On Wed, 2013-01-23 at 15:59 -0800, Roland McGrath wrote:
I'm not entirely sanguine about this direction of change, at
least not
without yet another switch making it optional. Sometimes what you want to
see from readelf is end-user information, where resolved addresses are the
most useful. But sometimes what you want is to really see what is in the
file (like when debugging encoders and decoders), where resolving the
addresses just hides the real details of the encoding.
Grin. Burned too many times by "bad producers"? ;)
But yeah, that would be a nice switch to have also in some other cases.
How about the attached patch that adds a -U, --unprocessed-values output
control?
Thanks,
Mark