kernel.git
10 years agoMERGE-via-mokopatches-tracking-hist-s3c64xx-dma-support stable-tracking github/stable-tracking
merge [Tue, 24 Feb 2009 01:33:05 +0000 (01:33 +0000)]
MERGE-via-mokopatches-tracking-hist-s3c64xx-dma-support
mokopatches-tracking-hist top was s3c64xx-dma-support / 2515f9a1d53d19b1e61d639875aedcbe7929666e ... parent commitmessage:
From: Ben Dooks <ben@simtec.co.uk>
S3C64XX: DMA support

Add support for the DMA blocks in the S3C64XX series
of CPUS, which are based on the ARM PL080 PrimeCell
system.

Unfortunately, these DMA controllers diverge from the
PL080 design by adding another DMA controller register
and configuration for OneNAND.

Signed-off-by: Ben Dooks <ben@simtec.co.uk>

10 years agoMERGE-via-mokopatches-tracking-hist-MERGE-via-master-MERGE-via-master-hist-1232625318...
merge [Fri, 6 Feb 2009 00:17:37 +0000 (00:17 +0000)]
MERGE-via-mokopatches-tracking-hist-MERGE-via-master-MERGE-via-master-hist-1232625318-1233879011
mokopatches-tracking-hist top was MERGE-via-master-MERGE-via-master-hist-1232625318-1233879011 / 1be1b01373f572a02c6f1f99863c8c11ed2f9f5b ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-master-MERGE-via-master-hist-1232625318

master top was MERGE-via-master-hist-1232625318 / dd4b117123ae66451695810017eb72fbdfc05df5 ... parent commitmessage:
From: merge <null@invalid>
MERGE-master-patchset-edits

10 years agoGTA02 build depends on S3C2410_CLOCK
Sven Rebhan [Wed, 28 Jan 2009 18:00:33 +0000 (18:00 +0000)]
GTA02 build depends on S3C2410_CLOCK

Signed-off-by: Sven Rebhan <odinshorse@googlemail.com>

10 years agoMERGE-via-mokopatches-tracking-hist-fix-stray-endmenu-patch
merge [Thu, 22 Jan 2009 13:47:43 +0000 (13:47 +0000)]
MERGE-via-mokopatches-tracking-hist-fix-stray-endmenu-patch
mokopatches-tracking-hist top was fix-stray-endmenu-patch / 3630e0be570de8057e7f8d2fe501ed353cdf34e6 ... parent commitmessage:
From: Andy Green <andy@openmoko.com>
fix-stray-endmenu.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoMERGE-via-mokopatches-tracking-hist-MERGE-via-master-MERGE-via-master-hist-1232625318...
merge [Thu, 22 Jan 2009 12:08:55 +0000 (12:08 +0000)]
MERGE-via-mokopatches-tracking-hist-MERGE-via-master-MERGE-via-master-hist-1232625318-1232625432
mokopatches-tracking-hist top was MERGE-via-master-MERGE-via-master-hist-1232625318-1232625432 / c25f214e3311c85714ad07a722c67de0aa5e2cc8 ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-master-MERGE-via-master-hist-1232625318

master top was MERGE-via-master-hist-1232625318 / dd4b117123ae66451695810017eb72fbdfc05df5 ... parent commitmessage:
From: merge <null@invalid>
MERGE-master-patchset-edits

10 years agoMERGE-via-mokopatches-tracking-hist-MERGE-via-master-MERGE-via-master-hist-1232618228...
merge [Thu, 22 Jan 2009 10:49:39 +0000 (10:49 +0000)]
MERGE-via-mokopatches-tracking-hist-MERGE-via-master-MERGE-via-master-hist-1232618228-1232619433
mokopatches-tracking-hist top was MERGE-via-master-MERGE-via-master-hist-1232618228-1232619433 / c7c5b42b74a7f8931ff9292cc839e1731479e1c5 ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-master-MERGE-via-master-hist-1232618228

master top was MERGE-via-master-hist-1232618228 / 83503deeba7defae0640c20ae68a1de14a1e567f ... parent commitmessage:
From: merge <null@invalid>
MERGE-master-patchset-edits

10 years agoGlamo: don't break kernel build if disabled
Werner Almesberger [Wed, 21 Jan 2009 19:54:28 +0000 (19:54 +0000)]
Glamo: don't break kernel build if disabled

If we disable CONFIG_MFD_GLAMO, a dependency on glamo_lcm_reset from
arch/arm/mach-s3c2440/mach-gta02.c:gta02_jbt6k74_reset cannot be
resolved.

There are many ways to fix this - this one does it with a bit of a
swagger ;-)

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agoAdd TPO L1K002 LCM device driver.
Matt Hsu [Tue, 20 Jan 2009 10:43:14 +0000 (10:43 +0000)]
Add TPO L1K002 LCM device driver.

This device is adopted as LCM of gta03.
The patch provides the initialization for this LCM driver IC.

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoS3C64XX: Add SPI-GPIO based driver.
Matt Hsu [Tue, 20 Jan 2009 10:43:11 +0000 (10:43 +0000)]
S3C64XX: Add SPI-GPIO based driver.

This driver drives SPI bus over generic GPIO on s3c64xx.

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoupdate GTA02 default configurations
Werner Almesberger [Mon, 19 Jan 2009 20:30:21 +0000 (20:30 +0000)]
update GTA02 default configurations

Add the settings introduced by the previous two patches to the default
configurations.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agoAR6k: don't reset on shutdown
Werner Almesberger [Mon, 19 Jan 2009 20:30:19 +0000 (20:30 +0000)]
AR6k: don't reset on shutdown

The AR6k driver can explicitly reset the module when shutting down the
device, which takes about two seconds and seems to be rather useless,
considering that we disable the entire SDIO function moments later.

