| Age | Commit message (Collapse) | Author |
|---|
|
Space transition between sleep-in and sleep-out of at least
120 ms as mentionned in the spec. Also, made sure the LCM
is moved into sleep state before the pixel clock is turned off.
To do so, I've modified glamo-fb to send FB_BLANK_POWERDOWN before
turning off the pixel clock. Also fixed various violation of coding
style. This patch should reduce the risk of white screen.
Signed-off-by: Nicolas Dufresne <nicolas.dufresne@gmail.com>
|
|
A network device is supposed to disable its internals in the "close"
function. Doing so affords us protection against various races,
including the ioctl vs. rfkill conflict reported by Michael.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
Reported-by: Michael Trimarch <michael@panicking.kicks-ass.org>
|
|
According to the PCF50606 manual, the issues resoved by commit
cc1663fc922c03feb0d7bbb8b18d62fbac0128de also exists there.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
Reported-by: Rask Ingemann Lambertsen <rask@sygehus.dk>
|
|
This is loosely based on a patch by Ivan Petrov.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
Reported-by: Ivan Petrov <ivan_p@hotbox.ru>
|
|
"pdata" is NULL on GTA01 and you oops.
Revised patch, this time with if(..) instead of #ifdef
Signed-off-by: Tim Niemeyer <tim.niemeyer@mastersword.de>
|
|
IDs 0 and 1 are AR6002, not AR6001. This is based on a patch by
Ivan Petrov, but it differs in leaving the IDs in
include/linux/mmc/sdio_ids.h
Signed-off-by: Werner Almesberger <werner@openmoko.org>
Reported-by: Ivan Petrov <ivan_p@hotbox.ru>
|
|
e2c0650efa751a6a2220618695fa41a2a5e7d23c introduced a platform-specific
hack to check if the MMC driver and hardware handle busy signaling from
the device properly.
Since this seems to be the case, we can revert this hack now.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
This patch cleans up a little after Ivan's netif_queue_stop fix.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
This patch resolves the following issue:
http://lists.openmoko.org/pipermail/openmoko-kernel/2009-March/009643.html
Changed: prevent rescheduling network queue at interface opened/connected.
Removed: wake network queue at transmit complete.
Added: wake network queue at packet queue limit not reached.
Signed-off-by: Ivan Petrov <ivan_p@hotbox.ru>
|
|
... and check the return value in ar6000_activate.
[ Werner: changed ___FUNCTION__ to __func__ ]
Signed-off-by: Michael Trimarchi <michael@panicking.kicks-ass.org>
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
This is done in a slightly nasty way. We should perhaps try to come up
with a better way of uniquely identifying the platform device to DRM.
Signed-off-by: Thomas White <taw@bitwiz.org.uk>
|
|
This adds some MMIO initialisation stuff to glamo-drm.c.
Signed-off-by: Thomas White <taw@bitwiz.org.uk>
|
|
This fixes the MMC buffer locations in glamo-mci.c, which were broken by the
reorganisation of Glamo's memory.
Signed-off-by: Thomas White <taw@bitwiz.org.uk>
|
|
This adds a missing statement to let glamo-drm.c know the core Glamo handle, so
it can manipulate registers such as the 2D and 3D clocks.
Signed-off-by: Thomas White <taw@bitwiz.org.uk>
|
|
This just cosmetically tidies up the declarations in glamo-core.h. We have to
be very clear about what lives where in the memory.
Signed-off-by: Thomas White <taw@bitwiz.org.uk>
|
|
This patch removes the platform devices corresponding to Glamo's 2D, 3D, JPEG and MPEG
engines. These will later be handled via DRM.
Signed-off-by: Thomas White <taw@bitwiz.org.uk>
|
|
|
|
Handle new memory layout.
The "not-needed" memory should not be taken by the framebuffer driver.
Use that for the DRM driver.
Add the cmdq platform device
Set aside 4k for hardware cursor, reject cursors that don't fit.
|
|
Handle correctly the proc/dri/0/name printing
Helper function to check the PLATFORM feature
|
|
|
|
|
|
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
platform_device_add_data copies the data, so no need for
kmalloc.
Signed-off-by: Tim Niemeyer <reddog@mastersword.de>
|
|
Watchdog and RTC entries were missing.
Signed-off-by: Tim Niemeyer <reddog@mastersword.de>
|
|
Signed-off-by: Tim Niemeyer <reddog@mastersword.de>
|
|
Signed-off-by: Werner Almesberger <werner@openmoko.org>
Reported-by: Tim Niemeyer <tim.niemeyer@mastersword.de>
|
|
[ Werner: Documentation/CodingStyle lines 166 and 448, put blank line after
local variables. ]
Signed-off-by: Tim Niemeyer <reddog@mastersword.de>
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
Signed-off-by: Tim Niemeyer <reddog@mastersword.de>
|
|
Get resume by AUX to work.
