Skip to content
This repository

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP

cm-11.0-XNPH25R-bacon-d22b777afa

Change CM recovery install script filename [2/2]
This is part 2/2 to maintain backwards compatibility with CWM's
verify_root_and_recovery() function. CWM checks if install-recovery.sh
exists and has an executable flag set, then offers to disable the script
for the user. CM now controls this with the persist.sys.recovery_update
property which is configurable via settings, so we don't need to
double-check this.

This changes the name of the recovery install script to
install-cm-recovery.sh.

Change-Id: I275dd358b46c626dfcb8fe02c583a308d5a89c56
(cherry picked from commit 6dc667c)

cm-11.0-XNPH22R-bacon-03d77315ea

Change CM recovery install script filename [2/2]
This is part 2/2 to maintain backwards compatibility with CWM's
verify_root_and_recovery() function. CWM checks if install-recovery.sh
exists and has an executable flag set, then offers to disable the script
for the user. CM now controls this with the persist.sys.recovery_update
property which is configurable via settings, so we don't need to
double-check this.

This changes the name of the recovery install script to
install-cm-recovery.sh.

Change-Id: I275dd358b46c626dfcb8fe02c583a308d5a89c56
(cherry picked from commit 6dc667c)

cm-10.2.1

Merge remote-tracking branch 'github/cm-10.2' into HEAD

cm-10.2.0

init: Remove $TERM from global environment
 * This is causing CTS failure due to GTest spewing colors.
 * We'll set it in mksh instead.

Change-Id: I377e30164019c01b0ad076c16e6ed3241e4d9940

cm-10.2-M1

charger: Allow home key on newer devices to wake device
 * On newer devices, the home key button is usually mapped
   to KEY_HOMEPAGE. Allow it to wake the device.

Change-Id: I6423e4c6fb7992c77dc052a9f0aadae1fa8055dd

cm-10.1.3

ueventd: allow platform devices to have just a /devices/ prefix - DO …
…NOT MERGE

When using device tree, platform devices may not have a
/devices/platform/ path prefix, but can be rooted in
/devices/.

Modify the platform device tracking code to store the device
path as well as the name. This way, when we create symlinks, we
can correctly skip the base platform device prefix and get to the
proper device node path.

Change-Id: I939ef8fbcb45c5c803cd9a054e40136a912efc72
Signed-off-by: Dima Zavin <dima@android.com>

cm-10.1.3-RC1

Fix failure to build from source with modern _host_ glibc/gcc.
One should not define __USE_GNU directly, and instead _GNU_SOURCE
should be defined as early as possible, before first (indirect)
inclusion of features.h

Change-Id: I4d109d4fa0c3d266b873d2a0a66a52e1a4faa234
Signed-off-by: Dmitrijs Ledkovs <dmitrijs.ledkovs@canonical.com>

cm-10.1.3-RC2

Fix failure to build from source with modern _host_ glibc/gcc.
One should not define __USE_GNU directly, and instead _GNU_SOURCE
should be defined as early as possible, before first (indirect)
inclusion of features.h

Change-Id: I4d109d4fa0c3d266b873d2a0a66a52e1a4faa234
Signed-off-by: Dmitrijs Ledkovs <dmitrijs.ledkovs@canonical.com>

cm-10.1.0-RC5

init: provide initial pemissions for ondemand governor
also set all governor permissions a bit late

Change-Id: I5c1445b31bcaf34a535efd7438fd612ada460ea2

cm-10.1.0-RC4

init: provide initial pemissions for ondemand governor
also set all governor permissions a bit late

Change-Id: I5c1445b31bcaf34a535efd7438fd612ada460ea2

cm-10.1.0

init: provide initial pemissions for ondemand governor
also set all governor permissions a bit late

Change-Id: I5c1445b31bcaf34a535efd7438fd612ada460ea2

cm-10.1.1

init: provide initial pemissions for ondemand governor
also set all governor permissions a bit late

Change-Id: I5c1445b31bcaf34a535efd7438fd612ada460ea2

cm-10.1.2

init: provide initial pemissions for ondemand governor
also set all governor permissions a bit late

Change-Id: I5c1445b31bcaf34a535efd7438fd612ada460ea2

cm-10.1.0-RC3

init: provide initial pemissions for ondemand governor
also set all governor permissions a bit late

Change-Id: I5c1445b31bcaf34a535efd7438fd612ada460ea2

cm-10.1.0-RC2

adb : add missing connection states
Both CS_RECOVERY and CS_SIDELOAD where not being checked by
connection_state_name which resulted in adb get-state returning
unknown when a device is in those modes.

