pyvorbis crashes on openmoko

0 views
Skip to first unread message

vanInwagen

unread,
Feb 23, 2009, 9:39:07 AM2/23/09
to PyOgg
Hello dear group,
I just tried to compile pyvorbis1.15a (with applied
patch:https://bugzilla.redhat.com/attachment.cgi?id=192181) and
pyogg1.3
on my openmoko running kernel 2.6.24 and OM2008.12 as distribution.

I've got this installed:
Python 2.5.2
libvorbis - 1.0.1-r2 -
libvorbis-dev - 1.0.1-r2 -
libvorbisidec1 - 20041119-r1 -
libogg-dev - 1.1.3-r0 -
libogg0 - 1.1.3-r0 -
...

The following glibc-related-packages are installed, if it's important:

glibc-binary-localedata-de-de - 2.5-r8 -
glibc-binary-localedata-en-gb - 2.6.1-r12 -
glibc-binary-localedata-en-us - 2.6.1-r12 -
glibc-extra-nss - 2.6.1-r12 -
glibc-gconv - 2.6.1-r12 -
glibc-gconv-iso8859-1 - 2.6.1-r12 -

Somewhere around here(>><<):

def open_ogg_file(self):

saveThread.track = open(saveThread.testfile, 'wb')
print 'opening file'

saveThread.vd = ogg.vorbis.VorbisInfo(channels = 1,
rate = 8000,
quality = 0.2).analysis_init()

saveThread.os = ogg.OggStreamState(5)
>>>>>>
map(saveThread.os.packetin, saveThread.vd.headerout())
<<<<<<
saveThread.og = saveThread.os.flush()
print saveThread.og
while saveThread.og:
saveThread.og.writeout(saveThread.track)
saveThread.og = saveThread.os.flush()
print 'header written'


which is at all just a copy of the delivered examples!!!
I get this:

*** glibc detected *** python: free(): invalid pointer: 0x002284a8 ***
Segmentation fault

doing the same thing in python-shell i get this:

>>> >>> import ogg
>>> >>> import ogg.vorbis
>>> >>> vd =
ogg.vorbis.VorbisInfo(channels=1,rate=8000,quality=0.2).analysis_init
()
>>> >>> os = ogg.OggStreamState(5)
>>> >>> map(os.packetin,vd.headerout())
*** glibc detected *** python: double free or corruption (out):
0x403afbd8 ***
Aborted

What's wrong with my installation? Where to find the bug/error/fault?
Help would be great!
Thanx
Matthias Felsche


vanInwagen

unread,
Feb 24, 2009, 4:15:35 AM2/24/09
to PyOgg
After applying the same patches as for pyvorbis under python 2.5, as
described in other threads in this group, for pyogg1.3 I at least got
my little program working without
any segmentation fault or something.
The rest may be my fault.
Good, to have talked to somebody! ;-)
Reply all
Reply to author
Forward
0 new messages