> What is the recommended procedure to regenerate them?
Not sure there is one.
> Shouldn't they be regenerated and committed in CVS?
No, because that changes the base requirements for all those packages.
DJ Delorie wrote:
I am on Fedora 7 with autoconf 2.61 with a checkout from
yesterday off the trunk. So I shouldn't have see it based
upon that requirement. What else could it be?
Did you re-generate all the configure's from all the configure.ac's?
The ones in CVS are all built with 2.59.
> I am on Fedora 7 with autoconf 2.61 with a checkout from
> yesterday off the trunk. So I shouldn't have see it based
> upon that requirement. What else could it be?
Did you re-generate all the configure's from all the configure.ac's?
The ones in CVS are all built with 2.59.
DJ Delorie wrote:
http://sourceware.org/ml/newlib/2006/msg00472.html
Shouldn't this patch already be in the top level
gcc/Makefile.in?
The right fix is to use autoconf 2.60 or later.
I am on Fedora 7 with autoconf 2.61 with a checkout from
yesterday off the trunk. So I shouldn't hav
> http://sourceware.org/ml/newlib/2006/msg00472.html
>
> Shouldn't this patch already be in the top level
> gcc/Makefile.in?
The right fix is to use autoconf 2.60 or later.
The patch you link to requires GNU make, and thus was rejected.
Hi,
I switched my testing from gcc 4.2.x to the svn
trunk so I could submit things. I ran into the
problem reported and fixed here:
http://sourceware.org/ml/newlib/2006/msg00472.html
Shouldn't this patch already be in the top level
gcc/Makefile.in?
--joel
