Mailing List Archive

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thu, 5 May 2005, Leif Sawyer wrote:

>
> Was exploring the differences between -r6 and -r7 in terms
> of the sparc changes, and noticed that the recent sunsu patch
> didn't make it in.
>
> the patch is from 4/20 by Dave Miller, and was reposted to G-S on 4/22
> by Ferris. [re: Weird mouse behaviour]
>

I have put 2.6.11-r6 (+irq patch) onto one U60-SMP, and 2.6.11-r7 (+
irq patch) onto one U60-SMP, one U2-SMP. Weird mouse behavior is present
on 33% of these systems (U60, 2.6.11-r7), which is an improvement from the
100% that I saw last fall. (Weird = must unplug to get it seen).

> Also, what was the outcome of the kernel/irq.c patch (should_forward=0) ?
> Is that something that we should be testing more 'actively'?
>

In my opinion, yes. It might help and shouldn't hurt. So far, the
systems are staying up, but I haven't really stress tested them beyond
some portage work and a run of the daily cron 'slocate' script.

Also, 'cvs -z0 up' for cvsroot/gentoo86 (the Gentoo cvs portage tree) ran
successfully on the U2 while at the same time a different system (SS20)
was hitting it with distcc compilation requests.

Both the slocate and the 'cvs up' had been pretty reliable "killer apps"
for the U2 when I was testing 2.6.6/7 kernels.

> I only ask because I'm gearing up to build -r7 and test it out, but
> before I do, I wanted to verify these last couple of patches and
> apply or not, depending on their status.
>
> Leif
>
I don't know status of the fix-the-mouse-timing patch. I recommend that
you try the (3 line change) irq.c patch. squash will have better
information on that one, though.
>

Regards,
Ferris
- --
Ferris McCormick (P44646, MI) <fmccor@gentoo.org>
Developer, Gentoo Linux (Sparc)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)

iD8DBQFCenYJQa6M3+I///cRAlFDAKCZDCl8JWMemxrKm5ncn1VEN9T7iwCgpTH1
x9kJ2pyIhgsBzUU565afpGE=
=Lpaq
-----END PGP SIGNATURE-----
--
gentoo-sparc@gentoo.org mailing list

On Thu, 2005-05-05 at 10:06 -0800, Leif Sawyer wrote:
> the patch is from 4/20 by Dave Miller, and was reposted to G-S on 4/22
> by Ferris. [re: Weird mouse behaviour]

I use a USB mouse, so I haven't tested it out and haven't merged it into
g-d-s. If I get any positive feedback on its effect, I'll include it in
the next 2.6.11-rX, but I think 2.6.12 might be out before we need one.

> Also, what was the outcome of the kernel/irq.c patch (should_forward=0) ?
> Is that something that we should be testing more 'actively'?

This has greatly improved my u2's stability when smp and has been merged
for the next g-d-s-2.6.11 release. I still have an oops when halting
the system, but I haven't had time to look into that more closely. This
patch is also merged into mainline.

> I only ask because I'm gearing up to build -r7 and test it out, but
> before I do, I wanted to verify these last couple of patches and
> apply or not, depending on their status.

If you need their functionality, then feel free to apply them manually.

--Jeremy