Closed jamuir closed 5 years ago
Seeing the same spam on Discovery and Akatsuki. The former one being without opengapps (or anything else). Both run a locally compiled Omnirom 9 image on kernel 4.9 and blobs v5.
If I remember correctly, this is one of the In-Band Switching
messages that should be captured and dealt with elsewhere, but instead end up in the Bluetooth hal. On the other hand, the HAL seems to vote for turning the UART clock on and off (at least according to the logs), which is what's supposed to happen. In that case, these are just debug messages that should be hidden.
thanks for confirming you see the same logs.
It's possible they are normal debug messages. However, in that case, I don't understand why they would be absent from my vanilla maple build (although you said you see them in a vanilla discovery build).
@jerpelea : are those repeated log messages normal? do you see them?
Noticed these props in crosshatch sources: https://android.googlesource.com/device/google/crosshatch/+/master/device.mk#364 Since these props are not set currently, not sure what the hal implementation defaults to.
@jamurir I will silence them in v6
Thanks. Don't forget about oreo (the logs I posted are from 8.1 with oem v16).
It would be nice to know what is triggering these logs (i.e. why do we see them in some builds and not others).
On Fri, Feb 1, 2019, 2:48 AM Alin Jerpelea <notifications@github.com wrote:
@jamurir I will silence them in v6
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/sonyxperiadev/bug_tracker/issues/318#issuecomment-459636000, or mute the thread https://github.com/notifications/unsubscribe-auth/AFFdoXN0UUqTjtTkGMQRxN0cvROChk_rks5vI_FMgaJpZM4aZACy .
@jerpelea also fix bluetooth on suzu please
@jerpelea : do you know what is triggering these logs? why do we see them in some builds but not others?
@jamuir I cant replicate reliably
ok. thanks for your reply.
Here is what I've found with oreo builds:
@jerpelea : note that the logs stop when the screen turns off.
@jamuir can you try disabling bluetooth scanning? it's an option under location
good idea.
Unfortunately, it is already disabled. Toggling it on and off does not change the log spew.
Are the logs supposed to disappear if android.hardware.bluetooth@1.0-impl-qti.so
successfully acquires a wake-lock? Maybe that is the cause.
@oshmoun: thanks also for your earlier suggestion re setting ro.vendor.bluetooth.emb_wp_mode=false
and ro.vendor.bluetooth.wipower=false
. Unfortunately, it did not change anything.
@jamuir actually, the hal is successfully acquiring a wakelock
i can see hal_bluetooth_lock
being acquired here
@jerpelea do you have a debug version of android.hardware.bluetooth@1.0-impl-qti.so
I could try? I could provide more logs to try and figure out the root cause.
I will see what I can do
After some more digging, I think the bt hal is working normally.
The images I examined (the ones with the bt log spam) had no network connectivity. I just noticed that after I connect to wifi, the bt activity stops.
What I think is happening is that one the google apps is repeatedly doing bt scans, possibly in an attempt to determine a location. After the network comes up, the bt scans stop (possibly because it now has a better way to determine location).
If you leave network disconnected and toggle "Location" off and on (under Settings \ Security & Location \ Location), then you can make the bt logs stop and start.
@jerpelea I don't think there is anything broken here (and there is no need to create a debug hal). I will close.
@MarijnS95 I don't understand why you are seeing the bt log spam without Google Play services. Maybe you have some other location provider in your Pie images.
@jamuir I cannot find a reliable wifi+location on/off state to trigger the spam. Both devices have bluetooth scanning "turned off".
Unfortunately, they are very inconsistent on Akatsuki; sometimes don't see them for over 5 minutes
The logs occur pretty consistently with NuPlayerDriver: notifyListener_l
on Discovery, though.
It would indeed be very interesting to know who and why is requesting these consecutive sleeps and wakes. Best guess is that it's the BT chip itself (or something along the pipeline going there). This could be normal behaviour though.
Platform: yoshino Device: maple (dual-sim) Kernel version: 4.4.168-g44235a04f24d Android version: 8.1 odm image: 8.1.0_4.4_yoshino_v16
Description
/odm/lib64/hw/android.hardware.bluetooth@1.0-impl-qti.so
appears to be stuck in some type of loop and emits the same sequence of logcat messages about every 4-5 seconds (see below).I see this behaviour only after flashing an open-gapps package on top of a sony image ~(and then wiping userdata & cache)~. I do not see this behaviour in a vanilla sony build. It may be that the additional apps are exposing some type of race condition in the qti bluetooth implementation.
Full logcat here: logcat-full.txt
Symptoms The main symptom is excessive logcat spam. I have also noticed some bt audio issues on maple (sound dropping sporadically for 1-2 seconds), but it's unclear if they are related to this.
How to reproduce see above.
(edit: wiping userdata and cache is not necessary)