[ clisp-Bugs-1714737 ] Bad restarts when .clisprc signals errors.

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

Bugs item #1714737, was opened at 2007-05-08 02:30
Message generated for change (Comment added) made by sds
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=101355&aid=1714737&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: Open
Resolution: None
Priority: 5
Private: No
Submitted By: szergling (szergling)
Assigned to: Bruno Haible (haible)
Summary: Bad restarts when .clisprc signals errors.

Initial Comment:

When .clisprc signals any error (the ones I've checked are:
non-matching parens, (error "Any error") forms, and load errors
[loading files]), choosing the ABORT restart causes a segfault. This
is repeatable. I have enclosed a few sample sessions (it shouldn't be
required).

This occurs on cygwin with clisp-2.41 (not prepackaged - compiled
myself), and with GNU/Linux clisp-2.41 as well.



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

>Comment By: Sam Steingold (sds)
Date: 2008-01-07 01:12

Message:
Logged In: YES 
user_id=5735
Originator: NO

this bug is similar to 1506316 in that it is exhibited only by the base
linkset, not the boot linkset.
(1506316 is exhibited only by the boot, not the base linkset)

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

Comment By: szergling (szergling)
Date: 2007-05-31 07:14

Message:
Logged In: YES 
user_id=1630354
Originator: YES

> My Ubuntu-Dapper clisp-2.38 offers a single restart, whereas your
samples
> show 3-5 restarts. How comes? (BTW, 2.38 crashes when I use that one
> restart).

You got me there. I don't know why. The one I showed came from my cygwin
test.

I can confirm that clisp encounters that same segfault behaviour on my
Ubuntu
(I believe it's the Breezy release), this time with 2 abort restarts. Some
tests
are enclosed (repl dump). I usually ignore the number of restarts offered
(only 
need 1 right), so I won't know what to do about them...

File Added: clisp-bug.txt

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

Comment By: Jörg Höhle (hoehle)
Date: 2007-05-23 05:40

Message:
Logged In: YES 
user_id=377168
Originator: NO

>"occurs ... with GNU/Linux clisp-2.41 as well"
Hmm, this does not crash for me on Linux (Ubuntu Dapper, April CVS
build).

BTW, mine says
lisp.run -M lispinit.mem
;; Loading file /home/hoehle/.clisprc ...
*** - READ: input stream
       #<INPUT BUFFERED FILE-STREAM CHARACTER #P"/home/hoehle/.clisprc"
@2>
      ends within an object. Last opening parenthesis probably in line 1.
Break 1 [1]> 

but it does not offer any restart. How comes you see many restarts?
>SKIP           :R1      skip (ERROR Will this error cause a segfault?)
>STOP           :R2      stop loading file /cygdrive/c/home/.clisprc
>ABORT          :R3      ABORT
>ABORT          :R4      ABORT
>ABORT          :R5      ABORT

My Ubuntu-Dapper clisp-2.38 offers a single restart, whereas your samples
show 3-5 restarts. How comes? (BTW, 2.38 crashes when I use that one
restart).
Might be related to bug #1220548 or bug #1448744 i.e. bogus stack
unwinding and error-handlers may not be fully setup within .clisprc?
E.g. ":r1" crashes, while "abort" does not (unwind_upto() is used
differently, for unknown reason).

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

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=101355&aid=1714737&group_id=1355