Hi, I recently decided to revamp my main Gentoo installation. I decided to have a go from scratch using the one of the new 1.4_rc2 LiveCDs, speficially the Pentium III GRP image. I had also decided to repartition using LVM. I booted the CD, created a large partition and turned it into a physical volume. I made a volume group from this, then started populating it with logical volumes - one for swap, one for /tmp, one for /var and so on. After I had created about 4, I realised that the order of the last 2 was not to my taste and I wanted to swap the latter with each other in terms of their ordering, for performance reasons. No problem I thought, I'll just use lvrename to rename them appropriately, rather than deleting any and creating them again. That seemed to work, but when I typed lvscan again to see the results of my handiwork I got a nasty oops - some of which is reproduced here: Unable to handle kernel NULL pointer dereference at virtual address 00000021 Printing EIP 8018ec90 *pde=00000000 Oops: 0000 CPU: 0 EIP: 0010:[<8018ec90>] Tainted: PF EFLAGS: 00010202 eax: 00000001 ebx: 0000003e ecx: 00000000 edx: 00000001 esi: 81367c8a edi: 8fca5f57 ebp: 8fca5f18 esp: 8fca5e9c Process lvscan (pid: 3025, stackpage = 8fca5000) Surely I did nothing wrong? Upon rebooting, the volume was inactive. I activated it and saw that the adjustment did seem to have been committed. I tried to rename another volume and the crash occurred again, so it was reproducible. Also, worthy of note is that I hadn't used hdparm to apply any potentially destabilising optimisations on the first attempt. I think the chances of any crashing needs to be utterly minimised on the final boot/install CD - this is only my second ever Oops since I have used Linux (RedHat 4.0), and both times it has been with a Gentoo boot CD. Furthemore, evmsn simply doesn't work. It says it requires an ioctl of version 11.0.0 but that the kernel features 12.0.0, thefore it refuses to run. That means I can't set it up, although I noticed it picked up my logical volumes at kernel level during boot. An observation too: I noticed a prepatch kernel is being used on the CD. Wouldn't it be safer just to pick a stock release kernel? And (although I don't use JFS so it bothers me not) doesn't the inclusion of XFS patches make it impossible to setup a system safely from scratch with JFS? Aside from that, nice work :-) Hardware ======== Machine: Compaq Deskpro EN/P500 CPU: PIII 500Mhz (Coppermine) Chipset: Intel 440BX RAM: 256Mb PC100
Re: EVMS problem, it transpires that this is the one open bug on EVMS 1.2.1, see http://sourceforge.net/tracker/index.php?func=detail&aid=659452&group_id=25076&atid=383342
reopen if this still exists
Moving these so we can remove the "Install CD" component from "Gentoo Linux". I apologize to everyone for this spam, but according to the bugzilla developers, this is the only reasonable way to do this.
