fs corruption with 2.4.20 IDE+md+LVM
Quoted 3 lines
I observed filesystem corruption on my home workstation recently. I was running kernel 2.4.20 (built myself with gcc 2.95.4), and ext3 with the default journaling mode (ordered?).
Hello,
Same problem here. I have software raid-1 on 2 IDE Seagate 80G, kernel 2.4.20aa1 built with gcc-3.2, all filesystems are ext2, no LVM. FS corruption after running Cerberus test for about 8 hours.
Quoted 2 lines
I will also point out that kernel 2.4.20-ac1 and 2.4.21-pre6 will not boot on my machine; they kernel panic when detecting my IDE devices.
I can confirm. Kernel 2.4.21-pre2 does not boot from a RAID device (/dev/md0).
Signature · 1 line
D.V.
Quoted 9 lines · Dmitry Volkoff
Quoted 3 lines
I observed filesystem corruption on my home workstation recently. I was running kernel 2.4.20 (built myself with gcc 2.95.4), and ext3 with the default journaling mode (ordered?).
Hello,
Same problem here. I have software raid-1 on 2 IDE Seagate 80G, kernel 2.4.20aa1 built with gcc-3.2, all filesystems are ext2, no LVM. FS corruption after running Cerberus test for about 8 hours.
glad to know I'm not the only one.
someone pointed out to me in a private e-mail, that the corruption may be related to my VIA KT133 chipset. (they had a similar problem).
Quoted 5 lines
Quoted 2 lines
I will also point out that kernel 2.4.20-ac1 and 2.4.21-pre6 will not boot on my machine; they kernel panic when detecting my IDE devices.
I can confirm. Kernel 2.4.21-pre2 does not boot from a RAID device (
/dev/md0).
sorry about the thinko in my mail. I meant 2.4.21-pre1. Glad to know I'm not crazy, but hopefully confirmation means it'll get fixed before 2.4.21-final.
<flamebait>
maybe I just missed the arguments since I wasn't reading LKML at the time; but *why* is IDE being revamped in the middle of a "stable" kernel series? however better it may be, I don't regard the existing situation as being bad enough to justify the risk.
</flamebait>
Carl Soderstrom.
Signature · 3 lines
Systems Administrator Real-Time Enterprises www.real-time.com
Quoted 6 lines · Carl Wilhelm Soderstrom
<flamebait>
maybe I just missed the arguments since I wasn't reading LKML at the time; but *why* is IDE being revamped in the middle of a "stable" kernel series? however better it may be, I don't regard the existing situation as being bad enough to justify the risk.
</flamebait>
You are reporting problems in 2.4.20. 2.4.20 doesn't have the revamped IDE...
The IDE is getting updated because
- Lots of new controllers dont work with the old code
- Lots of LBA48 problems exist with the older code
- SATA is right out with the older code
- Several existing controllers have weird bugs with the older code
I'd much prefer we didn't have to update the IDE too 8)
Quoted 1 line · Alan Cox
You are reporting problems in 2.4.20. 2.4.20 doesn't have the revamped IDE...
I know. which is why I put that comment after the section of my mail regarding md bugs in 2.4.21
Quoted 8 lines
The IDE is getting updated because
- Lots of new controllers dont work with the old code
- Lots of LBA48 problems exist with the older code
- SATA is right out with the older code
- Several existing controllers have weird bugs with the older codeI'd much prefer we didn't have to update the IDE too 8)
ok. I didn't see an extensive discussion of this on any of the kernel-digest forums (kerneltrap, kernel traffic).
I'll trust that you're doing the right thing, and try to avoid stepping in any other flamebait. ;)
Carl Soderstrom.
Signature · 3 lines
Systems Administrator Real-Time Enterprises www.real-time.com