linux-stable.git/tools/perf/arch, branch linux-5.6.y

tools: Fix off-by 1 relative directory includes

2020-03-06T11:36:46+00:00

This is currently working due to extra include paths in the build.

Committer testing:

  $ cd tools/include/uapi/asm/

Before this patch:

  $ ls -la ../../arch/x86/include/uapi/asm/errno.h
  ls: cannot access '../../arch/x86/include/uapi/asm/errno.h': No such file or directory
  $

After this patch;

  $ ls -la ../../../arch/x86/include/uapi/asm/errno.h
  -rw-rw-r--. 1 acme acme 31 Feb 20 12:42 ../../../arch/x86/include/uapi/asm/errno.h
  $

Check that that is still under tools/, i.e. hasn't escaped into the main
kernel sources:

  $ cd ../../../arch/x86/include/uapi/asm/
  $ pwd
  /home/acme/git/perf/tools/arch/x86/include/uapi/asm
  $

Signed-off-by: Ian Rogers 
Tested-by: Arnaldo Carvalho de Melo 
Acked-by: Jiri Olsa 
Cc: Adrian Hunter 
Cc: Alexander Shishkin 
Cc: Alexios Zavras 
Cc: Andi Kleen 
Cc: Greg Kroah-Hartman 
Cc: Igor Lubashev 
Cc: Kan Liang 
Cc: Mark Rutland 
Cc: Mathieu Poirier 
Cc: Namhyung Kim 
Cc: Nick Desaulniers 
Cc: Peter Zijlstra 
Cc: Stephane Eranian 
Cc: Thomas Gleixner 
Cc: Wei Li 
Link: http://lore.kernel.org/lkml/20200306071110.130202-2-irogers@google.com
Signed-off-by: Arnaldo Carvalho de Melo

perf arch powerpc: Sync powerpc syscall.tbl with the kernel sources

2020-02-18T16:36:57+00:00

Copy over powerpc syscall.tbl to grab changes from the below commits

  fddb5d430ad9 ("open: introduce openat2(2) syscall")
  9a2cef09c801 ("arch: wire up pidfd_getfd syscall")

Now 'perf trace' on powerpc will be able to map from those syscall
strings to the right syscall numbers, i.e.

  perf trace -e pidfd*

Will include 'pidfd_getfd' as well as:

  perf trace open*

Will cover all 'open' variants.

Reported-by: Stephen Rothwell 
Reviewed-by: Ravi Bangoria 
Cc: Adrian Hunter 
Cc: Aleksa Sarai 
Cc: Al Viro 
Cc: Christian Brauner 
Cc: Jiri Olsa 
Cc: Namhyung Kim 
Cc: Naveen N. Rao 
Cc: Nicholas Piggin 
Cc: Sargun Dhillon 
Signed-off-by: Arnaldo Carvalho de Melo

perf auxtrace: Add auxtrace_record__read_finish()

2020-02-18T13:13:29+00:00

All ->read_finish() implementations are doing the same thing. Add a
helper function so that they can share the same implementation.

Signed-off-by: Adrian Hunter 
Reviewed-by: Leo Yan 
Tested-by: Leo Yan 
Reviewed-by: Mathieu Poirier 
Cc: Jiri Olsa 
Cc: Kim Phillips 
Cc: Wei Li 
Link: http://lore.kernel.org/lkml/20200217082300.6301-1-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo

perf arm-spe: Fix endless record after being terminated

2020-02-18T13:13:29+00:00

In __cmd_record(), when receiving SIGINT(ctrl + c), a 'done' flag will
be set and the event list will be disabled by evlist__disable() once.

While in auxtrace_record.read_finish(), the related events will be
enabled again, if they are continuous, the recording seems to be
endless.

If the event is disabled, don't enable it again here.

Based-on-patch-by: Wei Li 
Signed-off-by: Adrian Hunter 
Cc: Jiri Olsa 
Cc: Tan Xiaojun 
Cc: stable@vger.kernel.org # 5.4+
Link: http://lore.kernel.org/lkml/20200214132654.20395-5-adrian.hunter@intel.com
Signed-off-by: Arnaldo Carvalho de Melo

perf cs-etm: Fix endless record after being terminated

2020-02-18T13:13:29+00:00

In __cmd_record(), when receiving SIGINT(ctrl + c), a 'done' flag will
be set and the event list will be disabled by evlist__disable() once.