Change-Id: I00716024d6a0bdb68d6e2380c8cd7b5d056bd15f
Signed-off-by: trevd <trevd1234@gmail.com>

cm-10.1.0-RC1

adb : add missing connection states
Both CS_RECOVERY and CS_SIDELOAD where not being checked by
connection_state_name which resulted in adb get-state returning
unknown when a device is in those modes.

Change-Id: I00716024d6a0bdb68d6e2380c8cd7b5d056bd15f
Signed-off-by: trevd <trevd1234@gmail.com>

cm-10.1-M3

Fix the QCOM_BSP ifdef
We only want this when the full BSP is enabled.

Change-Id: I7bb7b0d274d1221c5bb566d4ac24419e1e134519

cm-10.1-M2

init: Check for symlinks with lstat() not open()
Using open() with an O_NOFOLLOW flag will result in a file falsely
being determined to be a symlink if the open fails for another
reason.

For example, a character device that isn't ready due to the
underlying hardware not being initialized yet will fail to open()

An example of this are any CG2900 device files prior to
execution of ste-cg29xx_ctrl

This uses lstat() instead of open() with O_NOFOLLOW to check
whether a file is a symlink or not.

Change-Id: Ie0d8d94b426c84c29b1b04d41a4cc15536cb307c

cm-10.1-M1

audio: fix build break for non QCOM
Output:
error: 'AUDIO_DEVICE_IN_ANC_HEADSET' was not declared in this scope
error: 'AUDIO_DEVICE_IN_FM_RX' was not declared in this scope
error: 'AUDIO_DEVICE_IN_FM_RX_A2DP' was not declared in this scope
error: 'AUDIO_DEVICE_IN_PROXY' was not declared in this scope

Change-Id: Ie18c623303b944544b68eecca6208607d54b1c9a

cm-7.1.0

Revert "Revert "cpu-features.h is only available for ARM""
This reverts commit aed4146.

cm-7.0.2.1

Support screenshot via gralloc framebuffer HAL
Squashed commit from nvidia's gb-tegra, to enable DDMS screenshots
on Tegra/Tegra2 devices. This has to be enabled with a proper gralloc
header that supports the read() function, like this:

BOARD_SUPPORTS_GRALLOC_FB_READ := ../../../device/lge/p990/gralloc.h

Original description follows:

As there is a hardware abstraction layer for framebuffer access, the
adbd screenshot utility should make use of it instead of reading from
/dev/graphics/fb0. This removes the need for hardware vendors to
implement the kernel framebuffer driver.

Adbd, being a statically linked executable, can not directly depend on
libhardware for accessing the gralloc module. To circumvent this a new
tool called "fbread" is created to dump the current contents of the
framebuffer into stdout. For backwards compatibility the fbread
utility will fall back to reading from /dev/graphics/fb0 if the
gralloc implementation does not support the read()
operation. Additionally the gralloc read code is only compiled when
GRALLOC_FB_READ_SUPPORTED, for source level backwards compatibility
with older versions of the gralloc interface.

The old create_subprocess() routine in adb/services.c created a
pseudoterminal, the old routine is renamed to create_terminal() to
better describe its purpose and a simpler implementation of
create_subprocess() that redirects the child's streams into a socket
directly is added.

(originally reviewed-on: http://git-master/r/#change,634)
(cherry picked from commit 9f7ec2a)

Change-Id: I1bb23439b60f555643f8aefac4f3f28f4a36601a
Reviewed-on: http://git-master/r/21763
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

Fix io redirection for create_subprocess.

Previously filedescriptors 0, 1 and 2 were dup'ed for the
fbread child process. Only STDOUT_FILENO is needed.

(originally reviewed-on: http://git-master/r/942)
(cherry picked from commit cb31fec)

Change-Id: I95ea35dd569590265c46cb2cef227e37a258838e
Reviewed-on: http://git-master/r/21764
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

fix 32-bit format channel orders

The reported color channel offsets were the wrong way around.

(originally reviewed-on: http://git-master/r/1222)
(cherry picked from commit 1634a74)

Change-Id: I8c8055bcea51efdc0e7834bd758721457864505d
Reviewed-on: http://git-master/r/21765
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

fbread: fix broken build

Change-Id: I87bf3321de9974b28f2f75c06f66e611ea2b4ce1
Reviewed-on: http://git-master/r/21766
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

cm-7.0.1

Support screenshot via gralloc framebuffer HAL
Squashed commit from nvidia's gb-tegra, to enable DDMS screenshots
on Tegra/Tegra2 devices. This has to be enabled with a proper gralloc
header that supports the read() function, like this:

BOARD_SUPPORTS_GRALLOC_FB_READ := ../../../device/lge/p990/gralloc.h

Original description follows:

As there is a hardware abstraction layer for framebuffer access, the
adbd screenshot utility should make use of it instead of reading from
/dev/graphics/fb0. This removes the need for hardware vendors to
implement the kernel framebuffer driver.

Adbd, being a statically linked executable, can not directly depend on
libhardware for accessing the gralloc module. To circumvent this a new
tool called "fbread" is created to dump the current contents of the
framebuffer into stdout. For backwards compatibility the fbread
utility will fall back to reading from /dev/graphics/fb0 if the
gralloc implementation does not support the read()
operation. Additionally the gralloc read code is only compiled when
GRALLOC_FB_READ_SUPPORTED, for source level backwards compatibility
with older versions of the gralloc interface.

The old create_subprocess() routine in adb/services.c created a
pseudoterminal, the old routine is renamed to create_terminal() to
better describe its purpose and a simpler implementation of
create_subprocess() that redirects the child's streams into a socket
directly is added.

(originally reviewed-on: http://git-master/r/#change,634)
(cherry picked from commit 9f7ec2a)

Change-Id: I1bb23439b60f555643f8aefac4f3f28f4a36601a
Reviewed-on: http://git-master/r/21763
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

Fix io redirection for create_subprocess.

Previously filedescriptors 0, 1 and 2 were dup'ed for the
fbread child process. Only STDOUT_FILENO is needed.

(originally reviewed-on: http://git-master/r/942)
(cherry picked from commit cb31fec)

Change-Id: I95ea35dd569590265c46cb2cef227e37a258838e
Reviewed-on: http://git-master/r/21764
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

fix 32-bit format channel orders

The reported color channel offsets were the wrong way around.

(originally reviewed-on: http://git-master/r/1222)
(cherry picked from commit 1634a74)

Change-Id: I8c8055bcea51efdc0e7834bd758721457864505d
Reviewed-on: http://git-master/r/21765
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

fbread: fix broken build

Change-Id: I87bf3321de9974b28f2f75c06f66e611ea2b4ce1
Reviewed-on: http://git-master/r/21766
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

cm-7.0.3

Support screenshot via gralloc framebuffer HAL
Squashed commit from nvidia's gb-tegra, to enable DDMS screenshots
on Tegra/Tegra2 devices. This has to be enabled with a proper gralloc
header that supports the read() function, like this:

BOARD_SUPPORTS_GRALLOC_FB_READ := ../../../device/lge/p990/gralloc.h

Original description follows:

As there is a hardware abstraction layer for framebuffer access, the
adbd screenshot utility should make use of it instead of reading from
/dev/graphics/fb0. This removes the need for hardware vendors to
implement the kernel framebuffer driver.

Adbd, being a statically linked executable, can not directly depend on
libhardware for accessing the gralloc module. To circumvent this a new
tool called "fbread" is created to dump the current contents of the
framebuffer into stdout. For backwards compatibility the fbread
utility will fall back to reading from /dev/graphics/fb0 if the
gralloc implementation does not support the read()
operation. Additionally the gralloc read code is only compiled when
GRALLOC_FB_READ_SUPPORTED, for source level backwards compatibility
with older versions of the gralloc interface.

The old create_subprocess() routine in adb/services.c created a
pseudoterminal, the old routine is renamed to create_terminal() to
better describe its purpose and a simpler implementation of
create_subprocess() that redirects the child's streams into a socket
directly is added.

(originally reviewed-on: http://git-master/r/#change,634)
(cherry picked from commit 9f7ec2a)

Change-Id: I1bb23439b60f555643f8aefac4f3f28f4a36601a
Reviewed-on: http://git-master/r/21763
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

Fix io redirection for create_subprocess.

Previously filedescriptors 0, 1 and 2 were dup'ed for the
fbread child process. Only STDOUT_FILENO is needed.

(originally reviewed-on: http://git-master/r/942)
(cherry picked from commit cb31fec)

Change-Id: I95ea35dd569590265c46cb2cef227e37a258838e
Reviewed-on: http://git-master/r/21764
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

fix 32-bit format channel orders

The reported color channel offsets were the wrong way around.

(originally reviewed-on: http://git-master/r/1222)
(cherry picked from commit 1634a74)

Change-Id: I8c8055bcea51efdc0e7834bd758721457864505d
Reviewed-on: http://git-master/r/21765
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

fbread: fix broken build

Change-Id: I87bf3321de9974b28f2f75c06f66e611ea2b4ce1
Reviewed-on: http://git-master/r/21766
Tested-by: Peter Zu <pzu@nvidia.com>
Reviewed-by: Daniel Willemsen <dwillemsen@nvidia.com>

cm-7.0.0

Build mkdir only for TINY_TOOLBOX
Change-Id: I1b074a81c18ec8241f77826b33049cad4dfdc562
Something went wrong with that request. Please try again.