htc-msm-2-6-32:leviathan-incoming.git
7 years agosoc:msm8k refactorized nexusone-huge-WIP
Denis 'GNUtoo' Carikli [Sun, 26 Sep 2010 21:49:59 +0000 (23:49 +0200)]
soc:msm8k refactorized

7 years agodrivers/usb/gadget/composite : fix compilation header issues
Denis 'GNUtoo' Carikli [Sun, 26 Sep 2010 21:49:27 +0000 (23:49 +0200)]
drivers/usb/gadget/composite : fix compilation header issues

7 years agosound: soc:msm: nexusone: somewhat working
Denis 'GNUtoo' Carikli [Wed, 22 Sep 2010 16:45:47 +0000 (18:45 +0200)]
sound: soc:msm: nexusone: somewhat working

7 years ago[sound] soc msm: try harder
Denis 'GNUtoo' Carikli [Wed, 22 Sep 2010 11:01:19 +0000 (13:01 +0200)]
[sound] soc msm:  try harder

7 years agoimport alsa and make it compile
Denis 'GNUtoo' Carikli [Wed, 22 Sep 2010 03:54:21 +0000 (05:54 +0200)]
import alsa and make it compile

7 years agonet: wireless: bcm4329: Fix interrupt enabling in case of error
Dmitry Shmidt [Thu, 16 Sep 2010 22:47:05 +0000 (15:47 -0700)]
net: wireless: bcm4329: Fix interrupt enabling in case of error

Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
7 years agonet: wireless: bcm4329: Update to Version 4.218.248.6
Greg Goldman [Wed, 15 Sep 2010 22:05:19 +0000 (15:05 -0700)]
net: wireless: bcm4329: Update to Version 4.218.248.6

Fix watchdog time rescheduling, fix first scan no-return

Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
7 years agonet: wireless: bcm4329: Enable packet filtering during low power mode
Greg Goldman [Tue, 14 Sep 2010 23:55:46 +0000 (16:55 -0700)]
net: wireless: bcm4329: Enable packet filtering during low power mode

Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
7 years agonet: wireless: bcm4329: Ignore error if scan results are empty
Greg Goldman [Mon, 13 Sep 2010 23:11:16 +0000 (16:11 -0700)]
net: wireless: bcm4329: Ignore error if scan results are empty

Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
7 years agonet: wireless: bcm4329: Add debug print for cscan failure
Dmitry Shmidt [Sun, 12 Sep 2010 21:23:20 +0000 (14:23 -0700)]
net: wireless: bcm4329: Add debug print for cscan failure

Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
7 years agonet: wireless: bcm4329: Fix roaming setting on resume
Dmitry Shmidt [Sun, 12 Sep 2010 21:18:23 +0000 (14:18 -0700)]
net: wireless: bcm4329: Fix roaming setting on resume

Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
7 years agobounce: fix cache maintenance for bounced highmem pages
Gary King [Mon, 6 Sep 2010 22:37:12 +0000 (15:37 -0700)]
bounce: fix cache maintenance for bounced highmem pages

the bounced page should be flushed after memcpy, to ensure that
architecture implementations can synchronize instruction and data
caches of the copied data.

Change-Id: I095a3463ec7fdaff3d0ff416f5648bc661fb5f8a
Signed-off-by: Gary King <gking@nvidia.com>
7 years agoThe eMMC spec 4.4 and 4.3 + additional feature chips has CSD structure
Kyungmin Park [Tue, 7 Sep 2010 18:06:32 +0000 (11:06 -0700)]
The eMMC spec 4.4 and 4.3 + additional feature chips has CSD structure
version 3 and version 3 have to check the CSD_STRUCTURE byte in the
EXT_CSD register.

Also fix EXT_CSD revision message.

Change-Id: Ic5d3a1893d6c5d238e1f59fa1f2307e75d76bf96
[akpm@linux-foundation.org: fix comment, per Chris Ball]
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Cc: Adrian Hunter <adrian.hunter@nokia.com>
Cc: Chris Ball <cjb@laptop.org>
Cc: <linux-mmc@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Colin Cross <ccross@android.com>
7 years agommc: subtract boot sectors from disk size for eMMC 4.3+ devices
Gary King [Fri, 16 Jul 2010 00:11:32 +0000 (17:11 -0700)]
mmc: subtract boot sectors from disk size for eMMC 4.3+ devices

the csd sector count reported by eMMC 4.3+ cards includes the boot
partition size; subtract this from the size reported to the disk
since the boot partition is inaccessible

Change-Id: I601b83aa0159b7aa446409ea8c945b256dd0b5b1
Signed-off-by: Gary King <gking@nvidia.com>
7 years agoSet the touch panel input gain correctly
Giorgos Mavrikas [Wed, 15 Sep 2010 23:07:37 +0000 (02:07 +0300)]
Set the touch panel input gain correctly