While in auxtrace_record.read_finish(), the related events will be
enabled again, if they are continuous, the recording seems to be
endless.

If the cs_etm event is disabled, we don't enable it again here.

Note: This patch is NOT tested since i don't have such a machine with
coresight feature, but the code seems buggy same as arm-spe and
intel-pt.

Tester notes:

Thanks for looping, Adrian.  Applied this patch and tested with
CoreSight on juno board, it works well.

Signed-off-by: Wei Li 
Reviewed-by: Leo Yan 
Reviewed-by: Mathieu Poirier 
Tested-by: Leo Yan 
Cc: Jiri Olsa 
Cc: Tan Xiaojun 
Cc: stable@vger.kernel.org # 5.4+
Link: http://lore.kernel.org/lkml/20200214132654.20395-4-adrian.hunter@intel.com
[ahunter: removed redundant 'else' after 'return']
Signed-off-by: Adrian Hunter 
Signed-off-by: Arnaldo Carvalho de Melo

perf intel-bts: Fix endless record after being terminated

2020-02-18T13:13:29+00:00

In __cmd_record(), when receiving SIGINT(ctrl + c), a 'done' flag will
be set and the event list will be disabled by evlist__disable() once.

While in auxtrace_record.read_finish(), the related events will be
enabled again, if they are continuous, the recording seems to be
endless.

If the intel_bts event is disabled, we don't enable it again here.

Note: This patch is NOT tested since i don't have such a machine with
intel_bts feature, but the code seems buggy same as arm-spe and
intel-pt.

Signed-off-by: Wei Li 
Cc: Jiri Olsa 
Cc: Tan Xiaojun 
Cc: stable@vger.kernel.org # 5.4+
Link: http://lore.kernel.org/lkml/20200214132654.20395-3-adrian.hunter@intel.com
[ahunter: removed redundant 'else' after 'return']
Signed-off-by: Adrian Hunter 
Signed-off-by: Arnaldo Carvalho de Melo

perf intel-pt: Fix endless record after being terminated

2020-02-18T13:13:29+00:00

In __cmd_record(), when receiving SIGINT(ctrl + c), a 'done' flag will
be set and the event list will be disabled by evlist__disable() once.

While in auxtrace_record.read_finish(), the related events will be
enabled again, if they are continuous, the recording seems to be endless.

If the intel_pt event is disabled, we don't enable it again here.

Before the patch:

  huawei@huawei-2288H-V5:~/linux-5.5-rc4/tools/perf$ ./perf record -e \
  intel_pt//u -p 46803
  ^C^C^C^C^C^C

After the patch:

  huawei@huawei-2288H-V5:~/linux-5.5-rc4/tools/perf$ ./perf record -e \
  intel_pt//u -p 48591
  ^C[ perf record: Woken up 0 times to write data ]
  Warning:
  AUX data lost 504 times out of 4816!

  [ perf record: Captured and wrote 2024.405 MB perf.data ]

Signed-off-by: Wei Li 
Cc: Jiri Olsa 
Cc: Tan Xiaojun 
Cc: stable@vger.kernel.org # 5.4+
Link: http://lore.kernel.org/lkml/20200214132654.20395-2-adrian.hunter@intel.com
[ ahunter: removed redundant 'else' after 'return' ]
Signed-off-by: Adrian Hunter 
Signed-off-by: Arnaldo Carvalho de Melo

perf tools: Add arm64 version of get_cpuid()

2020-02-12T13:36:51+00:00

Add an arm64 version of get_cpuid(), which is used for various annotation
and headers - for example, I now get the CPUID in "perf report --header",
as shown in this snippet:

  # hostname : ubuntu
  # os release : 5.5.0-rc1-dirty
  # perf version : 5.5.rc1.gbf8a13dc9851
  # arch : aarch64
  # nrcpus online : 96
  # nrcpus avail : 96
  # cpuid : 0x00000000480fd010

Since much of the code to read the MIDR is already in get_cpuid_str(),
factor out this code.

Tester notes:

I tested this patch on my new ARM64 Kunpeng 920 server.
[root@node1 zsk]# ./perf --version
perf version 5.6.rc1.g2cdb955b7252

Both perf list and perf stat can work.

