[RFA:] Take 2: CPP: Correct (add) condition for calling debug_end_source_file

[Hans-Peter Nilsson]

> From: Andreas Jaeger <aj@suse.de>
> Date: 12 Nov 2000 22:13:46 +0100

> >>>>> Andreas Jaeger writes:
> 
> >>>>> Neil Booth writes:
> >> Would someone approve this, please?  It looks good to me, and I think
> >> it fixes Andreas' -g -include bug.
> AJ> Did you try it?  I tried it this morning with the current CVS tree and
> AJ> couldn't bootstrap at all :-(.  But this might have been a different
> AJ> fix.  I'll try to test the patch tomorrow,
> 
> I tested the patch again:
> - it bootstraps now (so the problem I had this morning was elsewhere)

Let me just point out again that I did bootstrap and ran the
regression suite with and without this patch.  Right, as usual
that's an indication rather than a guarantee of innocence.  See
the original message.  No, I don't have the URL handy.

> - but -g -include still doesn't work :-(

I can confirm that the patch doesn't fix *that* bug.

brgds, H-P