This patch adds the configuration option CONFIG_AR6000_WLAN_RESET that
has to be set to maintain the previous default of resetting on shutdown.
The option is turned off by default.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agoAR6k: don't retrieve the debug log
Werner Almesberger [Mon, 19 Jan 2009 20:30:17 +0000 (20:30 +0000)]
AR6k: don't retrieve the debug log

The AR6k firmware maintains a log of debugging events that are retrieved
by the host on various occasions, including driver removal. Retrieval of
this data is very slow and seems to have little value.

This patch adds a new configuration option CONFIG_AR6000_WLAN_DEBUG to
enable retrieval of the debugging log. This option is turned off by
default.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agofix-pcap7200-touchscreen-filter-move.patch
Andy Green [Mon, 19 Jan 2009 01:09:04 +0000 (01:09 +0000)]
fix-pcap7200-touchscreen-filter-move.patch

pcap7200 driver also uses our filter chain stuff now, so it needed a little
fixup when the includes moved for upstream

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoFix touchscreen filter includes
Nelson Castillo [Mon, 19 Jan 2009 00:32:00 +0000 (00:32 +0000)]
Fix touchscreen filter includes

Fix #includes to make the kernel compile again.

Signed-off-by: Nelson Castillo <arhuaco@freaks-unidos.net>

10 years agoMove touchscreen filter includes
Nelson Castillo [Mon, 19 Jan 2009 00:31:57 +0000 (00:31 +0000)]
Move touchscreen filter includes

This patch just moves the touchscreen filter header files
around, with no modifications. The kernel will not compile
after this patch and another patch will be needed to modify
the actual #includes. I read it's better not to modify files
when you move them around.

Signed-off-by: Nelson Castillo <arhuaco@freaks-unidos.net>

10 years agoMiscellaneous touchscreen filter cleanup
Nelson Castillo [Mon, 19 Jan 2009 00:31:54 +0000 (00:31 +0000)]
Miscellaneous touchscreen filter cleanup

This is a cleanup patch fixing simple mistakes and coding conventions.
It also applies feedback we got from upstream.

Signed-off-by: Nelson Castillo <arhuaco@freaks-unidos.net>

10 years agouse msleep instead of mdelay in AR6k
Werner Almesberger [Mon, 19 Jan 2009 00:31:51 +0000 (00:31 +0000)]
use msleep instead of mdelay in AR6k

The AR6000 driver uses A_MDELAY with sometimes long delays at several
places. A_MDELAY is just another name for "mdelay", which is a busy
loop. All these delays appear to occur in a context where we can sleep,
so it's better to use "msleep" instead.

This includes the "big fat mdelay(2000)" Balaji found in the suspend
path:
http://lists.openmoko.org/pipermail/openmoko-kernel/2008-December/007171.html

Note that only the ones in DevPollMboxMsgRecv and ar6000_reset_device
are actually invoked in normal use, and only during initialization and
shutdown, respectively. So the impact on overall system performance is
small.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agoselecting CONFIG_S3C2410_PWM
Werner Almesberger [Mon, 19 Jan 2009 00:31:48 +0000 (00:31 +0000)]
selecting CONFIG_S3C2410_PWM

We need CONFIG_S3C2410_PWM for the GTA02, but we don't explicitly
select it, which can cause build failures.

This patch makes CONFIG_MACH_NEO1973_GTA02 select CONFIG_S3C2410_PWM.

I'm not sure if this is the right place, though. mach-s3c2410/pwm.c
was done for Openmoko but it looks generic enough that one may well
consider it to be part of the general machine code. In this case,
CONFIG_CPU_S3C2440 should select CONFIG_S3C2410_PWM.

Ben, may I ask you for guidance on this issue ? If it's okay to
select CONFIG_S3C2410_PWM on CONFIG_CPU_S3C2440, then I'll post a
revised patch.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agorfkill support for the AR6000 driver
Werner Almesberger [Fri, 16 Jan 2009 11:06:21 +0000 (11:06 +0000)]
rfkill support for the AR6000 driver

This patch adds rfkill support to the AR6000 driver. The driver does
not directly implement an rfkill device but uses the help of a special
platform device, such that the latter can retain rfkill state when the
AR6k driver is removed (e.g., because of suspend).

If an attempt is made to bring the driver up (module load, bind, or
resume) while rfkill is blocking, only the driver's data structures
are initialized, but function activation and most of the rest of the
setup is deferred until the rfkill block is removed.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agoGTA02-specific rfkill framework for WLAN
Werner Almesberger [Fri, 16 Jan 2009 11:06:18 +0000 (11:06 +0000)]
GTA02-specific rfkill framework for WLAN

The patch extends the GTA02 WLAN power management platform device to
register an rfkill handler and to provide a "dock" to which the AR6k
driver can connect to pick up rfkill state and transitions.

We need this indirection because suspend/resume causes removal of the
driver, which therefore cannot retain rfkill state on its own.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agoAdd id_table in the i2c driver struct of lp5521.
Matt Hsu [Tue, 13 Jan 2009 08:38:21 +0000 (08:38 +0000)]
Add id_table in the i2c driver struct of lp5521.

In new style i2c client, i2c driver should have id_table to match its device id.
Otherwise, the attachment of i2c driver would be failed.