Change-Id: I301d59d951283face27aae112287050e3e597769

7 years agoMerge "scripts: setlocalversion: remove version appendage for non-tagged builds"...
Steve Kondik [Sun, 5 Sep 2010 14:09:41 +0000 (14:09 +0000)]
Merge "scripts: setlocalversion: remove version appendage for non-tagged builds" into android-msm-2.6.35

7 years agogitignore: add compressed boot files
Patrick Jacques [Sun, 5 Sep 2010 10:25:07 +0000 (03:25 -0700)]
gitignore: add compressed boot files

Add:
arch/arm/boot/compressed/lib1funcs.S
arch/arm/boot/compressed/piggy.gzip
arch/arm/boot/compressed/vmlinux

Change-Id: I37e113a317bfcd3fc54c81bec1054859ec631119

7 years agoscripts: setlocalversion: remove version appendage for non-tagged builds
Patrick Jacques [Sun, 5 Sep 2010 09:08:17 +0000 (02:08 -0700)]
scripts: setlocalversion: remove version appendage for non-tagged builds

Remove the plus sign from LOCALVERSION if the repository is not
in a clean tagged state.

Change-Id: I538cdd94eaa3df9292136781e68294511f63c386

7 years agoRevert "net: wireless: bcm4329: Fix roaming failure case"
Steve Kondik [Fri, 3 Sep 2010 20:40:38 +0000 (16:40 -0400)]
Revert "net: wireless: bcm4329: Fix roaming failure case"

This reverts commit 3816f9dd8cf2a8eb203d0e5595bbabad15b54d04.

7 years agoAdd CM configs
Steve Kondik [Sun, 5 Sep 2010 04:20:20 +0000 (00:20 -0400)]
Add CM configs

7 years agommc_block: Allow more than 8 partitions per card
Colin Cross [Fri, 3 Sep 2010 19:41:21 +0000 (12:41 -0700)]
mmc_block: Allow more than 8 partitions per card

Set the GENHD_FL_EXT_DEVT flag, which will allocate minor numbers
in major 259 for partitions past disk->minors.

Also remove the use of disk_devt to determine devidx from md->disk.
md->disk->first_minor is always initialized from devidx and can
always be used to recover it.

Signed-off-by: Colin Cross <ccross@android.com>
7 years agoEFI/GPT: Register named partitions.
Todd Poynor [Fri, 27 Aug 2010 00:25:50 +0000 (17:25 -0700)]
EFI/GPT: Register named partitions.

Change-Id: If318512de67d3d152192aa66b074427f4e5b5c2e
Signed-off-by: Todd Poynor <toddpoynor@google.com>
7 years agofs: partitions: Add support for named partitions
Colin Cross [Wed, 7 Apr 2010 19:05:32 +0000 (12:05 -0700)]
fs: partitions: Add support for named partitions

Adds a new file in /sys/block/<block>/<partition> called partition_name
that contains the name of the partition, if specified by the partition
handler.

Change-Id: I6648ed95eabefd1d00edbfdfd99eeb971d15f4b3
Signed-off-by: Colin Cross <ccross@android.com>
7 years agoVIDEO: MSM: Fix memory leak in mdp_probe
Martin Etti [Fri, 3 Sep 2010 09:29:05 +0000 (02:29 -0700)]
VIDEO: MSM: Fix memory leak in mdp_probe

Fix for not releasing mdp when failing to get clk

Change-Id: I6973f6d4db7ef70d06301717525e56a6619bf177

7 years agoARM: MSM: Fix on build warning regarding VMALLOC_END type
Patrick Jacques [Fri, 3 Sep 2010 09:03:23 +0000 (02:03 -0700)]
ARM: MSM: Fix on build warning regarding VMALLOC_END type

Fix this warning:

        arch/arm/mm/init.c: In function 'mem_init':
        arch/arm/mm/init.c:644: warning: format '%08lx' expects type
        'long unsigned int', but argument 12 has type 'unsigned int'

And remove the useless white space.

Adjusted for MSM vmalloc, based on samsung's 8eebae246e5edf8

Change-Id: I0a36977e1fc0960bf31cdaeaeba853aa5b33c692
Reported-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
7 years agoadd unset_consumer_device_supply declaration
Prashant SOmashekar [Fri, 3 Sep 2010 15:12:21 +0000 (11:12 -0400)]
add unset_consumer_device_supply declaration

7 years agoremove unset_supplies and add further logic (taken from 2.6.34) for additions within...
Prashant SOmashekar [Fri, 3 Sep 2010 14:52:58 +0000 (10:52 -0400)]
remove unset_supplies and add further logic (taken from 2.6.34) for additions within consumer devices, to unset the device_supply/supplies and scrub.

