>>>> "SY" == Steve Youngs
<youngs(a)xemacs.org> writes:
SY> Steve T, is it possible that this was caused by the
SY> `split-string' update? Can you comment?
Not the split-string stuff, no. Changes in Lisp can't cause crashes
(although they can trigger existing crashes). Furthermore, I don't
see how that could possibly cause a problem here since all that's
changed is an extra call to (delete "" ...) in the default case.
Possibly it's the change which clears stale match data (ie, extra
registers after success, which is already in 21.5.11 IIRC, and all
registers on failure, new in 21.5.12). We now clear the (match-* 0)
data, which AFAIK nothing ever did before. If something at the C
level depends on that making sense, even if the last match failed,
that could cause a crash. But that shouldn't be fixed by reverting
subr.el only.
More probable, but still unlikely, is continuing regexp breakage: a
really nasty regexp (one of the regexp crashes in Gnus required more
than 24MB of stack), or breakage in Ben's new regexp allocation code.
Most likely, since changing a Lisp file in a random way changes
things, as does a debug build, I'll put my money on the bad bytecode
crash Vin has partially analyzed. (Sorry, I don't have a good xref at
hand.)
--
Institute of Policy and Planning Sciences
http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.