* early-init is before the global restorecon of /sys gets run, so
we need to set a context on these before changing attributes
Change-Id: I493a9bb0c1bd2ea83a815e07169b8b689cf76a5d
* Samsung loads the firmware through the front door. Pie init
runs faster than previously, so we need to wait to change
the ownership until some point later, after the firmware has
time to load, else the load gets blocked by a dac_override.
Change-Id: I37be9e22b49735388ce1f5b922132dbfd20e2dc3
Framework will now add interface to supplicant (via HIDL) when it
needs to control an interface, so don't specify them in the startup
params.
Bug: 69426063
Test: Device boots up and able to connect to wifi.
Change-Id: I817d28093f8b982e3806c212babd888c0f5eb7c0
Wpa_supplicant's random pool is not necessary on Android. Randomness
is already provided by the entropymixer service which ensures
sufficient entropy is maintained across reboots. Commit b410eb1913
'Initialize /dev/urandom earlier in boot' seeds /dev/urandom with
that entropy before either wpa_supplicant or hostapd are run.
Bug: 34980020
Test: Use wifi and wifi tethering on 8974 targets
Change-Id: Iaa2a8a42e4e3fd941c5be0f8fa691245cdb429e9
The kernel does not support aio with ffs.
Set the property so adb/mtp know to use the compatible
functions, and the same while in recovery.
Bug: 37916658
Test: Use adb / use adb in recovery
Change-Id: Id35a05fcb9952dcf7239bd15a61cc620537419bc
* Mounting /system partition is handled in kernel now,
however removing the entry from fstab caused issues building
the OTA. The workaround was to have a separate fstab, but turns out
that simply setting the recoveryonly flag does the trick because
those are then ignored during a normal Android boot.
Change-Id: I2944384d0a1c41bc9f9f51e2e29daff2bed0a0f4
* We're messing with ownership and permissions here, so let's go
ahead and fix the context of this near worthless thing.
* The kernel will create it, we also write it here which will create
if it doesn't exist, so this is more manageable than chasing the
type_transition path
* The file is already labeled as a wifi_data_file, so fix this to
eliminate the below denial
avc: denied { read } for name=".wifiver.info" dev="mmcblk0p26"
ino=12 scontext=u:r:hal_wifi_default:s0
tcontext=u:object_r:system_data_file:s0 tclass=file permissive=0
Change-Id: Ie736b68f7d4d8559237b5cce072c6bf26f7ac4e7
* In particular, the RIL on CDMA variants seems to only work reliably
on first boot after flash, when things are all slowed down while
dexopting. After that first boot, it's hit and miss whether a
particular boot will have function RIL
* Slow things down by limiting ourselves to a single CPU on boot,
bringing the rest online when boot has completed
Change-Id: Ie194740cf0487268dc0dbd3377bbb790cdd1b04d
* Importing this in init.qcom.rc, especially when not all builds
have it is hacky. We'll have those devices install the script
into vendor/etc/init and let init handle it automatically, just
like the primary rild.rc script.
* Update the package name to align with the new name of the script.
Change-Id: I8eaf76c5c4f0aa590ebb1a396afa837597ebc26d
* Yes, I could do the sepolicy that lets this run correctly, but why?
Even if the phone takes 0.5s longer to boot, will you really even
notice? We are going to disable the deferred initcalls in the
kernel and let it do its thing the more "normal" way.
Change-Id: I2a9c50f12f4309923f3cdf6e646cad8edf717560
* O split charger and healthd into different processes, with healthd
running only in an full Android environment and charger running in
charge-only and recovery. It also updated the healthd sepolicy and
created a separate charger one.
* We've been using the charger symlink for ages, but the healthd
domain is no longer appropriate.
* See for details:
LineageOS/android_system_sepolicy@dc08245cLineageOS/android_system_core@526f8cf3
Change-Id: I6f528ffdcb7b14f9a51fd4246770ceecf8fb968b
* klte-common has to support both single- and dual-SIM devices, and
there are multiple devices of each kind. It makes most sense to
provide fragments in the -common tree that these devices can inherit
to configure HIDL manifest, system properties, audio support, and
any additional services necessary for MSIM support.
Change-Id: Ic3b04fa500615c33cef9b0f0a999209150969123
* Currently, config.fs doesn't give the ability to add AIDs to Android
init such that they can actually be used in the init scripts.
* Replace all instances of qcom_diag, rfs and rfs_shared with the
appropriate IDs (2950, 2951, 2952).
Change-Id: Ie983fd582d4f81c17d730df17363417f87249837
* This removes /system entry from fstab, since this partition
is now early-mounted and the fstab entry is specified in device tree.
* At the same time, also include a fully populated fstab, in order to avoid
build breakage while generating recovery updater scripts.
Change-Id: I8e558458b984e97a45a920c3751a6b7c3f2ebc5c
This reverts commit b7f109ed9f.
* So, that was 100% my derp. The cpufreq directories come and go as
cores go on- and off-line.
* Now, I wish I had notes as to whether I made this change in the
first place because of an error in a log or just based on manual
inspection.
Change-Id: I2782fefa80560bf9e20001e72ad378553d787ac0
* Don't do this, it destroys performance!
* Android uses memory cgroups to manage ZRAM. Memcg allows you to
view stats as well as set options like swappiness on a per-cg
basis. In the main init.rc, the background cgroup already has it's
swappiness set to 100. This patch sets it globally, which will
aggressively swap even foreground tasks into ZRAM which has a huge
amount of overhead.
Change-Id: I9f39b830e5817f796a42c377696864aec9bf2f22
This will let the wifi hal run as wifi user and not need root
permissions to access this path for firmware reloads.
Change-Id: I2fcd62b1dd6cbe292d89cf97085c9549bf69ba23
This reverts commit 3c04b582d7.
* The real cause for the seeming Adaptive LMK misbehavior was a merge
derp in the kernel that did not take into account zcache pages. This
is corrected with LineageOS/android_kernel_samsung_msm8974 1355fda449
Change-Id: I9441538de0d2d0dfc775519a7e8e17b584a2b833
* This is being overly aggressive, so let's go ahead and disable by
default. It is still available in kernel, so those "advanced users"
can go ahead, use it, and tweak things to their heart's content.
Change-Id: Ife119211a49479e0edb5d0c8ac04773d01d10f28
* cpu0 is the only one with cpufreq/scaling_governor or
cpufreq/scaling_min_freq, so let's not try that for all 4 of them
Change-Id: I5820a5e436e231d77de94babbea347d63ff9853a
Do not grant DAC override permission which would allow this daemon
unix permissions to everything.
avc: denied { dac_override } for pid=2664 comm="rmt_storage" capability=1 scontext=u:r:rmt_storage:s0 tcontext=u:r:rmt_storage:s0 tclass=capability permissive=0
Add wakelock group to access:
/sys/power/wake_lock
-rw-rw---- 1 radio wakelock 4096 2017-06-28 00:37 wake_unlock
Change-Id: Ib02b4aedab479f5ad8aca3a2100b5c489397002a
* Commit 722ab7ae6d inadvertently
removed the entry in ueventd.qcom.rc that sets ownership and
permissions for the NFC device node used in kltedv and kltespr
devices
Change-Id: Iae2758603b9cb91aa011e105b0d964561d9e362d