7 years agoreduce boot up clock to 806
Prashant SOmashekar [Thu, 2 Sep 2010 22:38:27 +0000 (18:38 -0400)]
reduce boot up clock to 806

7 years agotps65023-regulator: clear the client data in i2c
Prashant SOmashekar [Thu, 2 Sep 2010 21:58:46 +0000 (17:58 -0400)]
tps65023-regulator: clear the client data in i2c

7 years agoarm: mm: Don't free prohibited memmap entries
Michael Bohan [Tue, 31 Aug 2010 04:43:20 +0000 (21:43 -0700)]
arm: mm: Don't free prohibited memmap entries

The VM subsystem assumes that there are valid memmap entries to
the bank end aligned to MAX_ORDER_NR_PAGES. It will try and read
these page structs, and so we cannot free any memmap entries that
it may inspect.

Signed-off-by: Michael Bohan <mbohan@codeaurora.org>
Signed-off-by: Daniel Walker <dwalker@codeaurora.org>
Change-Id: I695198cf78d69cb5b279adcf376c6458d340da91

7 years agoset -2000/+800 for BT_* in q6audio and board-mahimahi-audio
Prashant SOmashekar [Sat, 28 Aug 2010 17:26:26 +0000 (13:26 -0400)]
set -2000/+800 for BT_* in q6audio and board-mahimahi-audio

7 years agods2784: set current_avg_ua <=40 for 2nd block check
Prashant SOmashekar [Fri, 27 Aug 2010 11:28:16 +0000 (07:28 -0400)]
ds2784: set current_avg_ua <=40 for 2nd block check

7 years agoremove * 1000 from AXI decrease for acpuclock-arm11.c
Prashant SOmashekar [Wed, 25 Aug 2010 02:22:53 +0000 (22:22 -0400)]
remove * 1000 from AXI decrease for acpuclock-arm11.c

7 years agocharge battery to 100%; merged from od and increased 95 to 99 for percentage - commit...
Prashant SOmashekar [Wed, 25 Aug 2010 02:22:16 +0000 (22:22 -0400)]
charge battery to 100%; merged from od and increased 95 to 99 for percentage - commit: ffb0ef63ab15c0a03df6ab25dd7325d06f199e79 (git://github.com/xdabravoteam/cm-kernel.git)

7 years agovmscan: raise the bar to PAGEOUT_IO_SYNC stalls
Wu Fengguang [Sun, 22 Aug 2010 00:37:13 +0000 (17:37 -0700)]
vmscan: raise the bar to PAGEOUT_IO_SYNC stalls

Fix "system goes unresponsive under memory pressure and lots of
dirty/writeback pages" bug.

http://lkml.org/lkml/2010/4/4/86

In the above thread, Andreas Mohr described that

Invoking any command locked up for minutes (note that I'm
talking about attempted additional I/O to the _other_,
_unaffected_ main system HDD - such as loading some shell
binaries -, NOT the external SSD18M!!).

This happens when the two conditions are both meet:
 - under memory pressure
 - writing heavily to a slow device

OOM also happens in Andreas' system. The OOM trace shows that 3
processes are stuck in wait_on_page_writeback() in the direct reclaim
path. One in do_fork() and the other two in unix_stream_sendmsg(). They
are blocked on this condition:

(sc->order && priority < DEF_PRIORITY - 2)

which was introduced in commit 78dc583d (vmscan: low order lumpy reclaim
also should use PAGEOUT_IO_SYNC) one year ago. That condition may be too
permissive. In Andreas' case, 512MB/1024 = 512KB. If the direct reclaim
for the order-1 fork() allocation runs into a range of 512KB
hard-to-reclaim LRU pages, it will be stalled.

It's a severe problem in three ways.

Firstly, it can easily happen in daily desktop usage.  vmscan priority
can easily go below (DEF_PRIORITY - 2) on _local_ memory pressure. Even
if the system has 50% globally reclaimable pages, it still has good
opportunity to have 0.1% sized hard-to-reclaim ranges. For example, a
simple dd can easily create a big range (up to 20%) of dirty pages in
the LRU lists.

Secondly, once triggered, it will stall unrelated processes (not doing IO
at all) in the system. This "one slow USB device stalls the whole system"
avalanching effect is very bad.

Thirdly, once stalled, the stall time could be intolerable long for the
users.  When there are 20MB queued writeback pages and USB 1.1 is
writing them in 1MB/s, wait_on_page_writeback() will stuck for up to 20
seconds.  Not to mention it may be called multiple times.

So raise the bar to only enable PAGEOUT_IO_SYNC when priority goes below
DEF_PRIORITY/3, or 6.25% LRU size. As the default dirty throttle ratio is
20%, it will hardly be triggered by pure dirty pages. We'd better treat
PAGEOUT_IO_SYNC as some last resort workaround -- its stall time is so
uncomfortably long (easily goes beyond 1s).

