Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!lll-winken!elroy.jpl.nasa.gov!sdd.hp.com!zaphod.mps.ohio-state.edu!rpi!batcomputer!munnari.oz.au!manuel!tridge From: tridge@anu.oz.au (Andrew Tridgell) Newsgroups: comp.unix.ultrix Subject: Re: kernel panic, "unaligned access" Message-ID: <1991Mar6.234212.21548@newshost.anu.edu.au> Date: 6 Mar 91 23:42:12 GMT References: Sender: news@newshost.anu.edu.au Organization: Australian National University, Canberra, ACT, Australia Lines: 149 Just yesterday morning I also got a crash for no apparent reason. According to uerf it was also the result of an unaligned access. This is the first time in 2 years we've had a random crash, and I'd like to find out the cause. We have a DecStation 3100 running Ultrix 4.0. At the time the load average was below 2 and there were 4 users logged on. The last processes running were a login and a process spawned by cron that did an rsh on another machine (a sun). Neither the rsh nor the login completed. uerf shows - ********************************* ENTRY 188. ********************************* ----- EVENT INFORMATION ----- EVENT CLASS OPERATIONAL EVENT OS EVENT TYPE 250. ASCII MSG SEQUENCE NUMBER 47. OPERATING SYSTEM ULTRIX 32 OCCURRED/LOGGED ON Wed Mar 6 12:32:37 1991 EDT OCCURRED ON SYSTEM aerodec SYSTEM ID x82011601 HW REV: x1 FW REV: x16 CPU TYPE: R2000A/R3000 PROCESSOR TYPE KN01 MESSAGE ********************************* ENTRY 189. ********************************* ----- EVENT INFORMATION ----- EVENT CLASS ERROR EVENT OS EVENT TYPE 117. ERROR & STATUS REGS SEQUENCE NUMBER 46. OPERATING SYSTEM ULTRIX 32 OCCURRED/LOGGED ON Wed Mar 6 12:32:37 1991 EDT OCCURRED ON SYSTEM aerodec SYSTEM ID x82011601 HW REV: x1 FW REV: x16 CPU TYPE: R2000A/R3000 PROCESSOR TYPE KN01 ----- ERROR & STATUS REGS ----- CAUSE x30000010 EXCEPTION CODE ADR ERR (LOAD OR INST _FETCH) EPC x801151B0 STATUS x00000000 CURRENT INTERRUPT STATE DISABLED CURRENT MODE KERNEL PREVIOUS INTERUPT STATE DISABLED PREVIOUS MODE KERNEL OLD INTERUPT STATE DISABLED OLD MODE KERNEL CACHE STATE NORMAL VIRTUAL ADDRESS x2F320951 SP xFFFFDED4 ********************************* ENTRY 190. ********************************* ----- EVENT INFORMATION ----- EVENT CLASS ERROR EVENT OS EVENT TYPE 200. PANIC SEQUENCE NUMBER 48. OPERATING SYSTEM ULTRIX 32 OCCURRED/LOGGED ON Wed Mar 6 12:32:37 1991 EDT OCCURRED ON SYSTEM aerodec SYSTEM ID x82011601 HW REV: x1 FW REV: x16 CPU TYPE: R2000A/R3000 PROCESSOR TYPE KN01 PANIC MESSAGE unaligned access ********************************* ENTRY 191. ********************************* ----- EVENT INFORMATION ----- EVENT CLASS ERROR EVENT OS EVENT TYPE 117. ERROR & STATUS REGS SEQUENCE NUMBER 49. OPERATING SYSTEM ULTRIX 32 OCCURRED/LOGGED ON Wed Mar 6 12:32:37 1991 EDT OCCURRED ON SYSTEM aerodec SYSTEM ID x82011601 HW REV: x1 FW REV: x16 CPU TYPE: R2000A/R3000 PROCESSOR TYPE KN01 ----- ERROR & STATUS REGS ----- CAUSE x30000C10 EXCEPTION CODE ADR ERR (LOAD OR INST _FETCH) HW INTERRUPT 0 PENDING HW INTERRUPT 1 PENDING EPC x80099558 STATUS x0000C004 CURRENT INTERRUPT STATE DISABLED CURRENT MODE KERNEL PREVIOUS INTERUPT STATE ENABLED PREVIOUS MODE KERNEL OLD INTERUPT STATE DISABLED OLD MODE KERNEL HW INTERRUPT 4 ENABLED HW INTERRUPT 5 ENABLED CACHE STATE NORMAL VIRTUAL ADDRESS x003938D7 SP xFFFFDB24 Which isn't particularly useful for me, as I haven't had much experience reading such things (it doesn't go down very often) dbx gives : aero[9:38am]>dbx -k ./vmunix.0 ./vmcore.0 dbx version 2.0 Type 'help' for help. reading symbolic information ... [using memory image in ./vmcore.0] (dbx) where > 0 swtch.swtch() ["../../machine/mips/swtch.c":275, 0x8011adb0] 1 sleep_unlock(chan = 0x80167f9d = "^A", pri = 0, l = (nil)) ["../../sys/kern _synch.c":519, 0x800aa458] 2 kern_synch.sleep(chan = (nil), pri = -2146891488) ["../../sys/kern_synch.c" :385, 0x800a9ffc] 3 .block419 ["../../vm/vm_sched.c":371, 0x800e2f50] 4 sched() ["../../vm/vm_sched.c":371, 0x800e2f50] 5 .block217 ["../../sys/init_main.c":755, 0x800909a4] 6 .block216 ["../../sys/init_main.c":755, 0x800909a4] 7 main() ["../../sys/init_main.c":755, 0x800909a4] 8 entry.start() ["../machine/entry.s":110, 0x80030098] (dbx) Any ideas at to what caused this? Thanks! --- Andrew Tridgell tridge@aerodec.anu.edu.au