It would be interesting to see the output from a run
that uses
libusb-1.0.9 or better yet the code from libusb.git
Should I apt-get remove the libusb (libusbx) before I try this or
leave it installed?
Hi, Which version of Debian are you using?
I'm using debian 7 but upgraded. Looking at /etc/debian_version it
reports "jessie/sid".
ljones
On 8/5/13, Peter Stuge <peter(a)stuge.se> wrote:
lee jones wrote:
[ 66.714466] rtl_tcp[3670]: segfault at 0 ip
00007f5d329389f2 sp
00007fff71c8e780 error 6 in libusb-1.0.so.0.1.0[7f5d32930000+15000]
[ 87.168292] rtl_tcp[3812]: segfault at 0 ip 00007f0d2fc609f2 sp
00007f0d2e26cb50 error 6 in libusb-1.0.so.0.1.0[7f0d2fc58000+15000]
[ 398.814077] rtl_tcp[8760]: segfault at 0 ip 00007f623a1a09f2 sp
00007f62384acb40 error 6 in libusb-1.0.so.0.1.0[7f623a198000+15000]
This shows a crash inside the library. That is caused either by a bug
in the library or by a bug in the application.
Maybe this is a case of something not liking
libusb at all? BTW here's
some info on the version of libusb I have installed on my system:
# aptitude show libusb-1.0-0
Package: libusb-1.0-0
State: installed
Automatically installed: no
Multi-Arch: same
Version: 2:1.0.16-3
Priority: optional
Section: libs
Actually this is not libusb at all. You have libusbx installed, it
was changed to be the default in debian and several other
distributions immediately after a Red Hat employee changed
the default in Fedora and wrote a one-line blog post about it.
The libusbx community doesn't value code quality highly so it's
about equally likely that the problem is in libusbx as that it is in
the application calling libusbx.
It would be interesting to see the output from a run that uses
libusb-1.0.9 or better yet the code from libusb.git, but that
requires building it from source and running with the special-built
library. You could look at
https://libusb.org/wiki/debug for
copypaste instructions.
//Peter
--
:)
SP