7 years agomodified: arch/arm/mach-omap2/board-omap3beagle.c kernel_omap3
greatmay [Sun, 15 May 2011 01:40:29 +0000 (08:40 +0700)]
modified:   arch/arm/mach-omap2/board-omap3beagle.c
titipong  add pmem driver map vitrual memory
modified:   drivers/dsp/bridge/rmgr/drv_interface.c
titipong  patch function in C

7 years ago[BEAGLE] Fix build failure in pm34xx.c
Namhyung Kim [Mon, 1 Nov 2010 04:07:52 +0000 (12:07 +0800)]
[BEAGLE] Fix build failure in pm34xx.c

omap_wdt_suspend/resume() are compiled only when OMAP_WATCHDOG is
defined so calling them in omap_sram_idle() breaks build as
followings. Add #ifdefs to workaround this.

 LD      .tmp_vmlinux1
arch/arm/mach-omap2/built-in.o: In function `omap_sram_idle':
arch/arm/mach-omap2/pm34xx.c:394: undefined reference to
arch/arm/mach-omap2/pm34xx.c:579: undefined reference to
make[2]: *** [.tmp_vmlinux1] Error 1
make[1]: *** [sub-make] Error 2
make: *** [all] Error 2

Signed-off-by: Namhyung Kim <>
Signed-off-by: Jim Huang <>
7 years agomm: keep a guard page below a grow-down stack segment
Linus Torvalds [Fri, 13 Aug 2010 00:54:33 +0000 (17:54 -0700)]
mm: keep a guard page below a grow-down stack segment

This is a rather minimally invasive patch to solve the problem of the
user stack growing into a memory mapped area below it.  Whenever we fill
the first page of the stack segment, expand the segment down by one

Now, admittedly some odd application might _want_ the stack to grow down
into the preceding memory mapping, and so we may at some point need to
make this a process tunable (some people might also want to have more
than a single page of guarding), but let's try the minimal approach

Tested with trivial application that maps a single page just below the
stack, and then starts recursing.  Without this, we will get a SIGSEGV
_after_ the stack has smashed the mapping.  With this patch, we'll get a
nice SIGBUS just as the stack touches the page just above the mapping.

Requested-by: Keith Packard <>
Signed-off-by: Linus Torvalds <>
7 years agoASoC: TWL4030: Enable audio capture for both devkit8000 and beagleboard
Jeremy Chang [Wed, 4 Aug 2010 15:15:00 +0000 (23:15 +0800)]
ASoC: TWL4030: Enable audio capture for both devkit8000 and beagleboard

For beagleboard, REG_ANAMICL is 0x34 and REG_ANAMICR is 0x14
For devkit8000,  REG_ANAMICL is 0x31 and REG_ANAMICR is 0x00
To enable for both, merge the value of them

7 years agoASoC: TWL4030: Fix audio capture playback for devkit8000
Jeremy Chang [Wed, 4 Aug 2010 09:27:53 +0000 (17:27 +0800)]
ASoC: TWL4030: Fix audio capture playback for devkit8000

7 years agoDevkit8000: Enable config modules
Erin Yueh [Fri, 30 Jul 2010 10:23:39 +0000 (18:23 +0800)]
Devkit8000: Enable config modules

7 years ago[beagleboard, devkit8000] Enable BT HID support for Apple Wireless Keyboard
Erin Yueh [Fri, 30 Jul 2010 09:57:02 +0000 (17:57 +0800)]
[beagleboard, devkit8000] Enable BT HID support for Apple Wireless Keyboard

7 years ago[omap3_beagle_defconfig] Enable early suspend mechanism.
Jeremy Chang [Wed, 28 Jul 2010 07:41:42 +0000 (15:41 +0800)]
[omap3_beagle_defconfig] Enable early suspend mechanism.

7 years agoDevkit8000: Adjust ads7846 config to more sensitive value
Kan-Ru Chen [Thu, 22 Jul 2010 08:04:39 +0000 (16:04 +0800)]
Devkit8000: Adjust ads7846 config to more sensitive value

7 years agoInput: ads7846 - fix pressure reporting
Pavel Machek [Mon, 23 Nov 2009 16:17:38 +0000 (08:17 -0800)]
Input: ads7846 - fix pressure reporting

On Zaurus, hx4700 and others pressure is reported inverted -- the lighter
the pressure, the bigger numerical value.

Signed-off-by: Pavel Machek <>
Signed-off-by: Dmitry Torokhov <>
7 years ago[BEAGLE] Get rid of registeration of I2C-Bus2.
Matt Hsu [Wed, 14 Jul 2010 09:56:10 +0000 (17:56 +0800)]
[BEAGLE] Get rid of registeration of I2C-Bus2.

