1. During startup, judge whether to register the LDO driver
for the first time. If it is the first time, it is not
necessary to disable all LDOS;
2. When different pmics are used, different initialization
voltages are used for initialization;
Change-Id: Ic22e2c05fb9edcb48f93f51add369cb71997741f
Signed-off-by: huaqin79 <huaqin79@motorola.com>
Reviewed-on: https://gerrit.mot.com/2138715
SME-Granted: SME Approvals Granted
SLTApproved: Slta Waiver
Tested-by: Jira Key
Reviewed-by: Hua Tan <tanhua1@motorola.com>
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
When the regulator driver is loaded into the
kernel as KO, the LDO driver registration is
too slow, resulting in the LCD power on and
can not be controlled normally.
Change-Id: If1f62f5204a8eda235a5aea91d27b59f6021c704
Signed-off-by: huaqin79 <huaqin79@motorola.com>
Reviewed-on: https://gerrit.mot.com/2132717
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Hua Tan <tanhua1@motorola.com>
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
Upstream pty kernel driver already moved the tty_flip_buffer_push
Out of tty_flip_buffer_push, adjust the tty_diag driver to do the same
Otherwise there will be lockup watchdog panic generated
Change-Id: Ib1ea301cff482b1b6905f20ce99995fe2f5b5aef
Signed-off-by: a17671 <a17671@motorola.com>
Reviewed-on: https://gerrit.mot.com/2130815
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
dwc3 driver now replies on extcon suppliers registered from another ko
in timing sequence, dwc3 is probed much earlier than the ko registeration.
To work with the related ko driver correctly, add a delayed work item to
check the extcon status periodicly. The frequency is 1HZ and it just stops
whenever the required extcon suppliers defined in dtsi have been found.
Change-Id: Ia7c50165eab13b0274590c744f49061fa6ef7c03
Reviewed-on: https://gerrit.mot.com/2130472
SME-Granted: SME Approvals Granted
SLTApproved: Slta Waiver
Tested-by: Jira Key
Reviewed-by: Wei Xu <xuwei9@lenovo.com>
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
Due to high risk to update FFU FW on end users, we are reverting (CR)
This reverts commit cd15497445c03dc4b93e8aec22ef845c697fc99a.
Change-Id: Iddc18277e4193e18ad1844cd920343fdad63c0df
Reviewed-on: https://gerrit.mot.com/2002586
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Carlos Pinho <cpinho@motorola.com>
Submit-Approved: Carlos Pinho <cpinho@motorola.com>
Since the property ro.crypto.allow_encrypt_override wasn't set in Q
file data in SDcard mounted as internal storage is double encrypted.
So we need to revert the patch that skips encription if bio is
fscrypto or blk-crypto encrypted to avoid decryption errors.
Change-Id: Ic46973a092f00227e0fb2706788df4caa674607d
Signed-off-by: Luiz Cressoni <lutcho@motorola.com>
Reviewed-on: https://gerrit.mot.com/1952458
SME-Granted: SME Approvals Granted
SLTApproved: Slta Waiver
Tested-by: Jira Key
Reviewed-by: Ling Jin <lingjin@motorola.com>
Submit-Approved: Jira Key
For not implemented apis, there is huge WARN prints as below
Introduced by QC commit
795434e6 msm: ipa: Add ipa rm support for ipa_v2
This caused regression in Rndis reverse path (DUT to PC)
W Call trace:
W : ipa_pm_is_used+0x44/0x78
W : rndis_ipa_start_xmit+0x2a0/0x708
W : dev_hard_start_xmit+0x140/0x290
W : sch_direct_xmit+0xd4/0x2e8
W : __qdisc_run+0x348/0x790
W : net_tx_action+0x274/0x2a8
W : __do_softirq+0x18c/0x374
W : run_ksoftirqd+0x24/0x40
W : smpboot_thread_fn+0x20c/0x270
W : kthread+0x140/0x150
W : ret_from_fork+0x10/0x18
W : ---[ end trace 4cfa2ba718c7aa80 ]---
W : ------------[ cut here ]------------
W : ipa_pm_is_used not implemented for IPA ver 18
W : WARNING: CPU: 3 PID: 37 at
drivers/platform/msm/ipa/ipa_api.c:3806 ipa_pm_is_used+0x44/0x78
Right now replace them with pr_debug to not hamper throughput and it
just print as below when dynamic debug is enabled
E : ipa_pm_is_used not implemented for IPA ver 18
Tests: iperf.exe -s -u -p5002 -w64M -i1 -t15 ( PC as server)
Seen 350+ Mbits/sec from DUT to PC with only USB tethering.
