| Commit message (Collapse) | Author | Age | Files | Lines |
| | |
|
| |
|
|
| |
Change-Id: I14ae121c58c0f2812a6b137a196afb3976f6ed77
|
| |
|
|
| |
Change-Id: I76d2009505c645f332dd52326f4e1f166f9fbeaa
|
| |
|
|
| |
Change-Id: I1950b97d9e84126799cf7a8d9b01ac41d92c9b37
|
| |
|
|
|
|
|
|
| |
Fix the radio access family list to include CDMA/EvDo. Fixes not being
able to pick up 3G from some network modes. Others still worked fine
without it (Ex: CDMA/EvDo Auto worked, vs: CMDA / LTE+EvDo would not).
Change-Id: I8e1056f86e70d600cdb8cc0d6a816d9f1af0138f
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Two sets of build files live in this repo - t0lte.mk and common.mk.
Some devices, like the l900 include common.mk instead of using
t0lte.mk
common.mk tries to include t0lte-common-vendor.mk, whereas t0lte.mk
includes t0lte-vendor.mk. The extract-files.sh is included in
devices like the l900 which use common.mk, but the script only
creates t0lte-vendor.mk, leaving no way for t0lte-common-vendor.mk
to be created.
It seemed to me that the least impactful way to fix this is to
introduce an arg to extract-files.sh to indicate whether to write
t0lte-common-vendor.mk or t0lte-vendor.mk - this allows a device
like the l900 to pass the common flag in its extract-files.sh,
making it possible to make both sets of makefiles happy.
I don't know the history of why there is a split in these build
files, nor do I have all the devices in question, so I'm trying
to keep the impact minimal since I only have an l900 to test with
This, along with the l900 patch I also submitted allows you to get a
working l900 build using the extract-files.sh as directed on the
wiki. Previously, you ended up with no radio due to missing blobs.
Change-Id: I2ad5ac86b767325ee26b79e4706741c43489e478
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
SELinux was blocking logcat from functioning properly. Logcat could
not be run unless you restarted adb as root, and even then was
pretty much useless since the vast majority of messages were being
blocked from reaching logcat.
After some research, I tried out the logging rules that i9300 uses
for cm13, which shares the same kernel.
These rules restore logcat to properly working on my l900, making
it much easier to troubleshoot & debug issues
Change-Id: I845a27169f27d03013ec8912123e2cb2b84a1347
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
-Add radio access family property. Dialog crashes without this.
-Set CarrierConfig settings based off of the existing Telephony
settings. It seems not all places are using CarrierConfig yet
so I'm leaving the old one. This fixes some settings such as
world phone not being picked up, allowing access to the expanded
network types dialog
Tested on my l900, the menu toggles no longer crash.
Note: there is an upstream problem where you cannot see the global
network types until you've opened the dialog at least one time
(per boot).
Change-Id: Ieecd0375c8ab9b63417c1960f0e217b3d719c15d
|
| |
|
|
| |
Change-Id: Ib1cf888da2db19f65a4ebd76ce9df2cf53a6e5a1
|
| |
|
|
|
|
|
|
|
| |
Relabel qmiproxy, at_distributor, smdexe, and diag_uart_log so init
won't get denied with system_file:file { execute_no_trans }. Clean
up rules so we don't have to set things to permissive. RIL is
working when SELinux is enforcing with this.
Change-Id: I174010d1546207037e1907d711e7f7c21871ee9e
|
| |
|
|
| |
Change-Id: Ie99a80a538ba3bfb15b76023e90fc0421178ecfb
|
| |
|
|
| |
Change-Id: Ic9a27889102b8d397f6edb4c3da1ae1918668a48
|
| |
|
|
|
|
| |
Fix Bluetooth and stop Bluetooth Share from crashing
Change-Id: I71b1f545fe4124b377f4848855886c979c51ce08
|
| |
|
|
| |
Change-Id: I68afaa40d89115ffdea0b5b61c10bbf80c61c98b
|
| |
|
|
| |
Change-Id: Icf5b722e09778366d78e3fcb65aa666ca013c3ed
|
| |
|
|
|
|
| |
Update to latest(final) official firmware
Change-Id: I31256d82de0649c98bad7f9a2ca8675147d652f3
|
| |
|
|
| |
Change-Id: I2e8451989b8ce5f5b76a9d82b4c3edc3782e3c88
|
| |
|
|
|
|
| |
Seems to fix the screen randomly starting to flash
Change-Id: Ieef0502086684e123d8a2a0b1c764aaa914d5b6e
|
| |
|
|
|
|
|
|
|
|
| |
The KK ks crashed when setting ril.ks.status property, we work around
this by using the D2 ks (which does not set this property) and wrapping it
with a bash script because D2's ks does not support some flags passed
by T0's qcks. We use the bash script to set the ril.ks.status property so qcks
knows it can continue.