Fix the the merge mistake.

7 years agoMTD: silence ecc errors on mtdblock0
Steve Sakoman [Mon, 26 Apr 2010 18:17:26 +0000 (11:17 -0700)]
MTD: silence ecc errors on mtdblock0

mtdblock0 is the x-load partition, which uses hw ecc
this confuses linux, which uses sw ecc
this patch silences ecc error messages when linux peeks into mtdblock0
* not for upstream submission *

7 years agomodedb.c: add proper 720p60 mode
Koen Kooi [Mon, 8 Mar 2010 13:38:31 +0000 (14:38 +0100)]
modedb.c: add proper 720p60 mode

7 years agoFix for bus width which improves SD card's peformance.
Kishore Kadiyala [Wed, 17 Feb 2010 14:04:47 +0000 (19:34 +0530)]
Fix for bus width which improves SD card's peformance.

This patch fixes bus width which improves peformance for SD cards.
OMAP-MMC controller's can support maximum bus width of '8'.
when bus width is mentioned as "8" in controller data,the SD
stack will check whether bus width is "4" and if not it will
set bus width to "1" and there by degrading peformance.

Signed-off-by: Kishore Kadiyala <>
7 years agommc: don't display single block read console messages
Steve Sakoman [Tue, 5 Jan 2010 03:20:25 +0000 (19:20 -0800)]
mmc: don't display single block read console messages

mmc: don't display single block read console messages

7 years agodrivers: input: touchscreen: ads7846: return ENODEV if device is not found
Steve Sakoman [Tue, 15 Dec 2009 23:24:10 +0000 (15:24 -0800)]
drivers: input: touchscreen: ads7846: return ENODEV if device is not found

Signed-off-by: Steve Sakoman <>
7 years agoMTD: NAND: omap2: proper fix for subpage read ECC errors
Steve Sakoman [Thu, 31 Dec 2009 15:05:02 +0000 (07:05 -0800)]
MTD: NAND: omap2: proper fix for subpage read ECC errors

7 years agodevkit8000: Enable wakeup from touchscreen
Jim Huang [Fri, 9 Jul 2010 03:26:56 +0000 (11:26 +0800)]
devkit8000: Enable wakeup from touchscreen

This patch enables the wakeup capabilities of ads7846 touchscreen
driver.  ads7846 driver can now wakeup the system from suspend on

7 years agoInput: ads7846 - add wakeup support
Ranjith Lohithakshan [Thu, 25 Feb 2010 04:24:53 +0000 (09:54 +0530)]
Input: ads7846 - add wakeup support

Add wakeup support to the ads7846 driver. Platforms can enable wakeup
capability by setting the wakeup flag in ads7846_platform_data. With this
patch the ads7846 driver can be used to wake the system from suspend.

Signed-off-by: Ranjith Lohithakshan <>
7 years agodevkit8000: enabled MMC_UNSAFE_RESUME
Jim Huang [Fri, 9 Jul 2010 03:14:24 +0000 (11:14 +0800)]
devkit8000: enabled MMC_UNSAFE_RESUME

7 years agobeagleboard: enabled MMC_UNSAFE_RESUME
Jim Huang [Fri, 9 Jul 2010 03:14:05 +0000 (11:14 +0800)]
beagleboard: enabled MMC_UNSAFE_RESUME

7 years ago[devkit8000_defconfig] Enable media and usb video
Jeremy Chang [Wed, 7 Jul 2010 06:19:28 +0000 (14:19 +0800)]
[devkit8000_defconfig] Enable media and usb video

7 years ago[BEAGLE] Add checking of external pull-up for i2c bus2.
Matt Hsu [Thu, 8 Jul 2010 09:11:41 +0000 (17:11 +0800)]
[BEAGLE] Add checking of external pull-up for i2c bus2.

If the SDA/SCL on I2C bu2 are not pulled-up externally,
I2C bus2 should not be registered.
Without external pull-up reistor, it would result in choas if
I2C controller tries to talk to slave devices.

Signed-off-by: Matt Hsu <>
7 years ago[MUX] Add I2C2_SCL_GPIO168 mux configs for checking external pull-up.
Matt Hsu [Thu, 8 Jul 2010 09:08:46 +0000 (17:08 +0800)]
[MUX] Add I2C2_SCL_GPIO168 mux configs for checking external pull-up.

Signed-off-by: Matt Hsu <>
7 years agoDevkit8000: modedb: Swap timing parameter for 480x272 display
Kan-Ru Chen [Wed, 7 Jul 2010 07:07:27 +0000 (15:07 +0800)]
Devkit8000: modedb: Swap timing parameter for 480x272 display

7 years ago[i2c] enable support of i2c port 2
Tick Chen [Wed, 7 Jul 2010 06:19:02 +0000 (14:19 +0800)]
[i2c] enable support of i2c port 2
Run the i2c interface with 400khz clock

7 years agoMerge branch 'die-id' into omap3-2.6.32
Kan-Ru Chen [Wed, 7 Jul 2010 04:10:27 +0000 (12:10 +0800)]
Merge branch 'die-id' into omap3-2.6.32

7 years agoomap: Devkit8000: Use DIE id to initialize dm9000 MAC address
Kan-Ru Chen [Mon, 5 Jul 2010 03:47:25 +0000 (11:47 +0800)]
omap: Devkit8000: Use DIE id to initialize dm9000 MAC address

The devkit8000 debug boards often come with empty EEPROM thus without
valid ethernet MAC address. The DIE id to MAC formula is copied from

7 years agoomap: Use omap_get_die_id() to get the DIE ids
Kan-Ru Chen [Mon, 5 Jul 2010 16:52:15 +0000 (00:52 +0800)]
omap: Use omap_get_die_id() to get the DIE ids

7 years agoomap: Add new interface omap_get_die_id
Kan-Ru Chen [Mon, 5 Jul 2010 16:15:13 +0000 (00:15 +0800)]
omap: Add new interface omap_get_die_id

Allow DIE id to be get and used by others.

7 years ago[Devkit8000] Use DIE ID to initialize dm9000 mac address
Kan-Ru Chen [Mon, 5 Jul 2010 03:47:25 +0000 (11:47 +0800)]
[Devkit8000] Use DIE ID to initialize dm9000 mac address

7 years ago[Devkit8000] Add usb_mass_storage function to android gadget
Kan-Ru Chen [Fri, 2 Jul 2010 09:17:49 +0000 (17:17 +0800)]
[Devkit8000] Add usb_mass_storage function to android gadget

7 years ago[Devkit8000] Update defconfig
Kan-Ru Chen [Fri, 2 Jul 2010 08:38:42 +0000 (16:38 +0800)]
[Devkit8000] Update defconfig

7 years ago[Devkit8000] Add android composite gadget support
Kan-Ru Chen [Fri, 2 Jul 2010 08:34:55 +0000 (16:34 +0800)]
[Devkit8000] Add android composite gadget support

7 years ago[Devkit8000] Use TPS65930 config instead of TWL4030
Kan-Ru Chen [Fri, 2 Jul 2010 08:34:31 +0000 (16:34 +0800)]
[Devkit8000] Use TPS65930 config instead of TWL4030

7 years ago[Devkic8000] modedb: Add panel config
Kan-Ru Chen [Fri, 2 Jul 2010 07:59:23 +0000 (15:59 +0800)]
[Devkic8000] modedb: Add panel config

7 years agosound: Add devkit8000 soc support
Kan-Ru Chen [Fri, 2 Jul 2010 06:47:37 +0000 (14:47 +0800)]
sound: Add devkit8000 soc support



7 years agoDevkit8000: Port to 2.6.32
Kan-Ru Chen [Mon, 24 May 2010 09:06:33 +0000 (17:06 +0800)]
Devkit8000: Port to 2.6.32

7 years agoDevkit8000: Add back ledB support
Kan-Ru Chen [Fri, 23 Apr 2010 08:04:17 +0000 (16:04 +0800)]
Devkit8000: Add back ledB support

7 years agoDevkit8000: Add proper active_low setup to gpio_keys
Kan-Ru Chen [Wed, 21 Apr 2010 07:19:04 +0000 (15:19 +0800)]
Devkit8000: Add proper active_low setup to gpio_keys

7 years agoDevkit8000: Replace hard coded gpio number with define
Kan-Ru Chen [Wed, 21 Apr 2010 07:18:35 +0000 (15:18 +0800)]
Devkit8000: Replace hard coded gpio number with define

7 years agoDevkit8000: Report keyboard event instead of button event on gpio_keys
Kan-Ru Chen [Wed, 21 Apr 2010 07:17:44 +0000 (15:17 +0800)]
Devkit8000: Report keyboard event instead of button event on gpio_keys

7 years agoDevkit8000: Adjust touchscreen sensitivity for 7 inch panel
Kan-Ru Chen [Wed, 21 Apr 2010 03:11:53 +0000 (11:11 +0800)]
Devkit8000: Adjust touchscreen sensitivity for 7 inch panel

7 years agoDevkit8000: Enable twl4030 pwrbutton in defconfig
Kan-Ru Chen [Wed, 21 Apr 2010 02:50:01 +0000 (10:50 +0800)]
Devkit8000: Enable twl4030 pwrbutton in defconfig

7 years agoDevkit8000: Enable boot_key gpio button
Kan-Ru Chen [Wed, 21 Apr 2010 02:46:37 +0000 (10:46 +0800)]
Devkit8000: Enable boot_key gpio button

7 years agoDevkit8000: Remove unused settings.
Kan-Ru Chen [Tue, 13 Apr 2010 09:08:32 +0000 (17:08 +0800)]
Devkit8000: Remove unused settings.

TFT LCD doesn't need to set the AC bias.

7 years agoEnable DVI-D output for Devkit8000
Kan-Ru Chen [Wed, 7 Apr 2010 08:40:18 +0000 (16:40 +0800)]
Enable DVI-D output for Devkit8000

Devkit8000 has different DVI reset pin with the BeagleBoard. On Devkit8000
the TWL4030 GPIO_7 is assigned to do the job.

Signed-off-by: Kan-Ru Chen <>
7 years agoDevkit8000: Add board specific MUX settings.
Kan-Ru Chen [Tue, 6 Apr 2010 10:36:44 +0000 (18:36 +0800)]
Devkit8000: Add board specific MUX settings.

Currently the kernel assumes the bootloader will setup the MUX properly.
However this is not always true, so we have to get this done in the kernel.

7 years agoDevkit8000: WIP Adjust the touchscreen sensitivity.
Kan-Ru Chen [Tue, 6 Apr 2010 01:34:34 +0000 (09:34 +0800)]
Devkit8000: WIP Adjust the touchscreen sensitivity.

7 years agoDevkit8000: Use generic DSS2 driver instead of new driver for every panel
Kan-Ru Chen [Mon, 22 Mar 2010 02:48:45 +0000 (10:48 +0800)]
Devkit8000: Use generic DSS2 driver instead of new driver for every panel

Remove the special driver for 4.3 inch panel. Instead we use the
generic driver and modedb database to specify the timing

7 years agoDevkit8000: Let the user gpio-key send BTN_BACK event
Kan-Ru Chen [Wed, 10 Mar 2010 04:07:21 +0000 (12:07 +0800)]
Devkit8000: Let the user gpio-key send BTN_BACK event

7 years agoAdd Devkit8000 default config
Kan-Ru Chen [Thu, 25 Feb 2010 10:01:34 +0000 (18:01 +0800)]
Add Devkit8000 default config

7 years agoAdd minimal support for DevKit8000
Kan-Ru Chen [Fri, 12 Feb 2010 03:29:14 +0000 (11:29 +0800)]
Add minimal support for DevKit8000

Reuse patch from Thomas Weber.

These patches add board support for the Timll DevKit8000.

The DevKit8000 is a beagle board clone from Timll, sold by The DevKit8000 has RS232 serial port, LCD, DVI-D,
S-Video, Ethernet, SD/MMC, keyboard, camera, SPI, I2C, USB and
JTAG interface.

8 years agoAdd back ledB gpio config.
Kan-Ru Chen [Fri, 2 Apr 2010 05:49:50 +0000 (13:49 +0800)]
Add back ledB gpio config.

8 years ago[INPUT] Multiply XYZ output with mg_wight and declare them as signed value.
Matt Hsu [Thu, 10 Dec 2009 09:33:23 +0000 (17:33 +0800)]
[INPUT] Multiply XYZ output with mg_wight and declare them as signed value.

Signed-off-by: Matt Hsu <>
8 years agoRegister LIS302DL on I2C Bus2 for the Beagleboard.
Matt Hsu [Wed, 2 Dec 2009 09:40:34 +0000 (17:40 +0800)]
Register LIS302DL on I2C Bus2 for the Beagleboard.

Signed-off-by: Matt Hsu <>
8 years ago[MUX] Change the mux config of GPIO_136_OUT as INPUT.
Matt Hsu [Wed, 12 May 2010 09:25:26 +0000 (17:25 +0800)]
[MUX] Change the mux config of GPIO_136_OUT as INPUT.

Signed-off-by: Jim Huang <>
8 years ago [Input] Init commit for 3-axis accelerometer device LIS302DL.
Matt Hsu [Wed, 2 Dec 2009 08:30:57 +0000 (16:30 +0800)]
 [Input] Init commit for 3-axis accelerometer device LIS302DL.

Signed-off-by: Matt Hsu <>
8 years ago[Input] Add support for LIS302DL to Kconfig and Makefile.
Matt Hsu [Wed, 12 May 2010 09:04:41 +0000 (17:04 +0800)]
[Input] Add support for LIS302DL to Kconfig and Makefile.

Signed-off-by: Jim Huang <>
8 years agoSolve videobuf related linking error for UVC driver.
Jeremy Chang [Mon, 30 Nov 2009 07:23:12 +0000 (15:23 +0800)]
Solve videobuf related linking error for UVC driver.

8 years ago[OMAP3] Fix the regression of USB device detection if MUX implementation is enabled
Matt Hsu [Wed, 12 May 2010 08:59:30 +0000 (16:59 +0800)]
[OMAP3] Fix the regression of USB device detection if MUX implementation is enabled

The MUX logic changes the HSUSB2 pins function as GPIOs.
This breaks the USB EHCI function such as the USB mouse would not be detecte
Fix this issue simply by getting rid of this config.

Signed-off-by: Jim Huang <>
8 years ago[HSMMC] Having new debug routine for omap HSMMC.
Matt Hsu [Wed, 12 May 2010 08:49:22 +0000 (16:49 +0800)]
[HSMMC] Having new debug routine for omap HSMMC.

This patch leverages MMC_DEBUG_CHANNEL parameter to provide
MMC related debug routine which could be specified to
the MMC slot you are interested.

Signed-off-by: Jim Huang <>
8 years ago[HSMMC] Correct support voltage in the platform data of HSMMC2
Matt Hsu [Fri, 31 Jul 2009 05:44:15 +0000 (13:44 +0800)]
[HSMMC] Correct support voltage in the platform data of HSMMC2

Previous ocr_mask ranges the support voltage only 1v6~1v95.
This caused an error of "host doesn't support this operating voltage"
in the initial communication between host and card.
It's fixed by providing appropriate ocm_mask.

Signed-off-by: Matt Hsu <>
8 years agoRemove GPIO config specific for EHCI USB port
Matt Hsu [Wed, 12 May 2010 08:34:02 +0000 (16:34 +0800)]
Remove GPIO config specific for EHCI USB port

Signed-off-by: Jim Huang <>
8 years ago[OMAP2] Beagleboard: Add support of MMC2 slot
Matt Hsu [Wed, 12 May 2010 08:30:43 +0000 (16:30 +0800)]
[OMAP2] Beagleboard: Add support of MMC2 slot

MMC2 slot is allocated for BRCM4325 WLAN device.

Signed-off-by: Matt Hsu <>
Signed-off-by: Jim Huang <>
8 years agoMove MMC2 clock initialization form twl_mmc23_set_power to twl4030_mmc_init.
Matt Hsu [Fri, 10 Jul 2009 04:05:31 +0000 (12:05 +0800)]
Move MMC2 clock initialization form twl_mmc23_set_power to twl4030_mmc_init.

Since MMC2 internal clock enable is requested to have regulator.
In some cases, such as beagleboard doesn't have support of
regulator (vmmc2). Therefore, moving internal clock enabling
from set_power to mmc_init makes more sense.

Signed-off-by: Matt Hsu <>
8 years agoAdd support for debugging MMC channel individually.
Matt Hsu [Wed, 12 May 2010 08:19:55 +0000 (16:19 +0800)]
Add support for debugging MMC channel individually.

In some stituation, you may be interested in debugging
messages for one MMC host. This patch enable the MMC debugging
output form specific MMC channel.

Per host controller driver also can leverage MMC_DEBUG_CHANNEL and
manipulate your debugging routine.

Signed-off-by: Matt Hsu <>
Signed-off-by: Jim Huang <>
8 years ago[ARM] omap2: omap3beagle: Enable Android Gadget and adb transport
Jim Huang [Tue, 11 May 2010 15:05:17 +0000 (23:05 +0800)]
[ARM] omap2: omap3beagle: Enable Android Gadget and adb transport

USB Ethernet is substituted by Android USB gadget.

Below is "lsusb -v" output for reference:
Bus 002 Device 020: ID 18d1:9018
Device Descriptor:
  bLength                18
  bDescriptorType         1
  bcdUSB               2.00
  bDeviceClass            0 (Defined at Interface level)
  bDeviceSubClass         0
  bDeviceProtocol         0
  bMaxPacketSize0        64
  idVendor           0x18d1
  idProduct          0x9018
  bcdDevice            2.16
  iManufacturer           1 0xlab
  iProduct                2 0xdroid gadget
  iSerial                 3 20090427
  bNumConfigurations      1

Signed-off-by: Jeremy Chang <>
8 years agoUSB: gadget: android: adb: Fix build
Jeremy Chang [Tue, 11 May 2010 14:48:22 +0000 (22:48 +0800)]
USB: gadget: android: adb: Fix build

8 years agoomap3_beagle_defconfig: update config file.
Jeremy Chang [Mon, 10 May 2010 08:36:08 +0000 (16:36 +0800)]
omap3_beagle_defconfig: update config file.

8 years agoRevert "[ARM] omap2: HACK don't comple generic usb-ehci support"
Jeremy Chang [Fri, 7 May 2010 16:07:58 +0000 (00:07 +0800)]
Revert "[ARM] omap2: HACK don't comple generic usb-ehci support"

This reverts commit 2a6cafe0328ae4f5846d54e25253efe222e01fc6.

8 years agoFix some usb build error.
Jeremy Chang [Fri, 7 May 2010 15:36:03 +0000 (23:36 +0800)]
Fix some usb build error.

8 years agoRevert "usb: ehci-omap: add support for more port configurations"
Jeremy Chang [Fri, 7 May 2010 10:15:03 +0000 (18:15 +0800)]
Revert "usb: ehci-omap: add support for more port configurations"

This reverts commit e1a5debc84007d94c83bb4065d90f1953c4bf48b.

8 years agoUpdate beagle board file for DSS model.
Jeremy Chang [Thu, 6 May 2010 09:11:25 +0000 (17:11 +0800)]
Update beagle board file for DSS model.

8 years agoRevert "[ARM] omap: HACK: Switch sholes to the omap serial driver"
Jeremy Chang [Tue, 4 May 2010 12:20:31 +0000 (20:20 +0800)]
Revert "[ARM] omap: HACK: Switch sholes to the omap serial driver"

This reverts commit 3e9adf317ef6930917923788033b67f669fa01e5.

8 years agoRevert "omap_serial: Use dedicated workqueue thread for tty RX."
Jeremy Chang [Fri, 7 May 2010 15:43:28 +0000 (23:43 +0800)]
Revert "omap_serial: Use dedicated workqueue thread for tty RX."

This reverts commit 397a4c8b4f26f32e0d1bd85a0a1011970d6e7d7f.

8 years agoRevert "[ARM] sholes_defconfig: Enable CONFIG_PROFILING"
Rebecca Schultz Zavin [Thu, 6 May 2010 23:47:47 +0000 (16:47 -0700)]
Revert "[ARM] sholes_defconfig: Enable CONFIG_PROFILING"

This reverts commit 2bab5cbbd6a9af4e28d4468e8934f74b685dedc4.

8 years agoMerge commit 'android-2.6.32' into android-omap-2.6.32
Rebecca Schultz Zavin [Thu, 6 May 2010 23:46:01 +0000 (16:46 -0700)]
Merge commit 'android-2.6.32' into android-omap-2.6.32

Change-Id: I6893b425a415372a8300b5ff0acffd3a1baa3d6d

8 years agostaging: android: lowmemorykiller: Remove bitrotted codepath
San Mehat [Thu, 6 May 2010 22:43:46 +0000 (15:43 -0700)]
staging: android: lowmemorykiller: Remove bitrotted codepath

Now that we're murder-synchronous, this code path will never be
called (and if it does, it doesn't tell us anything useful other
than we killed a task that was already being killed by somebody
else but hadn't gotten its' signal yet)

Signed-off-by: San Mehat <>
8 years agostaging: android: lowmemorykiller: Fix task_struct leak
San Mehat [Thu, 6 May 2010 22:40:07 +0000 (15:40 -0700)]
staging: android: lowmemorykiller: Fix task_struct leak

As it turns out, the CONFIG_PROFILING interfaces leak a
task struct if the notifier chain returns NOTIFY_OK.. doh.

This patch reworks lowmemkiller to use the new generic task
free notifier chain.

Signed-off-by: San Mehat <>
8 years agosched: Add a generic notifier when a task struct is about to be freed
San Mehat [Thu, 6 May 2010 22:37:55 +0000 (15:37 -0700)]
sched: Add a generic notifier when a task struct is about to be freed

This patch adds a notifier which can be used by subsystems that may
be interested in when a task has completely died and is about to
have it's last resource freed.

  The Android lowmemory killer uses this to determine when a task
it has killed has finally given up its goods.

Signed-off-by: San Mehat <>
8 years agobinder: Fix memory corruption via page aliasing
Christopher Lais [Sat, 1 May 2010 20:51:48 +0000 (15:51 -0500)]
binder: Fix memory corruption via page aliasing

binder_deferred_release was not unmapping the page from the buffer
before freeing it, causing memory corruption.  This only happened
when page(s) had not been freed by binder_update_page_range, which
properly unmaps the pages.

This only happens on architectures with VIPT aliasing.

To reproduce, create a program which opens, mmaps, munmaps, then closes
the binder very quickly.  This should leave a page allocated when the
binder is released.  When binder_deferrred_release is called on the
close, the page will remain mapped to the address in the linear
proc->buffer.  Later, we may map the same physical page to a different
virtual address that has different coloring, and this may cause
aliasing to occur.

PAGE_POISONING will greatly increase your chances of noticing any

Change-Id: I6941bf212881b8bf846bdfda43d3609c7ae4892e

Signed-off-by: Christopher Lais <>
8 years ago[ARM] omap: pm: HACK: Disable watchdog in pm34xx suspend
Mike Chan [Mon, 22 Mar 2010 22:51:54 +0000 (15:51 -0700)]
[ARM] omap: pm: HACK: Disable watchdog in pm34xx suspend

Disable hardware watchgod in the cpu suspend code, right before wfi
instead of driver suspend hooks. This allows us to catch any lockups in
the suspend / resume path that are caused by driver suspend / resume hooks.

Hard lockups turn into reboot in production, so users do not have to pull
their battery.

Change-Id: Ie36df35153adf5da79681b1064dc83ef31f1023c
Signed-off-by: Mike Chan <>
8 years agonet: ppp_deflate: Switch z_decomp_alloc to vmalloced decompress buffers
Rebecca Schultz Zavin [Thu, 29 Apr 2010 20:36:55 +0000 (13:36 -0700)]
net: ppp_deflate: Switch z_decomp_alloc to vmalloced decompress buffers

On systems with high memory pressure, allocating an order 4 contiguous region
for decompression can fail at run time.  Instead use virtually contigous memory
allocated via vmalloc.

Change-Id: I7821882bd183a1de0d73ab2932f6e6978e4f0970
Signed-off-by: Rebecca Schultz Zavin <>
8 years ago[ARM] sholes_defconfig: Enable CONFIG_PROFILING
Rebecca Schultz Zavin [Wed, 5 May 2010 19:49:44 +0000 (12:49 -0700)]
[ARM] sholes_defconfig: Enable CONFIG_PROFILING

Change-Id: I927809e055478a64b349f71d8f7ddde2909604a2
Signed-off-by: Rebecca Schultz Zavin <>
8 years agoMerge commit 'android-2.6.32' into android-omap-2.6.32
Rebecca Schultz Zavin [Wed, 5 May 2010 19:40:11 +0000 (12:40 -0700)]
Merge commit 'android-2.6.32' into android-omap-2.6.32

Change-Id: I4ca7c1b2a360cd21aa138d1ee54c075ee3ed3e23

8 years agostaging: android: lowmemkiller: Substantially reduce overhead during reclaim
San Mehat [Wed, 5 May 2010 18:38:42 +0000 (11:38 -0700)]
staging: android: lowmemkiller: Substantially reduce overhead during reclaim

This patch optimizes lowmemkiller to not do any work when it has an outstanding
kill-request. This greatly reduces the pressure on the task_list lock
(improving interactivity), as well as improving the vmscan performance
when under heavy memory pressure (by up to 20x in tests).

Note: For this enhancement to work, you need CONFIG_PROFILING

Signed-off-by: San Mehat <>
8 years ago[ARM] Do not call flush_cache_user_range with mmap_sem held
Dima Zavin [Wed, 28 Apr 2010 03:57:04 +0000 (20:57 -0700)]
[ARM] Do not call flush_cache_user_range with mmap_sem held

We can't be holding the mmap_sem while calling flush_cache_user_range
because the flush can fault. If we fault on a user address, the
page fault handler will try to take mmap_sem again. Since both places
acquire the read lock, most of the time it succeeds. However, if another
thread tries to acquire the write lock on the mmap_sem (e.g. mmap) in
between the call to flush_cache_user_range and the fault, the down_read
in do_page_fault will deadlock.

Also, since we really can't be holding the mmap_sem while calling
flush_cache_user_range AND vma is actually unused by the flush itself,
get rid of vma as an argument.

Change-Id: If55409bde41ad1060fa4fe7cbd4ac530d4d9a106
Signed-off-by: Dima Zavin <>
8 years agosound: oss: omap34xx_audio_driver: Fix cleanup on error.
Rebecca Schultz Zavin [Wed, 5 May 2010 00:05:36 +0000 (17:05 -0700)]
sound: oss: omap34xx_audio_driver: Fix cleanup on error.

Change-Id: I835f7677e3551ebd7bf71299cf3692b0122c1d5d

8 years agoDSPBRIDGE: Don't bother trying to flush a null ptr, it's log spammy and holds the...
Rebecca Schultz Zavin [Tue, 4 May 2010 22:30:20 +0000 (15:30 -0700)]
DSPBRIDGE: Don't bother trying to flush a null ptr, it's log spammy and holds the mmapsem

Change-Id: Ia302c8205e92256bca9b0e8fdfc0293f11cdac2e
Signed-off-by: Rebecca Schultz Zavin <>
8 years agoUSB: android gadget: mass storage: Fix format issue for Vista Host
Velempati Chiranjeevi [Wed, 21 Oct 2009 04:23:10 +0000 (09:53 +0530)]
USB: android gadget: mass storage: Fix format issue for Vista Host

As part of formating the mass storage device, Host sends the INQUIRY
scsi command. As per the standard, the command length for this
command should be 6 bytes, whereas the Vista host sends 12 bytes.

