[BUG] Regression in 2.14.90 (relative to 2.13.90)

Carlo Wood carlo@alinoe.com
Sat Nov 15 16:44:00 GMT 2003


On Fri, Nov 14, 2003 at 09:54:45AM +0000, Nick Clifton wrote:
> Hi Carlo,
> 
> >> FWIW I tried compiling the test case attached to the above message
> >> using a version of g++ built from today's sources and running on an
> >> x86 linux box, and then I used three different versions of readelf to
> >> try to get different results - I failed - all three produced the same
> >> output:
> >
> > Of course, the bug is not in readelf but in ld.
> > You will have to install the complete binutils before compiling
> > the testcase.  Then you can use a readelf from any version to
> > see what the .debug_info is.
> 
> Ah - but that is the point - I *did* use a copy of ld built from
> today's binutils sources.  And this version of ld produced a binary
> that appears to be correct.

The correct output would give an entry with PC = 0.  And you get
(just like me) two entries with both a non-zero PC, which is the
bug.  If you would *compile/link* the test case with binutils
2.3.90, then you would see different results, which is correct.

Note that you said that you tried to use different versions of
readelf only in order to get different results, which won't work.

> Would it be possible for you to check to see if you can still
> reproduce the problem using today's binutils sources ?

I can and apparently so can you.

-- 
Carlo Wood <carlo@alinoe.com>



More information about the Binutils mailing list