Stock PM is causing null pointer dereferences, so just update it
to a sane one and bring up cnss/gnss/ril.
Change-Id: I62bde6bdd31e637e7ae268dfe2b9ad8b65f7317d
This is a squash of the following:
Author: Bruno Martins <bgcngm@gmail.com>
Date: Tue Jun 19 23:55:25 2018 +0100
taimen: Fix typo from CodeAurora in blob_fixup
* Let com.qualcomm.qcrilmsgtunnel break free
Change-Id: Ide5759eef7e3100ac5b50f5dc95a3f90119e592f
Author: Michael Bestas <mkbestas@lineageos.org>
Date: Sat Dec 19 02:53:35 2020 +0200
taimen: Switch to standalone extract utils
Align with templates while we are at it
Change-Id: Ife50623a3d1b8733f5cc76995c982d8950d39ad4
Author: Rashed Abdel-Tawab <rashed@linux.com>
Date: Tue May 14 00:36:02 2019 -0700
taimen: Import and tailor {extract-files,setup-makefiles}.sh
* Partial pick of Id399faa404207ce018f0f0187c9576b61f5d4dc9.
Change-Id: I866e015e3ba717ae7fc585dc56f118f3e251154c
Change-Id: I231d822f364f57314009be4d08af993b97bf5705
* There is zero reason to ever build without blobs. We've even seen
this situation with official builds from our servers. It's always
better for a build to fail than it is for it to produce something
that has no chance at working.
Change-Id: I4968795670c91f691e9ecdc0e4af62e16ba3a93a
This is a squash of the following:
Author: LuK1337 <priv.luk@gmail.com>
Date: Sun Sep 27 16:12:10 2020 +0200
wahoo: gpt-utils: Drop unnecessary include
* When using generated_kernel_headers including
stdio.h will cause errors due to redefined macros.
Change-Id: I2f753a64f760b636be1325873e17697466fbf33d
Author: LuK1337 <priv.luk@gmail.com>
Date: Mon Sep 9 22:38:02 2019 +0200
wahoo: gpt-utils: Drop unnecessary include
* When using generated_kernel_headers including
asm/byteorder.h will cause a warning due to
unused static functions.
Change-Id: I2a74ec92487c13c4645048c96da8cb92b3cbd9f0
Author: Rashed Abdel-Tawab <rashed@linux.com>
Date: Sun Sep 29 15:37:38 2019 -0400
wahoo: Switch gpt-utils to generated_kernel_headers
Change-Id: Ia6e2f7e140e6ffbd3a93dd823f836aaf399d03fb
Change-Id: Ia1f608778b5b37fec770abbf185c5e313cd5ba16
When markBoolSuccessful is invoked, we update
the partition table. These writes should be
synced before merge operation is resumed post OTA.
If not, any crash before these writes are landed to
backing storage will lead to incorrect switching of
slots.
BUG: 175711601
Test: Verify slot switching correctly after crash when merge in progress
(on redbull)
Signed-off-by: Akilesh Kailash <akailash@google.com>
Change-Id: I4d30342b44aaeb7a4fdc915cc1e53ffb20c17a2f
In favor of using boot control HAL under recovery.
Bug: 34254109
Test: TreeHugger
This is a combination of 2 commits.
This is the 1st commit message:
Revert^2 "Build libgptutils with Android.bp.""
This reverts commit 215189f0b9ff44d28c2b4ace0abb26854b201271.
The previous landing broke checkbuild targets due to namespace issue. As
libgptutils.msm8998 was exposed to global namespace (previously guarded
via conditionals in Android.mk), but had a missing dependency on
nonexposed msm8998 kernel headers.
The relanding adds a Soong namespace of hardware/qcom/msm8998, to ensure
modules only getting built as needed.
Bug: 34254109
Test: Sideload on taimen.
Test: `m -j checkbuild` with aosp_x86-eng
Change-Id: I4a7c4cedcf58c7735172d989912e4f7afad1eaed
Merged-In: I4a7c4cedcf58c7735172d989912e4f7afad1eaed
(cherry picked from commit e74c7611d2cbea03ec8ec1559787a7a57b2bf767)
This is the commit message #2:
Revert^2 "Deprecate PRODUCT_STATIC_BOOT_CONTROL_HAL.""
This reverts commit d83c2cc22b3f037e94cd4ad7e19f5985da13a8b8.
The previous landing broke checkbuild targets due to namespace issue.
See the commit message for hardware/qcom/msm8998 change.
Bug: 34254109
Test: Build and boot taimen on device to home screen.
Test: Sideload on taimen.
Change-Id: I775026345eb5fc0ec580ba4ca0282835106a1d35
Merged-In: I775026345eb5fc0ec580ba4ca0282835106a1d35
(cherry picked from commit 4fa8e490208bef42faccf204381dc35ad1cfed46)
This commit replaces sparse_crc32() with crc32() from libz.so because
libsparse.so won't be vendor available.
Bug: 38303202
Test: android builds
Change-Id: Ib318da74b5d55f729cbd40d9a62da53e6316fe7b
libgptutils is only used by vendor. Move it to
vendor partition so they can work with generic AOSP
system image.
Bug: 37927875
Test: These libs are under /vendor/lib(64) instead of
/system/lib(64) and the resulted image works on the device.