Signed-off-by: Tim Niemeyer <reddog@mastersword.de>
|
|
The kernel would fail to link because cfb_fillrect, cfb_copyarea and
cfb_imageblit weren't build. This patch fixes it.
Signed-off-by: Rask Ingemann Lambertsen <rask@sygehus.dk>
|
|
This patch fixes the issues of RNDIS on the Openmoko Freerunner (and
subsequently Windows networking with the Freerunner). It arises from the
case where a stall is being issued by the upper level RNDIS driver to
the s3c2410_udc.c in the 2.6.28 series driver and it not being handled
correctly. This patch changes the RNDIS driver to send zero-length
packets instead of a stall, a condition that allowed the udc driver to
function correctly in 2.6.24 series kernels.
Further investigation into the udc driver is recommended. I am currently
looking into it to be able to use the Android ADB gadget driver, which
currently has stability issues using the underlying udc on both Linux
and Windows (although, it's possible that the Android ADB gadget has
problems).
Traces and logs of the USB traffic are available upon request. Thanks to
Aric at SDG Systems for all of his work on this matter.
Regards,
Brian Code
From de386af349bbfe1ad6d45c810185123975888d8d Mon Sep 17 00:00:00 2001
From: Aric D. Blumer <aric@sdgsystems.com>
Date: Mon, 16 Mar 2009 16:04:02 -0600
Subject: [PATCH] Zero packet length detection for RNDIS fix. Migrated from 2.6.24 series kernel
Signed-off-by: Aric D. Blumer <aric@sdgsystems.com>
Signed-off-by: Brian Code <Brian.Code@koolu.com>
|
|
If no battery is connected, we periodically get a burst of HDQ error
messages (at least on 3D7K), interrupting whatever we're doing on
the console.
This patch reduces this to only one message per sequence of errors,
and one more message if communication with HDQ is successful later.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
This patch explicitly enabled to "camif" clock, to make the driver
work again with Qi commit a24b5fcf84d2cf633a3f660edd23fa4c2a3da231.
Note that this is not fully tested since the Qi change also broke
the LCM driver.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
Signed-off-by: Andy Green <andy@openmoko.com>
|
|
This dirty hack adds camera activation/deactivation on open/close.
This code will have to move to a machine/platform area later. The
general driver is the wrong place for it.
It also removes the unconditional activation and moves the domain
setting to the rest of the power control code. (Thanks, Andy !)
Note that opening the camera now always takes about 20 seconds
until the settings are downloaded, not just on the first open.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
Filter chains should be completely opaque to the drivers that use it.
We fix this with this patch.
~ Make the "filter chain" a new object.
~ We can build with CONFIG_TOUCHSCREEN_FILTER=n with no problems in
a cleaner way.
~ Update s3c2410_ts.c to use the filter_chain object.
~ Cleanups.
Signed-off-by: Nelson Castillo <arhuaco@freaks-unidos.net>
|
|
~ Make a few symbols constant.
~ Export symbols explicitly.
~ Move ts_filter.c to ts_filter_chain.c (this will make sense later).
Signed-off-by: Nelson Castillo <arhuaco@freaks-unidos.net>
|
|
This patch defines a ts_filter_configuration structure to
avoid using void* in the filter initialization, fixing another
upstream correction. This also makes the initialization more readable.
Tested in GTA02/rev6.
Other changes:
~ Comment filter configuration structures.
~ ts_filter.c:ts_filter_chain_create improved.
~ Small cleanups.
~ More TODOs/FIXMEs.
~ Updated GTA02 filter configuration.
~ Updated GTA01 filter configuration.
~ Updated mach-s3c2410/include/mach/ts.h for the new ts. configuration
structure.
~ Updated all the filters to use the new configuration structure.
~ Removed MAX_TS_FILTER_CHAIN constant that is no longer needed.
No more evil casts left it seems.
Signed-off-by: Nelson Castillo <arhuaco@freaks-unidos.net>
|
|
This patch turns upstream feedback into API modifications and code
improvements. There will be more patches implementing upstream
corrections but this one is the that will make most of the invasive
changes and make the most important improvements to the API.
Tested in a GTA02/rev06.
The goals of this patch are:
* Replace recursive calls with iteration.
* General code improvements.
* Make ts_filter_mean.c a reference for the rest of the filters.
* Make the (almost)minimum number of changes to the other filters so
that they compile and work, patches for cleaning these up will
come next.
* Filters should do what they were doing before.
Some important changes:
* Move "struct ts_filter tsf" in the private structures to force
a crash (or break things) if we forget to remove an open-coded cast.
* ts_filter.c/ts_filter.h
~ API modifications.
* s3c2410_ts.c:
~ Use the new API.
~ Cleanups.
* ts_filter_mean.c
~ Replace with a simple mean.
~ Use as a reference for the new API.
~ Move private structure from the .h to the .c.
* ts_filter_group.c
~ Update to use the new API.
* ts_filter_median.c
~ Update to use the new API.
* ts_filter_linear.c
~ Remove functions that are no longer needed.
