serial console fails after intr_establish

Frank Wille <[email protected]>
Newsgroups gmane.os.netbsd.ports.powerpc
Organization Privates Internet Ostwestfalen/Lippe e.V.
Message-ID <[email protected]>
Hi!

I'm seeing a strange bug on NetBSD/sandpoint since I tested recent
kernels on it (NetBSD6 works). The serial console stops working at
the first intr_establish() call, and continues working after going
to multiuser with interrupts enabled.

Booting on a LinkStation HD-HLAN looks like this on the serial console:

NetBSD 6.99.27 (KUROBOX) #4: Sat Nov 16 12:06:14 CET 2013
	[email protected]:/home/frank/netbsd/current/src/objsand/sys/arch/sandpoint/compile/KUROBOX
Model: kurobox
total memory = 65536 KB
avail memory = 59436 KB
OpenPIC Version 1.2: Supports 1 CPUs and 26 interrupt sources.
mainbus0 (root)
cpu0 at mainbus0: 8245 (Revision 0.4), ID 0 (primary)
cpu0: HID0 0x90c000<DOZE,DPM,ICE,DCE>, powersave: 1
cpu0: 195.12 MHz
eumb0 at mainbus0
com0 at eumb0 unit 1: ns16550a, working fifo
com0: console?Sat Nov 16 12:07:51 CET 2013
Starting root file system check:
[...]

There is a strange character after "com0: console" which makes we think
that the port has been reinitialized at a different speed?

But this happens definitely during a call to intr_establish(), and there
is nothing which should change the port speed.

When I remove the intr_establish() in sandpoint/com_eumb.c the
output continues, until the next intr_establish()!

Has anybody an idea what might have changed to cause such a behaviour?

-- 
Frank Wille
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.