Jaunty - experimenting with static pata/sata drivers

As a final step towards laying the foundation for discontinuing the use
of initramfs in the most common boot scenarios, I've been experimenting
with building all PATA/SATA drivers into the kernel. It works on the few
machines that I've tested, so I'm looking for a wider audience for
further testing. I have binary debs at

http://people.ubuntu.com/~rtg/2.6.28-4.6-build-in-ata

In order to boot without initramfs you'll have to edit
/boot/grub/menu.lst to comment out the initrd statement, and also change
the kernel root to reference your root device. For example, my menu.lst
clause looks like this:

title           Ubuntu jaunty (development branch), kernel 2.6.28-4-generic
root            (hd0,0)
kernel          /boot/vmlinuz-2.6.28-4-generic root=/dev/sda1 ro

quiet
#kernel         /boot/vmlinuz-2.6.28-4-generic
root=UUID=86186ea6-6c7f-481a-9d3c-ed8011d58e42 ro
#initrd         /boot/initrd.img-2.6.28-4-generic

I imagine there is a better way to do this. Passing 'noinitrd' as a boot
parameter didn't appear to work.

If you want to build your own kernel, I have a build-in-ata topic branch
in my Jaunty git repository at kernel.ubuntu.com. For example,

git clone git://kernel.ubuntu.com/rtg/ubuntu-jaunty
git checkout -b build-in-ata origin/build-in-ata

Please pull db139d606ea25e059 from drm-(intel)-next

 	Hi!

   Please pull this for jaunty, it'll allow building 
xserver-xorg-video-intel against the kernel drm headers. After this we 
should be able to drop the headers from libdrm-dev.

---------- Forwarded message ----------
Date: Thu, 18 Dec 2008 09:02:28 -0800 (PST)
Subject: [Bug 19132] 2.5.99.1 fails to build using drm headers from 2.6.28

http://bugs.freedesktop.org/show_bug.cgi?id=19132

Eric Anholt <eric <at> anholt.net> changed:

            What    |Removed                     |Added
----------------------------------------------------------------------------
            Severity|critical                    |blocker
              Status|NEW                         |RESOLVED
          Resolution|                            |FIXED

--- Comment #3 from Eric Anholt <eric <at> anholt.net>  2008-12-18 09:02:26 PST ---
In drm-intel-next:

commit db139d606ea25e0590373d5ce6bfc42ce317a61c
Author: Eric Anholt <eric <at> anholt.net>
Date:   Wed Dec 17 22:32:14 2008 -0800

     drm/i915: Add missing userland definitions for gem init/execbuffer.

Re: Use modular config for AGP settings - otherwise fglrx wil

realKano <at> directbox.com wrote:
>> I reverted all of the i386 AGP drivers back to being built as modules,
>> but not CONFIG_AGP as that affects DRI. Please build and test to make
>> sure this change is sufficient.
> 
> Now I know, why it did not work...
> 
> [   77.626663] fglrx: module license 'Proprietary. (C) 2002 - ATI Technologies, Starnberg, GERMANY'
taints kernel.
> [   77.659673] [fglrx]   vendor: 1002 device: 9587 count: 1
> [   77.660383] [fglrx] ioport: bar 1, base 0xd000, size: 0x100
> [   77.660400] [fglrx] Maximum main memory to use for locked dma buffers: 1645 MBytes.
> [   77.660818] [fglrx] PAT is enabled successfully!
> [   77.660836] [fglrx:firegl_init_module] *ERROR* firegl_stub_register failed
> 
> In order to make fglrx work you usually have to unload radeon + drm module. BUT if you compile DRM statically
that will never work.
> 

http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-jaunty.git;a=commit;h=7e9c4e3936e05214eb6cd615da07e5b0edfb0399

--

-- 
Tim Gardner tim.gardner <at> canonical.com

Re: Please pull db139d606ea25e059 from drm-(intel)-next

Timo Aaltonen wrote:
>  	Hi!
> 
>    Please pull this for jaunty, it'll allow building 
> xserver-xorg-video-intel against the kernel drm headers. After this we 
> should be able to drop the headers from libdrm-dev.
> 
> 
> ---------- Forwarded message ----------
> Date: Thu, 18 Dec 2008 09:02:28 -0800 (PST)
> Subject: [Bug 19132] 2.5.99.1 fails to build using drm headers from 2.6.28
> 
> http://bugs.freedesktop.org/show_bug.cgi?id=19132
> 
> 
> Eric Anholt <eric <at> anholt.net> changed:
> 
>             What    |Removed                     |Added
> ----------------------------------------------------------------------------
>             Severity|critical                    |blocker
>               Status|NEW                         |RESOLVED
>           Resolution|                            |FIXED
> 
> 
> 
> 
> --- Comment #3 from Eric Anholt <eric <at> anholt.net>  2008-12-18 09:02:26 PST ---
> In drm-intel-next:
> 
> commit db139d606ea25e0590373d5ce6bfc42ce317a61c

Uploaded Jaunty 2.8.28-4.6

https://launchpad.net/ubuntu/+source/linux

This upload should address
https://bugs.launchpad.net/ubuntu/+source/libdrm/+bug/308387 (libdrm
compile problems).

rtg
--

-- 
Tim Gardner tim.gardner <at> canonical.com

Tim Gardner <timg <at> tpi.com>

Jaunty config changes: Build AGP/DRM as modules

From 23d0a09b06e5013a06bc701ece4284947c94bde6 Mon Sep 17 00:00:00 2001
From: Tim Gardner <tim.gardner <at> canonical.com>
Date: Sat, 3 Jan 2009 12:29:46 -0700
Subject: [PATCH] UBUNTU: Build all i386/amd64 AGP/DRM components as modules.

