m68k-rtems-binutils-2.11-3.i386.rpm

[David J. Fiddes]

Hi,

>A few days ago I posted of a possible optimization bug for the m68332
>target in m68k-rtems-gcc-gcc2.95.2newlib1.9.0-2.i386.rpm. After further
>testing, it now appears the problem resides in binutils-2.11-3.
>
>binutils-2.11-3 is erroneously compiling m68k code, and the problem
>seems to be aggravated by compiling without optimization. A sample code
>segment is included below.

Sorry. I should have spotted your previous email...

Last week Peter Jakubek [pjak at snafu.de] discovered a number of serious bugs
in the m68k support in binutils-2.11 and the current CVS source. These seem
to mostly be related to the relocation code in gas/bfd which was rewritten
between 2.10 and 2.11 I think. The threads of interest are:

http://sources.redhat.com/ml/binutils/2001-05/msg00471.html
http://sources.redhat.com/ml/binutils/2001-05/msg00418.html
http://sources.redhat.com/ml/binutils/2001-05/msg00315.html

Thanks to the efforts of Alan Modra and Peter the current CVS version of
binutils seems to be working properly now. It could do with some testing
though. It works for us with the ColdFire but I don't think anyone has tried
it on any other m68k family members. I'm also starting to work on testcases
for the bugs to try to ensure that they don't happen again...

hope this helps,

Dave