Re: GEOM_MBR breaks my kernel
Poul-Henning Kamp wrote: In message [EMAIL PROTECTED], walt writes: I've been unable to boot any kernel I've built since about March 11 and I've narrowed it down to the GEOM_MBR option. With GEOM_MBR I get a kernel page fault error when trying to mount the root filesystem at boot time. Can you get us the messages and a traceback ? Well, no. I've been trying to find a kernel configuration that will allow me to reproduce the bug AND generate a traceback, but so far I can't find one. The problem is that just adding GEOM_MBR to a GENERIC kernel doesn't produce the bug. My normal custom kernel doesn't contain the debugging stuff, and if I start changing things the bug doesn't show. The only semi-interesting result I've come up with is this: I normally use only the 'cpu I686_CPU' flag because I have an Athlon cpu. But if I also include the 'cpu I586_CPU' flag the bug completely changes: the machine boots and the filesystems mount just fine but about ten seconds after I start X running the machine panics and reboots shortly thereafter. The panic message doesn't appear on the screen because the console is not visible at that point. Does this suggest a gcc problem? I've never really understood how more than one 'cpu' flag can be included in the kernel config file, so I'm not sure what actually changes when I do that. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
GEOM_MBR breaks my kernel
I've been unable to boot any kernel I've built since about March 11 and I've narrowed it down to the GEOM_MBR option. With GEOM_MBR I get a kernel page fault error when trying to mount the root filesystem at boot time. ad0: 76319MB WDC WD800JB-00CRA1 [155061/16/63] at ata0-master UDMA100 ad1: 76319MB ST380021A [155061/16/63] at ata0-slave UDMA100 acd0: CDROM LTN301 at ata1-master PIO4 MBREXT Slice 5 on ad0s2: [0] f:00 typ:7 s(CHS):6/1/66 e(CHS):250/254/255 s:64 l:12161141 [1] f:00 typ:5 s(CHS):251/0/193 e(CHS):255/254/255 s:12161205 l:20980890 MBREXT Slice 6 on ad0s2: [0] f:00 typ:11 s(CHS):251/1/193 e(CHS):255/254/255 s:63 l:20980827 [1] f:00 typ:5 s(CHS):255/0/193 e(CHS):255/254/255 s:33142095 l:16787925 MBREXT Slice 7 on ad0s2: [0] f:00 typ:131 s(CHS):255/1/193 e(CHS):255/254/255 s:63 l:16787862 [1] f:00 typ:5 s(CHS):255/0/193 e(CHS):255/254/255 s:49930020 l:2088450 MBREXT Slice 8 on ad0s2: [0] f:00 typ:130 s(CHS):255/1/193 e(CHS):255/254/255 s:63 l:2088387 [1] f:00 typ:0 s(CHS):0/0/0 e(CHS):0/0/0 s:0 l:0 MBREXT Slice 5 on ad1s1: [0] f:00 typ:11 s(CHS):255/1/193 e(CHS):255/254/255 s:63 l:20980827 [1] f:00 typ:0 s(CHS):0/0/0 e(CHS):0/0/0 s:0 l:0 Information from DOS bootblock is: The data for partition 1 is: sysid 11 (0x0b),(DOS or Windows 95 with 32 bit FAT) start 63, size 4208967 (2055 Meg), flag 80 (active) beg: cyl 0/ head 1/ sector 1; end: cyl 261/ head 254/ sector 63 The data for partition 2 is: sysid 15 (0x0f),(Extended DOS (LBA)) start 4209030, size 52018470 (25399 Meg), flag 0 beg: cyl 262/ head 0/ sector 1; end: cyl 1023/ head 254/ sector 63 The data for partition 3 is: sysid 165 (0xa5),(FreeBSD/NetBSD/386BSD) start 139926150, size 16370235 (7993 Meg), flag 0 beg: cyl 1023/ head 0/ sector 1; end: cyl 1023/ head 254/ sector 63 The data for partition 4 is: UNUSED # /dev/ad0s3c: type: ESDI sectors/cylinder: 16065 cylinders: 9729 sectors/unit: 156301488 8 partitions: #size offsetfstype [fsize bsize bps/cpg] a: 15321596 10485764.2BSD0 0 0 # (Cyl. 65*- 1018*) b: 10485760 swap# (Cyl.0 - 65*) c: 163702350unused0 0 # (Cyl.0 - 1018) Warning, partition c doesn't cover the whole unit! Note that I put the swap partition first. Could that cause this problem? To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: GEOM_MBR breaks my kernel
In message [EMAIL PROTECTED], walt writes: I've been unable to boot any kernel I've built since about March 11 and I've narrowed it down to the GEOM_MBR option. With GEOM_MBR I get a kernel page fault error when trying to mount the root filesystem at boot time. Can you get us the messages and a traceback ? -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 [EMAIL PROTECTED] | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: GEOM_MBR breaks my kernel
On Fri, Mar 14, 2003 at 03:37:53PM +0100, Poul-Henning Kamp wrote:
In message [EMAIL PROTECTED], walt writes:
I've been unable to boot any kernel I've built since about March 11
and I've narrowed it down to the GEOM_MBR option.
With GEOM_MBR I get a kernel page fault error when trying to
mount the root filesystem at boot time.
Can you get us the messages and a traceback ?
I saw the same thing on my system. I don't have the exact message
or traceback around, but the problem was essentially a null pointer
dereference while in kernel mode.
I was able to locate the offending line in the source:
In devfs_allocv:
if (de-de_dirent-d_type == DT_CHR) {
dev = *devfs_itod(de-de_inode);
if (dev == NULL)
return (ENOENT);
} else {
...
The first comparison causes the problem, since de-de_dirent == NULL.
The problem did not exist until I turned WITNESS INVARIANTS off
(in a kernel with all GEOM_* stuff enabled).
Let me know if you need more information. If you need the traceback,
I'd appreciate if you told me how to get it written to a file. :)
Krzysztof
To Unsubscribe: send mail to [EMAIL PROTECTED]
with unsubscribe freebsd-current in the body of the message
