uncorrectable DMA error AFAR on -CURRENT with Netra X1

[Kevin Day]

Hey, I'm trying to get a Netra X1 going with -CURRENT. I've got the if_dc 
fixes in place to make it endian safe, and it's mostly working. (Only one 
of the two ethernet ports works, and it thinks the MAC address is all 0's). 
Ignoring those, the system seems to work okay until it's put under some 
kind of load.

A few minutes of heavy disk and network activity will eventually cause:

panic: pcib: uncorrectable DMA error AFAR 0x606d1108 AFSR 0x4100ff0020800000
panic messages:
---
panic: pcib: uncorrectable DMA error AFAR 0x606d1108 AFSR 0x4100ff0020800000

syncing disks, buffers remaining... 1402 1402 panic: mi_switch: switch in a 
critical section
Uptime: 36m52s
Dumping 128 MB (1 chunks)


I've got two crashdumps from this, but haven't been able to find much 
meaningful other than a common set in the backtrace:

(kgdb) bt
#0  0x00000000c00b71a8 in doadump () at ../../../kern/kern_shutdown.c:239
#1  0x00000000c00b7884 in boot (howto=260) at ../../../kern/kern_shutdown.c:370
#2  0x00000000c00b7c88 in panic (fmt=0xc0208bd8 "mi_switch: switch in a 
critical section") at ../../../kern/kern_shutdown.c:548
#3  0x00000000c00bf890 in mi_switch () at ../../../kern/kern_synch.c:477
#4  0x00000000c00b7504 in boot (howto=256) at ../../../kern/kern_shutdown.c:311
#5  0x00000000c00b7c88 in panic (fmt=0xc021fdd8 "%s: uncorrectable DMA 
error AFAR %#lx AFSR %#lx") at ../../../kern/kern_shutdown.c:548
#6  0x00000000c01e6df0 in psycho_ue (arg=0xfffff80060438200) at 
../../../sparc64/pci/psycho.c:738
#7  0x00000000c0039368 in intr_fast () at 
../../../sparc64/sparc64/interrupt.S:189


Anything higher up seems to be different on each crash, which makes sense, 
but doesn't help much in narrowing it down.

Does anyone have any insight on what to look at next? Or does anyone want 
to take a look at the crash dumps? I'm already using "options OFW_NEWPCI" 
since without that my console didn't seem to be getting interrupts (input 
and output were very very slow) and I was getting alot of dropped packets.



Copyright (c) 1992-2003 The FreeBSD Project.
Copyright (c) 1979, 1980, 1983, 1986, 1988, 1989, 1991, 1992, 1993, 1994
         The Regents of the University of California. All rights reserved.
FreeBSD 5.1-CURRENT #1: Wed Jul 23 02:10:49 GMT 2003
     toasty@:/usr/src/sys/sparc64/compile/SERVER
Preloaded elf kernel "/boot/kernel/kernel" at 0xc0362000.
Timecounter "tick"  frequency 400000000 Hz
real memory  = 134217728 (128 MB)
avail memory = 119930880 (114 MB)
cpu0: Sun Microsystems UltraSparc-IIe Processor (400.00 MHz CPU)
nexus0: <OpenFirmware Nexus device>
pcib0: <U2P UPA-PCI bridge> on nexus0
pcib0: Sabre, impl 0, version 0, ign 0x7c0, bus A
DVMA map: 0x60000000 to 0x63ffffff
pci0: <OFW PCI bus> on pcib0
isab0: <PCI-ISA bridge> at device 7.0 on pci0
isa0: <ISA bus> on isab0
pci0: <old, non-VGA display device> at device 3.0 (no driver attached)
pci0: <old, non-VGA display device> at device 3.0 (no driver attached)
dc0: <Davicom DM9102A 10/100BaseTX> port 0x10000-0x100ff at device 12.0 on pci0
dc0: Ethernet address: 00:00:00:00:00:00
miibus0: <MII bus> on dc0
ukphy0: <Generic IEEE 802.3u media interface> on miibus0
ukphy0:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
dc1: <Davicom DM9102A 10/100BaseTX> port 0x10100-0x101ff mem 0x2000-0x20ff 
at device 5.0 on pci0
dc1: Ethernet address: 00:00:00:00:00:00
miibus1: <MII bus> on dc1
ukphy1: <Generic IEEE 802.3u media interface> on miibus1
ukphy1:  10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
pci0: <serial bus, USB> at device 10.0 (no driver attached)
atapci0: <AcerLabs Aladdin UDMA66 controller> port 
0x10220-0x1022f,0x10208-0x1020b,0x10210-0x10217,0x10218-0x1021b,0x10200-0x10207 
at device 13.0 on pci0
ata2: at 0x10200 on atapci0
ata3: at 0x10210 on atapci0
Timecounters tick every 10.000 msec
ad0: 19092MB <ST320413A> [38792/16/63] at ata2-master UDMA66
Mounting root from ufs:/dev/ad0a
WARNING: / was not properly dismounted
warning: no time-of-day clock registered, system time will not be set 
accurately

-- Kevin