htc-msm-2-6-32:leviathan-incoming.git
8 years agoMake xf86-video-msm from codeaurora recognize the board pmem-free
David Lanzendörfer [Thu, 11 Feb 2010 19:20:32 +0000 (20:20 +0100)]
Make xf86-video-msm from codeaurora recognize the board

8 years agoModified the includes for pmem to make it build
David Lanzendörfer [Thu, 11 Feb 2010 13:17:45 +0000 (14:17 +0100)]
Modified the includes for pmem to make it build
without android_pmem

8 years agoRemoved include/linux/android_pmem.h and replaced it by
David Lanzendörfer [Thu, 11 Feb 2010 13:16:48 +0000 (14:16 +0100)]
Removed include/linux/android_pmem.h and replaced it by
the temporary dummy file mach/arm/mach-msm/include/mach/pmem.h

8 years agoAdding android pmem replacement
David Lanzendörfer [Thu, 11 Feb 2010 13:05:36 +0000 (14:05 +0100)]
Adding android pmem replacement
* Makes every android pmem function which isnt yet removed from the code becoming a dummy function
* Needs to be removed as soon as every driver is either totaly removed, replaced or cleaned up

8 years agoRemoved pmem
David Lanzendörfer [Sat, 6 Feb 2010 17:42:37 +0000 (18:42 +0100)]
Removed pmem

8 years agoAdding kgsl_drm.h
David Lanzendörfer [Thu, 4 Feb 2010 13:08:19 +0000 (14:08 +0100)]
Adding kgsl_drm.h

8 years agoFixing structs
David Lanzendörfer [Thu, 4 Feb 2010 10:12:47 +0000 (11:12 +0100)]
Fixing structs

8 years agoExtending msm_mdp.h
David Lanzendörfer [Thu, 4 Feb 2010 03:53:30 +0000 (04:53 +0100)]
Extending msm_mdp.h

8 years agoAdded missing macros from codeaurora for (hopefully finaly) enabling compilation...
David Lanzendörfer [Thu, 4 Feb 2010 00:30:19 +0000 (01:30 +0100)]
Added missing macros from codeaurora for (hopefully finaly) enabling compilation for xf86-video-msm
from codeaurora
Just copy'n and paste because to join you have to pay 1k$ but its GPL so free distribution under GPL
is allowd ;P

8 years agoAdded missing defines from codeaurora
David Lanzendörfer [Wed, 3 Feb 2010 19:31:57 +0000 (20:31 +0100)]
Added missing defines from codeaurora

8 years agoAdding missing defines from codeaurora
David Lanzendörfer [Wed, 3 Feb 2010 14:53:34 +0000 (15:53 +0100)]
Adding missing defines from codeaurora

8 years agomsm_camera, msm_v4l2 and msm_vfe7x(camera sensor): make it compile(still oops at...
Denis 'GNUtoo' Carikli [Sat, 30 Jan 2010 23:56:54 +0000 (00:56 +0100)]
msm_camera, msm_v4l2 and msm_vfe7x(camera sensor): make it compile(still oops at runtime)

msm_vfe7x.c and msm_camera lacked the linux/sched.h include so I added it
  Else it would fail to compile because TASK_NORMAL and TASK_UNINTERRUPTIBLE are
  undeclared

I also changed VID_TYPE_CAPTURE into V4L2_CAP_VIDEO_CAPTURE because:
*even if including videodev2.h was correct,it didn't resolve VID_TYPE_CAPTURE
 So I went to look at VID_TYPE_CAPTURE in videodev2.h and just before it I saw that:
  These defines are V4L1 specific and should not be used with the V4L2 API!
  They will be removed from this header in the future.
So I updated it with V4L2_CAP_VIDEO_CAPTURE because of that:
 /* Values for 'capabilities' field */
 #define V4L2_CAP_VIDEO_CAPTURE          0x00000001  /* Is a video capture device */

Signed-off-by: Denis 'GNUtoo' Carikli <GNUtoo@no-log.org>
8 years agoFixing usb gadget as module
David Lanzendörfer [Sat, 30 Jan 2010 23:20:20 +0000 (00:20 +0100)]
Fixing usb gadget as module

8 years agoDepency for backlight added
David Lanzendörfer [Fri, 29 Jan 2010 23:58:28 +0000 (00:58 +0100)]
Depency for backlight added

