On Tue, Apr 24, 2001 at 10:53:26AM -0400, Sam Steingold wrote:
> > ,----
> > | Compiling file /usr/local/src/clisp/with-gcc-wall/bindings/linuxlibc6/linux.lisp ...
> > | *** - COMMON-LISP:EVAL: the function delete-package is undefined
> > `----
> I fixed and tested this.
Confirmed here. Thanks.
> > and the hang/hog happened. It tool a "killall lisp.run" to get my
> > prompt back.
> I do not observe this, either before or after my fix.
The fix you mention above? It would seem unlikely that it would have
anything to do with it. As I think I've already demonstrated the hang
problem seems to be when an `error' is thrown and clisp has been called as a
> I would appreciate it if could create a reproducible test case, i.e., a
> stand-alone file foo.lisp such that
> $ clisp -norc -c foo.lisp
See <20010423171551.S28537@...> for one example of a
reproducible test case. If it's compilation you're interested in then this:
,----[ foo.lisp ]
is the smallest possible lisp file that causes the hang when compiled.
IOW, what I'm observing isn't a compiler problem, it's a problem with clisp
hanging when an `error' is thrown.
Note that 2.25.1 doesn't demonstrate this version, clisp built from current
As I asked in <20010423212446.C28537@...>, are there some
obvious environmental differences I should be looking for? Failing that, is
there some useful way I can check what clisp is doing at the point that it
hangs during an error?