Signed-off-by: John Garry 
Tested-by: Shaokun Zhang 
Cc: Alexander Shishkin 
Cc: Jiri Olsa 
Cc: Mark Rutland 
Cc: Namhyung Kim 
Cc: Peter Zijlstra 
Cc: Will Deacon 
Cc: linux-arm-kernel@lists.infradead.org
Cc: linuxarm@huawei.com
Link: http://lore.kernel.org/lkml/1576245255-210926-1-git-send-email-john.garry@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo

tools include UAPI: Sync x86's syscalls_64.tbl, generic unistd.h and fcntl.h to pick up openat2 and pidfd_getfd

2020-02-11T19:41:06+00:00

  fddb5d430ad9 ("open: introduce openat2(2) syscall")
  9a2cef09c801 ("arch: wire up pidfd_getfd syscall")

We also need to grab a copy of uapi/linux/openat2.h since it is now
needed by fcntl.h, add it to tools/perf/check_headers.h.

  $ diff -u tools/perf/arch/x86/entry/syscalls/syscall_64.tbl arch/x86/entry/syscalls/syscall_64.tbl
  --- tools/perf/arch/x86/entry/syscalls/syscall_64.tbl	2019-12-20 16:43:57.662429958 -0300
  +++ arch/x86/entry/syscalls/syscall_64.tbl	2020-02-10 16:36:22.070012468 -0300
  @@ -357,6 +357,8 @@
   433	common	fspick			__x64_sys_fspick
   434	common	pidfd_open		__x64_sys_pidfd_open
   435	common	clone3			__x64_sys_clone3/ptregs
  +437	common	openat2			__x64_sys_openat2
  +438	common	pidfd_getfd		__x64_sys_pidfd_getfd

   #
   # x32-specific system call numbers start at 512 to avoid cache impact
  $

Update tools/'s copy of that file:

  $ cp arch/x86/entry/syscalls/syscall_64.tbl tools/perf/arch/x86/entry/syscalls/syscall_64.tbl

See the result:

  $ diff -u /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c.before /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c
  --- /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c.before	2020-02-10 16:42:59.010636041 -0300
  +++ /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c	2020-02-10 16:43:24.149958337 -0300
  @@ -346,5 +346,7 @@
   	[433] = "fspick",
   	[434] = "pidfd_open",
   	[435] = "clone3",
  +	[437] = "openat2",
  +	[438] = "pidfd_getfd",
   };
  -#define SYSCALLTBL_x86_64_MAX_ID 435
  +#define SYSCALLTBL_x86_64_MAX_ID 438
  $

Now one can use:

  perf trace -e openat2,pidfd_getfd

To get just those syscalls or use in things like:

  perf trace -e open*

To get all the open variant (open, openat, openat2, etc) or:

  perf trace pidfd*

To get the pidfd syscalls.

Cc: Adrian Hunter 
Cc: Aleksa Sarai 
Cc: Al Viro 
Cc: Christian Brauner 
Cc: Jiri Olsa 
Cc: Namhyung Kim 
Cc: Sargun Dhillon 
Signed-off-by: Arnaldo Carvalho de Melo

perf parse: Refactor 'struct perf_evsel_config_term'

2020-01-30T10:55:02+00:00

The struct perf_evsel_config_term::val is a union which contains fields
'callgraph', 'drv_cfg' and 'branch' as string pointers.  This leads to
the complex code logic for handling every type's string separately, and
it's hard to release string as a general way.

This patch refactors the structure to add a common field 'str' in the
'val' union as string pointer and remove the other three fields
'callgraph', 'drv_cfg' and 'branch'.  Without passing field name, the
patch simplifies the string handling with macro ADD_CONFIG_TERM_STR()
for string pointer assignment.

This patch fixes multiple warnings of line over 80 characters detected
by checkpatch tool.

Signed-off-by: Leo Yan 
Reviewed-by: Andi Kleen 
Acked-by: Jiri Olsa 
Cc: Adrian Hunter 
Cc: Alexander Shishkin 
Cc: Ian Rogers 
Cc: Mark Rutland 
Cc: Mathieu Poirier 
Cc: Mike Leach 
Cc: Namhyung Kim 
Cc: Peter Zijlstra 
Cc: Suzuki Poulouse 
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lore.kernel.org/lkml/20200117055251.24058-1-leo.yan@linaro.org
Signed-off-by: Arnaldo Carvalho de Melo