linux.git/tools/perf, branch v4.18-rc4

Merge tag 'perf-urgent-for-mingo-4.18-20180625' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux into perf/urgent

2018-06-26T06:37:57+00:00

Pull perf/urgent fixes from Arnaldo Carvalho de Melo:

perf bench: (Jiri Olsa):

- Fix NUMA report output code handling of less than 1s runtimes.

perf script: (Ravi Bangoria)

- Add missing output fields in a 'perf script -h' hint.

- Fix crash because of missing evsel->priv.

- Fix crash caused by accessing feat_ops[HEADER_LAST_FEATURE], which
  is just a end of features header marker.

perf stat: (Thomas Richter)

- Remove duplicate event counting

perf test:

- Wire parsing error handling in 'parse events' test (Jiri Olsa)

- Fix 'session topology' test on s/390 (Thomas Richter)

eBPF: (Yonghong Song)

- Fix a clang 7.0 compilation error when building perf linking
  with libclang

intel-pt: (Adrian Hunter)

- Fix packet decoding of CYC packets.

Copies of kernel files: (Arnaldo Carvalho de Melo)

- Synchronize drm/drm.h UAPI

- Update x86's syscall_64.tbl, adding support for 'io_pgetevents' and 'rseq'
  in 'perf trace'.

- Update powerpc uapi/asm/unistd.h, adding support for the 'rseq' syscall.

- Update if_link.h and bpf.h, no effect on tool features.

PowerPC: (Sandipan Das)

- Fix crash if callchain is empty.

s/390: (Thomas Richter)

- Support random socked_id assignment in the perf header.

- Support s390 random socket_id assignment in perf.data file.

- Make PMU alias definitions taken from sysfs and JSON files comparable
  by normalizing them wrt spaces and newlines.

Signed-off-by: Arnaldo Carvalho de Melo 
Signed-off-by: Ingo Molnar

perf tools: Fix crash caused by accessing feat_ops[HEADER_LAST_FEATURE]

2018-06-25T14:59:37+00:00

perf_event__process_feature() accesses feat_ops[HEADER_LAST_FEATURE]
which is not defined and thus perf is crashing. HEADER_LAST_FEATURE is
used as an end marker for the perf report but it's unused for perf
script/annotate. Ignore HEADER_LAST_FEATURE for perf script/annotate,
just like it is done in 'perf report'.

Before:
  # perf record -o - ls | perf script
  
  Segmentation fault (core dumped)
  #

After:
  # perf record -o - ls | perf script
  
  Segmentation fault (core dumped)
  ls 7031 4392.099856:  250000 cpu-clock:uhH:  7f5e0ce7cd60
  ls 7031 4392.100355:  250000 cpu-clock:uhH:  7f5e0c706ef7
  #

Signed-off-by: Ravi Bangoria 
Cc: Alexander Shishkin 
Cc: Andi Kleen 
Cc: David Ahern 
Cc: David Carrillo-Cisneros 
Cc: Jin Yao 
Cc: Jiri Olsa 
Cc: Namhyung Kim 
Fixes: 57b5de463925 ("perf report: Support forced leader feature in pipe mode")
Link: http://lkml.kernel.org/r/20180625124220.6434-4-ravi.bangoria@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf script: Fix crash because of missing evsel->priv

2018-06-25T14:59:37+00:00

'perf script' in piped mode is crashing because evsel->priv is not set
properly. Fix it.

Before:

  # perf record -o - -- ls | perf script
  
    Segmentation fault (core dumped)
  #

After:

  # perf record -o - -- ls | perf script
  
  ls 2282 1031.731974:  250000 cpu-clock:uhH:  7effe4b3d29e
  ls 2282 1031.732222:  250000 cpu-clock:uhH:  7effe4b3a650
  #

Signed-off-by: Ravi Bangoria 
Tested-by: Arnaldo Carvalho de Melo 
Cc: Alexander Shishkin 
Cc: Andi Kleen 
Cc: David Ahern 
Cc: David Carrillo-Cisneros 
Cc: Jin Yao 
Cc: Jiri Olsa 
Cc: Namhyung Kim 
Fixes: a14390fde64e ("perf script: Allow creating per-event dump files")
Link: http://lkml.kernel.org/r/20180625124220.6434-3-ravi.bangoria@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf script: Add missing output fields in a hint

2018-06-25T14:59:37+00:00

A few fields are missing in a perf script -F hint. Add them.