Change-Id: Id04f7a09fdc8e60de79f83b8fd0c2504fd2e2ebe
Change-Id: I7f3f8fe845d35a05492ffc4785787622c7851b79
Author: Thierry Strudel <tstrudel@google.com>
Date: Thu May 4 18:15:01 2017 -0700
gpt-utils: PTN_SWAP_LIST add the missing new partitions
abl, dtbo, vbmeta were not part of the list updated by bootctl HAL.
This will fix the issue but bootctl HAL should be rewritten to not
detect A/B partition candidate at runtime.
Bug: 37501587
Change-Id: I64140d9fce8d142e16a12c9778f1b10394f9dc76
Signed-off-by: Thierry Strudel <tstrudel@google.com>
Author: Patrick Tjin <pattjin@google.com>
Date: Tue Jul 25 19:26:1 2017 -0700
Add aes as A/B OTA partition
Bug: 37554629
Change-Id: I162150922683000dbfc0beceab807d46bd51325e
Author: Patrick Tjin <pattjin@google.com>
Date: Wed May 10 15:36:34 2017 -0700
Add laf partition to swap list.
Bug: 37743272
Change-Id: If9521bbd29af9439aa01ac72c8bb8cbc0bf674b6
Change-Id: I8aa1386206e6b675a4b7d1e696841f23f5955c2d
Currently all writes to the dev interface are done
by WriteStringToFd that uses char internally.
The current values seem to expect to be written as s32,
this sadly won't happen and triggers a special case in
the write handling of pm_qos_power_write:
it expects char buffers to be encoded as base16 and
decodes them.
This means the current 44 is actually 0x44 -> 68 seen by
the kernel.
Luckily it seems like both accepted values for this node
don't hit the threshold to enter C2, so it was never
noticed in real usage and didn't effect the device
C-States handling during hints.
Signed-off-by: Luca Stefani <luca.stefani.ge1@gmail.com>
Change-Id: Ic544d4dcaa1edc3de913aed737baf1af88a45360
Today, the dark boot follows the Dark Theme mode in Settings. Now
the UI team decide to always use dark boot after the device is
provisioned. So users will see fewer disturbances when a scheduled
update is taken at night
Bug: 181339788
Test: flash a Pixel -> boot -> dark theme is set;
boot again -> observe dark boot
Change-Id: Id4989e4e78471bcefc2730ba7d9adc36f5ac9c75
For non-REL branch, move VNDK APEX to /vendor partition so we don't need
to update vendor.img prebuilt everytime there is a change in the VNDK
libraries.
For REL branch, the API/ABI surfaces of VNDK libraries are frozen so
don't need to move to /vendor partition.
Bug: 140136207
Test: Build on REL and non-REL branch
Change-Id: Ibce24465b546c52bc447b4b28a474de2b4b53792
These debuf.sf.early_* properties have been inactive because they are
all eclipsed by the settings in wahoo/device.mk. Specifically,
// wahoo/device.mk
PRODUCT_PROPERTY_OVERRIDES := a=10
// taimen/device.mk
include wahoo/device.mk
PRODUCT_PROPERTY_OVERRIDES += a=20
With above, PRODUCT_PROPERTY_OVERRIDES becomes "a=10 a=20" and the build
system has chosen "a=10" via the uniq-pairs-by-first-component macro.
This was because PRODUCT_* lists were designed with an assumption that
those mk files are inherited (without being included) in which case the
values from more specific *.mk file are 'prepended' (not appaended).
Since the settings in taimen/device.mk has been obsolete, let's remove
them. This problem was actually found with
I9c073a21c8257987cf2378012cadaeeeb698a4fb where duplicated sysprop
assignments are prohibited, which is an attempt to make the sysprop
settings be agnostic to the confusing ordering behavior imposed by the
product inheritance mechanism.
Bug: 117892318
Bug: 158735147
Test: m
Exempt-From-Owner-Approval: cherry-pick master
Merged-In: I6d1dd46edba0ad69586791935fca0da484ef2746
(cherry picked from commit 695f1fbe29f2ea08ed3a5cefb0fa487bea9144f4)
Change-Id: I6d1dd46edba0ad69586791935fca0da484ef2746
These debuf.sf.early_* properties have been inactive because they are
all eclipsed by the settings in wahoo/device.mk. Specifically,
// wahoo/device.mk
PRODUCT_PROPERTY_OVERRIDES := a=10
// taimen/device.mk
include wahoo/device.mk
PRODUCT_PROPERTY_OVERRIDES += a=20
With above, PRODUCT_PROPERTY_OVERRIDES becomes "a=10 a=20" and the build
system has chosen "a=10" via the uniq-pairs-by-first-component macro.
This was because PRODUCT_* lists were designed with an assumption that
those mk files are inherited (without being included) in which case the
values from more specific *.mk file are 'prepended' (not appaended).
Since the settings in taimen/device.mk has been obsolete, let's remove
them. This problem was actually found with
I9c073a21c8257987cf2378012cadaeeeb698a4fb where duplicated sysprop
assignments are prohibited, which is an attempt to make the sysprop
settings be agnostic to the confusing ordering behavior imposed by the
product inheritance mechanism.
Bug: 117892318
Bug: 158735147
Test: m
Exempt-From-Owner-Approval: cherry-pick master
Merged-In: I6d1dd46edba0ad69586791935fca0da484ef2746
(cherry picked from commit 695f1fbe29f2ea08ed3a5cefb0fa487bea9144f4)
Change-Id: I6d1dd46edba0ad69586791935fca0da484ef2746