Note:
I might leave some TODOs and FIXMEs with this patch. Most of them
will be removed shortly.
Signed-off-by: Nelson Castillo <arhuaco@freaks-unidos.net>
|
|
Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>
|
|
This patch does some light cleanup on the camera driver:
- removed unused return values or unused initializations
- remove backslashes at regular line ends (Python-style ?)
- removed redundant casts
- remove redundant parentheses
- wrapped long lines
- general whitespace cleanup
- fixed broken logic in s3c_camif_v4l2_g_input and s3c_camif_v4l2_g_output
- let mutex that really really wants to be a semaphore have its way
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
The PLL converts a 26.6 MHz input clock to 87.8 MHz, which is later
divided by two and becomes PCLK. (See comments below for details.)
All this looks like a textbook example for platform stuff, but let's
take it one step a time.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
This patch does some light cleanup on the S5K4BA driver:
- change structure initializers from field: to .field = syntax
- exlicitly name fields in structure initializers
- added proper error handling to sensor_read
- use ARRAY_SIZE
- remove redundant parentheses and braces
- change // to /* ... */
- remove non-ASCII characters
- general whitespace cleanup
- removed an unused variable warning
(4xa_sensor.h is only partially cleaned up.)
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
This patch does some cleanup on the camera interface driver:
- remove backslashes at regular line ends (Python-style ?)
- removed unused return values or unused initializations
- combined duplicated or otherwise highly redundant code
- remove redundant parentheses
- wrapped long lines
- general whitespace cleanup
- s3c_camif_input_msdma_codec, s3c_camif_input_msdma_preview,
s3c_camif_set_target_format: fixed clearing of old register values
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
Signed-off-by: Andy Green <andy@openmoko.com>
|
|
X-Git-Url: http://git.openmoko.org/?p=kernel.git;a=commitdiff_plain;h=b55b8e56a4a7e43b7243be48f77a326236a37c68
USB: Change s3c2410_ohci into s3c_ohci and change gta02 to use it
Signed-off-by: Balaji Rao <balajirrao@openmoko.org>
|
|
http://lists.openmoko.org/pipermail/openmoko-kernel/2009-March/009387.html
)
Change accelerometers to use ABS events rather than REL events.
[Obviously if this patch is accepted we need to tell developers about
it. I have a number of other improvements to the accelerometers I
hope to deliver over the next couple of weeks. They will have minimal
or zero disruption to current code. ]
REL events should be used when there is no absolute reference, and
only changes are meaningful. The classic example is a "mouse" where
the absolute position of the device is not measurable and not
particularly meaning, but change in position from one time to the next
is interesting.
With REL events, a value of '0' is not reported, as 'not change' is
not interesting.
With REL events, the expectation is that successive values will be
eventually summed (possibly with acceleration and clipping
adjustments) to get a usable value.
ABS events should be used when there is an absolute references against
which things that be measured.
With ABS events, the 'current value' is meaningful and can be read
(EVIOCGABS).
With ABS events, the value '0' is very meaningful and is reported.
However if consecutive values are the same, the value is only reported
once.
ABS events can be used as-is or compared with previous events to get
some measure of change.
An obvious example is a touchscreen where each measure in
independently meaningful.
Acceleration is an absolute value as it is measuring against a frame
of reference. '0' acceleration is just as meaningful as any other
value, and finding the 'current' acceleration is each direction is a
potentially useful thing to do.
The Freerunner accelerometers currently report REL events. This is
wrong. So this patch changes them to report ABS events.
With this patch, the min/max/level/fuzz values are left at zero. It
might be useful to make use of these in a subsequent patch.
min/max/level can be used to calibrate the accelerometers if accuracy
is important.
fuzz could possibly be used in conjunction with the 'threshold' sysfs
value to get less frequent, lower-precision reports.
This may well break some applications that read accelerometer data.
This cannot be helped, but it is quite easy to write code that copes
with the incorrect EV_REL events as well as the more correct and
useful EV_ABS events.
Signed-off-by: NeilBrown <neilb@suse.de>
|
|
The camera driver reduced all IO drivers from their reset defaults of
6 mA to only 2 mA, which caused severe signal distortion at higher
speeds. This patch sets them to 8 mA and also removes apparently
useless repetitions of the setting.
Note that the correct setting of the I2C pullups still needs to be
verified.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
|
- provide clean sysfs to control RGB channel directly.
They are looked like the following:
channel_mode
channel_pwm
channel_cur
- add platform data to export RGB channel accordingly.
Signed-off-by: Matt Hsu <matt_hsu@openmoko.org>
|
|
A few dirty hacks to make the camera driver work:
- because V4L no longer guarantees that minor numbers provided by the
client are actually used, the preview/codec selection mechanism falls
apart. We work around this by defaulting to preview when we don't
know better.
- power up the camera in platform code, not cleanly via power control
device.
Signed-off-by: Werner Almesberger <werner@openmoko.org>
|
