Re: All filesystems hang under long periods of heavy load (read andwrite) on a filesystem

[Ged Haywood]

Hi Marcelo,

On Mon, 29 Dec 2003, Marcelo Tosatti wrote:

> > there is certainly something nasty in the released 2.4.23 IDE code...
> 
> The attached patch from Daniel Lux should fix it. Its has just been
> applied to the 2.4 BK tree.
> 
> Please try it.

Thanks very much for the patch, I can see how that piece of code could
well be causing the problem.

Between my message and your reply, someone else came up with another
clue for me and I found that enabling DMA for the drive appears if not
to fix the problem at least to avoid it fairly well.  I think using
DMA for most of the transfers makes it less likely that the lockup
will be triggered.  At least this gets me out of the immediate bind.

It's a busy machine and it's a little painful to keep rebooting it.
I'm going to put an identical one together for testing.  When I've
done that I'll test the patch to destruction.

It will be a couple of weeks before I can get to the bench to do it,
but then it should only take a few hours to be sure it's a good fix.
I'll certainly get back to you with the results.

Thanks again for the pointer.

73,
Ged.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/