From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0 Description of problem: Here we go, this will be interesting.. Yesterdays rawhide ( 10/11/2004), along with many rawhides for the past few weeks. Following the instruction listed on colins charles site ( http://www.bytebot.net/geekdocs/ibook/fedorappc.html ), the system was able to be installed. At the next fsck of the system, it comes up with an error. This can be forced on the first install (by jumping to a terminal before hitting "reboot", and here are the "typed" results. e2fsck /dev/hda2 e2fsck 1.35 (28-Feb-2004) / was not cleanly unmounted, check forced. Pass 1: Checking inodes, blocks, and sizes Inode 7 has illegal blocks(s). Clear<y> y Illegal block #262156 (4294967295) in inode 7. CLEARED. Illegal block #252177 (4294967295) in inode 7. CLEARED. .... Too many illegal blocks in inode 7. Clear inode<y> If i clear the inode, this pattern repeats, and it seems if I continue to answer yes, or force fsck to finish with -y, it will end up ruining access to files on the disk. I have reproduced this problem on 9 different installs, with different disk layouts, all ext3 (now that I think about it, i should maybe try ext2 or xfs) I have tried a different hard drive without any luck, I dont think this is hardware issue. This hardware is an ibook G4 1Ghz machine. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. Install FC rawhide on ppc 2. Fsck the system. 3. watch the burn Actual Results: fsck errors, as per above. Expected Results: fsck to complete, without bringing the system into a hellish nosedive. Additional info: cant reproduce on x86, no problem on warty ubuntu, im available for testing ! I'd just love to get this fixed.
Funny, I had this plaguing me when I performed an install using kernel-2.6.8-1.541.ppc.rpm. This was a bad kernel release with the ext3 resize patch doing weird things I installed FC3 on the Mac, clean install (i.e. removing /, /home partitions), and so far, knocking wood, nothing bad has happened. I figure that the .541 kernel borked ext3 even from creation, and no matter what rawhide I upgraded to afterward, it was just going to suck for me
Wade, is this still evident with FC-3? I'm on the tree that we're about to release, and I can't quite notice this. This is 2.6.9-1.667
Just re-installed the "just released" public test for FC3, problem has disappeared This kernel is 2.6.9-1.700 from DWMW2's repository.
this was actually a problem in e2fsprogs with online resizing, which only manifested on ppc. There should be an update out that fixes this.
Well, I guess its "closed" then ?