8 years agoSynaptics patch
David Lanzendörfer [Fri, 29 Jan 2010 23:36:56 +0000 (00:36 +0100)]
Synaptics patch

8 years agoregistering backlight as REAL backlight (needs perhaps some cleanup... e.g. by GNUtoo)
David Lanzendörfer [Fri, 22 Jan 2010 18:35:57 +0000 (19:35 +0100)]
registering backlight as REAL backlight (needs perhaps some cleanup... e.g. by GNUtoo)

8 years agoPrevent LCD from crazy darkness (needs additional hacking)
David Lanzendörfer [Thu, 21 Jan 2010 01:16:52 +0000 (02:16 +0100)]
Prevent LCD from crazy darkness (needs additional hacking)

8 years agoFixed sound listing in alsamixer
David Lanzendörfer [Thu, 21 Jan 2010 01:15:33 +0000 (02:15 +0100)]
Fixed sound listing in alsamixer

8 years ago* Removed functions which are not needed anymore
David Lanzendörfer [Wed, 20 Jan 2010 02:34:37 +0000 (03:34 +0100)]
* Removed functions which are not needed anymore
* Put the device ids into an array for simpler lookup and implementation of additional switches

8 years agoFixed sound routing method
David Lanzendörfer [Wed, 20 Jan 2010 00:51:54 +0000 (01:51 +0100)]
Fixed sound routing method

8 years agoalsa-soc: msm: fixed where SNDRV_PCM_IOCTL_WRITEI_FRAMES blocked mplayer + others...
gnutoo [Sun, 17 Jan 2010 20:04:33 +0000 (21:04 +0100)]
alsa-soc: msm: fixed where SNDRV_PCM_IOCTL_WRITEI_FRAMES blocked mplayer + others minor fixes

Minor fixes:
*removed debugging
*changed frame = prtd->out + prtd->out_tail; into frame = &prtd->out[prtd->out_tail];

For more background details see the following threads in alsa-devel:
[PATCH] MSM DSP
[PATCH] Sound: MSM soc : imported alsa for the MSM from codeaurora

The fix idea was from Thingol in #alsa
So many thanks to him,I spent so much time trying to fix it and,
in 2 irc conversation he found how to fix it.
(altough my previous mails and understanding of the bug may have
helped him)

Basically for an unknown reason the alsa driver had that:
prtd->out[0].used = BUF_INVALID_LEN;
prtd->out_head = 1; /* point to second buffer on startup */
Removing it made the kernel panic.

So:
When the sound card is opened the fist buffer is assigned to
BUF_INVALID_LEN and is skipped(prtd->out_head = 1).

Then at a point alsa_send_buffer (in msm-pcm.c) is called.
But because mplayer uses a bigger buffer it goes this way:
The first run goes fine:
rc = wait_event_interruptible(the_locks.write_wait, (frame->used == 0)
|| (prtd->stopped)); is passed.
Then in "xfer = count > frame->size ? frame->size : count;",the minimum
between frame size and count is assigned to xfer.
Then in copy_from_user(frame->data, buf, xfer) the audio frame is copied
from userspace.
Then xfer is assigned to frame->used
Then "prtd->out_head ^= 1;" needs some little explanations:
There are 2 buffers prtd->out[0] and prtd->out[1]
And both out_head and out_tail can only be 0 or 1
So when prtd->out_head is 0,prtd->out_head ^= 1 changes to 1 into a 0
and vice versa.
So...prtd->out_head was 1 because it was assigned to 1 in msm_pcm_open
in msm7k-pcm.c
Now it becomes 0

Then it does that:
count -= xfer;
which doesn't zero  count it because mplayer used some big buffer size

Then it sends the data to the dsp...and doesn't exit the "while (count >
0) {" loop because count is not 0

And Finally it reaches that line a second time:
"rc = wait_event_interruptible(the_locks.write_wait, (frame->used == 0)
 || (prtd->stopped));"
 But...frame->used was, and is still BUF_INVALID_LEN so it blocks.

