Re: current broken on 486

["William D. Jones" <thor0505%comcast.net@localhost>]

I have not been actively following -current for a while (as indicated by my 
backlog!), but was this issue ever fixed?

-----Original Message----- 
From: Felix Deichmann
Sent: Wednesday, December 07, 2016 3:22 PM
To: port-i386%netbsd.org@localhost
Subject: current broken on 486

Hi,

I tested current against a 486DX2/66 with 48 MB RAM and a 486DX/33 with
24 MB RAM, and it is broken.
Tried a GENERIC_TINY kernel, a GENERIC_TINY kernel with options DDB and
DIAGNOSTIC, and also GENERIC (on the DX2 with more RAM only).

In all cases the machines panic right after "cpu0 at mainbus0" with a
"uvm_fault" and "fatal page fault in supervisor mode".
Unfortunately I cannot get into ddb, the machines will hang or reboot
after printing the lines as in this log via serial console:

http://www.netbsd.org/~flxd/486_console_20161207.txt

Following the addresses after "eip" with nm or the map file gives:

nm:
0xc0119b5f cpu_get_tsc_freq
0xc012828d lapic_initclocks

netbsd.map:
c0119bac t cpu_attach
c0128338 t i82489_icr_wait

HTH (but don't understand why they differ?)...
NetBSD 7.0.1 works on aforementioned machines.

Felix

--
William D. Jones
thor0505%comcast.net@localhost