Signed-off-by: Ravi Bangoria 
Cc: Alexander Shishkin 
Cc: Andi Kleen 
Cc: David Ahern 
Cc: David Carrillo-Cisneros 
Cc: Jin Yao 
Cc: Jiri Olsa 
Cc: Namhyung Kim 
Link: http://lkml.kernel.org/r/20180625124220.6434-2-ravi.bangoria@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf bench: Fix numa report output code

2018-06-25T14:59:37+00:00

Currently we can hit following assert when running numa bench:

  $ perf bench numa mem -p 3 -t 1 -P 512 -s 100 -zZ0cm --thp 1
  perf: bench/numa.c:1577: __bench_numa: Assertion `!(!(((wait_stat) & 0x7f) == 0))' failed.

The assertion is correct, because we hit the SIGFPE in following line:

  Thread 2.2 "thread 0/0" received signal SIGFPE, Arithmetic exception.
  [Switching to Thread 0x7fffd28c6700 (LWP 11750)]
  0x000.. in worker_thread (__tdata=0x7.. ) at bench/numa.c:1257
  1257 td->speed_gbs = bytes_done / (td->runtime_ns / NSEC_PER_SEC) / 1e9;

We don't check if the runtime is actually bigger than 1 second,
and thus this might end up with zero division within FPU.

Adding the check to prevent this.

Signed-off-by: Jiri Olsa 
Cc: Alexander Shishkin 
Cc: David Ahern 
Cc: Namhyung Kim 
Cc: Peter Zijlstra 
Link: http://lkml.kernel.org/r/20180620094036.17278-1-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo

perf stat: Remove duplicate event counting

2018-06-25T14:59:37+00:00

'perf stat' shows a mismatch in perf stat regarding counter names on
s390:

Run command:

   [root@s35lp76 perf]# ./perf stat -e tx_nc_tend  -v --
                ~/mytesttx 1 >/tmp/111
   tx_nc_tend: 1 573146 573146
   tx_nc_tend: 1 573146 573146

   Performance counter stats for '/root/mytesttx 1':

                 3      tx_nc_tend

       0.001037252 seconds time elapsed

   [root@s35lp76 perf]#

shows transaction counter tx_nc_tend with value 3 but it was triggered
only once as seen by the output of mytesttx.

When looking up the event name tx_nc_tend the following function
sequence is called:

parse_events_multi_pmu_add()
+--> perf_pmu__scan() being called with NULL argument
     +--> pmu_read_sysfs() scans directory ../devices/ for
                           all PMUs
          +--> perf_pmu__find() tries to find a PMU in the
                           global pmu list.
               +--> pmu_lookup() called to read all file
                                 entries when not in global
                                 list.

pmu_lookup() causes the issue. It calls
+---> pmu_aliases() to read all the entries in the PMU directory.
                    On s390 this is named
                    /sys/devices/cpum_cf/events.
      +--> pmu_aliases_parse() reads all files and creates an
                       alias for each file name.

                       So we end up with first entry created by
                       reading the sysfs file
                       [root@s35lp76 perf]# cat /sys/devices/cpum_cf
                                                /events/TX_NC_TEND
                       event=0x008d
                       [root@s35lp76 perf]#

                       Debug output shows this entry
                       tx_nc_tend -> 'cpum_cf'/'event=0x008d
                       '/
                       After all files in this directory have been
                       read and aliases created this function is called:
      +--> pmu_add_cpu_aliases()
                       This function looks up the CPU tables
                       created by the json files.
                       With json files for s390 now available all
                       the aliases are added to
                       the PMU alias list a second time.
                       The second entry is added by
                       reading the json file converted by jevent
                       resulting in file pmu-events/pmu-events.c:

                       {
                         .name = "tx_nc_tend",
                         .event = "event=0x8d",
                         .desc = "Unit: cpum_cf Completed TEND \
                                  instructions \
                                  in non-constrained TX mode",
                         .topic = "extended",
                         .long_desc = "A TEND instruction has \
                                       completed  in a \
                                       non-constrained \
                                       transactional-execution mode",
                         .pmu = "cpum_cf",
                        },

                        Debug output shows this entry
                        tx_nc_tend -> 'cpum_cf'/'event=0x8d'/

Function pmu_aliases_parse() and pmu_add_cpu_aliases() both use
__perf_pmu__new_alias() to add an alias to the PMU alias list. There is
no check if an alias already exist

So we end up with 2 entries for tx_nc_tend in the PMU alias list.

