re: port-macppc/54331: macppc MP kernels fail to boot successfully in -current (8.99.49)

[matthew green <mrg%eterna.com.au@localhost>]

The following reply was made to PR port-macppc/54331; it has been noted by GNATS.

From: matthew green <mrg%eterna.com.au@localhost>
To: "David H. Gutteridge" <david%gutteridge.ca@localhost>
Cc: gnats-bugs%netbsd.org@localhost
Subject: re: port-macppc/54331: macppc MP kernels fail to boot successfully in -current (8.99.49)
Date: Tue, 13 Aug 2019 12:46:06 +1000

 > I built an MP+DEBUG+DIAGNOSTIC+LOCKDEBUG 9.99.7 kernel with the patches
 > included, and it's hanging after USB devices are attached. I can unplug
 > and reattach USB devices and it continues to handle those events fine,
 > but it never actually mounts (or prompts to mount) the hard drive. I
 > don't see anything useful in the output to indicate why. I'm not able
 > to get into ddb as far as I know. That is, typing something on the USB
 > keyboard that's attached doesn't result in anything happening.
 
 does it still have the cpu failure attach messages in early boot?
 
 here's a simple way to see if i'm right about config_pending not
 being reduced at this point:
 
 in config_finalize() there is this loop:
 
 	while (config_pending != 0)
 		cv_wait(&config_misc_cv, &config_misc_lock);
 
 if you convert it to a cv_timedwait() that sleeps for hz (or
 even hz*2 or hz*5) and then prints config_pending, eg:
 
 	while (config_pending != 0) {
 		cv_timedwait(&config_misc_cv, &config_misc_lock, hz);
 		printf("%s: config_pending still %d\n", __func__,
 		    config_pending);
 	}
 
 (untested/compiled).  this would at least give us any idea on
 where to look next.
 
 
 .mrg.