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoReturn-Path: <openmoko-kernel-bounces@lists.openmoko.org>
Andy Green [Mon, 12 Jan 2009 17:39:02 +0000 (17:39 +0000)]
Return-Path: <openmoko-kernel-bounces@lists.openmoko.org>
Received: from mail.openmoko.org ([unix socket])
by mail.openmoko.org (Cyrus v2.1.18-IPv6-Debian-2.1.18-5.1) with LMTP; Mon, 12 Jan 2009 17:42:13 +0000
X-Sieve: CMU Sieve 2.2
Return-path: <openmoko-kernel-bounces@lists.openmoko.org>
Received: from sita.openmoko.org ([88.198.124.203])
by mail.openmoko.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32)
(Exim 4.63)
(envelope-from <openmoko-kernel-bounces@lists.openmoko.org>)
id 1LMQnk-00023C-W3
for andy@imap.openmoko.org; Mon, 12 Jan 2009 17:42:13 +0000
Received: from localhost ([127.0.0.1] helo=sita.openmoko.org)
by sita.openmoko.org with esmtp (Exim 4.63)
(envelope-from <openmoko-kernel-bounces@lists.openmoko.org>)
id 1LMQnH-00021T-Ub; Mon, 12 Jan 2009 18:41:43 +0100
Received: from aakash.openmoko.org ([124.219.5.209])
by sita.openmoko.org with esmtp (Exim 4.63)
(envelope-from <matt_hsu@openmoko.org>)
id 1LMQnC-00021F-QB; Mon, 12 Jan 2009 18:41:42 +0100
Received: from [172.16.23.143] (helo=abacus-om.tw.openmoko.com
ident=Debian-exim)
by aakash.openmoko.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32)
(Exim 4.63) (envelope-from <matt_hsu@openmoko.org>)
id 1LMQhU-000398-Ik; Tue, 13 Jan 2009 01:35:54 +0800
Received: from matt by abacus-om.tw.openmoko.com with local (Exim 4.69)
(envelope-from <matt_hsu@openmoko.org>)
id 1LMRF1-0006Dl-2R; Tue, 13 Jan 2009 02:10:23 +0800
From: Matt Hsu <matt_hsu@openmoko.org>
To: openmoko-kernel@lists.openmoko.org
Date: Tue, 13 Jan 2009 02:10:23 +0800
Message-Id: <1231783823-23893-1-git-send-email-matt_hsu@openmoko.org>
X-Mailer: git-send-email 1.5.6.5
X-Spam-Checker-Version: SpamAssassin 3.2.3 (2007-08-08) on sita.openmoko.org
X-Spam-Level:
X-Spam-Status: No, score=-2.6 required=5.0 tests=AWL,BAYES_00 autolearn=ham
version=3.2.3
Subject: [PATCH] This patch fixs compiling warnings of pcap7200_ts driver.
Cc: matt_hsu@openmoko.com
X-BeenThere: openmoko-kernel@lists.openmoko.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Discussion regarding the OpenMoko Linux Kernel and boot loader
<openmoko-kernel.lists.openmoko.org>
List-Unsubscribe: <http://lists.openmoko.org/mailman/listinfo/openmoko-kernel>,
<mailto:openmoko-kernel-request@lists.openmoko.org?subject=unsubscribe>
List-Archive: <http://lists.openmoko.org/pipermail/openmoko-kernel>
List-Post: <mailto:openmoko-kernel@lists.openmoko.org>
List-Help: <mailto:openmoko-kernel-request@lists.openmoko.org?subject=help>
List-Subscribe: <http://lists.openmoko.org/mailman/listinfo/openmoko-kernel>,
<mailto:openmoko-kernel-request@lists.openmoko.org?subject=subscribe>
Sender: openmoko-kernel-bounces@lists.openmoko.org
Errors-To: openmoko-kernel-bounces@lists.openmoko.org

In addition, it adds handling in case of any failures after creating sysfs.

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoadd-missing-pcap7200-guts.patch
Matt Hsu [Mon, 12 Jan 2009 09:01:54 +0000 (09:01 +0000)]
add-missing-pcap7200-guts.patch

Patch by hand vs stg add issue...

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoAdd coniditional check for callback function of lp5521 platform_data
Matt Hsu [Sat, 10 Jan 2009 17:34:28 +0000 (17:34 +0000)]
Add coniditional check for callback function of lp5521 platform_data

Avoid to break the support of N80, it's needed to have this check.

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoRemove lp5521 device from drivers/i2c to drivers/leds
Matt Hsu [Sat, 10 Jan 2009 17:34:26 +0000 (17:34 +0000)]
Remove lp5521 device from drivers/i2c to drivers/leds

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoThis patch adds platform_data and callback of suspend/resume
Matt Hsu [Thu, 8 Jan 2009 18:22:11 +0000 (18:22 +0000)]
This patch adds platform_data and callback of suspend/resume

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoRemove machine-dependent included file of lp5521.
Matt Hsu [Thu, 8 Jan 2009 18:22:08 +0000 (18:22 +0000)]
Remove machine-dependent included file of lp5521.

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agointroduce lp5521 driver from linux-omap repository
Matt Hsu [Thu, 8 Jan 2009 18:22:06 +0000 (18:22 +0000)]
introduce lp5521 driver from linux-omap repository

10 years agoadd support for pcap7200 capacitive touch device.
Matt Hsu [Mon, 29 Dec 2008 11:55:59 +0000 (11:55 +0000)]
add support for pcap7200 capacitive touch device.

Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>

10 years agoFix ts_filter_linear sysfs path
Nelson Castillo [Mon, 15 Dec 2008 07:55:15 +0000 (07:55 +0000)]
Fix ts_filter_linear sysfs path

With this patch we can set the calibration data in:

/sys/class/i2c-adapter/i2c-0/0-0073/s3c2440-ts/calibration/[0-6]

We can do it in two ways:

1) send "struct platform_device *pdev" to the filters so that they
   can register sysfs objects.

2) modify the TS driver to add calibration logic, not using the same
   API the filters use but a different one for calibration.

