Re: abort_on_gc in Fgarbage_collect

[Gerd Moellmann]

justus schwartz <address@hidden> writes:

> sure, but if i use this patch "old", which is needed in line 798
> (set_buffer_internal (old);) and is defined in PRINTDECLARE isn't available
> anymore. 
> 
> from the comment to PRINTDECLARE i think, it should be used at the beginning 
> of
> a function?

At the beginning of a block, actually.

The problem with what's in CVS is that PRINTFINISH unbinds to a place
set in PRINTDECLARE, which means that inhibit_modification_hooks is
unbound too after PRINTFINISH.

Alas, I can't work on this further at present.  Can you perhaps fix
the problem with the clue above?

Or, maybe the change to use abort_on_gc should be reverted entirely,
and the GCPROs restored; see my other mail.  I can't guess because I
don't know what the new code is trying to fix.