Date: Sun, 22 Feb 1998 01:36:41 -0500 From: "matthew c. mead" <mmead@calvin.math.vt.edu> To: Greg Lehey <grog@lemis.com>, questions@FreeBSD.ORG Subject: Re: new 2.2.5 installation randomly (and constantly) panics Message-ID: <19980222013641.59923@math.vt.edu> In-Reply-To: <19980222165501.24195@freebie.lemis.com>; from Greg Lehey on Sun, Feb 22, 1998 at 04:55:01PM %2B1030 References: <19980221202539.15260@math.vt.edu> <19980222120905.35825@freebie.lemis.com> <19980222010353.57237@math.vt.edu> <19980222165501.24195@freebie.lemis.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Feb 22, 1998 at 04:55:01PM +1030, Greg Lehey wrote:
> Ugh. A VM problem. Does this always look the same? The important
> part of this dump are the frames 11 (i586_copyin) to 5
> (vm_pager_has_page). Possibly frames 14 to 12 are also of importance.
Ok... I built after config -g, size /kernel and size
/sys/compile/GOOF/kernel are identical - here's a kgdb output:
IdlePTD 1e0000
current pcb at 1bf010
panic: general protection fault
#0 boot (howto=256) at ../../kern/kern_shutdown.c:266
266 dumppcb.pcb_cr3 = rcr3();
(kgdb) where
#0 boot (howto=256) at ../../kern/kern_shutdown.c:266
#1 0xf01105d2 in panic (fmt=0xf018aa05 "general protection fault")
at ../../kern/kern_shutdown.c:390
#2 0xf018b556 in trap_fatal (frame=0xefbffb90) at ../../i386/i386/trap.c:742
#3 0xf018ae06 in trap (frame={tf_es = 16, tf_ds = 16, tf_edi = 3,
tf_esi = -266245280, tf_ebp = -272630776, tf_isp = -272630856,
tf_ebx = 0, tf_edx = -199155712, tf_ecx = 0, tf_eax = 14, tf_trapno = 9,
tf_err = 30492, tf_eip = -272630680, tf_cs = 8, tf_eflags = 66178,
tf_esp = -190351586, tf_ss = -272630664}) at ../../i386/i386/trap.c:440
#4 0xefbffc68 in ?? ()
#5 0xf017f70f in vm_pager_has_page (object=0xf0d33f80, offset=0,
before=0xefbffc6c, after=0xefbffc68) at ../../vm/vm_pager.c:209
#6 0xf0176b51 in vm_fault_additional_pages (m=0xf0216b60, rbehind=3,
rahead=4, marray=0xefbffcf8, reqpage=0xefbffcd4)
at ../../vm/vm_fault.c:1101
#7 0xf0175faf in vm_fault (map=0xf0d35a00, vaddr=135004160,
fault_type=1 '\001', change_wiring=0) at ../../vm/vm_fault.c:417
#8 0xf018afd8 in trap_pfault (frame=0xefbffd78, usermode=0)
at ../../i386/i386/trap.c:633
#9 0xf018ad1f in trap (frame={tf_es = 16, tf_ds = 16, tf_edi = -259688448,
tf_esi = 135004160, tf_ebp = -272630160, tf_isp = -272630368,
tf_ebx = 2048, tf_edx = 135006208, tf_ecx = 1792, tf_eax = -199155712,
tf_trapno = 12, tf_err = 0, tf_eip = -266822348, tf_cs = 8,
tf_eflags = 66054, tf_esp = -60814, tf_ss = -65536})
at ../../i386/i386/trap.c:311
#10 0xf0189d34 in fastmove ()
#11 0xf0189c82 in i586_copyin ()
#12 0xf0125347 in sosend (so=0xf0d35600, addr=0x0, uio=0xefbfff34, top=0x0,
control=0x0, flags=0) at ../../kern/uipc_socket.c:445
#13 0xf0119ef5 in soo_write (fp=0xf0d34b80, uio=0xefbfff34, cred=0xf0d33980)
at ../../kern/sys_socket.c:82
#14 0xf0117823 in write (p=0xf0c18400, uap=0xefbfff94, retval=0xefbfff84)
at ../../kern/sys_generic.c:263
#15 0xf018b7ef in syscall (frame={tf_es = 39, tf_ds = 39, tf_edi = 135004160,
tf_esi = 870118, tf_ebp = -272642004, tf_isp = -272629788,
tf_ebx = 134960252, tf_edx = 0, tf_ecx = 0, tf_eax = 4, tf_trapno = 22,
tf_err = 7, tf_eip = 134841233, tf_cs = 31, tf_eflags = 531,
tf_esp = -272642056, tf_ss = 39}) at ../../i386/i386/trap.c:890
#16 0x8098391 in ?? ()
#17 0x359f in ?? ()
#18 0x8467 in ?? ()
#19 0x2009 in ?? ()
#20 0x1095 in ?? ()
> > The hardware in the machine is an ISA vga card, a dec 21040 based
> > ethernet card, an NCR 8510 PCI scsi card, an Adaptec UltraSCSI 2940,
> > and a P90 cpu (clocked to 100 - has been that way for about 2 years
> > without problems).
> i586_copyin has been introduced since 2.1.7. I wouldn't put it beyond
> the realms of possibility for it to trigger an instability in your
> overclocking that so far has gone untriggered. How easy would it be
> to turn the machine back to 90 MHz? If that doesn't help, it would be
> instructive to build a kernel for a 486 and see if that works at
> either frequency. As you might expect, a 486 kernel doesn't use
> i586_<mumble>.
But if I build a kernel for a 486 CPU, will it run on a 586 CPU? I was
under the impression it wouldn't, but I don't know why. I'll try that
assuming it should run and see if things improve. If it'll save me a
trip, it's worth it. :)
-matt
--
Matthew C. Mead Virginia Tech Mathematics Department
460 McBryde Hall
Blacksburg, VA 24061-0123
mmead@math.vt.edu (540)231-2643 FAX: (540)231-5960
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980222013641.59923>