Re: COUNTER-[PATCH] 64 bit formatting directives on MingW32

[Danny Smith <danny_r_smith_2001 at yahoo dot co dot nz>]

 --- Earl Chew <earl_chew@agilent.com> wrote: > Danny Smith wrote:
> > This patch will break bootstrap because of format warnings. The
> > HOST_WIDE_INT_PRINT changes are wrong (ming32 WIDE_INT is simply long).
> > Also this ugliness should be hidden away from the rest of the world.
> > 
> > Here is a counter-patch, which has been in my local sandbox for awhile without
> > causing problems. Bootstrap with todays CVS in progress.
> 
> I'll try your patch shortly, but I think I'll have a problem.
> 
> I encountered my original problem with building powerpc-wrs-vxworks
> hosted on MingW32.
> 
> When running the resulting cross-compiler, the backend wanted
> to use HOST_WIDE_INT_* to print a 64 bit quantity. I don't have
> the exact details at hand, but the comments on hwint.h suggest:
> 
> > /* Set HOST_WIDE_INT.  This should be the widest efficient host
> >    integer type.  It can be 32 or 64 bits, except that if we are
> >    targeting a machine with 64-bit size_t then it has to be 64 bits.
> 
> So it would appear that it might be possible that HOST_WIDE_INT_*
> needs to format a 64 bit quantity.
> 
> If this is indeed true, I think your patch needs some additions
> to accommodate this case.
> 

Yes it does.  There was a reason I hadn't submitted earlier, and
 you've just reminded me of it.

Danny

> Earl
>  

http://mobile.yahoo.com.au - Yahoo! Mobile
- Check & compose your email via SMS on your Telstra or Vodafone mobile.