Now...The fix moves the alsa_audio_configure:
If the buffer was big enough,the audio was not unblocked after the
 first buffer was sent:
It worked only if the buffer was consumed imediately in
 alsa_send_buffer and didn't made it loop on count.
Now it configures the audio at the end of the loop,which
 change frame->used and makes it become something else than
 BUF_INVALID_LEN,which unblocks it correctly without making the
 driver crash or whatever

Signed-off-by: gnutoo <gnutoo@gnutoo-desktop.(none)>
8 years agoAdded additional headers to headers_install
David Lanzendörfer [Sun, 17 Jan 2010 18:39:00 +0000 (19:39 +0100)]
Added additional headers to headers_install

8 years agomsmfb: added refresh thread to support X's fbdev (disabled by default)
David Lanzendörfer [Mon, 11 Jan 2010 14:47:31 +0000 (15:47 +0100)]
msmfb: added refresh thread to support X's fbdev (disabled by default)

rebased patch from  Martijn Stolk <linuxtogo@netripper.nl>
from the original commit message:
"As the touchscreen is now interrupt based instead of polling, the screen does not get updated in a regular fashion. The refresh thread has been added to compensate for this. As it's not a very
nice solution,it is disabled by default and can be enabled at compile-time by setting CONFIG_MSM_FB_REFRESH.

Note that this refresh thread is not neccesary for Android.
"

8 years agoAlsa fix to get mplayer working
David Lanzendörfer [Sat, 9 Jan 2010 23:48:31 +0000 (00:48 +0100)]
Alsa fix to get mplayer working

8 years agofixed new lines
David Lanzendörfer [Sat, 9 Jan 2010 23:20:33 +0000 (00:20 +0100)]
fixed new lines

8 years agofixed memset oops
David Lanzendörfer [Sat, 9 Jan 2010 23:20:11 +0000 (00:20 +0100)]
fixed memset oops

8 years agofixed building of msm wifi
David Lanzendörfer [Sat, 9 Jan 2010 18:22:45 +0000 (19:22 +0100)]
fixed building of msm wifi

8 years agoadded msm wifi patch from bob copeland
David Lanzendörfer [Sat, 9 Jan 2010 18:21:02 +0000 (19:21 +0100)]
added msm wifi patch from bob copeland

8 years agomade GNU conformity modification start
David Lanzendörfer [Sat, 9 Jan 2010 17:25:05 +0000 (18:25 +0100)]
made GNU conformity modification start

8 years agomodded some stuff (new lines and restructuration for readabelidy) master
David Lanzendörfer [Fri, 8 Jan 2010 19:02:21 +0000 (20:02 +0100)]
modded some stuff (new lines and restructuration for readabelidy)

8 years agofixed building of the qdsp5 alsa driver
David Lanzendörfer [Fri, 8 Jan 2010 18:04:42 +0000 (19:04 +0100)]
fixed building of the qdsp5 alsa driver

8 years agoimported qdsp5 base functions into the alsa driver itself (needs cleanup)
David Lanzendörfer [Fri, 8 Jan 2010 17:27:13 +0000 (18:27 +0100)]
imported qdsp5 base functions into the alsa driver itself (needs cleanup)

8 years agofixed building
David Lanzendörfer [Mon, 4 Jan 2010 21:02:35 +0000 (22:02 +0100)]
fixed building

8 years agoadded starting driver
David Lanzendörfer [Mon, 4 Jan 2010 09:07:59 +0000 (10:07 +0100)]
added starting driver

8 years agoFixed building of ethernet gadget for use as module and use as static built-in
David Lanzendörfer [Tue, 22 Dec 2009 12:07:58 +0000 (13:07 +0100)]
Fixed building of ethernet gadget for use as module and use as static built-in

8 years agofixed usb gadget driver
David Lanzendörfer [Mon, 14 Dec 2009 17:51:04 +0000 (18:51 +0100)]
fixed usb gadget driver

8 years agofixed oops on bootup
David Lanzendörfer [Mon, 14 Dec 2009 17:50:48 +0000 (18:50 +0100)]
fixed oops on bootup

8 years agochanged something
David Lanzendörfer [Mon, 14 Dec 2009 17:50:27 +0000 (18:50 +0100)]
changed something

8 years agosysfs: Fix sysfs owner and group.
Arve Hjønnevåg [Thu, 17 Dec 2009 05:15:49 +0000 (21:15 -0800)]
sysfs: Fix sysfs owner and group.

First call to sysfs_setattr ignored the requested owner and group.