I did prefer the first alternative. Is it OK to trust pdev to the
filters? We could also share pdev's kobject only.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoAdd Linear Filter
Nelson Castillo [Mon, 15 Dec 2008 07:55:12 +0000 (07:55 +0000)]
Add Linear Filter

This filter allows us to modify the coordinates of the
touchscreen output with the calibration values for the device.

It adds a /sys/ts-calibration directory with 7 files (0..6) that
you can use to change the calibration constants.

This path is not correct. Please do not rely on it because we
will change it really soon.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoCheck MAX_TS_FILTER_CHAIN in ts_filter_destroy_chain
Nelson Castillo [Mon, 15 Dec 2008 07:55:09 +0000 (07:55 +0000)]
Check MAX_TS_FILTER_CHAIN in ts_filter_destroy_chain

We need to check for MAX_TS_FILTER_CHAIN so that we
don't overrun the array of ts_filter pointers when we
have MAX_TS_FILTER_CHAIN filters.

- Check the boundary
- Small cleanups

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoMERGE-via-mokopatches-tracking-hist-MERGE-via-master-hist-1229249926
merge [Sun, 14 Dec 2008 10:25:51 +0000 (10:25 +0000)]
MERGE-via-mokopatches-tracking-hist-MERGE-via-master-hist-1229249926
mokopatches-tracking-hist top was MERGE-via-master-hist-1229249926 / 8f700d425b380707d4f2b7052c1b1f7e77c8b7d3 ... parent commitmessage:
From: merge <null@invalid>
MERGE-mokopatches-tracking-patchset-edits

10 years agofix-jbt6k74-display-onoff-gone.patch
Andy Green [Sun, 14 Dec 2008 10:25:19 +0000 (10:25 +0000)]
fix-jbt6k74-display-onoff-gone.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoMERGE-via-mokopatches-tracking-hist-MERGE-via-master-MERGE-via-master-hist-1229078937...
merge [Fri, 12 Dec 2008 11:49:03 +0000 (11:49 +0000)]
MERGE-via-mokopatches-tracking-hist-MERGE-via-master-MERGE-via-master-hist-1229078937-1229079488
mokopatches-tracking-hist top was MERGE-via-master-MERGE-via-master-hist-1229078937-1229079488 / 7eb66508f95eeebbd7ad3487c5183b76524d4765 ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-master-MERGE-via-master-hist-1229078937

master top was MERGE-via-master-hist-1229078937 / ff0b5902f29135a782a3bfb68e3429b86669aea4 ... parent commitmessage:
From: merge <null@invalid>
MERGE-master-patchset-edits

10 years agoconfig-kill-evbug.patch
Andy Green [Thu, 11 Dec 2008 17:13:48 +0000 (17:13 +0000)]
config-kill-evbug.patch

Reported-by: Klaus Kurzmann <mok@mnet-online.de>
Signed-off-by: Andy Green <andy@openmoko.com>

10 years agointroduce-gta02-packaging-defconfig
Andy Green [Wed, 10 Dec 2008 11:03:42 +0000 (11:03 +0000)]
introduce-gta02-packaging-defconfig

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoconfig-gta02-closer-to-packaging-config.patch
Andy Green [Wed, 10 Dec 2008 11:03:40 +0000 (11:03 +0000)]
config-gta02-closer-to-packaging-config.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoconfig-android-compat.patch
Andy Green [Wed, 10 Dec 2008 10:12:55 +0000 (10:12 +0000)]
config-android-compat.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoFix s3c2410_ts FIFO allocation
Nelson Castillo [Mon, 8 Dec 2008 22:47:03 +0000 (22:47 +0000)]
Fix s3c2410_ts FIFO allocation

When I added the FIFO improving the interrupts handlers I introduced a bug.
The FIFO is allocated after the interrupts are requested. This makes the kernel
crash if the touchscreen generates activity before the allocation takes place.
This patch fixes the bug. I reproduced it and tested the fix in a GTA02.

- Fix bug
- Fix a typo

Reported-by: Andy Green <andy@openmoko.com>
Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoconfig-touchscreen-group.patch
Andy Green [Mon, 8 Dec 2008 11:18:56 +0000 (11:18 +0000)]
config-touchscreen-group.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoRemove skip filter
Nelson Castillo [Mon, 8 Dec 2008 10:15:54 +0000 (10:15 +0000)]
Remove skip filter

With more reliable points median and mean filters perform a better job.
We no longer need this filter.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agos3c2410_ts.c misc cleanup
Nelson Castillo [Mon, 8 Dec 2008 10:15:51 +0000 (10:15 +0000)]
s3c2410_ts.c misc cleanup

- Remove unneeded touchscreen state
- Remove old compile-command line
- Fix spaces in Changelog
- Added "improve interrupt handling" to Changelog

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoReplace "variance" filter with "group" filter
Nelson Castillo [Mon, 8 Dec 2008 10:15:48 +0000 (10:15 +0000)]
Replace "variance" filter with "group" filter

This filter is useful to reject samples that are not reliable. We consider
that a sample is not reliable if it deviates form the Majority.

This filter performs better than the variance filter thus we can replace it.
I will remove the skip filter later.

I haven't seen a bad click in a while. Somebody wake me up.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agojbt6k74_cleanup_no_deep_sleep.patch
nicolas dufresne [Fri, 5 Dec 2008 11:08:41 +0000 (11:08 +0000)]
jbt6k74_cleanup_no_deep_sleep.patch

Improve jbt6k74 driver changes to avoid WSOD so they don't stomp
on QVGA / VGA mode situation

10 years agojbt6k74_no_deep_sleep.patch
Nicolas Dufresne [Fri, 5 Dec 2008 10:43:31 +0000 (10:43 +0000)]
jbt6k74_no_deep_sleep.patch

This patch from

https://docs.openmoko.org/trac/ticket/1841

defeats deep sleep on the LCM ASIC and is reported to stop the WSOD
behaviour on LCMs that exhibit it.

