Re: mi/2549: data-list-register-values igonores the format parameter

[Nick Roberts <nickrob@snap.net.nz>]

 > I have the following trace from a debug session:
 > 
 > 226,777 52-data-list-register-values --thread 1 d 3
 > 226,778 53-data-list-register-values --thread 1 t 3
 > 226,778 54-data-list-register-values --thread 1 N 3
 > 226,779 55-data-list-register-values --thread 1 o 3 0 1 2 4 5 6
 > 226,781 52^done,register-values=[{number="3",value="-1209315340"}]
 > 226,781 (gdb) 
 > 226,782 53^done,register-values=[{number="3",value="-1209315340"}]
 > 226,782 (gdb) 
 > 226,782 54^done,register-values=[{number="3",value="-1209315340"}]
 > 226,782 (gdb) 
 > 226,783 55^done,register-values=[{number="3",value="-1209315340"},{number="0",value="-1073743420"},{number="1",value="-1073743552"},{number="2",value="1"},{number="4",value="0xbffff920"},{number="5",value="0xbffff928"},{number="6",value="-1207956256"}]
 > 226,783 (gdb) 

FWIW I can see this regression too.  I think it's due to this change:


    2008-10-28  Tom Tromey  <tromey@redhat.com>

            ...
   	    (val_print): Remove format, deref_ref, pretty arguments; add
	    options.  Update.
            ...


The test mi2-regs.exp only tests this command on Solaris, so running the
testsuite on any other architecture wouldn't reveal the problem.

I've cc'ed the GDB mailing list rather than GNATS as it's a recent change and
this list seems to get more attention.

-- 
Nick                                           http://www.inet.net.nz/~nickrob