The bar is only raised for (order < PAGE_ALLOC_COSTLY_ORDER) allocations,
which are easy to satisfy in 1TB memory boxes. So, although 6.25% of
memory could be an awful lot of pages to scan on a system with 1TB of
memory, it won't really have to busy scan that much.

Reported-by: Andreas Mohr <andi@xxxxxxxx>
Signed-off-by: Mel Gorman <mel@xxxxxxxxx>
Signed-off-by: Wu Fengguang <fengguang.wu@xxxxxxxxx>
7 years agoDon't force V4L2 on for msm_camera_old
Prashant SOmashekar [Fri, 20 Aug 2010 23:16:51 +0000 (19:16 -0400)]
Don't force V4L2 on for msm_camera_old

7 years agomsm_kgsl: Add writecombine flag to map vmalloced memory
Shubhraprakash Das [Fri, 20 Aug 2010 19:38:26 +0000 (15:38 -0400)]
msm_kgsl: Add writecombine flag to map vmalloced memory

Adding writecombine flag to map vmalloc memory into user space
improves performance

Signed-off-by: Shubhraprakash Das <sadas@quicinc.com>
7 years agoremove case CMD_AXI_CFG_O1_AND_O2 from 720P_CAMERA (msm_vfe8x.c)
Prashant Somashekar [Fri, 20 Aug 2010 03:15:46 +0000 (23:15 -0400)]
remove case CMD_AXI_CFG_O1_AND_O2 from 720P_CAMERA (msm_vfe8x.c)

7 years agoqdsp6: Fix allocation in the encoder/decoder.
Steve Kondik [Thu, 19 Aug 2010 22:02:00 +0000 (18:02 -0400)]
qdsp6: Fix allocation in the encoder/decoder.

7 years agoincrease voltage to 975 for 245 and 256; increase voltage to 1000 for 384; increase...
Prashant Somashekar [Wed, 18 Aug 2010 22:55:02 +0000 (18:55 -0400)]
increase voltage to 975 for 245 and 256; increase voltage to 1000 for 384; increase voltage to 1250 for 844 (acpuclock-qsd8x250).  set b frame path to MSM_FRAME_ENC for 720P_CAMERA (msm_vfe8x.c)

7 years agofix board-sapphire.c
Prashant Somashekar [Wed, 18 Aug 2010 06:12:29 +0000 (02:12 -0400)]
fix board-sapphire.c

7 years agofix once more camera.h
Prashant Somashekar [Wed, 18 Aug 2010 06:07:04 +0000 (02:07 -0400)]
fix once more camera.h

7 years agoone more fix camera.h
Prashant Somashekar [Wed, 18 Aug 2010 06:02:32 +0000 (02:02 -0400)]
one more fix camera.h

7 years agofix msm_camera.h
Prashant Somashekar [Wed, 18 Aug 2010 05:59:46 +0000 (01:59 -0400)]
fix msm_camera.h

7 years agofix msm_q6venc.c
Prashant Somashekar [Wed, 18 Aug 2010 05:50:45 +0000 (01:50 -0400)]
fix msm_q6venc.c

7 years agoone more fix: dal.h
Prashant Somashekar [Wed, 18 Aug 2010 05:49:58 +0000 (01:49 -0400)]
one more fix: dal.h

7 years agofix dal.h
Prashant Somashekar [Wed, 18 Aug 2010 05:49:18 +0000 (01:49 -0400)]
fix dal.h

7 years agofix board.h
Prashant Somashekar [Wed, 18 Aug 2010 05:47:53 +0000 (01:47 -0400)]
fix board.h

7 years agofix io.c
Prashant Somashekar [Wed, 18 Aug 2010 05:35:30 +0000 (01:35 -0400)]
fix io.c

7 years agoidle-v7.S
Prashant Somashekar [Wed, 18 Aug 2010 05:29:34 +0000 (01:29 -0400)]
idle-v7.S

7 years agoinput: synaptics_i2c_rmi: duplicated_filter to throttle touch events
Patrick Jacques [Tue, 17 Aug 2010 07:41:24 +0000 (00:41 -0700)]
input: synaptics_i2c_rmi: duplicated_filter to throttle touch events

Create multitouch threshold to prevent event hub from being flooded.
This attempts to (partially) address Android issue #7836:
http://code.google.com/p/android/issues/detail?id=7836

Depends on CONFIG_TOUCHSCREEN_DUPLICATED_FILTER=y

Originally authored by HTC for their Hero 2.6.29 kernel, ported to
kernel-hero by netarchy/ninpo, adapted to .34 by kernelzilla.

