[6965] in Athena Bugs
decmips 7.2P: emacs
daemon@ATHENA.MIT.EDU (daemon@ATHENA.MIT.EDU)
Sun Jan 27 20:37:57 1991
To: bugs@ATHENA.MIT.EDU
Date: Sun, 27 Jan 91 20:37:32 EST
From: Raul Acevedo <acevedo@ATHENA.MIT.EDU>
System name: tailgunner
Type and version: KN01 7.2P
Display type: PM-MONO
What were you trying to do?
---------------------------
Run some processes inside of emacs with `start-process'.
What's wrong:
-------------
Emacs has a tendency to, under some circumstances, to hang in disk wait
for a couple minutes, on the decmips only. It will invariably do this if
you try to startup a process, and the program doesn't exist:
(start-process "foo" nil "/this-doesnt-exist")
The other circumstance under which it occurs is not as reliable; that
is, it doesn't always happen. I *think* it happening has something to do
with either fprintf or output to stderr being used in the program
called, but I'm not sure. At any rate, the following program:
#include<stdio.h>
main()
{
fprintf(stderr, "Hi there, Raul!\n");
}
started in emacs with:
(start-process "foo" (current-buffer) "a.out")
will 50% time work, 50% will hang emacs in disk wait:
(start-process "foo" (current-buffer) "/mit/acevedo/a.out")
#<process foo>
Hi there, Raul!
Process foo finished
(start-process "foo" (current-buffer) "/mit/acevedo/a.out")
#<process foo>
Process foo finished
Above, the first one worked, and the second one hung.
Another reliable way to get this, for those who use my eolcr emacs mode,
is to simply run klook. Half the time you use it, this will happen.
(This is how I first took notice of this.)
What should have happened:
--------------------------
If the executable for `start-process' is not found, I should get a
`program not found' error right away. And whatever the problem is in the
other case, emacs shouldn't hang in disk wait.
Please describe any relevant documentation references:
------------------------------------------------------
man emacs
