Re: -m68020 -fnobitfields != -mcpu32 ?

• Date: Tue, 15 Oct 1996 19:30:38 -0700
• From:
  heard@vvnet.com (C. M. Heard/VVNET, Inc.)
• Reply-To:
  heard@vvnet.com (C. M. Heard/VVNET, Inc.)
• Cc: crossgcc@cygnus.com
• In-Reply-To: <Pine.LNX.3.91.961016083201.426A-100000@moshpit.air.net.au>

> I actually use the -Wa,-m68360 linker command to check for unsupported
> instructions.  A couple of bitfield instructions are generated when I 
> recompile newlib patched for rtems.  I guess I'll just have to upgrade to 
> gcc 2.7.2.1.

Before you do so, please read the attached note 
which announced the release of gcc 2.7.2.1.

/cmh

--------------------------------------------------------------------------

gcc-2.7.2.1.tar.gz, 7,132,291 bytes and gcc-2.7.2-2.7.2.1.diff.gz, 18,029
bytes, are now available on prep.ai.mit.edu and on the mirror sites below.

This package contains compilers for the C, C++ and Objective-C
programming languages.

This version of GCC fixes some major bugs in GCC 2.7.2 that prevent it from
compiling some common packages on major targets.  Most notably, it disables
part of a loop optimization that can generate incorrect code under obscure
circumstances that happen to occur in Linux and fixes problems compiling
Ghostscript on some machines (such as Alpha).

Most bugs reported in GCC 2.7.2, which are less severe, have not been fixed
in 2.7.2.1, but will be fixed in the next major release of GCC, 2.8.0.

You should not switch to this version if you are not having any problems
with GCC 2.7.2.  Though this new version was tested by a large number of
GCC pre-testers and the changes between it and GCC 2.7.2 are small and
believed safe, it is nevertheless the case that GCC 2.7.2 has been much
more extensively tested than 2.7.2.1.

As usual, report bugs in the C++ portion of GCC to bug-g++@prep.ai.mit.edu
and the rest of the compiler to bug-gcc@prep.ai.mit.edu.

• Re: -m68020 -fnobitfields != -mcpu32 ?
• From: Timothy Potter <timbo@moshpit.air.net.au>

• Prev: Re: -m68020 -fnobitfields != -mcpu32 ?
• Next: Re: -m68020 -fnobitfields != -mcpu32 ?
• Index: Cross gcc Mailing list
• Thread: Cross gcc Mailing list