I don't know whether this is the same bug or not, but I am seeing some
crashing issues with xemacs21-nomule.
It's actually taking my whole machine (? XSession? I can't even ssh in!)
down, not just throwing a segfault.
I compiled from the ubuntu source, and noticed a bunch of warnings about
integer and pointer size mismatches. This reminds me of compiling xemacs
from scratch on the old DEC Alpha machines, and I am immediately
suspicious about the current 64-bit cleanliness of the codebase.
The XEmacs people ("upstream") *claim* to have 64-bit clean code, but
that assertion is from several years ago, and there has been much added
to the codebase since then (with a relative sparsity of 64 bit machines
for testing since the Alphas faded away into obscurity).
It's been a long time since I have hacked on XEmacs code so I don't
really know where to start looking, but I would personally be looking
for something along the lines of architecture dependent definitions of
various int and void* types, and then pay particular attention to the
compiler warnings to see if the plain old (int) or (void*) casts have
architecture-dependent macro substitutions defined for them...
--
Xemacs21-mule crashes on 8.04 amd64
https://bugs.launchpad.net/bugs/224896
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