7 years agomahimahi: Enable full light sensor range
Jonas Larsson [Sat, 14 Aug 2010 18:30:36 +0000 (20:30 +0200)]
mahimahi: Enable full light sensor range

7 years agomahimahi: Fix the memory layout.
Steve Kondik [Mon, 16 Aug 2010 16:44:41 +0000 (12:44 -0400)]
mahimahi: Fix the memory layout.

Yeah, this was totally wrong. Thanks Defer and Kali- for spotting this.

7 years agoFix atomic functions in bfq
Steven Barrett [Tue, 10 Aug 2010 02:12:16 +0000 (21:12 -0500)]
Fix atomic functions in bfq

7 years agoblock: introduce the BFQ I/O scheduler
paolo [Fri, 2 Jul 2010 14:53:47 +0000 (16:53 +0200)]
block: introduce the BFQ I/O scheduler

Add the BFQ-v1-r1 I/O scheduler to 2.6.35.
The general structure is borrowed from CFQ, as much of the code. A (bfq_)queue
is associated to each task doing I/O on a device, and each time a scheduling
decision has to be taken a queue is selected and it is served until it expires.

The main differences are:
     - slices are given in the service domain: tasks are assigned budgets,
       measured in number of sectors. Once got the disk, a task must
       however consume its assigned budget within a configurable maximum time
       (by default, the maximum possible value of the budgets is automaticall
       computed to comply with this timeout). This allows the desired latency
       vs "throughput boosting" tradeoff to be set.

    - Budgets are scheduled according to a variant of WF2Q+, implemented
      using an augmented rb-tree to take eligibility into account while
      preserving an O(log N) overall complexity.

    - BFQ supports full hierarchical scheduling, exporting a cgroups
      interface.  Each node has a full scheduler, so each group can
      be assigned its own ioprio and an ioprio_class.

    - If the cgroups interface is used, weights can be explictly assigned,
      otherwise ioprio values are mapped to weights using the relation
      weight = IOPRIO_BE_NR - ioprio.

    - ioprio classes are served in strict priority order, i.e., lower
      priority queues are not served as long as there are higher priority
      queues.  Among queues in the same classes the bandwidth is distributed
      in proportion to the weights of each queue.

Regarding what has not changed it is worth noting:
    - the handling of cfq_io_contexts to associate queues to tasks.
      Much
      of the code has been reused just renaming it.  (There is room for
      code sharing with CFQ but we wanted to minimize the impact of this
      patch.)

    - The handling of async queues.

    - The handling of idle windows.

    - The handling of merging.

    - The heuristics to assert that a task is worth an idle window (with
      minor modifications to hw_tag/CIC_SEEKY detection).

Signed-off-by: Fabio Checconi <fabio@gandalf.sssup.it>
Signed-off-by: Paolo Valente <paolo.valente@unimore.it>
Ported the following commits to bfq:
    - cfq-iosched: compact io_context radix_tree (80b15c7389caa)
    - cfq-iosched: remove dead_key from cfq_io_context (bca4b914b5da3d8e)

7 years agoblock: add cgroups, kconfig and build bits for BFQ
Paolo Valente [Fri, 26 Feb 2010 21:19:41 +0000 (22:19 +0100)]
block: add cgroups, kconfig and build bits for BFQ

Add a Kconfig option and do the related Makefile changes to compile
the BFQ I/O scheduler.  Also let the cgroups subsystem know about the
BFQ I/O controller.

Signed-off-by: Fabio Checconi <fabio@gandalf.sssup.it>
Signed-off-by: Paolo Valente <paolo.valente@unimore.it>
7 years agoblock: prepare I/O context code for BFQ
Prashant Somashekar [Wed, 18 Aug 2010 05:13:39 +0000 (01:13 -0400)]
block: prepare I/O context code for BFQ

7 years agomahimahi: Give 1MB back to pmem_adsp for decoder
Steve Kondik [Sat, 14 Aug 2010 10:12:06 +0000 (06:12 -0400)]
mahimahi: Give 1MB back to pmem_adsp for decoder

7 years agomahimahi: Rearrange memory, kang another 12MB
Steve Kondik [Thu, 12 Aug 2010 01:54:43 +0000 (21:54 -0400)]
mahimahi: Rearrange memory, kang another 12MB

7 years agomsm: camera: Updated VFE driver code from HTC
Steve Kondik [Thu, 12 Aug 2010 01:08:27 +0000 (21:08 -0400)]
msm: camera: Updated VFE driver code from HTC

7 years agomsm: Update vdec/venc drivers (HTC)
Prashant Somashekar [Wed, 18 Aug 2010 04:49:41 +0000 (00:49 -0400)]
msm: Update vdec/venc drivers (HTC)

7 years agomahimahi: pmem_camera isn't needed anymore, kang 8MB back.
Prashant Somashekar [Wed, 18 Aug 2010 04:36:58 +0000 (00:36 -0400)]
mahimahi: pmem_camera isn't needed anymore, kang 8MB back.

