Re: FS Corruption in 2.1.109

Christopher Seawood (
Fri, 24 Jul 1998 11:31:19 -0700 (PDT)

On Wed, 22 Jul 1998, Linus Torvalds wrote:

> The tests are roughly:
> - can you disable IDE-DMA with hdparm? Does the problem go away?

I disabled it with hdparm but the problems didn't seem to go away. It
could be because half the time, the drive would not keep dma turned off
between reboots even though I had set the keepsettings flag to 1.

> - can you try to pinpoint which kernel this started at for you?

The first kernel I noticed it was 2.1.103. 2.1.102 didn't build properly
w/o patches so I never bothered with it. 2.1.101 has always worked fine.

> - use an older gcc instead of the egcs-1.0.3a compiler that you seem to
> be using. Just to get one variation out of the picture.

I compiled 2.1.109 against gcc and it did not seem to make a

> I don't think this is a filesystem problem, or I'd have been getting a lot
> more reports about it (and judging by how easily it happens to you I
> should have seen it myself). But there have been occasional grumblings
> about IDE DMA..

I disabled IDE-DMA in the kernel and everything seems to be working fine
now. Multiple rm/untar tests yielded no errors. I've compiled mozilla &
wine a few times and installed rh5.1 on the spare drive (while still
running 2.1.109). Nothing. Incidentally, a guy in our local LUG has
experienced the same problem with his Tyan TomcatIII mb under 2.1.108. He
said he disabled ide-dma (but didn't say where...kernel or drive) and his
problems went away.

- cls

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to
Please read the FAQ at