Please try r8473 on bleeding_edge. It should fix the issue.

Thanks,
Vitaly

On Jun 29, 6:27 am, tatwai chong <[email protected]> wrote:
> Thanks for your help!
> the following is the output after applied the patch:
> $ ../tools/ll_prof.py --disasm-top=10
> V8 log: v8.log, v8.log.ll (no snapshot)
> Perf trace file: perf.data
> Generated code architecture: ia32
>
> /lib/modules/2.6.38-8-generic/kernel/sound/pci/hda/snd-hda-
> intel.koltek 2f 6c 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38
> 2d 38 2d 67 65 6e 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64
> 2f 70 63 69 2f 68 64 61 2f 73 6e 64 2d 68 64 61 2d 69 6e 74 65 6c 2e
> 6b 6f 00 6c 74 65 6b
> /lib/modules/2.6.38-8-generic/kernel/sound/pci/hda/snd-hda-
> codec.koltek 2f 6c 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38
> 2d 38 2d 67 65 6e 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64
> 2f 70 63 69 2f 68 64 61 2f 73 6e 64 2d 68 64 61 2d 63 6f 64 65 63 2e
> 6b 6f 00 6c 74 65 6b
> /lib/modules/2.6.38-8-generic/kernel/sound/core/snd-hwdep.kodec 2f 6c
> 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e
> 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64 2f 63 6f 72 65 2f
> 73 6e 64 2d 68 77 64 65 70 2e 6b 6f 00 64 65 63
> /lib/modules/2.6.38-8-generic/kernel/sound/core/snd-pcm.koodec 2f 6c
> 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e
> 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64 2f 63 6f 72 65 2f
> 73 6e 64 2d 70 63 6d 2e 6b 6f 00 6f 00 64 65 63
> /lib/modules/2.6.38-8-generic/kernel/sound/core/snd-rawmidi.koi 2f 6c
> 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e
> 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64 2f 63 6f 72 65 2f
> 73 6e 64 2d 72 61 77 6d 69 64 69 2e 6b 6f 00 69
> /lib/modules/2.6.38-8-generic/kernel/sound/core/seq/snd-seq.koi 2f 6c
> 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e
> 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64 2f 63 6f 72 65 2f
> 73 65 71 2f 73 6e 64 2d 73 65 71 2e 6b 6f 00 69
> /lib/modules/2.6.38-8-generic/kernel/sound/core/snd-timer.kohel 2f 6c
> 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e
> 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64 2f 63 6f 72 65 2f
> 73 6e 64 2d 74 69 6d 65 72 2e 6b 6f 00 68 65 6c
> /lib/modules/2.6.38-8-generic/kernel/drivers/gpu/drm/drm.kodevi 2f 6c
> 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e
> 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 64 72 69 76 65 72 73 2f 67 70 75
> 2f 64 72 6d 2f 64 72 6d 2e 6b 6f 00 64 65 76 69
> /lib/modules/2.6.38-8-generic/kernel/sound/core/snd.kom 2f 6c 69 62 2f
> 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e 65 72 69
> 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64 2f 63 6f 72 65 2f 73 6e 64
> 2e 6b 6f 00 6d
> /lib/modules/2.6.38-8-generic/kernel/sound/core/snd-page-alloc.ko.ko
> 2f 6c 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67
> 65 6e 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 73 6f 75 6e 64 2f 63 6f 72
> 65 2f 73 6e 64 2d 70 61 67 65 2d 61 6c 6c 6f 63 2e 6b 6f 00 2e 6b 6f
> /lib/modules/2.6.38-8-generic/kernel/drivers/hwmon/asus_atk0110.koko
> 2f 6c 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67
> 65 6e 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 64 72 69 76 65 72 73 2f 68
> 77 6d 6f 6e 2f 61 73 75 73 5f 61 74 6b 30 31 31 30 2e 6b 6f 00 6b 6f
> /lib/modules/2.6.38-8-generic/kernel/drivers/char/ppdev.ko0110. 2f 6c
> 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e
> 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 64 72 69 76 65 72 73 2f 63 68 61
> 72 2f 70 70 64 65 76 2e 6b 6f 00 30 31 31 30 2e
> /lib/modules/2.6.38-8-generic/kernel/drivers/acpi/video.kot.ko 2f 6c
> 69 62 2f 6d 6f 64 75 6c 65 73 2f 32 2e 36 2e 33 38 2d 38 2d 67 65 6e
> 65 72 69 63 2f 6b 65 72 6e 65 6c 2f 64 72 69 76 65 72 73 2f 61 63 70
> 69 2f 76 69 64 65 6f 2e 6b 6f 00 74 2e 6b 6f
> [vdso]8 5b 76 64 73 6f 5d 00 38
> Traceback (most recent call last):
>   File "../tools/ll_prof.py", line 904, in <module>
>     library_repo.Load(mmap_info, code_map, options)
>   File "../tools/ll_prof.py", line 697, in Load
>     if not os.path.exists(mmap_info.filename):
>   File "/usr/lib/python2.7/genericpath.py", line 18, in exists
>     os.stat(path)
> TypeError: must be encoded string without NULL bytes, not str
>
> Tat-Wai
>
> On 6月28日, 下午11時03分, Vitaly Repeshko <[email protected]> wrote:
>
>
>
>
>
>
>
> > Could you please apply the patch below and provide its output to help
> > me debug:
>
> > --- a/tools/ll_prof.py
> > +++ b/tools/ll_prof.py
> > @@ -634,6 +634,8 @@ class TraceReader(object):
> >      # Read null-padded filename.
> >      filename = self.trace[offset + self.header_size +
> > ctypes.sizeof(mmap_info):
> >                            offset + header.size].rstrip(chr(0))
> > +    if chr(0) in filename:
> > +      print filename, " ".join("%02x" % ord(x) for x in filename)
> >      mmap_info.filename = filename
> >      return mmap_info
>
> > On Jun 27, 6:45 am, tatwai chong <[email protected]> wrote:
>
> > > 1. Profile Richards of v8 benchmark:
> > > $ perf record -R -e cycles -c 10000 -f ../shell_g run.js --ll-prof
> > > Richards: 8296
> > > ----
> > > Score (version 6): 8296
> > > [ perf record: Woken up 46 times to write data ]
> > > [ perf record: Captured and wrote 11.464 MB perf.data (~500863
> > > samples) ]
>
> > > 2. Use ll_prof.py to process collected data
> > > $ ../tools/ll_prof.py --disasm-top=10
>
> > > but ...
> > > V8 log: v8.log, v8.log.ll (no snapshot)
> > > Perf trace file: perf.data
> > > Generated code architecture: ia32
>
> > > Traceback (most recent call last):
> > >   File "../tools/ll_prof.py", line 902, in <module>
> > >     library_repo.Load(mmap_info, code_map, options)
> > >   File "../tools/ll_prof.py", line 695, in Load
> > >     if not os.path.exists(mmap_info.filename):
> > >   File "/usr/lib/python2.7/genericpath.py", line 18, in exists
> > >     os.stat(path)
> > > TypeError: must be encoded string without NULL bytes, not str
>
> > > Does anyone know what problem?
> > > Thanks.
>
> > > Regards,
> > > Tat-Wai

-- 
v8-users mailing list
[email protected]
http://groups.google.com/group/v8-users

Reply via email to