In my opinion those messages (03:08: rw=0, want=541478279, limit=1855469) hint to a defective L2 cache, RAM, hard disk, filesystem or maybe CPU. Messages like this occur for example when using a ramdisk image that is too large for the ramdisk size the kernel is configured for.
You might run fsck in order to check the filesystem but if your L2 cache/CPU/RAM is defective it can destroy the filesystem. The kernel itself can't autodetect the L2 cache settings - if the Sonnet board doesn't automatically activate the cache on power up it should be disabled. With the following command you can verify whether the cache is enabled or not: cat /proc/sys/kernel/l2cr If it is enabled try disabling it completely using the command: echo '0x00000000' > /proc/sys/kernel/l2cr in order to see if the error messages still occur. In order to see if errors really only occur in Linux you can use Gauge Pro's cyclic RAM test for MacOS. You can download it here: http://eshop.macsales.com/Tech/FTP/newertech/GaugePRO11.sit A possible L2 cache or CPU failure was probably caused by overheating when you ran the PB with the lid closed - and at that time the power saving features of the CPU weren't enabled causing the CPU to always draw much current and therefore produce much heat. Now that the power saving features are enabled in the kernel it might be possible to run the PB with the lid closed. Tobias -----Original Message----- Date: Sat, 27 May 2006 22:09:12 +0200 Subject: Re: [Nubus-pmac-users] again new kernel images (SCSI and interrupt fixes) From: Peter To: nubus-pmac-users@lists.sourceforge.net On Sat, 2006-05-27 at 19:06 +0200, Tobias Netzel wrote: > First you should try with L2 cache disabled -> no "l2cr=0x......." > option to the kernel. I left out the l2cr=.. argument, and it booted alright. Interestingly, it looks like it detected the cache anyway, because my bogomips are at the same value as with the L2 cache anabled: pb1400:~# linuxinfo Linux pb1400 2.4.32 #649 Do Mai 25 01:55:56 CEST 2006 One PowerBook,NuBus processor, 665.19 total bogomips, 60M RAM System library 2.3.2 Does that mean that your newest kernels automaticly pick up the right value? Well, it seems like I'm back in the race for now :-) The only thing that bothers me now are regular messages of this kind: attempt to access beyond end of device 03:08: rw=0, want=541478279, limit=1855469 Thanks for your advise sofar, Peter > Then you can try without PCMCIA card inserted. (Do you also have two of > them?) > > It seems to have to do with the PMU....everything has to do with the > PMU. > > Tobias ------------------------------------------------------- All the advantages of Linux Managed Hosting--Without the Cost and Risk! Fully trained technicians. The highest number of Red Hat certifications in the hosting industry. Fanatical Support. Click to learn more http://sel.as-us.falkag.net/sel?cmd=lnk&kid=107521&bid=248729&dat=121642 _______________________________________________ Nubus-pmac-users mailing list Nubus-pmac-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nubus-pmac-users ------------------------------------------------------- All the advantages of Linux Managed Hosting--Without the Cost and Risk! Fully trained technicians. The highest number of Red Hat certifications in the hosting industry. Fanatical Support. Click to learn more http://sel.as-us.falkag.net/sel?cmd=lnk&kid=107521&bid=248729&dat=121642 _______________________________________________ Nubus-pmac-users mailing list Nubus-pmac-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nubus-pmac-users
