein@studium.uni-erlangen.de>
Acked-by: default avatarJiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Richter <tmricht@linux.ibm.com>
Link: http://lkml.kernel.org/r/20190219154515.3954-1-jonas.rabenstein@studium.uni-erlangen.de


Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      7a663c0f
  11. Feb 19, 2019