7 years agocamera: Add generic msm_flashlight driver.
Prashant Somashekar [Wed, 18 Aug 2010 03:51:49 +0000 (23:51 -0400)]
camera: Add generic msm_flashlight driver.

7 years agocamera: Add HTC AWB calibration.
Prashant Somashekar [Wed, 18 Aug 2010 03:46:42 +0000 (23:46 -0400)]
camera: Add HTC AWB calibration.

7 years agocamera: Porting of HTC camera code for 720P.
Prashant Somashekar [Wed, 18 Aug 2010 03:43:58 +0000 (23:43 -0400)]
camera: Porting of HTC camera code for 720P.

7 years agoremove V4L bindings, msm Makefile
Prashant Somashekar [Tue, 17 Aug 2010 04:27:18 +0000 (00:27 -0400)]
remove V4L bindings, msm Makefile

7 years agodrivers: usb: gadget: mass_storage: Fix section mismatch warning
Prashant Somashekar [Tue, 17 Aug 2010 04:15:22 +0000 (00:15 -0400)]
drivers: usb: gadget: mass_storage: Fix section mismatch warning

7 years agomsm: camera: add support for building old camera driver
Prashant Somashekar [Sun, 15 Aug 2010 18:33:05 +0000 (14:33 -0400)]
msm: camera: add support for building old camera driver

7 years agoAdd PCA963X LED driver.
Prashant Somashekar [Sun, 15 Aug 2010 18:13:39 +0000 (14:13 -0400)]
Add PCA963X LED driver.

7 years agorevert to using f_mass_storage_tmp
Prashant Somashekar [Sun, 15 Aug 2010 18:11:39 +0000 (14:11 -0400)]
revert to using f_mass_storage_tmp

7 years agoUpdate 32A/EBI1 patch from Bcrook and Pershoot.
Prashant Somashekar [Sun, 15 Aug 2010 16:38:53 +0000 (12:38 -0400)]
Update 32A/EBI1 patch from Bcrook and Pershoot.

7 years agoAdd TURBO_MODE support from desirec kernel. + oc/uv stuff for acpuclock-arm11.c
Prashant Somashekar [Sun, 15 Aug 2010 15:50:54 +0000 (11:50 -0400)]
Add TURBO_MODE support from desirec kernel. + oc/uv stuff for acpuclock-arm11.c

7 years agomsm: sapphire: Fix myTouch 3G LE Fender/1.2 zones
Prashant Somashekar [Sun, 15 Aug 2010 15:43:20 +0000 (11:43 -0400)]
msm: sapphire: Fix myTouch 3G LE Fender/1.2 zones

7 years agoarm: msm: smd: convert unsigned addr to unsigned long
Daniel Walker [Wed, 17 Mar 2010 16:58:29 +0000 (09:58 -0700)]
arm: msm: smd: convert unsigned addr to unsigned long

"unsigned" translates to "unsigned int", but this value holds an
address. We always want to use unsigned long for addresses since
it will change size to fit the machine.

This just convert the one address holder to unsigned long.

Signed-off-by: Daniel Walker <dwalker@codeaurora.org>
7 years agovdec: Add VDEC_IOCTL_GETVERSION
Steve Kondik [Sat, 14 Aug 2010 10:25:22 +0000 (06:25 -0400)]
vdec: Add VDEC_IOCTL_GETVERSION

7 years agoyaffs: sync with yaffs repo
Patrick Jacques [Wed, 11 Aug 2010 21:19:43 +0000 (14:19 -0700)]
yaffs: sync with yaffs repo

Merge recent changes by Charles Manning for yaffs and
2.6.34 kernel, including:

commit 10baaa5fc285ffe29fa5c6d0ab0c58bcce78e3e0
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs: Skip checkpoint writing if mounted readonly

commit 4c78eb9aa16a108e258ab6185f37a51c3b413cb8
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs: Batter read-only support handling

commit ca0474074161712816ce1ead07dcc83c4e2ebe29
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs: Fix return value for yaffs direct sync

commit 66dfee0783bfda493d1b8f8f900906f103b23c6b
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs: Add new follow link code for 2.6.34 support

commit 107910497be0a0d0e4a8ab357c403fc2a234323b
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs  Fix yaffs1 soft delete issue

commit 570d9b811895119f6b3411678cbdcaa27226f403
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs  Changes to direct tests

commit 9d8bb3ae8a8362d1899d6565ca6ca1217fb172f0
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Get fuzz tester to run further

commit 2d221c0839d2e8733f52ee31e72e7f81faecb6cc
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs: Fix issues kicked up by fuzz testing

commit d22f3456a462be87d8fa7c69c06bd22a95522a0f
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Implement phase 1 fuzz testing.

