Messages in this thread | | | Date | Thu, 23 Mar 2000 09:17:41 -0800 | From | Simon Kirby <> | Subject | Re: [2.3.99pre2] [CORRUPTION] Doh! Corruption problems again... |
| |
On Tue, Mar 21, 2000 at 08:37:53AM -0800, Simon Kirby wrote:
> Well, I've traced it down to between 2.3.48pre4 and 2.3.48: > > 2.3.44-pre8 : GOOD > 2.3.47 : GOOD > 2.3.48-pre1 : Can't mount hdb partitions > 2.3.48-pre2 : Can't mount hdb partitions > 2.3.48-pre3 : Doesn't boot > 2.3.48-pre4 : GOOD (heavily tested) > 2.3.48 : BAD
Update: I just reproduced the corruption problem in 2.3.48pre4! It's a lot harder to reproduce than in 2.3.48, and I could only reproduce it with all drives in PIO mode (DMA mode doesn't seem to show it). I might be able to reproduce it in DMA mode if I wait a few hours, but takes a few runs in PIO mode as it is before it shows anything (usually).
If I apply the IDE changes between 2.3.48pre4 and 2.3.48 on top of 2.3.48pre4, the problem does not seem to get as easy to reproduce as in vanilla 2.3.48, so it must be something else. I will try some other suspicious parts.
Hmm, this is annoying. Now I'm wondering if the plugging fixes that went into 2.3.46 didn't fix everything...
Also, Andre: I tried filling an entire partition with "HelloWorld" (I used "yes 'HelloWorld' > bigfile" until the drive filled up and yes exited with an error) and then tried to reproduce the corruption to see if it was actually that the drive wasn't writing all of the blocks...but I didn't see _any_ traces of "HelloWorld" anywhere, just other kernel file data munged in with the kernel files I was copying.
Simon-
[ Stormix Technologies Inc. ][ NetNation Communcations Inc. ] [ sim@stormix.com ][ sim@netnation.com ] [ Opinions expressed are not necessarily those of my employers. ]
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |