CAP_LARGE_READX and samba performance

[Maulik Desai]

Hi,

What does supporting CAP_LARGE_READX capability buy? I'm looking at ways to
maximize
samba performance, and was wondering if this would help. I understand that
this will only help in
case of clients speaking NT LM 0.12 dialect and only while reading large
files from server (using 
read_and_x command).

The cifs spec says that:
"If CAP_LARGE_READX was indicated by the server in the negotiate protocol
response, 
the request's MaxCount field may exceed the negotiated buffer size if Fid
refers to a disk 
file. The server may arbitrarily elect to return fewer than MaxCount bytes
in response. "

Does this mean that a client would ask to read more bytes (per each
read_and_x request) if this
capability is supported by samba server?

smbd/negprot.c has a comment:
/*
  other valid capabilities which we may support at some time...
                     CAP_LARGE_READX|CAP_STATUS32|CAP_LEVEL_II_OPLOCKS;
 */

I searched the samba & M$ cifs archives, but couldn't find any relevant
information.

If it is worthwhile to add this capability, I can help out to the best of my
capacity in order 
to explore further and possibly implement it.

TIA,
-Maulik