Contribute
Register

[Guide] Dell XPS 13 9360 on MacOS Sierra 10.12.x - LTS (Long-Term Support) Guide

Status
Not open for further replies.
Do you use Chrome?
Did you install Android File Transfer?

Android File Transfer - no

Issue existed before Chrome was installed, and exists with web-usb disabled as well.
 
What evidence do you see in ioreg that the Logitech drivers are hooking USB?
With more details, I will add it the FAQ.

Waiting 10sec to upload firmware will likely cause the native BT drivers to fail...

With drivers installed (LogitechUnifying.kext and LogitechHIDDevices.kext in L/E), every time I plug any mouse or HID wireless receiver into USB, the device starts working, then freezes for two seconds, then starts working again.
The same freezing behaviour happens just after a wake from sleep.
My idea is that this driver messes with the ports, maybe to increase polling rate from 125hz.
With driver removed, I get 10X slower mouse and it does not freeze upon insertion or wake from sleep.
The kexts listed are bundles with Logitech Control Center pref-pane, which I uninstalled, leaving only the two kexts.
Since the same issue happens with AFT or chrome, I think that every driver that blocks or polls the USB ports upon wake make the uploader fail.
As a proof, with only the BT injector (no firmware loaded), bluetooth worked with limited functionality, but 100% consistently, even with thise Logitech drivers enabled.
It seems like there's a race condition, or the uploader tries to reset the device while the USB port is unresponsive, then resulting in 2 device instances.
This happens with your uploader, and also with EmilyDinesh one (closed source).
I think this race condition should be investigated a bit.
I can help you troubleshoot the issue.
If you have a old Unifying wireless receiver somewhere, I can send you/tell you how to extract the drivers so you can hopefully reproduce this issue.

Can you confirm that your Bluetooth does not crash after wake (from 5min+ of unplugged sleep) if the Logitech drivers are not present?

Because I had the same problem (excluding the duplicating BCM20702A0 device, and also strange that your webcam is not present there - have you disabled it?), without ever having any 3rd party peripheral drivers installed.

Camera disabled in BIOS.
Look in L/E for third party extensions, do a complete cleanup. We identified ATF, Chrome and likely LCC as "usb-hookers", but the issue may happen with other drivers polling USB at wake.
Hopefully, we can fix the BT uploader to stop avoiding these apps as a workaround.
 
With drivers installed (LogitechUnifying.kext and LogitechHIDDevices.kext in L/E), every time I plug any mouse or HID wireless receiver into USB, the device starts working, then freezes for two seconds, then starts working again.
The same freezing behaviour happens just after a wake from sleep.
My idea is that this driver messes with the ports, maybe to increase polling rate from 125hz.
With driver removed, I get 10X slower mouse and it does not freeze upon insertion or wake from sleep.
The kexts listed are bundles with Logitech Control Center pref-pane, which I uninstalled, leaving only the two kexts.
Since the same issue happens with AFT or chrome, I think that every driver that blocks or polls the USB ports upon wake make the uploader fail.
As a proof, with only the BT injector (no firmware loaded), bluetooth worked with limited functionality, but 100% consistently, even with thise Logitech drivers enabled.
It seems like there's a race condition, or the uploader tries to reset the device while the USB port is unresponsive, then resulting in 2 device instances.
This happens with your uploader, and also with EmilyDinesh one (closed source).
I think this race condition should be investigated a bit.
I can help you troubleshoot the issue.
If you have a old Unifying wireless receiver somewhere, I can send you/tell you how to extract the drivers so you can hopefully reproduce this issue.

With AFT or Chrome it is obvious they are hooking USB in ioreg.
But I didn't see any evidence in the ioreg you attached.

I use a Logitech mouse (and sometimes keyboard) on my systems, but I didn't install any drivers for them since they work OOB.
 
With AFT or Chrome it is obvious they are hooking USB in ioreg.
But I didn't see any evidence in the ioreg you attached.

I use a Logitech mouse (and sometimes keyboard) on my systems, but I didn't install any drivers for them since they work OOB.
Is it possible that they hook them and then release them?
What I've been able to understand is that LCC drivers force a reinit on the mouse in these two seconds. From that moment until the power is cut, the mouse sends data to the receiver with a more granular resolution (used SteerMouse to debug).
 
Is it possible that they hook them and then release them?

It is possible, but then there would be little evidence of the hooking...
And you said with certainty that the drivers were hooking the port associated with the bluetooth controller, so I thought you had some evidence of this.

What I've been able to understand is that LCC drivers force a reinit on the mouse in these two seconds. From that moment until the power is cut, the mouse sends data to the receiver with a more granular resolution (used SteerMouse to debug).

Re-init on the mouse only should not affect other devices...
 
It is possible, but then there would be little evidence of the hooking...
And you said with certainty that the drivers were hooking the port associated with the bluetooth controller, so I thought you had some evidence of this.

Sorry for not weighing my words correctly.
The only evidence I have is that, without these LCC drivers, bluetooth works everytime, and the uploader never fails on wake.
If I reinstall them, bluetooh starts crashing on wake almost 1/3rd of the times.
I think that, since these drivers also support some Bluetooth Logitech peripherals, they may also poll the bluetooth adapter, thus making it unresponsive to other processes that are trying to talk with it (the fw uploader, in our case).

BT unavailable after wake is an issue I've been plagued with since 10.10, when I ran OSX on a totally different machine (an Ivy Bridge HP Envy DV7) and the only thing that did not change is... this Logitech mouse.
 
Sorry for not weighing my words correctly.
The only evidence I have is that, without these LCC drivers, bluetooth works everytime, and the uploader never fails on wake.
If I reinstall them, bluetooh starts crashing on wake almost 1/3rd of the times.
I think that, since these drivers also support some Bluetooth Logitech peripherals, they may also poll the bluetooth adapter, thus making it unresponsive to other processes that are trying to talk with it (the fw uploader, in our case).

BT unavailable after wake is an issue I've been plagued with since 10.10, when I ran OSX on a totally different machine (an Ivy Bridge HP Envy DV7) and the only thing that did not change is... this Logitech mouse.

I'll add a general note in the FAQ advising not to install Logitech provided drivers for USB mice.
Fortunately, USB mice work just fine without special drivers...
 
I'll add a general note in the FAQ advising not to install Logitech provided drivers for USB mice.
Fortunately, USB mice work just fine without special drivers...

Well, it depends. My mouse (Anywhere MX) has a smooth scroll wheel. Without proprietary drivers, the mouse sends wrong data from the scroll wheel, making the scroll wheel almost unusable. With drivers, it's pixel-smooth.
 
Well, it depends. My mouse (Anywhere MX) has a smooth scroll wheel. Without proprietary drivers, the mouse sends wrong data from the scroll wheel, making the scroll wheel almost unusable. With drivers, it's pixel-smooth.

My M705 has a (selectable between smooth/ratchet) scroll wheel. It works fine without special drivers.
 
Status
Not open for further replies.
Back
Top