This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH, libgfortran] Remove AM_MAKEFLAGS.


> 	I'm fairly confident that the AM_MAKEFLAGS change was fixing the
> symptom and not the problem.
> 
> David

David,

I looked around some more and I think there is a bug but I think it is
in GNU make and I don't think this is an HP-UX or IA64 specific problem.

I tried to find out when the setting of AM_MAKEFLAGS in library builds
started and it looks like zlib/Makefile.am was the first, back in 1999
(version 1.5) and everyone since has copied that code.  The comment in
CVS/ChangeLog isn't very descriptive so I am still not sure exactly what
the bug is.

I think my change to set AM_MAKEFLAGS should stay because every other
library's Makefile.am that I see has same code in it (boehm-gc, fastjar,
libffi, libjava, libmudflap, libstdc++v3, zlib) and because I do not
believe a f95 cross compiler that includes libgfortran can be built on
any platform without this change.

You mentioned that the problem you are having is that "The combination
of the environment and the libgfortran object files list is larger than
AIX ARG_MAX."  Is there some other way to fix this?  Can we do something
with the libgfortran object files list?  Is it the object file list that
makes libgfortran different than the other librarires which also set
AM_MAKEFLAGS?

Steve Ellcey
sje@cup.hp.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]