Having set up the PMU alias list for this PMU now
parse_events_multi_add_pmu() reads the complete alias list and adds each
alias with parse_events_add_pmu() to the global perfev_list.  This
causes the alias to be added multiple times to the event list.

Fix this by making __perf_pmu__new_alias() to merge alias definitions if
an alias is already on the alias list.  Also print a debug message when
the alias has mismatches in some fields.

Output before:

  [root@s35lp76 perf]# ./perf stat -e tx_nc_tend  -v \
                        -- ~/mytesttx 1 >/tmp/111
  tx_nc_tend: 1 551446 551446

   Performance counter stats for '/root/mytesttx 1':

                   3      tx_nc_tend

         0.000961134 seconds time elapsed

  [root@s35lp76 perf]#

Output after:

  [root@s35lp76 perf]#  ./perf stat -e tx_nc_tend  -v \
                        -- ~/mytesttx 1 >/tmp/111
  tx_nc_tend: 1 551446 551446

   Performance counter stats for '/root/mytesttx 1':

                   1      tx_nc_tend

         0.000961134 seconds time elapsed

  [root@s35lp76 perf]#

Signed-off-by: Thomas Richter 
Reviewed-by: Hendrik Brueckner 
Reviewed-by: Jiri Olsa 
Cc: Heiko Carstens 
Cc: Martin Schwidefsky 
Link: http://lkml.kernel.org/r/20180615101105.47047-3-tmricht@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf alias: Rebuild alias expression string to make it comparable

2018-06-25T14:59:37+00:00

PMU alias definitions in sysfs files may have spaces, newlines and
numbers with leading zeroes. Some alias definitions may also appear in
JSON files without spaces, etc.

Scan alias definitions and remove leading zeroes, spaces, newlines, etc
and rebuild string to make alias->str member comparable.

s390 for example  has terms specified as event=0x0091 (read from files
..//events/ and terms specified as event=0x91 (read from JSON
files).

Signed-off-by: Thomas Richter 
Reviewed-by: Hendrik Brueckner 
Cc: Heiko Carstens 
Cc: Jiri Olsa 
Cc: Martin Schwidefsky 
Link: http://lkml.kernel.org/r/20180615101105.47047-2-tmricht@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf alias: Remove trailing newline when reading sysfs files

2018-06-25T14:59:37+00:00

Remove a trailing newline when reading sysfs file contents such as
/sys/devices/cpum_cf/events/TX_NC_TEND.  This shows when verbose option
-v is used.

Output before:

  tx_nc_tend -> 'cpum_cf'/'event=0x008d
  '/

Output after:

  tx_nc_tend -> 'cpum_cf'/'event=0x8d'/

Signed-off-by: Thomas Richter 
Reviewed-by: Hendrik Brueckner 
Cc: Heiko Carstens 
Cc: Jiri Olsa 
Cc: Martin Schwidefsky 
Link: http://lkml.kernel.org/r/20180615101105.47047-1-tmricht@linux.ibm.com
Signed-off-by: Arnaldo Carvalho de Melo

perf tools: Fix a clang 7.0 compilation error

2018-06-25T14:59:37+00:00

Arnaldo reported the perf build failure with latest llvm/clang compiler
(7.0).

