[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: lynx-dev LYMainLoop.c -- gcc signal 11, egcs-1.1.2
From: |
brian j pardy |
Subject: |
Re: lynx-dev LYMainLoop.c -- gcc signal 11, egcs-1.1.2 |
Date: |
Sun, 17 Oct 1999 22:03:15 -0400 |
On Setting Orange, the 71st of Bureaucracy, 3165, Klaus Weide wrote:
> On Sun, 17 Oct 1999, brian j pardy wrote:
> > On Sun, Oct 17, 1999, T.E.Dickey wrote:
> > [...]
> > > > Any suggestions on which functions to start with? Probably the main
> > > > loop. Also, I just noticed in what I have up there -- it says "cpp"
> > >
> > > yes - or the functions "handle_LYK_xxx" that I split-out of the main
> > > loop. That would be possibly better (since chopping out the main
> > > loop would leave a lot of orphaned private functions, while chopping
> > > out the functions themselves would just leave a lot of implicit
> > > declarations).
> >
> > I've narrowed it down to within handle_LYK_digit().
>
> Hmm. Is the GCC_UNUSED in
>
> BOOLEAN *, try_internal GCC_UNUSED)
>
> maybe confusing the compiler? The variable is unused in the case
> where you have no problem, but is used in the case where you do
> do have a problem.
>
> I don't know whether the various gccs take that only as a directive
> to suppress some warnings or as more serious information.
Eliminating the GCC_UNUSED seems to make no difference (BTW Leonid, I
tried cold rebooting with a 15min downtime in between, it didn't
change anything). I'm still playing with it, but I'm about to write
it off as egcs bogosity and leave it at that.
> > It looks like the
> > code in the #ifndef DONT_TRACK_INTERNAL_LINKS (nasty double negatives
> > again) is causing it.
>
> Yes, I have done it too... It should probably be changed to a
> positive form ('TRACK_INTERNAL_LINKS') if you (and others) find
> that would help.
A quick check of src/ (I'm assuming it's nowhere in WWW/) shows 25
uses in the negative form (#ifndef), and only one #ifdef. Might be a
good idea.
> It seems it is clear that this is an error in the compiler, not
> in the code itself (it may only 'trigger' that error) - right?
That's what I would say. It could be anything, especially since Tom
seems to be having no problems with the same version. I don't think
it's Lynx.
--
Memory fault -- brain fried