Change-Id: I8343f818c65c54f6dbb55533cd34b31eeba5b554
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years agomsm_serial_hs: fix lockup on 2.6.32
Arve Hjønnevåg [Thu, 17 Dec 2009 05:15:07 +0000 (21:15 -0800)]
msm_serial_hs: fix lockup on 2.6.32

Change-Id: I38cc48fea1e5876b68f43194f505461373682602

8 years agorfkill: Introduce CONFIG_RFKILL_PM and use instead of CONFIG_PM to power down
Nick Pelly [Fri, 5 Dec 2008 01:37:05 +0000 (17:37 -0800)]
rfkill: Introduce CONFIG_RFKILL_PM and use instead of CONFIG_PM to power down

Some platforms do not want to power down rfkill devices on suspend.

Change-Id: I62a11630521c636d54a4a02ab9037a43435925f5
Signed-off-by: Nick Pelly <npelly@google.com>
8 years ago[ARM] msm: sdcc: Make slot status irq be a resource
Dima Zavin [Sat, 23 May 2009 03:21:50 +0000 (20:21 -0700)]
[ARM] msm: sdcc: Make slot status irq be a resource

Also, convert all SDCC IRQ resources to be named. No longer pass status_irq
in the platform_data

Change-Id: I0a53fd544742191ebf6fd14c9b5feac154a11371
Signed-off-by: Dima Zavin <dima@android.com>
8 years ago[ARM] msm_fb: Fix framebuffer console
Arve Hjønnevåg [Tue, 15 Dec 2009 06:20:03 +0000 (22:20 -0800)]
[ARM] msm_fb: Fix framebuffer console

Don't allow non panning updates to bypass the wait for the panel to turn on.

Change-Id: I32eedf37dd1b0d34cdcad782027b635e5c65f691

8 years agoconsoleearlysuspend: Fix for 2.6.32
Arve Hjønnevåg [Tue, 15 Dec 2009 06:14:52 +0000 (22:14 -0800)]
consoleearlysuspend: Fix for 2.6.32

vt_waitactive now needs a 1 based console number

Change-Id: I07ab9a3773c93d67c09d928c8d5494ce823ffa2e

8 years agomsm_serial_debugger: fix to build without CONFIG_PREEMPT
Arve Hjønnevåg [Tue, 15 Dec 2009 04:36:23 +0000 (20:36 -0800)]
msm_serial_debugger: fix to build without CONFIG_PREEMPT

Change-Id: I71e115a26142cccd809aa979dfa9541f842ae680

8 years agokeychord: fix to build without CONFIG_PREEMPT
Arve Hjønnevåg [Tue, 15 Dec 2009 04:27:04 +0000 (20:27 -0800)]
keychord: fix to build without CONFIG_PREEMPT

Change-Id: I911f13aeda4224b6fa57863bc7e8972fec8837fb

8 years agomt9t013: fix to build on 2.6.32
Arve Hjønnevåg [Tue, 15 Dec 2009 04:21:44 +0000 (20:21 -0800)]
mt9t013: fix to build on 2.6.32

Change-Id: I42dfb9f0f92a626510b6f5c186ee613ce79578c9

8 years agosched: Fix task priority bug
Peter Zijlstra [Fri, 4 Dec 2009 08:59:02 +0000 (09:59 +0100)]
sched: Fix task priority bug

83f9ac removed a call to effective_prio() in wake_up_new_task(), which
leads to tasks running at MAX_PRIO.

This is caused by the idle thread being set to MAX_PRIO before forking
off init. O(1) used that to make sure idle was always preempted, CFS
uses check_preempt_curr_idle() for that so we can savely remove this bit
of legacy code.

Reported-by: Mike Galbraith <efault@gmx.de>
Tested-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1259754383.4003.610.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
8 years agoRevert "Revert "sched: Set correct normal_prio and prio values in sched_fork()""
Arve Hjønnevåg [Tue, 15 Dec 2009 04:06:56 +0000 (20:06 -0800)]
Revert "Revert "sched: Set correct normal_prio and prio values in sched_fork()""

This reverts commit 8bf14ef1d2bb205e11e00e1aa05eb80aa43f9243.

8 years ago[ARM] msm: acpuclock: Fix cpu ramp down and up around power collapse.
Arve Hjønnevåg [Sat, 12 Dec 2009 05:16:18 +0000 (21:16 -0800)]
[ARM] msm: acpuclock: Fix cpu ramp down and up around power collapse.

