[CFT] NDIS optional header length related fixups

[Andrew R. Reiter]

Due to a lack of response from most parties, I created a PR for the below.  
It can be viewed at:

http://www.freebsd.org/cgi/query-pr.cgi?pr=83477

Thanks,
Andrew

On Wed, 29 Jun 2005, Andrew R. Reiter wrote:

:Calling NDIS -CURRENT users,
:
:Attached is a patch that should fix any possible issues with mis-calculating
:offsets or sizes when dealing with anything 'image_optional_header' related in
:the PE loading code.  The reason for the patch is that the optional header can
:have a varying length due to the lack of requiring the existence of all the
:'image_data_directory's to exist within a binary.  As far as I can tell, most
:drivers tend to include all, but due to the basic idea that there can be less
:than IMAGE_DIRECTORY_ENTRIES_MAX data directories in the optional header, we
:should at least make an attempt at preemptively catch any bugs that might arise
:due to improper pointer calculation.
:
:If you could, please give it a run in your tree!
:
:The patch is also located at:
:  http://www.watson.org/~arr/ndis_opthdrsz.diff
:
:I guess let me know if anyone has any problems with this working (or other).
:
:Cheers,
:Andrew
:
:--
:Andrew R. Reiter
:arr at watson.org

--
Andrew R. Reiter
arr at watson.org