On Mon, 7 Jun 2004, David M. Cooke wrote:
> On Mon, Jun 07, 2004 at 06:55:09PM +0000, Faheem Mitha wrote: > > Dear People, > > > > See http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=252517 > > > > I'd like to ask for help in debugging this. Firstly, can anyone > > reproduce this on sarge? > > > > When I run inside gdb I don't get any information about the segfault, > > presumably because I don't have python2.3 compiled with debugging support > > > > I tried to compile python2.3 with debugging support. I made what I > > thought were appropriate changes in rules (see patch below). > > > > However, when I try to do *anything* with the resulting executable, I > > get errors, eg. something about cannot import site (whatever that > > means), and I cannot load the numarray shared libraries. > > With a debug build of python, *all* the C extensions you use must be > recompiled. The debug build changes the layout of PyObject, so non-debug > C extensions access the fields of that wrong. > > [This of course makes a debug build pretty hard to use...] Thanks for the reply. Firstly, can you or someone else confirm that the problem is indeed reproducible? John Hasler wrote that he could not reproduce the problem on unstable, but I just upgraded a test installation to unstable, and I was still able to reproduce it. It is only a few minutes work. Thanks in advance. Faheem.