Change-Id: I67d9df7c1f77326f6172595967798e82c7f85175
Signed-off-by: Srihari Sathyanarayana <sriharis@motorola.com>
Reviewed-on: https://gerrit.mot.com/1877836
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Vijayakumar Gn <vijaygn@motorola.com>
Reviewed-by: Ling Jin <lingjin@motorola.com>
Submit-Approved: Jira Key
Signed-off-by: Robert Cabral <robertc@motorola.com>
Reviewed-on: https://gerrit.mot.com/1924169
1. Fix NTP-32 support
When connecting a CDC-NCM gadget to an host that uses the NTP-32 mode,
or that relies on the default CRC setting, the current implementation gets
confused, and does not expect the correct signature for its packets.
Fix this, by ensuring that the ndp_sign member in the f_ncm structure
always contain a valid value.
https://lore.kernel.org/patchwork/patch/1062690/
2. Add OS descriptor support
To be able to use the default USB class drivers available in Microsoft
Windows, we need to add OS descriptors to the exported USB gadget to
tell the OS that we are compatible with the built-in drivers.
Copy the OS descriptor support from f_rndis into f_ncm. As a result,
using the WINNCM compatible ID, the UsbNcm driver is loaded on
enumeration without the need for a custom driver or inf file.
https://lore.kernel.org/patchwork/patch/1062691/
3. Use "ncm0" interface name
The default interfacename "usb0" is confused. Use "ncm0" better.
Change-Id: I36d085c519f8205d7f9e0a1d9c669d3fb6623483
Reviewed-on: https://gerrit.mot.com/1872002
SME-Granted: SME Approvals Granted
SLTApproved: Slta Waiver
Tested-by: Jira Key
Reviewed-by: Yonghua Yan <yanyh2@motorola.com>
Reviewed-by: Yuecai Liu <liuyc20@motorola.com>
Reviewed-by: Zonghua Liu <a17671@motorola.com>
Reviewed-by: Huosheng Liao <liaohs@motorola.com>
Submit-Approved: Jira Key
(cherry picked from commit edda0121735f5436ef2287379fb6b57325ef04f4)
Reviewed-on: https://gerrit.mot.com/1916170
Reviewed-by: Xuejin Wang <wangxjn@motorola.com>
Tested-by: M(CR) SU <suzh1@lenovo.com>
This gerrit disables CONFIG_SLUB_DEBUG for bengal user
builds. This configs tends to badly impact performance
on user builds. Therefore, it should be disabled.
Change-Id: I09c64131492af6217d3f645f0870b0c93b981492
Reviewed-on: https://gerrit.mot.com/1913031
SME-Granted: SME Approvals Granted
SLTApproved: Slta Waiver
Tested-by: Jira Key
Reviewed-by: Rafael Ortolan <rafones@motorola.com>
Reviewed-by: Fernanda Schmidt <fschmidt@motorola.com>
Reviewed-by: Carlos Pinho <cpinho@motorola.com>
Submit-Approved: Jira Key
This gerrit disables CONFIG_DEBUG_PANIC_ON_OOM in userdebug builds
(for cebu device). It prevents Kernel Panic generation when Kernel
memory allocation fails, due to an Out of Memory (OOM) detection.
As OOM condition isn't so unusual (specially for heavy users), it
won't be a good experience for users to face kernel panics everytime
this OOM condition is satisfied.
Reference gerrit: https://gerrit.mot.com/#/c/1732467/
Change-Id: I7c1d017daa1671eef9b723adb72391901a2aed27
Reviewed-on: https://gerrit.mot.com/1905120
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Tiago Ganselli <ganselli@motorola.com>
Reviewed-by: Joo Costa <jmcosta@motorola.com>
Reviewed-by: Fernanda Schmidt <fschmidt@motorola.com>
Reviewed-by: Rafael Ortolan <rafones@motorola.com>
Reviewed-by: Carlos Pinho <cpinho@motorola.com>
Submit-Approved: Jira Key
Issue: At mot4250 android R is not supporting inline encryption
This causes recovery mode with init_user0_failed log
when upgrade from Q to R
Fix: Include inlinecrypt at fstab and config.gz
Change-Id: I2b20a7cbd3670e6b9bf5c9885a14a1b0e4ec2083
Reviewed-on: https://gerrit.mot.com/1896972
SME-Granted: SME Approvals Granted
SLTApproved: Slta Waiver
Tested-by: Jira Key
Reviewed-by: Ling Jin <lingjin@motorola.com>
Submit-Approved: Jira Key
PROPAGATED_from (CR)
Add device tree property xhci-hw-lpm-disable support for
xhci hardware lpm feature disabling.
For some connected devices the host phone, they may be some
detection issue if xhci hardware lpm feature is enabled.
For this case, add this property in device tree of usb controller
to disable xhci hardware lpm.