Change-Id: I2e43df297489ed5a4d533beb0a92713d21864df8
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm: timer: Fix sleep duration from idle.
Arve Hjønnevåg [Sat, 5 Dec 2009 06:53:27 +0000 (22:53 -0800)]
[ARM] msm: timer: Fix sleep duration from idle.

Don't read TIMER_MATCH_VAL. This register may return an old value which
would cause the calculated sleep duration to be wrong.

Change-Id: Iaf079b28b47a677d485135912aa8b66b183cacf4
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years agoRevert "sched: Set correct normal_prio and prio values in sched_fork()"
Arve Hjønnevåg [Fri, 11 Dec 2009 07:24:34 +0000 (23:24 -0800)]
Revert "sched: Set correct normal_prio and prio values in sched_fork()"

This reverts commit f83f9ac2632732bd1678150b5a03d152f912fe72.

8 years agoserial debug fixes
Arve Hjønnevåg [Fri, 11 Dec 2009 06:04:27 +0000 (22:04 -0800)]
serial debug fixes

Change-Id: If15c0d0aedbf8ad443bb9c8d2dd47b109d593eb2

8 years agomsm_fb: fix for 2.6.32.
Arve Hjønnevåg [Fri, 11 Dec 2009 02:40:19 +0000 (18:40 -0800)]
msm_fb: fix for 2.6.32.

Change-Id: I5e8a2ef7d0f7a86883e0f77c7e68251b2a3c4584

8 years agomsm_serial_debugger: fix for 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 01:36:58 +0000 (17:36 -0800)]
msm_serial_debugger: fix for 2.6.32

Change-Id: I3c512234519f7a0d9f8bcb5718bef1fc2c446884

8 years agoboard-sapphire-rfkill: fix to build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 01:36:30 +0000 (17:36 -0800)]
board-sapphire-rfkill: fix to build on 2.6.32

Change-Id: I6ed063945b0461292aa784788a7e879db9e2e922

8 years agofunction/mass_storage: Fix to build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 01:22:47 +0000 (17:22 -0800)]
function/mass_storage: Fix to build on 2.6.32

Change-Id: If78296c2e464d4a70045fb676427f3f508fa7da6

8 years agomsm_serial_hs: fix to build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 01:13:01 +0000 (17:13 -0800)]
msm_serial_hs: fix to build on 2.6.32

Change-Id: I4c2555eb7cb9a333a2d85136afe6292f7db9b9f6

8 years agomsm_serial: fix to build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 01:06:42 +0000 (17:06 -0800)]
msm_serial: fix to build on 2.6.32

Change-Id: Ib97cd6d7e127573ebf4eae7f2a4a215e03dcb74f

8 years agomsm_rmnet: fix to build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 01:03:42 +0000 (17:03 -0800)]
msm_rmnet: fix to build on 2.6.32

Change-Id: Ic6a4903dd12ea83723354d00f639ae2f9375167f

8 years agopca963x: build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 00:54:05 +0000 (16:54 -0800)]
pca963x: build on 2.6.32

Change-Id: I89ff2705f8bd2de6eba90ef25928d81febfbc6c7

8 years agoakm8976: build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 00:53:46 +0000 (16:53 -0800)]
akm8976: build on 2.6.32

Change-Id: Iff5e3753fe770984198231b640190cc8e423010e

8 years agohci_ll.c: fix to build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 00:50:38 +0000 (16:50 -0800)]
hci_ll.c: fix to build on 2.6.32

Change-Id: I0b394dab02500c76ea8a1429747bd23616a39c52

8 years agohtc_pwrsink: Fix to build on 2.6.32
Arve Hjønnevåg [Fri, 11 Dec 2009 00:46:29 +0000 (16:46 -0800)]
htc_pwrsink: Fix to build on 2.6.32

Change-Id: I80f71adde17b3a1a99e91055c947a419ab71ba1f

8 years agoakm8976: disable_irq -> disable_irq_nosync
Arve Hjønnevåg [Sat, 18 Apr 2009 01:24:07 +0000 (18:24 -0700)]
akm8976: disable_irq -> disable_irq_nosync

Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years agoInput: gpio_input: disable_irq -> disable_irq_nosync
Arve Hjønnevåg [Sat, 18 Apr 2009 01:24:59 +0000 (18:24 -0700)]
Input: gpio_input: disable_irq -> disable_irq_nosync

Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm: mdp: disable_irq -> disable_irq_nosync
Arve Hjønnevåg [Sat, 18 Apr 2009 01:26:51 +0000 (18:26 -0700)]
[ARM] msm: mdp: disable_irq -> disable_irq_nosync

Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm: dma: disable_irq -> disable_irq_nosync
Arve Hjønnevåg [Sat, 18 Apr 2009 01:21:34 +0000 (18:21 -0700)]
[ARM] msm: dma: disable_irq -> disable_irq_nosync

Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm: trout: Pass wakeup irq to serial debugger
Arve Hjønnevåg [Thu, 10 Dec 2009 08:03:36 +0000 (00:03 -0800)]
[ARM] msm: trout: Pass wakeup irq to serial debugger

Change-Id: I695226bbd595d35ed6f9e29f09d13741c345de76

8 years agohack: serial debugger baud rate is wrong on dream
Arve Hjønnevåg [Thu, 10 Dec 2009 08:02:49 +0000 (00:02 -0800)]
hack: serial debugger baud rate is wrong on dream

Change-Id: I494319bff007f033003836ff21249cf19d3bbe27

8 years ago[ARM] msm: Switch to svc mode before calling fiq handler
Arve Hjønnevåg [Fri, 5 Jun 2009 05:55:24 +0000 (22:55 -0700)]
[ARM] msm: Switch to svc mode before calling fiq handler

Also add 8k fiq stack.

Change-Id: Ia08fb31d9d235a7fed1c77911db82b2b6470b678
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years agomsm_serial_debugger 2.6.32 fixes
Arve Hjønnevåg [Thu, 10 Dec 2009 07:42:23 +0000 (23:42 -0800)]
msm_serial_debugger 2.6.32 fixes

Change-Id: I3b5c05e007240ffb479911c360a4adeb723d163d

8 years ago[ARM] msm_serial_debugger: Add some debugger commands
Arve Hjønnevåg [Wed, 10 Jun 2009 03:34:00 +0000 (20:34 -0700)]
[ARM] msm_serial_debugger: Add some debugger commands

Add allregs to dump registers for all modes.
Add bt to get a stackstrace.

Change-Id: Ia85e72b6c8243eba38a04cf4f6cc9cba5342a6de
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm: Fix register dump in fiq debugger
Arve Hjønnevåg [Fri, 17 Apr 2009 05:53:47 +0000 (22:53 -0700)]
[ARM] msm: Fix register dump in fiq debugger

Change-Id: Iff5cd48291c9b09aace30220c4229c157a7db1d0
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm_serial_debugger: Add option to never disable wakeup IRQ.
Arve Hjønnevåg [Fri, 2 Oct 2009 22:25:43 +0000 (15:25 -0700)]
[ARM] msm_serial_debugger: Add option to never disable wakeup IRQ.

This works better if the radio ignores the uart clock request while
power collapsed.

Change-Id: Ib0989e714e883b3667c9ecc4cfd1ebfe014a35df
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm_serial_debugger: Add option to keep serial debugger active from boot.
Arve Hjønnevåg [Fri, 2 Oct 2009 04:19:21 +0000 (21:19 -0700)]
[ARM] msm_serial_debugger: Add option to keep serial debugger active from boot.

If msm_serial_debugger.no_sleep=1 is added to the kernel command line, or
MSM_SERIAL_DEBUGGER_NO_SLEEP is selected in the config, the serial debugger
is activated on boot and stays active until it receives a sleep command.

Change-Id: Ibf84435af8203360ee808fd903dd6322cf5d9d17
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm: serial_debugger: move the clock disable to after we enable the fiq
Dima Zavin [Fri, 14 Aug 2009 03:57:57 +0000 (20:57 -0700)]
[ARM] msm: serial_debugger: move the clock disable to after we enable the fiq

If the uart fifo is not empty when we enable the fiq, the handler will try to
empty it and hang since the clock disabled prior to enabling the fiq.

Signed-off-by: Dima Zavin <dima@android.com>
8 years ago[ARM] msm_serial_debugger: Add wakeup irq and disable uart clock when idle
Arve Hjønnevåg [Tue, 4 Aug 2009 04:07:31 +0000 (21:07 -0700)]
[ARM] msm_serial_debugger: Add wakeup irq and disable uart clock when idle

The serial debugger is now inactive by default so we can enter low power
modes. Hit enter twice to activate it for 5 seconds.

Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm_serial_debugger: Support 19.2MHz clock.
Arve Hjønnevåg [Thu, 25 Jun 2009 00:50:40 +0000 (17:50 -0700)]
[ARM] msm_serial_debugger: Support 19.2MHz clock.

Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years agodisable staging/dream
Arve Hjønnevåg [Thu, 10 Dec 2009 07:29:56 +0000 (23:29 -0800)]
disable staging/dream

Change-Id: I8f2fdae60761ba773fd0ee5b49a624f3d1798fd6

8 years agomsm_nand.c fix to build on 2.6.32
Arve Hjønnevåg [Thu, 10 Dec 2009 07:27:41 +0000 (23:27 -0800)]
msm_nand.c fix to build on 2.6.32

Change-Id: Ice6b449ae0b1a3c06646c30448995ccc7769461f

8 years agoboard-sapphire-mmc from 2.6.29
Arve Hjønnevåg [Thu, 10 Dec 2009 07:26:59 +0000 (23:26 -0800)]
board-sapphire-mmc from 2.6.29

Change-Id: I74c9d09c243536856f88188d71bc5cd9b7917851

8 years agomsm_serial_debugger build on 2.6.32
Arve Hjønnevåg [Thu, 10 Dec 2009 07:26:18 +0000 (23:26 -0800)]
msm_serial_debugger build on 2.6.32

Change-Id: I05b34d35181fbbe5d3b204a6a7a43a634d539654

8 years agoboard-sapphire-panel.c from 2.6.29
Arve Hjønnevåg [Thu, 10 Dec 2009 06:46:11 +0000 (22:46 -0800)]
board-sapphire-panel.c from 2.6.29

Change-Id: I28f707096722c744a2d6818b08eeb885e70a2452

8 years agohack: mmc files from 2.6.29
Arve Hjønnevåg [Thu, 10 Dec 2009 06:45:03 +0000 (22:45 -0800)]
hack: mmc files from 2.6.29

Change-Id: I5f1791108fb0577b6d520e2d357b32aef9498510

8 years agoboard-trout-panel from android-2.6.29
Arve Hjønnevåg [Thu, 10 Dec 2009 06:40:24 +0000 (22:40 -0800)]
board-trout-panel from android-2.6.29

Change-Id: I16c9290dfa3dd0bf7825d14445e7e7767c57756d

8 years ago[ARM] msm: Adding display drivers.
Rebecca Schultz [Thu, 20 Nov 2008 00:24:43 +0000 (16:24 -0800)]
[ARM] msm: Adding display drivers.

Adds mddi, mdp and msm_fb drivers.

Change-Id: I6da4c81f2cbc2777f1c99808d07fdf4256e42c84
Signed-off-by: Rebecca Schultz <rschultz@google.com>
8 years agoasm/mach/mmc.h is gone
Arve Hjønnevåg [Thu, 10 Dec 2009 05:55:17 +0000 (21:55 -0800)]
asm/mach/mmc.h is gone

Change-Id: Ieac3299a24db23afb2b4e5a3e5d904b585850931

8 years agoadd msm_fb.h from android-msm-2.6.29
Arve Hjønnevåg [Thu, 10 Dec 2009 05:49:16 +0000 (21:49 -0800)]
add msm_fb.h from android-msm-2.6.29

Change-Id: I2bfbd1b8b2157357fee2332333e5dfbe4b525fbf

8 years ago[ARM] msm: trout-rfkill: Build on 2.6.32
Arve Hjønnevåg [Thu, 10 Dec 2009 05:46:35 +0000 (21:46 -0800)]
[ARM] msm: trout-rfkill: Build on 2.6.32

Change-Id: Ibc87559f527b0e2de416d76e6cd12ae96e4ed792

8 years ago[ARM] msm: timer: Build on 2.6.32.
Arve Hjønnevåg [Thu, 10 Dec 2009 05:18:28 +0000 (21:18 -0800)]
[ARM] msm: timer: Build on 2.6.32.

