unofficial mirror of guile-devel@gnu.org 
 help / color / mirror / Atom feed
* gcc 4 stack overflow in head
@ 2005-07-12  1:33 Kevin Ryde
  2005-08-10 22:37 ` Marius Vollmer
  0 siblings, 1 reply; 3+ messages in thread
From: Kevin Ryde @ 2005-07-12  1:33 UTC (permalink / raw)


The head built with debian i386 gcc 4 bombs with a stack overflow on
startup.  Under gdb it looks like it really has used up its 80kbytes,
but I can't tell where or why.  The same built with gcc 3.3 is ok, so
something evil has happened.


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: gcc 4 stack overflow in head
  2005-07-12  1:33 gcc 4 stack overflow in head Kevin Ryde
@ 2005-08-10 22:37 ` Marius Vollmer
  2005-08-12  1:00   ` Kevin Ryde
  0 siblings, 1 reply; 3+ messages in thread
From: Marius Vollmer @ 2005-08-10 22:37 UTC (permalink / raw)


Kevin Ryde <user42@zip.com.au> writes:

> The head built with debian i386 gcc 4 bombs with a stack overflow on
> startup.  Under gdb it looks like it really has used up its 80kbytes,
> but I can't tell where or why.  The same built with gcc 3.3 is ok, so
> something evil has happened.

I have seen this intermittently as well (for example, compiling
without optimization would give a stack overflow when launching into
the repl, but compiling with optimization would not).  Maybe the
default stack limit is dangerously close to what we need to survive...

-- 
GPG: D5D4E405 - 2F9B BCCC 8527 692A 04E3  331E FAF8 226A D5D4 E405


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: gcc 4 stack overflow in head
  2005-08-10 22:37 ` Marius Vollmer
@ 2005-08-12  1:00   ` Kevin Ryde
  0 siblings, 0 replies; 3+ messages in thread
From: Kevin Ryde @ 2005-08-12  1:00 UTC (permalink / raw)
  Cc: guile-devel

Marius Vollmer <mvo@zagadka.de> writes:
>
> I have seen this intermittently as well (for example, compiling
> without optimization would give a stack overflow when launching into
> the repl, but compiling with optimization would not).  Maybe the
> default stack limit is dangerously close to what we need to survive...

Looks like gcc 4 unoptimized uses a lot more stack in ceval than
previous versions, in the head 976 bytes vs 284 with gcc 3.3.  Or in
1.6 similar 688 up from 288.  Maybe every variable gets its own stack
slot, even when they're in separate blocks.  I guess it's a case of
don't do that (unoptimize), or increase the stack limit if you do.


_______________________________________________
Guile-devel mailing list
Guile-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/guile-devel


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2005-08-12  1:00 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2005-07-12  1:33 gcc 4 stack overflow in head Kevin Ryde
2005-08-10 22:37 ` Marius Vollmer
2005-08-12  1:00   ` Kevin Ryde

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).