(pick from https://gerrit.mot.com/#/c/1426964/)
Change-Id: I3907c945dd83fdefdda597975318e45d8be047e1
Signed-off-by: yanyh2 <yanyh2@motorola.com>
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Zonghua Liu <a17671@motorola.com>
Reviewed-by: Lu Chai <chailu1@motorola.com>
Reviewed-by: Yuecai Liu <liuyc20@motorola.com>
Reviewed-by: Huosheng Liao <liaohs@motorola.com>
Submit-Approved: Jira Key
Signed-off-by: Robert Cabral <robertc@motorola.com>
Reviewed-on: https://gerrit.mot.com/1858485
cherry-pick: (CR) kernel:ram overwritten by TZ logs to cause random panics
in panic RAMDump file, we can see the memory in kernel was overwritten by TZ logs
the tzlog strings corrupted the memory address to cause abort on VM to PHY-M
similar panic issue was frequently reported in User build which disable
DebugFS along with tzlog driver which depends on DebugFS. Add the tzlog driver
back, but remove the dependency of DebugFS, instead, use ProcFS for output
Change-Id: If6f9f0bf0fe3ce08c868582a457b207cee07923d
Signed-off-by: wangwang1 <wangwang1@lenovo.com>
Reviewed-on: https://gerrit.mot.com/1836697
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
Reviewed-on: https://gerrit.mot.com/1836767
Vibrator on phone main board can disturb camera AF VCM when it's vibrating.
To avoid the camera lens str(CR) noise during vibrating, power on AF and
push lens to certain position (for caprip DAC 154) can greatly reduce the
noise.This change is for notifying camera actuator that vibrator will start
or stop work, then camera can respond accordingly..
Change-Id: Ie26d6f811d6628f51045a853dc3e90f422481da9
Signed-off-by: huangdra <huangdra@motorola.com>
Reviewed-on: https://gerrit.mot.com/1836813
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
This is a workarround to reduce the vibrator strength for long vib.
Add attribute "qcom,vib-dis-short-long" in device-tree to enable the
function.
Now it is applied in caprip product.
Change-Id: I41c04e0dfb14c184f999d7d0a4121c10dd763cd5
Signed-off-by: panzh2 <panzh2@lenovo.com>
Reviewed-on: https://gerrit.mot.com/1805268
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Wang Wang <wangwang1@lenovo.com>
Submit-Approved: Jira Key
We often see lost ram is still very high even we close the camera app.
That's because camera's ion memory goes into ion cache after use, not
buddy system. And this cache was not counted in proc/meminfo/memAvailable,
so it can not reflect the true memAvailable in this state. It may affects
some components l(CR) camera or lmkd which need evaluate true memAvailable
in system. Take camera as example, while launching camera, camera app needs
calcuate the available memory then call framework api to kill processes,
without this patch, we will see overkill.
This patch is only for 4.14 & 4.19, on kernel 5.4, ion cache will
be counted in memAvailable, so memAvailable is always trustable.
Change-Id: I4a9678530d387bd8da404d797f7d6e33bd258ac4
Signed-off-by: huangzq2 <huangzq2@motorola.com>
Reviewed-on: https://gerrit.mot.com/1792299
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
(cherry picked from commit a9acaa732c7fcb5e9ec9eea08bb6bb390c316722)
Reviewed-on: https://gerrit.mot.com/1795665
This reverts commit 3a7c2ddda918ea0ba239c51d74984a499c392942.
This change has been included in new base .so revert it
Change-Id: I50210a029c2f69b8561a7cae2f4b87b190b8a842
Reviewed-on: https://gerrit.mot.com/1791620
SME-Granted: SME Approvals Granted
SLTApproved: Slta Waiver
Submit-Approved: Jira Key
Tested-by: Jira Key
Reviewed-by: gaoyx9 <gaoyx9@motorola.com>
porting (CR)
1. Print active wakeup source before entering suspend
2. Use to monitor the kernel long term wake lock issue
Change-Id: Icee9744da0c46fbb56d516e996eed7115f754931
Signed-off-by: xieqiong2 <xieqiong2@lenovo.com>
Reviewed-on: https://gerrit.mot.com/1788058
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
In P, run_cmd(poweroff_cmd) can work to trigger /sbin/poweroff. Now in
R, it can not work but still return success. We need to use
kernel_power_off() to trigger power off for factory kill.
The change is referring to (CR).
Change-Id: Ia18faee3138259c7ee35946313e9b0dac75dd244
Signed-off-by: xieqiong2 <xieqiong2@lenovo.com>
Reviewed-on: https://gerrit.mot.com/1763914
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Xiangpo Zhao <zhaoxp3@motorola.com>
Submit-Approved: Jira Key
cherry-pick from:
(CR): arm64/defconfig: disable debug fs on user build
according to android R,
disable debug fs for user build
Change-Id: I985888644c92e14205f7260c423bb88bad026dd9
Signed-off-by: zhaoxp3 <zhaoxp3@motorola.com>
Reviewed-on: https://gerrit.mot.com/1752415
SLTApproved: Slta Waiver
SME-Granted: SME Approvals Granted
Tested-by: Jira Key
Reviewed-by: Wang Wang <wangwang1@lenovo.com>
Submit-Approved: Jira Key
Reviewed-on: https://gerrit.mot.com/1753930
Reviewed-by: Jianguo Liu <liujg6@motorola.com>
huaqin79