commit 497c3cfe39bd95fe6e5a4e9a5938fddf97bbed11
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Update RAM simulator to support start and end blocks

commit 2597593cb367f99f68bc3251c27e8b694c7c2a51
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs: Fix AutoUnicode handling

commit a6fbf618f9e46629a254f1ad246f4ca93a897881
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Fix yaffs direct creating opening files with no name

commit 505255d500d864c1689a6752a728b97859605fd0
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Add another test to basic tests

commit 993de0d40a67e466cc22de0df144a6dc363e3193
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Keep unused linux alloactor code up to date

commit 2da43953d71ac4430ef38dd81990baba9393b1d1
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Fix in-kernel makefile to use correct allocator

commit 32b9c53162d647a640c2c506c76502dbbc594902
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Fix xattrib headers for use with unicode

commit 227cbf526a2c0228495ff2405c866a93c4cd576b
Author: Charles Manning <cdhmanning@gmail.com>

    yaffs Move yaffscfg.h to where it should be

And more from http://yaffs.net/gitweb?p=yaffs2/.git

7 years agotcp: bind() fix when many ports are bound
Patrick Jacques [Sun, 8 Aug 2010 20:39:34 +0000 (13:39 -0700)]
tcp: bind() fix when many ports are bound

Port autoselection done by kernel only works when number of bound
sockets is under a threshold (typically 30000).

When this threshold is over, we must check if there is a conflict before
exiting first loop in inet_csk_get_port()

Change inet_csk_bind_conflict() to forbid two reuse-enabled sockets to
bind on same (address,port) tuple (with a non ANY address)

Same change for inet6_csk_bind_conflict()

author Eric Dumazet <eric.dumazet@gmail.com>
commit fda48a0d7a8412cedacda46a9c0bf8ef9cd13559

7 years agos5k3e2fx: Add ioctl for reading fuse id
Steve Kondik [Sat, 7 Aug 2010 05:50:05 +0000 (01:50 -0400)]
s5k3e2fx: Add ioctl for reading fuse id

7 years agobcm4329: disable combo scan, for now.
Steve Kondik [Sat, 31 Jul 2010 00:13:43 +0000 (20:13 -0400)]
bcm4329: disable combo scan, for now.

7 years agoFix bluetooth sleep bug (from kernelzilla). Included in official kernel here http...
kmobs [Tue, 27 Jul 2010 03:09:17 +0000 (22:09 -0500)]
Fix bluetooth sleep bug (from kernelzilla).  Included in official kernel here git./?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=485f1eff73a7b932fd3abb0dfcf804e1a1f59025

7 years agocpufreq: interactive: New 'interactive' governor
Mike Chan [Tue, 22 Jun 2010 18:26:45 +0000 (11:26 -0700)]
cpufreq: interactive: New 'interactive' governor

New interactive governor.

This governor is designed for latency sensitive workloads, UI interaction for
example.

Advantages:
+ significantly more responsive to ramp cpu up when required (UI interaction)
+ more consistent ramping, existing governors do their cpu load sampling in a
workqueue context, the 'interactive' governor does this in a timer context, which
gives more consistent cpu load sampling.
+ higher priority for cpu frequency increase, rt_workqueue is used for scaling
up, giving the remaining tasks the cpu performance benefit, unlike existing
governors which schedule rampup work to occur after your performance starved
tasks have completed.

Existing governors sample cpu load at a particular rate, typically
every X ms. Which can lead to under powering UI threads when the user has
interacted with an idle system until the next sample period happns.

The 'interactive' governor has a different approach. Instead of sampling the cpu
at a specified rate, the governor will scale the cpu frequency up when coming
out of idle. When the cpu comes out of idle, a timer is configured to fire
within 1-2 ticks. If the cpu is 100% busy from exiting idle to when the timer
fires then we assume the cpu is underpowered and ramp to MAX speed.

If the cpu was not 100% busy, then the governor evaluates the cpu load over the
last 'min_sample_rate' (default 50000 uS) to determine the cpu speed to ramp down
to.

There is only one tuneable for this governor:
/sys/devices/system/cpu/cpufreq/interactive/min_sample_rate:
The minimum ammount of time to spend at the current frequency before
ramping down. This is to ensure that the governor has seen enough
historic cpu load data to determine the appropriate workload.
Default is 5000 uS.

Signed-off-by: Mike Chan <mike@android.com>
Change-Id: I686d2f57b0ed9cbb73217403b7438be5719588d2

7 years ago[ARM] msm: kgsl: Remove support for mmap from the kgsl driver.
Ken Sumrall [Fri, 23 Jul 2010 00:36:52 +0000 (17:36 -0700)]
[ARM] msm: kgsl: Remove support for mmap from the kgsl driver.