   $ make LIBCLANGLLVM=1 -C tools/perf/
   
    CC       /tmp/tmp.t53Qo38zci/tests/kmod-path.o
   util/c++/clang.cpp: In function ‘std::unique_ptr >
       perf::getBPFObjectFromModule(llvm::Module*)’:
   util/c++/clang.cpp:150:43: error: no matching function for call to
       ‘llvm::TargetMachine::addPassesToEmitFile(llvm::legacy::PassManager&,
        llvm::raw_svector_ostream&, llvm::TargetMachine::CodeGenFileType)’
               TargetMachine::CGFT_ObjectFile)) {
                                             ^
   In file included from util/c++/clang.cpp:25:0:
   /usr/local/include/llvm/Target/TargetMachine.h:254:16: note: candidate:
       virtual bool llvm::TargetMachine::addPassesToEmitFile(
       llvm::legacy::PassManagerBase&, llvm::raw_pwrite_stream&,
       llvm::raw_pwrite_stream*, llvm::TargetMachine::CodeGenFileType, bool,
       llvm::MachineModuleInfo*)
     virtual bool addPassesToEmitFile(PassManagerBase &, raw_pwrite_stream &,
                  ^~~~~~~~~~~~~~~~~~~
  /usr/local/include/llvm/Target/TargetMachine.h:254:16: note:
      candidate expects 6 arguments, 3 provided
  mv: cannot stat '/tmp/tmp.t53Qo38zci/util/c++/.clang.o.tmp': No such file or directory
  make[7]: *** [/home/acme/git/perf/tools/build/Makefile.build:101:
      /tmp/tmp.t53Qo38zci/util/c++/clang.o] Error 1
  make[6]: *** [/home/acme/git/perf/tools/build/Makefile.build:139: c++] Error 2
  make[5]: *** [/home/acme/git/perf/tools/build/Makefile.build:139: util] Error 2
  make[5]: *** Waiting for unfinished jobs....
    CC       /tmp/tmp.t53Qo38zci/tests/thread-map.o

The function addPassesToEmitFile signature changed in llvm 7.0 and such
a change caused the failure. This patch fixed the issue with using
proper function signatures under different compiler versions.

Reported-by: Arnaldo Carvalho de Melo 
Signed-off-by: Yonghong Song 
Tested-by: Arnaldo Carvalho de Melo 
Cc: Alexei Starovoitov 
Cc: Daniel Borkmann 
Cc: Jiri Olsa 
Cc: Martin KaFai Lau 
Cc: Wang Nan 
Link: http://lkml.kernel.org/r/20180616174739.1076733-1-yhs@fb.com
Signed-off-by: Arnaldo Carvalho de Melo

perf tools: Update x86's syscall_64.tbl, adding 'io_pgetevents' and 'rseq'

2018-06-25T14:59:36+00:00

This updates the tools/perf/ copy of the system call table for x86 which makes
'perf trace' become aware of the new 'io_pgetevents' and 'rseq' syscalls, no
matter in which system it gets built, i.e. older systems where the syscalls are
not available in the running kernel (via tracefs) or in the system headers will
still be aware of these syscalls/.

These are the csets introducing the source drift:

  05c17cedf85b ("x86: Wire up restartable sequence system call")
  7a074e96dee6 ("aio: implement io_pgetevents")

This results in this build time change:

  $ diff -u /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c.old /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c
  --- /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c.old	2018-06-15 11:48:17.648948094 -0300
  +++ /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c	2018-06-15 11:48:22.133942480 -0300
  @@ -332,5 +332,7 @@
          [330] = "pkey_alloc",
          [331] = "pkey_free",
          [332] = "statx",
  +       [333] = "io_pgetevents",
  +       [334] = "rseq",
   };
  -#define SYSCALLTBL_x86_64_MAX_ID 332
  +#define SYSCALLTBL_x86_64_MAX_ID 334
  $

This silences the following tools/perf/ build warning:

  Warning: Kernel ABI header at 'tools/perf/arch/x86/entry/syscalls/syscall_64.tbl' differs from latest version at 'arch/x86/entry/syscalls/syscall_64.tbl'

Cc: Adrian Hunter 
Cc: Christoph Hellwig 
Cc: David Ahern 
Cc: Jiri Olsa 
Cc: Mathieu Desnoyers 
Cc: Namhyung Kim 
Cc: Thomas Gleixner 
Cc: Wang Nan 
Link: https://lkml.kernel.org/n/tip-tfvyz51sabuzemrszbrhzxni@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo