With new strategy default values for lmkd properties should work and no
property overrides are necessary.
Bug: 132642304
Test: lmkd_unit_test, ACT memory pressure tests
Change-Id: I9cb0eb2d49fdbf9a76ff790ed19b4b5beb4cccb7
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
E OpenGLRenderer: Unable to match the desired swap behavior.
Coauthored-by: pix106 <sbordenave@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
move persist.log.tag.OpenGLRenderer=S
Only for kernel 5.4+
CRs-Fixed: 3243885
Change-Id: I64f82a25768b561bb343fdb7686b7d6c020d024a
Signed-off-by: Immanuel Raj <iamimmanuelraj@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
* imsrcsd is a key IMS HAL service bundle, please don't do the thing like "ginkgo did it, so we copy"
* uceservice(User Capability Exchange) HIDL would never be found with this, so that no capability exchange if you disable that and will get a "always retry" logspam
* com.qualcomm.qti.imscmservice HIDL interface would never be found
* vendor.qti.ims.factory HIDL interface would never be found
* when "rearranging prop" please do not add/remove unexpected props
* "NV" subscription type is missing
* as persist.vendor.dpm.feature=11,
In binary 1011, the "0" means NSRM not enabled, the NSRM prop should not be added even for rearrange
* set back dpm.feature to 1, we don't need 0x8 TCP Connection Manager, 0x4 Network Socket Request Manager(NSRM), 0x2 Connection Tracking on an Android tablet
* dpm.feature=0x1 Fast Dormancy is also matched up with miui config
Signed-off-by: pix106 <sbordenave@gmail.com>
From latest CLO tag available LA.UM.11.2.1.r1-03000-sdm660.0
Signed-off-by: Shridhan Varadkar <shridhan98@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
* Seems to work just fine with 4.19 version of jaswayne
* Nuked mi_thermald
Signed-off-by: clarencelol <clarencekuiek@icloud.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
Google has mentioned that 512 KB performs worse under heavy memory pressure.
Therefore, hardcode 128 KB read-ahead for all which is optimal.
Also, remove the unused block paths.
Signed-off-by: Tashfin Shakeer Rhythm <tashfinshakeerrhythm@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
CAF's CPU boost has been disabled in kernel. So, this is useless.
Signed-off-by: Tashfin Shakeer Rhythm <tashfinshakeerrhythm@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
rtg boost and pl have been removed from schedutil in kernel. So, this is useless.
Signed-off-by: Tashfin Shakeer Rhythm <tashfinshakeerrhythm@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
This is now obsolete, enabled by default in AOSP as of commit
108be5bcfad8fdd6474cc88b15f9a21419667189 in frameworks/base.
Change-Id: I96868c259318f597c583ee6e07cf98687d0ef285
Signed-off-by: Immanuel Raj <iamimmanuelraj@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
- SkiaVK works, in some cases HwUI works better, but mostly shuttery and microlags, so let's go for SkiaGL Threaded
Signed-off-by: Immanuel Raj <iamimmanuelraj@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
* We don't need to manually set this anymore.
Change-Id: I1239ba541636e53f8ea4373d3815024f2cd60824
Signed-off-by: Immanuel Raj <iamimmanuelraj@gmail.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
- This flag turns off GPU/CPU intensive UI effects and
reduces the RAM footprint.
- Using this is a better compromise than the 'low_ram' flag
as this does not disable features in the rom but still
reduces RAM usage and gives better smoothness in the UI.
Signed-off-by: drkphnx <dark.phnx12@gmail.com>
Signed-off-by: 7Soldier <reg.fm4@gmail.com>
Change-Id: Ic0e7a55a7c6e760e8a5172c6de4ccffcbe28d2fa
Signed-off-by: DarkJoker360 <simoespo159@gmail.com>
Signed-off-by: clarencelol <clarencekuiek@icloud.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
The extra 2-3 frames benefit jank without adding much input latency.
Signed-off-by: clarencelol <clarencekuiek@icloud.com>
Signed-off-by: pix106 <sbordenave@gmail.com>
