objdump --full-contents truncates addresses to long

[Ian Lance Taylor]

Alexandre Oliva <aoliva@redhat.com> writes:

> Truncating to long is not a good idea for a 32-x-64 objdump.  Is there
> any reason for us to not use bfd_printf_vma?  Perhaps bfd_printf_vma
> should be made more flexible, accepting arguments specifying the a
> minimum/maximum width to be printed, and whether to pad with zeros?
> Other than the ugly implications of this change to the testsuite, that
> I'll look into addressing if this patch is approved, is this ok to
> install along with the testsuite fixes?

Doesn't this make the output particularly ugly when dumping object
files?  Perhaps the code could use bfd_sprintf_vma and strip off
excess leading zeroes, while retaining even spacing?

Ian