PERFORCE change 110969 for review

[M. Warner Losh]

In message: <200612031915.kB3JFLbh077129 at repoman.freebsd.org>
            Hans Petter Selasky <hselasky at freebsd.org> writes:
: http://perforce.freebsd.org/chv.cgi?CH=110969
: 
: Change 110969 by hselasky at hselasky_mini_itx on 2006/12/03 19:14:41
: 
: 	Fix a bug in the uplcom probe routine where a revision of 0xFFFF
: 	was not treated as a special value. The old version of uplcom.c has
: 	the check, so probably a misunderstanding happened.
: 
: Affected files ...
: 
: .. //depot/projects/usb/src/sys/dev/usb/uplcom.c#14 edit
: 
: Differences ...
: 
: ==== //depot/projects/usb/src/sys/dev/usb/uplcom.c#14 (text+ko) ====
: 
: @@ -424,7 +424,8 @@
:  	    while(up->vendor) {
:  	        if ((up->vendor == uaa->vendor) &&
:  		    (up->product == uaa->product) &&
: -		    (up->release <= uaa->release)) {
: +		    ((up->release <= uaa->release) ||
: +		     (up->release == 0xFFFF))) {
:  		    return up;
:  		}
:  		up++;

What really needs to happen here is that we probe the plcom chip for
its properties, ala the linux driver rather than having a fixed (and
often wrong) table of hard codings.  I looked into this a while ago,
but got distracted...

Warner