Re: buffer cache patches - anybody willing to summarize?

Jeff Wiegley (
Mon, 04 Aug 1997 05:49:35 -0700


You certainly know more than I and if you think they should be added
then I would agree. (besides I like to see linux running on everything
under the sun too ;-)

I just thought that the newer drivers are more of an enhancement/feature
fix than a bug fix and I think it would be a good idea to keep the work
load for the new 2.0.31 guy (whomever that may be) to a minimum.

Yes, the shafted networking would certainly be a .31 candidate.

There was a lot of talk about a problem/patch web site last week. Did
anything come of that or are we still discussing the possibility of a
patch maintainer? I'm like everyone else: I know enough to write a
non-interrupt driver character device driver but beyond that I'm
clueless. I would be willing to maintain a patch site and
collect/organize patches though.

I think I could squeeze a few hours a week into doing this.
I might even be able to provide a machine to do it on here at work if
the bosses agreed to loan me the bandwidth which probably shouldn't be
too high. It does look though that already does this to
some degree.

- Jeff

Alan Cox wrote:
> > Linus: just fix the buffer cache problem please! This seems to be the
> > bug that is causing all the lock-ups and memory oops and such.
> No. The fact the networking is totally shafted also causes a lot of oopses
> > 1) ask the driver maintainer for the latest stable driver patch against
> > 2.0.31 when it is released, or...
> The new tulip should be in the 2.0.31 kernel.
> > 2) go purchase hardware that is known to be supported.
> Not possible, the only chip that the old old driver supported is
> replaced now.
> > 2.0 tree. Security and fatal bug fixes *only*. other than that 2.0
> > supports what it supports and you should wait for 2.2/3.0 to include new
> > hardware support.
> Both the AHA29xx and tulip are pretty fatal if you have these boards. Neither
> of these are "new things" they are bug fixes. Also just about everyone with
> a tulip is already using the new driver.
> Alan