The mmap(2) functionality of kgsl is no longer used, and it had a
bug that would crash the kernel when mmap(2) was called with
"odd" parameters.  So we have removed the mmap functionality.
If for some reason in the future you want to resurrect this,
make sure memdesc is properly initialized in all paths through
the kgsl_mmap() routine.

Signed-off-by: Ken Sumrall <ken@android.com>
7 years ago[ARM] mahimahi: add support for Sony TFT panel
Roy Chiang [Thu, 10 Jun 2010 21:29:52 +0000 (14:29 -0700)]
[ARM] mahimahi: add support for Sony TFT panel

Signed-off-by: Iliyan Malchev <malchev@google.com>
7 years agoUpdate to compcache to 0.6.2
ABAAKOUK Mehdi [Mon, 19 Jul 2010 14:52:57 +0000 (16:52 +0200)]
Update to compcache to 0.6.2

7 years ago[ARM] msm: disable the allocator for pmem_camera
Iliyan Malchev [Thu, 15 Jul 2010 15:28:29 +0000 (08:28 -0700)]
[ARM] msm: disable the allocator for pmem_camera

Signed-off-by: Iliyan Malchev <malchev@google.com>
7 years agoUpdate tiwlan1251 to latest version from Froyo.
Steve Kondik [Wed, 30 Jun 2010 02:25:58 +0000 (22:25 -0400)]
Update tiwlan1251 to latest version from Froyo.

7 years agocyanogen_msm: Update configuration for Froyo
Steve Kondik [Wed, 30 Jun 2010 02:25:05 +0000 (22:25 -0400)]
cyanogen_msm: Update configuration for Froyo

7 years agoFixes HW keyboard layout detection with DangerSPL
philenotfound [Sat, 19 Jun 2010 18:52:56 +0000 (20:52 +0200)]
Fixes HW keyboard layout detection with DangerSPL

7 years agocyanogen_msm: Enable AUFS module
Steve Kondik [Mon, 21 Jun 2010 21:47:52 +0000 (17:47 -0400)]
cyanogen_msm: Enable AUFS module

7 years agoexport cap_file_mmap() for aufs
Dominic Duklas [Sat, 13 Feb 2010 20:13:24 +0000 (20:13 +0000)]
export cap_file_mmap() for aufs
Signed-off-by: Dominic Duklas <dodo1122@zen-kernel.org>
7 years agoexpost do_splice_from
Dominic Duklas [Sun, 24 Jan 2010 21:59:16 +0000 (21:59 +0000)]
expost do_splice_from

7 years agocyanogen_mahimahi: Enable AUFS module.
Steve Kondik [Mon, 21 Jun 2010 21:02:15 +0000 (17:02 -0400)]
cyanogen_mahimahi: Enable AUFS module.

7 years agoallow aufs to be compiled as a module
Dominic Duklas [Sun, 24 Jan 2010 19:04:47 +0000 (19:04 +0000)]
allow aufs to be compiled as a module

7 years agoIncrease camera clock to improve recording framerate.
Steve Kondik [Fri, 18 Jun 2010 16:17:03 +0000 (12:17 -0400)]
Increase camera clock to improve recording framerate.

7 years ago[ARM] msm: usb: Reset the PHY before suspend on Scorpion.
Steve Kondik [Fri, 18 Jun 2010 14:25:36 +0000 (10:25 -0400)]
[ARM] msm: usb: Reset the PHY before suspend on Scorpion.

7 years agosapphire: Fix gamma/interlace issues on sapphire
Steve Kondik [Tue, 15 Jun 2010 15:08:00 +0000 (11:08 -0400)]
sapphire: Fix gamma/interlace issues on sapphire

Code from HTC via cadetnudt@XDA and shinshin@Hiapk

7 years agoRevert "board-sapphire: Set mux to debug uart"
Steve Kondik [Tue, 15 Jun 2010 15:01:13 +0000 (11:01 -0400)]
Revert "board-sapphire: Set mux to debug uart"

This reverts commit 438319d29fa95f4cdffdfda20a5ca7d141685cb4.

7 years agoRevert "[ARM] msm: usb: reset the PHY before suspend"
Steve Kondik [Tue, 15 Jun 2010 14:21:26 +0000 (10:21 -0400)]
Revert "[ARM] msm: usb: reset the PHY before suspend"

This reverts commit b223d9c1359e68bb9573161c3cc2de5fbe2c0dc3.

7 years agoDisable MTD_LAZYECCSTATS.
Steve Kondik [Wed, 9 Jun 2010 23:40:13 +0000 (19:40 -0400)]
Disable MTD_LAZYECCSTATS.

7 years agocyanogen_msm: turn LBDAF back on.
Steve Kondik [Wed, 9 Jun 2010 21:27:32 +0000 (17:27 -0400)]
cyanogen_msm: turn LBDAF back on.