Re: port-mac68k/37611: macs with cuda adb don't boot -current

[Scott Reynolds <scottr%NetBSD.org@localhost>]

The following reply was made to PR port-mac68k/37611; it has been noted by 
GNATS.

From: Scott Reynolds <scottr%NetBSD.org@localhost>
To: gnats-bugs%NetBSD.org@localhost,
 Martin Husemann <martin%duskware.de@localhost>
Cc: 
Subject: Re: port-mac68k/37611: macs with cuda adb don't boot -current
Date: Sun, 30 Mar 2008 21:18:44 -0500

 I've looked at this several times with a varying degree of  
 thoroughness, but haven't come with anything conclusive.
 
 While there are certainly similarities to a condvar or a semaphore,  
 this is a somewhat unique situation in that the ADB completion code is  
 shared by both the MRG (ROM) and direct drivers. Further, the flag is  
 set in an interrupt context, either directly or indirectly, so certain  
 accommodations must be made.
 
 Simply put, it shouldn't be an issue to pass the address with  
 __UNVOLATILE since it seems to only be dereferenced in compiled code  
 within the function where it's declared volatile.
 
 This doesn't answer the question as to why it's broken; there could  
 certainly be something that I'm missing. I am interested in knowing  
 which machine you're having trouble with, though.
 
 --scott