Change-Id: Icc1a64b93edc38bc477acc953595280929a46549
|
| |
|
|
| |
Change-Id: I8d04edbda8c92d2580246a8bd31ba9fc05668d6e
|
| |
|
|
| |
Change-Id: Id8caf67ee8d544fc1fab0fb8b9d13b1f2cb07204
|
| |
|
|
| |
Change-Id: I40d1732b8bd7b8bf9ca18095848c27454ba6adc2
|
| |
|
|
|
|
|
|
|
|
| |
AAPT ignores densities in PRODUCT_AAPT_CONFIG.
The use of PRODUCT_AAPT_PREF_CONFIG for density is
encouraged, as AAPT is able to determine the fallback
density to use if a resource of the specified density
does not exist.
Change-Id: If2b43a44593bc234118045d03bd48cd9820dc05a
|
| |
|
|
| |
Change-Id: I668e299e7c6d9927144e3eedf59d559dfa8d0b23
|
| |
|
|
| |
Change-Id: Ic67147f2587dfaca34f734611040708536c76625
|
| |
|
|
| |
Change-Id: I96cf82160723617f76b98e30f145006a80c9f228
|
| |
|
|
| |
Change-Id: Idd531709bb024c2cdcc4d2aeadb4722f6a999ec7
|
| |
|
|
| |
Change-Id: I0950653a34a4053126848ab987335d833a2e7a37
|
| |
|
|
|
|
|
|
| |
Will follow with policy updates
This reverts commit 8e368fa918f244e214ee8bd53ce332ce6ad74663.
Change-Id: I58247300df68442709b44623e29b1bee0c6d5496
|
| |
|
|
| |
Change-Id: Ia70e48b6f2102c94e605838dc587ec8dc7495346
|
| |
|
|
| |
Change-Id: I53ed06f9832239eb8f980230861dd9f2c97d7657
|
| |
|
|
| |
Change-Id: If01f049ce7582820cfa68fa62b171a5ef9e1344d
|
| |
|
|
| |
Change-Id: I9b64aaa7cce0c8dc71374df1c3c9caac1cb051a5
|
| |
|
|
| |
Change-Id: I12c645ec5fd6e5d8c91deaab41bff683048ca659
|
| |
|
|
| |
Change-Id: I7859cf116f298c4a518d97c8c823ad8b08b064a6
|
| |
|
|
| |
Change-Id: Icf65691f123dd940cd86a26c7e355adf4a4b8f29
|
| |
|
|
| |
Change-Id: I54455a76f8eddc66f02b4e0599ab6cfa1dd733a8
|
| |
|
|
| |
Change-Id: Id7f00d4313a980aa95fce0eeab23a874ead5a1fe
|
| |
|
|
|
|
|
|
|
| |
Message Wait Indicator Quirk is Sprint specific. Lets not disable it by default for all CDMA devices
tolte >> l900
Change-Id: I1e5883cdea18f5266bb372bfffe01d65575dcd10
Signed-off-by: TheBr0ken <jtoro2716@gmail.com>
|
| |
|
|
| |
Change-Id: I8347a8ee8834f1b8ec4c3a246b17c8bef157bb3b
|
| |
|
|
| |
Change-Id: I164b875318379e024502d14f8f77771184694c0f
|
| |
|
|
| |
Change-Id: Ib0710f5deacca309491e27777efaeb414ffbc27b
|
| |
|
|
| |
Change-Id: I38444fd7b48db422ce4c4cc984ac0e88b6d00ea3
|
| |
|
|
| |
Change-Id: I97c427fb494396057bdff518dd538b641c23fb78
|
| |
|
|
| |
Change-Id: Ia82cd6cbe8cd9fe16ec0758d87e207bb6b98b6dc
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* "//Accel sensor is first on and last off"
This is totally wrong. The since we are using a single sysfs path
to enable and disable all of the sensors, the kernel driver uses
bitwise operations on the input value to know which sensors should
be enabled. The current logic wrongly assumes that the sensors are
enabled and disabled in a specific order and if accelerometer was
disabled, all of the other sensors were being disabled too. This
caused inconsistent sensor states between the sensor HAL and the
kernel driver and most of the time, the sensors failed to work
properly and drained battery.
Simplify the code to use bitwise operations instead of arithematic
operations to keep things clean. Additionally, since two or more
sensors can be enabled or disabled at the same time, we use a lock
to prevent race conditions.
Change-Id: Ifc3fab1456d4061f58d52f2ecda822c6157aae6e
|
| |
|
|
| |
Change-Id: I344d31bf5f25093bcf28a14b813a4a4aa832cff9
|
| |
|
|
|
|
|
| |
-selinux policies still need work
-switch fstab to by-name
Change-Id: I397931d0aa2ea7887774007acf40c9f03e66fb14
|
| |
|
|
| |
Change-Id: I5e8a81771b4bd4d18a1dbc290d78bac731aa78f2
|
| |
|
|
| |
Change-Id: I94bed7e8e1a722cef01649930c82c84276720946
|