----- "marc verwerft" <[email protected]> schreef: > Hello, > > My previous firewall needs a replacement so I got hold of an > ACER aspire 9300 laptop. For a number of days I've tried installing > the 4.9 (i386) release. Unfortunatly, most things I try fail so I'm > looking for some help/advice. > > Upon booting from the 4.9 release CD I usually (9 out of 10) get a > panic screen. I was lucky enough to install 4.9 on the laptop but when > booting from the HD, I also usually get a panic screen. Sometimes I > can boot and run OpenBSD without trouble but that's the exception, > not the rule. > > In order to exclude HW problems I ran memtest86 for 11 hours, did a > thorough check on the disk that only revealed 2 reallocated sectors > and installed a number of linux distros (ubuntu, mint, clearos, ...). > Every distro worked out of the box. > > Then I asked for some advice on daemonforums (1) and 2 people helped > me. First advice was to try out the -current or AMD release. Second > advice was to upgrade the BIOS of the laptop. > > In order to upgrade the laptop I've reinstalled XP and was able to > flash the BIOS. I then retried all 3 three OpenBSD variants. No > change in behaviour. > > When I boot 4.9 i386 a get a panic screen shortly after the 'boot>' > prompt. Doesn't matter whether I boot from CD or HD. > > When I boot 5.0-current (8 of sept) I don't get a panic screen. But > the install halts and nothing happens. Last line I see is: > entry point at 0x20012 > and that's it. > > The AMD release gets to the 'boot>' prompt. After a short wait it > shows a spinning cursor and 1 number (in the other releases it seems > like calculating and then displays something like > '5965416+946088 [61+227984+215947]=0x703e08' ). Here I only get > the first number. Then it crashes without a panic screen and the > system reboots automatically. > > I've included the dmesg output at the end and for those interested > there's an attachment with a picture from the panic screens on the > daemonforums (1). > > Any help greatly appreciated > > Marc > > 1: http://www.daemonforums.org/showthread.php?t=6312 > > Dmesg output: > > ...
I have narrowed down the problem a bit. The rest of this email is based upon working with i386-current of sep 8th, booting from HD. The problem is somewhere in the second stage boot (/boot). Whenever the laptop boots fine, after the 'boot>' a line appears that contains a calculation of the entry point like this: booting cd0a:/5.0/i386/bds.rd: 5965416+946088 [61+227984+215947]=0x703e08 entry point at 0x200120 Note that these are not the actual numbers since they fly by to quick to see. If the laptop panics then I don't see that calculation. I only see the first number (guess it is still 5965416) followed by an hourglass. My guess is that an incorrect entry point is calculated, the processor jumps to that address and finds an invalid opcode there. This is consistent with what I saw using i386/4.9 release where the entry point was calculated as above (0x200120) and where the panic screen showed: trap: 0(0): invalid opcode fault cn_tab=0x4b250 .... Code dump [0x200121]: <- Note the address!! .... I found an bug report that might be related: kernel/5879: 4.3 kernel crash - trap: 0(0): invalid opcode fault Unfortunately, it is from 2008. I've tried getting access to bug reports from the OpenBSD site but whenever I try http://www.openbsd.org/query-pr.html I get a 404 error back. Is there another way to query the bugs database? Help still appreciated ;-) Regards, Marc
