|
OpenBSD Installation and Upgrading Installing and upgrading OpenBSD. |
|
Thread Tools | Display Modes |
|
|||
Reporting Installation Kernel Panics on New Boxes
Greetings,
I've a new Laptop with a new Processor (Intel); and am running into Kernel Panics Failures during installations. I'm no kernel debugger, so "ddb>" commands are a bit foreign to me. I have downloaded a few man pages on it; but my questions are: Q1: Since the Laptop is "non-functional" due to the kernel panic (presuming sendbug doesn't work, since it kicked me out to the ddb prompt), how do I get the "trace" and "ps" outputs to bugs@openbsd.org? Q1A: Is it, "copy on paper by hand, and send the email on another OS"? Thanks. |
|
|||
Greetings:
I've contacted bugs@openbsd.org... From what I can tell, getting the dmesg is pretty ideal for a panicking machine. I understand the FAQs at openbsd.org address means of recording dmesg; and one bugs administrator commented that an USB Section may be appropriate. I took a gander at the install disc; and I noticed all the xx0x entities on /dev. Since it can read attached devices (one external USB Flash Drive as "sd1") - is there a way we can mount a usb drive to save dmesg to (e.g, a .txt file per the FAQ example for the floppy)? I'll need someone with a good grasp of mountpoints to help me out here - I keep getting "device not configured" whenever I try to copy over, mkdir, etc. an sd1x entity and mount the sd1 on the install disc's /dev directory Pardon the ignorance; but I'm a front end guy, not a Sysadmin ^_^ Handwriting panic, trace, and ps was my good deed of the day...ROTFL! Last edited by IronForge; 5th April 2013 at 03:42 AM. Reason: grammar |
|
|||
Quote:
If the usb thumb drive is msdos (aka vfat in linux) then you mount it as Code:
mount -t msdos /dev/sd1i /mnt Code:
dmesg > /mnt/dmesg.txt |
|
|||
Thanks, Shep.
The "extra' USB drive was read by the install disc as "sd1"; but there are no sd1x identifiers - only xxx0x identifiers on the install disc. I know - occicat might not like my approach; and may pounce on this thread; but I'm reading FAQ-14, fdisk, mount, and disklabel right now. From the hip, reading and multitasking (non PC chores), I can see that mounting the device by name or using disklabel may be a factor. I'm using a Linux LiveCD in the meanwhile; and have contemplated using a similar function so I can get this over to bugs@openbsd.org Again, thanks. |
|
|||
You did not mention which version of OpenBSD you tried to install. If it was the 5.2 release you could try to install a 5.3 snapshot:
Code:
98596 Apr 4 10:33 INSTALL.i386 22581 Apr 4 10:33 INSTALL.linux 2075 Apr 4 10:34 SHA256 56418201 Apr 4 10:33 base53.tgz 9595838 Apr 4 10:33 bsd 9623641 Apr 4 10:33 bsd.mp 6494632 Apr 4 10:33 bsd.rd 6639616 Apr 4 10:33 cd53.iso 70868 Apr 4 10:33 cdboot 2048 Apr 4 10:33 cdbr 3012608 Apr 4 10:33 cdemu53.iso 46063494 Apr 4 10:33 comp53.tgz 523640 Apr 4 10:33 etc53.tgz 1474560 Apr 4 10:33 floppy53.fs 1474560 Apr 4 10:33 floppyB53.fs 1474560 Apr 4 10:33 floppyC53.fs 2629519 Apr 4 10:33 game53.tgz 1607 Apr 4 10:34 index.txt 223625216 Apr 4 10:34 install53.iso 9984457 Apr 4 10:34 man53.tgz 78952 Apr 4 10:34 pxeboot 10735567 Apr 1 08:01 xbase53.tgz 64833 Apr 1 08:01 xetc53.tgz 39801283 Apr 1 08:01 xfont53.tgz 26952303 Apr 1 08:01 xserv53.tgz 4479364 Apr 1 08:01 xshare53.tgz
__________________
You don't need to be a genius to debug a pf.conf firewall ruleset, you just need the guts to run tcpdump |
|
|||
dmseg, trace, ps sent.
Thank you all.
J65nko: It was version 5.3 Per jggimi's instructions, I was able to run dmesg and attach it as a .txt file to bugs. Well, I hope they can incorporate the Laptop - I volunteered my laptop to assist in the effort, so I look forward to see this working on the new platforms. Again, thanks. |
|
|||
I had the same problem
Quote:
|
|
|||
kimotori, due to the lack of information provided by IronForge, it is not clear that the problem you experienced is the same as his.
|
|
||||
There was more detail on IronForge's problem posted to bugs@; a brief look at the trace output shows a failure to store AML code type 8; you may compare it to this problem report on tech@ which fails to store AML code type 100. Both IronForge's report and the report on tech@ were using a -current kernel from the last week of March 2013.
The CVS log for the module containing aml_store -- sys/dev/acpi/dsdt.c -- has had four patches since then which might resolve aml_store panics. Unfortunately, all I know of AML is how to spell it and then look it up in Wikipedia: Quote:
Last edited by jggimi; 28th June 2013 at 04:24 PM. Reason: clarity |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Chromium loses focus in text boxes | kly | FreeBSD Ports and Packages | 4 | 19th October 2011 02:53 PM |
Anyone with AMD64 random panics on i386? | indiocolifa | FreeBSD General | 13 | 24th March 2009 02:53 AM |
How to look for open slots in Sun Boxes | revzalot | Solaris | 1 | 19th July 2008 04:54 PM |
Constant Kernel Panics, Corupted file system reboots over and over | pseudonym | FreeBSD General | 24 | 14th July 2008 03:12 AM |
Bug reporting | jones | OpenBSD General | 4 | 3rd July 2008 02:51 PM |