lkml.org 
[lkml]   [1998]   [Jul]   [29]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: FS Corruption in 2.1.109 (fwd)


On Wed, 29 Jul 1998, Alan Cox wrote:
> >
> > It's been tested. Alan has numbers, and it does not happen only on UP.
> > Note also how he cannot even _enable_ DMA on 2.0.x.
>
> [That one is atypical btw in that respect]
>
> > No wonder 2.0.x doesn't corrupt things, it doesn't even try to use DMA.
>
> The numbers dont back that one up either.

Oh, I mean: "Oh his machine".

I'm sure DMA is enabled on a lot of machines. And DMA works beautifully in
2.1.x on a lot of machines too. It may just be that for some reasons 2.0.x
is less likely to use DMA than 2.1.x is.

For example, remind me what the default configuration is in 2.0.x? I know
that 2.1.x defaulted to having CONFIG_BLK_DEV_IDEDMA set to 'y', but in
2.0.x we have CONFIG_BLK_DEV_TRITON instead (again defaulting to 'y'), but
when the name changed maybe something else changed too?

For example, if I read the 2.0.x code right, then 2.0.x will enable DMA
only when it finds a TRITON chipset or with the specific case of a Promise
Technology IDE Ultra-DMA 33.

While the newer code seems to enable DMA whenever the drive reports
supporting it.

No, I didn't read the code all that clearly, so yes, please do set me
straight.

Anyway, the fact that on his machine he cannot even enable DMA at least to
me indicates that maybe DMA has NOT gotten all that much testing in 2.0.x,
because it only got enabled on a subset of the machines.

Or maybe not.

> I also don't believe it to be hardware. Some of these folks have Win98
> and FreeBSD running reliably in UDMA mode - two folks have gone and stress
> tested on both systems without problems.

Note that it could easily be some other timing thing that triggers it.
Which is definitely supported by the thing that it seems to have appeared
at least for one person between 2.1.101 and 2.1.103 even though there are
no real changes wrt either disk driver or filesystem in those patches as
far as I can see.

Linus


-
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.altern.org/andrebalsa/doc/lkml-faq.html

\
 
 \ /
  Last update: 2005-03-22 13:43    [from the cache]
©2003-2011 Jasper Spaans