[ clisp-Bugs-1310753 ] Problems with Gray streams causing stack overflow and segfau

[SourceForge.net <no...@so...>]

Bugs item #1310753, was opened at 2005-10-02 04:49
Message generated for change (Comment added) made by sds
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=101355&aid=1310753&group_id=1355

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: clisp
Group: segfault
>Status: Pending
>Resolution: Works For Me
Priority: 5
Submitted By: Marc Mertens (marcmertens)
Assigned to: Sam Steingold (sds)
Summary: Problems with Gray streams causing stack overflow and segfau

Initial Comment:
Hello, 

     I'm having some strange problems using Gray
streams on version 2.35 (I did not have these problems
on version 2.33, I did not tested it on version 2.34).
Basically I use Gray stream to interweave calling of a
function when the system waits on input. To simulate
the problem, start clisp and load the attached source
file (clisperror.lisp), call then (enable-listeners),
and then type in the following
(let ((som 0))
  (dotimes (tel 5 som)
           (setf som (+ 1 tel som))))
(The newlines are important, I don't have a problem
when there is only one new line in the expression). I
get then the following 


[3]> (let ((som 0))
(dotimes (tel 5 som)
(setf som (+ 1 te
*** - Lisp stack overflow. RESET
[4]>
*** - EVAL: variable DOTIMES has no value
The following restarts are available:
USE-VALUE      :R1      You may input a value to be
used instead of DOTIMES.
STORE-VALUE    :R2      You may input a new value for
DOTIMES.l
ABORT          :R3      ABORT
Break 1 [5]>          

In some cases I had a message about a undefined TEL and
then a segfault.

Is there something changed in Gray streams, so that the
assumption I made in 2.33 (where everything worked) is
not true anymore or is this a new bug.

PS. This problem is causing Jabberwocky my Lisp IDE to
fail in communication with a CLISP session.

Thanks in advance for any help

Marc Mertens


----------------------------------------------------------------------

>Comment By: Sam Steingold (sds)
Date: 2005-10-08 21:04

Message:
Logged In: YES 
user_id=5735

the attached code WFM.
please test and report back.

----------------------------------------------------------------------

Comment By: Sam Steingold (sds)
Date: 2005-10-08 20:25

Message:
Logged In: YES 
user_id=5735

what I see is usually related to unread-char errors.
indeed, you have no control over the original *standard-input*,
so you cannot just pass unread-char to it - 
it might have been read from between your read and unread.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=101355&aid=1310753&group_id=1355