10 years agobuild-fix-path-to-toolchain-one-at-last.patch
Andy Green [Fri, 5 Dec 2008 09:34:16 +0000 (09:34 +0000)]
build-fix-path-to-toolchain-one-at-last.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoMERGE-stable-tracking-patchset-edits
merge [Thu, 4 Dec 2008 21:59:59 +0000 (21:59 +0000)]
MERGE-stable-tracking-patchset-edits

10 years agoCleanup - Add internal functions for clearing filters
Nelson Castillo [Thu, 4 Dec 2008 21:29:21 +0000 (21:29 +0000)]
Cleanup - Add internal functions for clearing filters

This patch adds the following functions:

* ts_filter_mean_clear_internal
* ts_filter_median_clear_internal

The idea: avoid calling the clean function of other filters
          when we initialize one.

Also:

* modify messages for consistency.
* remove an unneeded else.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoAdd Variance filter
Nelson Castillo [Thu, 4 Dec 2008 21:29:20 +0000 (21:29 +0000)]
Add Variance filter

This filter is useful to reject clicks that are not reliable. We
 only care about what happens when we receive DOWN events for the fist time.
 If this filter does not reject the first samples then it will change
 its internal state to "passed" and the remaining samples
 will be sent to the next filter in the chain.

 First we collect N samples, then then we sort them. We discard the borders
 of the vector (with a window) and then compute the variance of the remaining
 set. If the computed variance is bigger than a threshold, we reject the click.

 The parameters of the filter are:

 * size
 * threshold /* maximum allowed variance */
 * window

 I'm using values I tuned by trial an error. I noticed we discard a lot
 of bugus clicks. In fact, it seems harder to get them now.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoworkaround-glamo-mmc-insanity-timeout-warning-only.patch
sprite_tm [Thu, 4 Dec 2008 21:16:51 +0000 (21:16 +0000)]
workaround-glamo-mmc-insanity-timeout-warning-only.patch

Modified version of one line patch from

https://docs.openmoko.org/trac/ticket/2078

by sprite_tm

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agojbt6k74_no_deep_sleep.patch
Nicolas Dufresne [Thu, 4 Dec 2008 18:54:14 +0000 (18:54 +0000)]
jbt6k74_no_deep_sleep.patch

This patch from

https://docs.openmoko.org/trac/ticket/1841

defeats deep sleep on the LCM ASIC and is reported to stop the WSOD
behaviour on LCMs that exhibit it.

10 years agoandy-tracking merge breakage
Werner Almesberger [Thu, 4 Dec 2008 09:07:33 +0000 (09:07 +0000)]
andy-tracking merge breakage

Fix assorted merge breakage in andy-tracking. With this patch,
0b519f14e15365213924505e2e3019c1ff159aa9 builds for GTA02 and it
even boots ! ;-)

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agoconfig-gta02-uplevel-s3c-fifo-option.patch
Andy Green [Wed, 3 Dec 2008 13:43:00 +0000 (13:43 +0000)]
config-gta02-uplevel-s3c-fifo-option.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agorepair-broken-include.patch
Andy Green [Wed, 3 Dec 2008 11:35:56 +0000 (11:35 +0000)]
repair-broken-include.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoMERGE-via-mokopatches-tracking-hist-MERGE-via-master-1228302402
merge [Wed, 3 Dec 2008 11:12:57 +0000 (11:12 +0000)]
MERGE-via-mokopatches-tracking-hist-MERGE-via-master-1228302402
mokopatches-tracking-hist top was MERGE-via-master-1228302402 / de9177f7bd127e9b6fa6213018c7c731b8ca0d0c ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-master-

master top was / 3838a80929f91d35c6d987e518bf9ea397f3e13c ... parent commitmessage:
From: Andy Green <andy@openmoko.com>
fix-wm8753-DBG.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoMERGE-via-mokopatches-tracking-hist-MERGE-via-master
merge [Tue, 2 Dec 2008 21:27:05 +0000 (21:27 +0000)]
MERGE-via-mokopatches-tracking-hist-MERGE-via-master
mokopatches-tracking-hist top was MERGE-via-master / c3ec0335ef762adf82547202a675bfb653eb99bb ... parent commitmessage:
From: merge <null@invalid>
MERGE-via-master-

master top was / 5221720118174d16dbfdbb63668d2b533ae50101 ... parent commitmessage:
From: Andy Green <andy@openmoko.com>
fix-wm8753-DBG.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoadd-skip_filter-s3c2410_ts.patch
Nelson Castillo [Tue, 2 Dec 2008 10:01:10 +0000 (10:01 +0000)]
add-skip_filter-s3c2410_ts.patch

Skip filter for touchscreen values.

Problem: The first and the last sample might be unreliable. We provide
this filter as a separate function in order to keep the event_send_timer_f
function simple. This filter:

* Skips NHEAD points after IE_DOWN
* Skips NTAIL points before IE_UP
* Ignores a click if we have less than (NHEAD + NTAIL + 1) points

Right now the filter is embedded in the driver. We have two #defines with
the parameters SKIP_NHEAD and SKIP_NTAIL. I didn't store their values in
struct skip_filter on purpose.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoclean-hif-warning.patch
Andy Green [Sun, 30 Nov 2008 10:23:27 +0000 (10:23 +0000)]
clean-hif-warning.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agofix-last_issued-ts_filter_median.patch
Nelson Castillo [Fri, 28 Nov 2008 22:54:01 +0000 (22:54 +0000)]
fix-last_issued-ts_filter_median.patch

- Decimation_below was not being used in the median filter. Fixed.