Bug: #312721

I guess all of these components have to be built as modules in order
to avoid breaking ATI graphics. While none of these components are
boot essential, loading them as modules will incrementally slow the
boot process.

Signed-off-by: Tim Gardner <tim.gardner <at> canonical.com>
---
 debian/config/amd64/config |    8 ++++----
 debian/config/i386/config  |    2 +-
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/debian/config/amd64/config b/debian/config/amd64/config
index 413781f..a4c735e 100644
--- a/debian/config/amd64/config
+++ b/debian/config/amd64/config
 <at>  <at>  -67,9 +67,9  <at>  <at>  CONFIG_AF_RXRPC=m
 # CONFIG_AF_RXRPC_DEBUG is not set
 CONFIG_AGP=y
 CONFIG_AGP_AMD64=y
-CONFIG_AGP_INTEL=y
-CONFIG_AGP_SIS=y
-CONFIG_AGP_VIA=y
+CONFIG_AGP_INTEL=m

Intel Firmware

In this bug, someone has pointed out that the Intel firmware requires an
EULA that we are not presenting:

     https://bugs.launchpad.net/ubuntu/+source/linux-firmware/+bug/179139

I am sure this is under control as we are so closly in touch with Intel.
Does anyone have any background on this?

-apw

Re: boot speedup

(``-_-´´) -- BUGabundo wrote:
> Since kernel 2.6.28-4-generic suffered many changes to speedup the
> boot, I took my last 2.6.28-3-generic and the 1st 2.6.28-4-generic
> bootchart pics, to show you guys. 
> http://fileland.bugabundo.net/galeria/v/BUGabundo/Linux/bootchart/jaunty-20081227-1.png.html
>  
> http://fileland.bugabundo.net/galeria/v/BUGabundo/Linux/bootchart/jaunty-20081228-1.png.html
> 
> 
> I haven't noticed many differences, other then some udev changing
> places, and a new sleep().
> 
> More bootcharts here 
> http://fileland.bugabundo.net/galeria/v/BUGabundo/Linux/bootchart
> 

Substantial reduction of boot times will not be had until there is
sufficient support in the kernel to allow booting without initramfs.

See Martin's comment:
https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/311730/comments/4

rtg
-- 
Tim Gardner tim.gardner <at> canonical.com

--

-- 
kernel-team mailing list
kernel-team <at> lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team

RE: Intel Firmware

We need to release the EULA together with the binary firmware in Ubuntu images to resolve this legal issue. 
Packaging the license files for ipw* and iwl* firmware in the "linux-firmware" is a proper solution.

- Yingying 

-----Original Message-----
From: kernel-team-bounces <at> lists.ubuntu.com [mailto:kernel-team-bounces <at> lists.ubuntu.com] On
Behalf Of Andy Whitcroft
Sent: 2009年1月4日 20:02
To: Ubuntu Kernel Team
Subject: Intel Firmware

In this bug, someone has pointed out that the Intel firmware requires an
EULA that we are not presenting:

     https://bugs.launchpad.net/ubuntu/+source/linux-firmware/+bug/179139


I am sure this is under control as we are so closly in touch with Intel.
Does anyone have any background on this?

-apw

-- 
kernel-team mailing list
kernel-team <at> lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team

--

-- 
kernel-team mailing list
kernel-team <at> lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team

Re: testing area for drivers/modules not upstream (was Dazuko Module Inclusion for Jaunty)

Olá Pete e a todos.

On Wednesday 17 December 2008 15:43:11 Pete Graner wrote:
> This is an area in the upstream kernel where non-upstreamed drivers can go. Info is here: http://lkml.org/lkml/2008/6/10/329

Nol is the maintainer of the Genesys Logic Webcam driver.
I've been testing it for a while in ibex and now in jaunty, and it is working quite well (as opposed to not
working at all), but it is still under heavy improvement.
Would it be appropriated to have the trunk version of the driver in "that" area, so it can have a large set of
users testing it while the developing of Jaunty for later inclusion upstream?

Genesys Logic Webcam 05e3:0503 Not Working  https://bugs.launchpad.net/bugs/215604
http://ubuntuforums.org/showthread.php?t=869263

--

-- 
Hi, I'm BUGabundo, and I am Ubuntu (whyubuntu.com)
(``-_-´´)	http://LinuxNoDEI.BUGabundo.net
Linux user #443786    GPG key 1024D/A1784EBB
My new micro-blog  <at>  http://BUGabundo.net
ps. My emails tend to sound authority and aggressive. I'm sorry in advance. I'll try to be more assertive as
time goes by...

Olá Pete e a todos.

On Wednesday 17 December 2008 15:43:11 Pete Graner wrote:
> This is an area in the upstream kernel where non-upstreamed drivers can go. Info is here: http://lkml.org/lkml/2008/6/10/329

Nol is the maintainer of the Genesys Logic Webcam driver.
I've been testing it for a while in ibex and now in jaunty, and it is working quite well (as opposed to not