When the command length of the command is not equal to the standard
length, the device sends a phase error as part of the status phase.
When the host receives a phase error, it re-enumerates, hence the

If the command is INQUIRY, and the command length is 12 bytes,
treating this as a good command and not sending the phase error
to the host fixes this issue.

Signed-off-by: Velempati Chiranjeevi <>
Signed-off-by: Mike Lockwood <>
8 years ago[ARM] omap-pm: resource: Prevent possible deadlock when alloc user dynamically
Mike Chan [Mon, 3 May 2010 23:50:21 +0000 (16:50 -0700)]
[ARM] omap-pm: resource: Prevent possible deadlock when alloc user dynamically

Prevent possible deadlock when allocating dynamically for a user struct.

Change-Id: I08a3997da5800497090be40df774461b9d575a5d
Signed-off-by: Mike Chan <>
8 years ago[ARM] omap: pm34xx: Read CORE and PER next states before io chain
Mike Chan [Mon, 3 May 2010 22:59:21 +0000 (15:59 -0700)]
[ARM] omap: pm34xx: Read CORE and PER next states before io chain

Change-Id: I18fd5346a75fa9851479edb46d4b95ee2eb31c28
Signed-off-by: Mike Chan <>
8 years ago[ARM] omap: pm34xx: Set IO-chain wakeup before PER goes into RET
Mike Chan [Mon, 3 May 2010 21:03:36 +0000 (14:03 -0700)]
[ARM] omap: pm34xx: Set IO-chain wakeup before PER goes into RET

Program io chain wakeup before we setting PER domain into RET.

Change-Id: Ib7f14f308a6e0b3fa169054a23cc5be5ccae62c5
Signed-off-by: Mike Chan <>
8 years ago[ARM] omap: pm34xx: Change prcm irq warn to printk_once
Mike Chan [Mon, 3 May 2010 20:38:02 +0000 (13:38 -0700)]
[ARM] omap: pm34xx: Change prcm irq warn to printk_once

Avoid log spam, switch to printk_once.

Change-Id: I7b4b0fd80140a59270be91c3c2379752e7278e52
Signed-off-by: Mike Chan <>
8 years ago[ARM] defconfig: sholes: Switch from CBQ -> HTB
San Mehat [Thu, 29 Apr 2010 21:09:23 +0000 (14:09 -0700)]
[ARM] defconfig: sholes: Switch from CBQ -> HTB

Signed-off-by: San Mehat <>
8 years agoMerge commit 'android-2.6.32' into android-omap-2.6.32
Rebecca Schultz Zavin [Mon, 26 Apr 2010 23:52:18 +0000 (16:52 -0700)]
Merge commit 'android-2.6.32' into android-omap-2.6.32

Change-Id: I3d82adedcbad435537c814720935bf84a4341883

8 years agolowmemorykiller: Don't try to kill the same pid over and over
San Mehat [Mon, 26 Apr 2010 22:11:04 +0000 (15:11 -0700)]
lowmemorykiller: Don't try to kill the same pid over and over

  Under certain circumstances, a process can take awhile to
handle a sig-kill (especially if it's in a scheduler group with
a very low share ratio). When this occurs, lowmemkiller returns
to vmscan indicating the process memory has been freed - even
though the process is still waiting to die. Since the memory
hasn't actually freed, lowmemkiller is called again shortly after,
and picks the same process to die; regardless of the fact that
it has already been 'scheduled' to die and the memory has already
been reported to vmscan as having been freed.

  Solution is to check fatal_signal_pending() on the selected
task, and if it's already pending destruction return; indicating
to vmscan that no resources were freed on this pass.

Signed-off-by: San Mehat <>
8 years ago[ARM] omap: resource: Properly initalize variable in update_resource_level
Mike Chan [Mon, 26 Apr 2010 18:31:17 +0000 (11:31 -0700)]
[ARM] omap: resource: Properly initalize variable in update_resource_level

Change-Id: I54aee1ab5fdf366eeb2e5f3afe1b3d843b28de22
Signed-off-by: Mike Chan <>
8 years ago[ARM] omap: sholes: Add disclamer for 700 and 800mhz, tweak voltage lvls
Mike Chan [Mon, 26 Apr 2010 17:58:13 +0000 (10:58 -0700)]
[ARM] omap: sholes: Add disclamer for 700 and 800mhz, tweak voltage lvls

These speeds are not officially supported by the sholes hardware,
so these are at your own risk. Also tweak voltage levels.

Change-Id: Ice8b35154b267460606418483d3e49bb355e2c13
Signed-off-by: Mike Chan <>