Change-Id: I30ff766c8e064932d65bac895f5bb529d89e0a20

8 years ago[ARM] msm: clock: initialize the hlist correctly.
Chia-chi Yeh [Mon, 7 Dec 2009 18:32:59 +0000 (02:32 +0800)]
[ARM] msm: clock: initialize the hlist correctly.

Signed-off-by: Chia-chi Yeh <chiachi@android.com>
8 years ago[ARM] msm: clock: Use hlist instead of list
Arve Hjønnevåg [Wed, 21 Oct 2009 00:29:24 +0000 (17:29 -0700)]
[ARM] msm: clock: Use hlist instead of list

Change-Id: Id0bb9f2a6fe73ef05a818e7e5bd627c0052493b0
Signed-off-by: Arve Hjønnevåg <arve@android.com>
8 years ago[ARM] msm: msm_serial_hs: Improve RX flush procedure during clock off sequence.
Nick Pelly [Fri, 20 Nov 2009 21:31:36 +0000 (13:31 -0800)]
[ARM] msm: msm_serial_hs: Improve RX flush procedure during clock off sequence.

Clock off is requested by msm_hs_request_clock_off. The new procedure is:
1) Immediately deassert RTS to prevent new RX data.
2) Wait for the TX pipeline to flush (TXLEV, TXEMT).
3) Issue a FORCE_STALE_EVENT command, and flush on RXSTALE.
5) After this flush completes, disable DLINK and flush one more time
   (discarding data).
6) Finally clock off the uart.

The key changes are to disable DLINK - which fixes the datamover lockup, and
the issuing of the FORCE_STALE_EVENT command - which prevents RX data loss
during clock off sequence.

Signed-off-by: Nick Pelly <npelly@google.com>
8 years ago[ARM] msm: pm: msm_sleep() hooks for clock.c
Mike Chan [Mon, 16 Nov 2009 21:23:40 +0000 (13:23 -0800)]
[ARM] msm: pm: msm_sleep() hooks for clock.c

Signed-off-by: Mike Chan <mike@android.com>
8 years ago[ARM] msm: qdsp5: add a NULL-pointer check on module->ops->event
Kuma Chang [Fri, 13 Nov 2009 00:41:01 +0000 (16:41 -0800)]
[ARM] msm: qdsp5: add a NULL-pointer check on module->ops->event

Signed-off-by: Iliyan Malchev <malchev@google.com>
8 years ago[ARM] msm: acpuclock: Do not ramp up when going to power collapse/wfi
Mike Chan [Sat, 14 Nov 2009 01:15:22 +0000 (17:15 -0800)]
[ARM] msm: acpuclock: Do not ramp up when going to power collapse/wfi

Power collapse and wfi speeds specified in the board files are the fastest
speeds one can go before entering the mode. If the acpu is running at a
slower rate do not clock up, this can cause PLL depencency issues.

Signed-off-by: Mike Chan <mike@android.com>
8 years agommc: msm_sdcc: Clean up clock management and add a 10us delay after enabling clocks
San Mehat [Sun, 8 Nov 2009 21:00:37 +0000 (13:00 -0800)]
mmc: msm_sdcc: Clean up clock management and add a 10us delay after enabling clocks

It appears that in some cases there may be a delay on the ARM9 in enabling our clock.
As a result, we may put the controller into a bad state. Delay 10us after enabling
clocks to let the peripheral settle. Note - this is all imperical.

Also ensure set_ios() callback grabs the host lock.

Signed-off-by: San Mehat <san@google.com>
8 years ago[ARM] msm: msm_serial_hs: Do not allow suspend while BT DMA could be active.
Nick Pelly [Thu, 5 Nov 2009 04:30:42 +0000 (20:30 -0800)]
[ARM] msm: msm_serial_hs: Do not allow suspend while BT DMA could be active.

Signed-off-by: Nick Pelly <npelly@google.com>
8 years ago[ARM] msm: i2c: Work around i2c controller bug.
Arve Hjønnevåg [Wed, 4 Nov 2009 05:26:19 +0000 (21:26 -0800)]
[ARM] msm: i2c: Work around i2c controller bug.

When writing I2C_WRITE_DATA right before the i2c clock line is scheduled
to go high again, the controller does not stretch the clock so the data
setup time is within the i2c spec.

Change-Id: Ibfa35a036454ac39a8bb94b0d1fc220976ace88f
Signed-off-by: Arve Hjønnevåg <arve@android.com>