- Decimation_below should be small. There's no need to change many
  points in the median filter if we're moving slow. In this way we
  will filter more noise.

- Now we deliver 2X input events -- 1 each 5 milliseconds.
  (side effect of the previous two changes)

- We should omit the last point before the UP event but I think this
  belongs to user-space.  I'm programming a tslib filter ("skip") and
  if it doesn't work will I'll send a patch later to skip the last
  point in the driver.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoPATCH-rfkill-fix-init-values-and-cleanup-compilation-warning.eml
Sean McNeil [Fri, 28 Nov 2008 09:02:35 +0000 (09:02 +0000)]
PATCH-rfkill-fix-init-values-and-cleanup-compilation-warning.eml

This applies cleanly to stable-tracking to fix errors on initialization.

Sean

10 years agoignore-very-short-clicks-s3c2410_ts.patch
Nelson Castillo [Thu, 27 Nov 2008 19:55:39 +0000 (19:55 +0000)]
ignore-very-short-clicks-s3c2410_ts.patch

With this patch the issue of the noisy clicks seems to be
alleviated. We ignore the first two points. If a click has less
than 3 points we ignore it.

On a normal click (I mean with at least 3 points) this patch
will delay the first DOWN event by 20 milliseconds.

I'm looking for a better solution but this one works and it should
make the driver more usable.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoconfig-gta02-netfilter-stuff.patch
Andy Green [Thu, 27 Nov 2008 14:24:32 +0000 (14:24 +0000)]
config-gta02-netfilter-stuff.patch

Reported-by: Timo Jyrinki
Signed-off-by: Andy Green <andy@openmoko.com>

10 years agofix the eint_unmask logic for s3c64xx
Matt Hsu [Thu, 27 Nov 2008 12:07:00 +0000 (12:07 +0000)]
fix the eint_unmask logic for s3c64xx

10 years agoconfig-gta02-uplevel.patch
Andy Green [Thu, 27 Nov 2008 08:32:31 +0000 (08:32 +0000)]
config-gta02-uplevel.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agocleanup-adc-s3c2410_ts.patch
Nelson Castillo [Thu, 27 Nov 2008 08:31:21 +0000 (08:31 +0000)]
cleanup-adc-s3c2410_ts.patch

- Factor the ADC conversion.

 - Include a FIFO that gives us more flexibility in the interrupt handlers.

 - Changed timer function. Now we can have more states for the touchscreen.

   The new states make the logic easier to understand.

   The state TS_STATE_PRESSED_PENDING will be useful to filter short clicks
   that fail to provide a valid point. We can check this with an
   additional filter.

   The state TS_STATE_RELEASE_PENDING helps us ignore short UP events.

 - Integrate Jitter Detection

   (preserving the threshold chosen by Tick -- 1/16 sec)
   http://lists.openmoko.org/pipermail/devel/2008-October/002822.html

 - Fix trivial compilation warning in a ts_filter_create_chain call.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agos3cmci-suspend-persist.patch
Werner Almesberger [Thu, 27 Nov 2008 08:31:20 +0000 (08:31 +0000)]
s3cmci-suspend-persist.patch

Allow s3cmci to preserve its state across suspend/resume without
letting the stack take down and re-scan the interface.

This can be useful for cards that implement their own low-power
mode and that have state that can be preserved and that should
be available quickly after resume. E.g., a WLAN card with an
active association.

This patch does not change the default behaviour. To keep the
interface alive across suspend/resume, set the "persist" option,
either with persist=1 (module loading) or s3cmci.persist=1 (on
the boot parameter line).

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agommc-export-flush-workqueue.patch
Werner Almesberger [Thu, 27 Nov 2008 08:31:20 +0000 (08:31 +0000)]
mmc-export-flush-workqueue.patch

Export mmc_flush_scheduled_work so that drivers that don't go through
mmc_suspend_host/mmc_resume_host can make sure the stack is quiescent.

Note: if we send this patch upstream, it's okay if upstream prefers to
use EXPORT_SYMBOL instead of EXPORT_SYMBOL_GPL. All other exports in
mmc/core/core.c are EXPORT_SYMBOL, but it would be rude to open
somebody else's hitherto internal interface and not use
EXPORT_SYMBOL_GPL.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agoandroid-neo1973-bt.patch
Sean McNeil [Thu, 27 Nov 2008 08:31:19 +0000 (08:31 +0000)]
android-neo1973-bt.patch

Signed-off-by: Sean McNeil <sean@mcneil.com>

10 years agoandroid-binder.patch
Sean McNeil [Thu, 27 Nov 2008 08:31:18 +0000 (08:31 +0000)]
android-binder.patch

Signed-off-by: Sean McNeil <sean@mcneil.com>

10 years agoar6k-break-down-insane-allocation.patch
Werner Almesberger [Wed, 26 Nov 2008 20:44:44 +0000 (20:44 +0000)]
ar6k-break-down-insane-allocation.patch

The Atheros WLAN stack kmallocs almost 64kB of contiguous kernel
memory for a structure containing almost entirely buffers.

As is commonly known , this kind of large allocation has a very
high risk of failing as kernel memory fragments during the life of
a system.

This patch allocates the buffers indiviudually, thus shrinking the
structure to a size below 4kB.

Note: this is untested. These buffers are only used with Atheros'
raw interface, which none of the code we have, including wmiconfig,
even seems to know about.

