Re: [PATCH V6 08/17] perf tools: Add Intel PT support

[Arnaldo Carvalho de Melo <acme@kernel.org>]

Em Thu, Jun 25, 2015 at 10:45:57AM -0300, Arnaldo Carvalho de Melo escreveu:
> Em Thu, Jun 25, 2015 at 04:37:01PM +0300, Adrian Hunter escreveu:
> > On 23/06/15 18:15, Arnaldo Carvalho de Melo wrote:
> > > Em Tue, Jun 23, 2015 at 09:29:34AM +0300, Adrian Hunter escreveu:
> > >> On 23/06/15 02:00, Arnaldo Carvalho de Melo wrote:
> > >>> [root@zoo ~]# rpm -q glibc-debuginfo
> > >>> glibc-debuginfo-2.20-8.fc21.x86_64
> > > 
> > >>> But then, it didn't even resolve the DSO, which it should, as I did manually :-/
> > > 
> > >>> Will continue investigating... Perhaps this is fixed in another patch? What I
> > >>> have test merged so far is at my tmp.perf/pt branch.
> > > 
> > >> I tried the same commands with perf tools from that branch (tmp.perf/pt) and
> > >> it seemed to work fine.
> > > 
> > >> One reason for not getting symbols is compiling perf tools without ELF support.
> > > 
> > > sure, but that is not the case here. But yeah, I'll try and triple check
> > > everything, next time will add the list of features detected to the
> > > problem report, so that you know which features were detected.
> > 
> > How is it going?
> 
> Yesterday was a holiday here, so no progress, will let you know when I
> make some.

So, now using the 4.1+ kernel, and the results seems to be the expected
ones for intel_bts//u:

[root@zoo ~]# perf record --per-thread -e intel_bts//u  ls
anaconda-ks.cfg  b  bin  lib64	libexec  new  old  perf.data
perf.data.old  stream_test  tg.run
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 1.274 MB perf.data ]
[root@zoo ~]# perf report
[root@zoo ~]# perf evlist
intel_bts//u
dummy:u
[root@zoo ~]# perf report --stdio
# To display the perf.data header info, please use --header/--header-only options.
#
# Total Lost Samples: 0
#
# Samples: 0  of event 'intel_bts//u'
# Event count (approx.): 0
#
# Overhead  Command  Shared Object  Symbol
# ........  .......  .............  ......

# Samples: 0  of event 'dummy:u'
# Event count (approx.): 0
#
# Overhead  Command  Shared Object  Symbol
# ........  .......  .............  ......

# Samples: 55K of event 'branches:u'
# Event count (approx.): 55012
#
# Overhead  Command  Shared Object       Symbol                                
# ........  .......  ..................  ......................................
#
    15.73%  ls       ld-2.20.so          [.] strcmp                            
    15.63%  ls       libc-2.20.so        [.] _dl_addr                          
    10.08%  ls       ld-2.20.so          [.] do_lookup_x                       
     7.00%  ls       ld-2.20.so          [.] _dl_name_match_p                  
     6.47%  ls       ld-2.20.so          [.] _dl_lookup_symbol_x               
     4.96%  ls       ld-2.20.so          [.] _dl_relocate_object               
     2.97%  ls       ls                  [.] quotearg_buffer_restyled          
     2.79%  ls       libc-2.20.so        [.] getenv                            
     1.95%  ls       ld-2.20.so          [.] _dl_cache_libcmp                  
     1.76%  ls       ld-2.20.so          [.] check_match.isra.0                
     1.64%  ls       libc-2.20.so        [.] __memmove_sse2                    
     1.47%  ls       ld-2.20.so          [.] _dl_map_object_deps               
     1.27%  ls       ld-2.20.so          [.] _dl_map_object_from_fd            
     1.17%  ls       ls                  [.] quote_name                        

---------------------------------------------------------------------

Will do the same tests with intel_pt as well, on a remote machine, add examples
to the changeset logs and everything going well, aim for pushing for Ingo soon,

Thanks,

- Arnaldo