Re: patch for bfd/ecofflink.c

[Nick Clifton <nickc at redhat dot com>]

Hi Nathan,

> Description:
> ------------
> Fix the reading of the debugging information of Tru64/Alpha
> binaries produced by recent Compaq compilers.  Reading of the debugging
> information for recent compilers was 1) just plain wrong at one
> place and

This part of the patch is OK.

> 2) confused because of very strange information generated by Compaq's
> compiler.  I have fixed the first added a simple hack for the second.  I
> have included extensive documentation as to why this is the best solution
> short of an extensive rewrite or another hack.

Not sure about this part of the patch.  I understand what you are
doing, but is it right to have such a horrible hack to cope with what
is obviously a very broken compiler ?  And since it is only
(presumably) the Compaq compiler that is broken in this way, shouldn't
the horrible hack only be enabled if that compiler is the potential
producer of the bfd, rather than for all bfds ?

I am not rejecting the patch - I am just asking for a discussion about
the horrible hack and whether it is suitable for inclusion into the
sources.

Cheers
        Nick