This may fix bug #2133. Code follows Atheros' style, so checkpatch
hates it.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years ago(no commit message)
Andy Green [Wed, 26 Nov 2008 14:56:12 +0000 (14:56 +0000)]

10 years agofix-touchscreen-up-down-issues.patch
Tick Chen [Wed, 26 Nov 2008 12:17:55 +0000 (12:17 +0000)]
fix-touchscreen-up-down-issues.patch

Port the fixes from stable branch

Signed-off-by: Tick Chen <tick@openmoko.com>

10 years agoar6k-disconnect-more-often.patch
Werner Almesberger [Wed, 26 Nov 2008 12:17:45 +0000 (12:17 +0000)]
ar6k-disconnect-more-often.patch

When changing wireless settings, the WMI is often not updated, because
we send a connect but no prior disconnect. The firmware seems to ignore
the connect if already associated.

The disconnect is omitted because the SIOCSIWESSID logic in the
Atheros driver seems to try to prevent multiple disconnects to be sent
to the WMI, which doesn't really seem to be necessary.

This patch makes sure we disconnect when a parameter change has happened.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agogta02-s3cmci-power-control.patch
Werner Almesberger [Sat, 22 Nov 2008 08:55:12 +0000 (08:55 +0000)]
gta02-s3cmci-power-control.patch

This patch gives s3cmci nominal control over power to the WLAN
module.

In face, it only gives it the ability to reset the WLAN module
after module removal/insertion or suspend/resume. Resetting the
module is necessary, because s3cmci requests a full device
initialization from the SD/MMC stack, including CMD5, and this
command is only allowed if the SDIO module has been reset.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agodeglitch-wlan-power-control.patch
Werner Almesberger [Sat, 22 Nov 2008 08:55:07 +0000 (08:55 +0000)]
deglitch-wlan-power-control.patch

gta02_wlan_power(1) causes a reset each time it's invoked, even if the
device was not turned off.

This causes problems if we use the set_power callback from s3cmci,
because it will try to turn on power even when the device is already
operational.

This patch makes sure we only reset the WLAN if the device was off or
if we don't know the current state.

Doing the "deglitching" in gta02_wlan_power instead of the set_power
handler has the advantage that we're synchronized with any other
agents that may change WLAN power.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agohif-fix-removal.patch
Werner Almesberger [Sat, 22 Nov 2008 08:55:04 +0000 (08:55 +0000)]
hif-fix-removal.patch

There are two entry points for removing the AR6k driver: HIFShutDownDevice
(for module unloading) and sdio_ar6000_remove (for shutdown and suspend).

This patch makes sure that shutdown/suspend notify the WLAN stack and
that module unloading calls sdio_ar6000_remove without recursion.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agos3c2410_ts-improve-interrupt-handling.patch
Nelson Castillo [Sat, 22 Nov 2008 08:55:00 +0000 (08:55 +0000)]
s3c2410_ts-improve-interrupt-handling.patch

Let's improve interrupt handling logic. We do not need to
to call the timer function (touch_timer_fire) from the
stylus_updown interrupt handler. With this patch the filters
work much better. We were getting many erratic results when
tapping that were making this driver unusable.

I tested using this program by Tick:

http://svn.openmoko.org/developers/tick/touchtest/touch_test.py

The driver is behaving much better now. Please help us test.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoavoid-double-kfree-ts_filter_destroy_chain.patch
Nelson Castillo [Sat, 22 Nov 2008 08:54:42 +0000 (08:54 +0000)]
avoid-double-kfree-ts_filter_destroy_chain.patch

The function ts_filter_mean_destroy is destroying all the
filters of the chain. At the same time each filter is trying
to remove other filters in the chain. Let's fix this by
making each filter ignore it is part of a chain when we are
freeing it.

The bug shows when s3c2410_ts.c is built as a module.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoMERGE-s3c64xx-fix-clock.patch
Andy Green [Fri, 21 Nov 2008 11:27:28 +0000 (11:27 +0000)]
MERGE-s3c64xx-fix-clock.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoThe S3C64XX series has a new TCFG divider setting to
Ben Dooks [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
The S3C64XX series has a new TCFG divider setting to
allow the clock directly through, which means that
we need to update the pwm-clock code to cope with
this.

Add <mach/pwm-clock.h> containing the specific code
to deal with the TCFG divider settings and provide
any other per-arch data that the pwm-clock driver
needs to function.

Signed-off-by: Ben Dooks <ben-linux@fluff.org>

10 years agolis302dl-ack-ff_wu-interrupts.patch
Simon Kagstrom [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
lis302dl-ack-ff_wu-interrupts.patch

Setup interrupt latching for thresholded operation

From: Simon Kagstrom <simon.kagstrom@gmail.com>

This patch, for stable-tracking, sets up interrupt latching for
thresholded operation, and also acks the interrupt in the handler. This
reduces the interrupt load, and thresholds < 36 are once again allowed
(though not of much use).

The patch also fixes some warnings in the code (including a potential
bug).

Thanks to Werner for pointing this out!

Signed-off-by: Simon Kagstrom <simon.kagstrom@gmail.com>

10 years agos3cmci-dont-leak-pdata.patch
Werner Almesberger [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
s3cmci-dont-leak-pdata.patch

When compiling s3cmci as a module and doing an insmod-rmmod-insmod
sequence, s3cmci_probe often oopses because the platform data contains
garbage.

The reason for this is that I'm not providing any platform data (removed
in gta02-mmc-mci.patch), and thus s3cmci tries to put its own defaults:

        if (!host->pdata) {
                pdev->dev.platform_data = &s3cmci_def_pdata;
                host->pdata = &s3cmci_def_pdata;
        }

s3cmci_def_pdata is inside the module, so the pointer in
pdev->dev.platform_data becomes invalid after the module is unloaded.

Since nothing else in s3cmci uses pdev->dev.platform_data, "patching"
it seems unnecessary. This patch removes it.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agotypo-debug-s3cmci-fmax.patch
Werner Almesberger [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
typo-debug-s3cmci-fmax.patch

Oops ! It compiled file as non-module, but not as module.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agobuild-create-module-tarball.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
build-create-module-tarball.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agodebug-s3cmci-fmax.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
debug-s3cmci-fmax.patch

This patch adds the module parameter f_max to s3cmci to limit the
maximum clock frequency of the SD interface.

E.g., when booting with s3cmci.f_max=10000000, the kernel should say
s3c2440-sdi s3c2440-sdi: f_max lowered from 50000000 to 10000000 Hz
and the interface frequency should be limited accordingly.

Setting the limit above the maximum frequency supported by the card
or the interface has no effect, so the option is ignored (and no
"f_max lowered" line is printed).

Setting the limit below the minimum frequency supported by the
interface sets it to that frequency.

Signed-off-by: Werner Almesberger <werner@openmoko.org>

10 years agofix-pcf50633-backlight-never-down.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
fix-pcf50633-backlight-never-down.patch

This solves the problem that the backlight never turns off
in stable-tracking

Signed-off-by: Andy Green <andy@openmoko.com>

10 years ago: lis302dl-port-work-from-stable.patch
Simon Kagstrom [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
: lis302dl-port-work-from-stable.patch

Merge in the accelerometer functionality from stable

From: Simon Kagstrom <simon.kagstrom@gmail.com>

This giant patch merges the functionality from stable. It also changes
the default threshold to 0 (generate interrupts on data). As the
level-triggered interrupts cause too frequent interruptions on low
thresholds, also set a lower limit to 36mg to avoid slowing down the
phone too much.

Also rearrange the probe function to disable interrupts only over device
I/O. This fixes the kernel errors when inserting the module.

Andy was right: In stable-tracking, no magic workaround is needed to
make suspend/resume work. I've tested the accelerometers under heavy
load (making the phone basically unusable) without getting lockups.

Signed-off-by: Simon Kagstrom <simon.kagstrom@gmail.com>

10 years ago: export-irq-functionality.patch
Simon Kagstrom [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
: export-irq-functionality.patch

Export some irq-related functionality to modules

From: Simon Kagstrom <simon.kagstrom@gmail.com>

This patch exports handle_level_irq and __set_irq_handler to modules.
It's needed to compile lis302dl as a module.

Signed-off-by: Simon Kagstrom <simon.kagstrom@gmail.com>

10 years agoMake the console turn off the backlight when blanking the screen.
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
Make the console turn off the backlight when blanking the screen.
Work in progress.

10 years agoSubject: [PATCH 2/2] cleanup-gta02_lis302dl_bitbang.patch
Nelson Castillo [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
Subject: [PATCH 2/2] cleanup-gta02_lis302dl_bitbang.patch

* Improve comment about the chip select issue with the accelerometers.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoSubject: [PATCH 1/2] fix-gta02_lis302dl_bitbang.patch
Nelson Castillo [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
Subject: [PATCH 1/2] fix-gta02_lis302dl_bitbang.patch

Make accelerometers work again in stable-tracking.
They stopped working with a previous commit.

(shifter >> 7) & 1); ...; shifter <<= 1;
  Had been changed to:
(shifter >> (7 - n)) & 1) /* assuming tx_bytes == 1 */

The problem: an unneeded shifter <<= 1 was still there.

We make a small change to the indexing code to make things work again.

Signed-off-by: Nelson Castillo <nelsoneci@gmail.com>

10 years agoconfig-gta02-s3c-mmc-for-wlan.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
config-gta02-s3c-mmc-for-wlan.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agofix-glamo-gpio-resume.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
fix-glamo-gpio-resume.patch

Glamo GPIO are not set correctly after resume / reset action.
This patch forces them to correct state for GTA02.

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoAdd default platform data for i2c to make GTA01 boot
Micael Henriksson [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
Add default platform data for i2c to make GTA01 boot

Signed-off-by: Micael Henriksson <micael.h@gmail.com>

10 years agobuild-introduce-device-specific-named-kernel-binary.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
build-introduce-device-specific-named-kernel-binary.patch

This patch changes the ./build script to generate a uImage.bin file with
the build device appended, eg, uImage-GTA02.bin.

The reason for this change is that it is possible to generate rootfs images
now that can target multiple (and probably increasing numbers of) devices
with the one image.  Therefore qi or other bootloader running on the device
is going to have to pick out the right kernel from the rootfs filesystem
for the device it finds it is running on from a set provided down /boot
by the rootfs generator.

/lib/modules for the kernels doesn't get in the way of each other because the
version-specific path is also extended to have the device name as part of it.

2.6.28-GTA02_stable-tracking_c77dda03a7c11666-mokodev

For legacy U-Boot support on GTA01 or GTA02, a symlink to uImage-GTA01.bin or
uImage-GTA02.bin will be needed.

The dfu utility script is updated to take an argument, eg, GTA02

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoconfig-gta02-uplevel-s3c-changes.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
config-gta02-uplevel-s3c-changes.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agoconfig-enable-storing-config-in-kernel.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
config-enable-storing-config-in-kernel.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agofix-remove-broken-faster-nand.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
fix-remove-broken-faster-nand.patch

Signed-off-by: Andy Green <andy@openmoko.com>

10 years agofix-dangling-drivers-sdio-Makefile-reference.patch
Andy Green [Fri, 21 Nov 2008 11:23:03 +0000 (11:23 +0000)]
fix-dangling-drivers-sdio-Makefile-reference.patch

Change to removing Atheros stack left a dangling reference to Makefile
in drivers/sdio that made errors on make distclean

Reported-by: Christopher Hall <hsw@openmoko.org>
Signed-off-by: Andy Green <andy@openmoko.com>