Re: Strange problem with USB device

Re: Strange problem with USB device

[Alan Stern]

On Tue, Jun 08, 2021 at 03:56:11PM -0500, Larry Finger wrote:
> On 6/8/21 1:53 PM, Alan Stern wrote:
> > I don't get it.  If this is a PCIe device, why should it appear
> > on a USB bus?  Wouldn't you expect it to show up as a PCI device
> > on a PCI bus instead?
> > 
> 
> I do not know the internal details, but Realtek packages a PCIe wifi
> device and a bluetooth USB device in the same package. Intel does the
> same thing on my Wireless 7260.
> 
> My lsusb shows:
> Bus 003 Device 002: ID 8087:8000 Intel Corp. Integrated Rate Matching Hub
> Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
> Bus 001 Device 002: ID 8087:8008 Intel Corp. Integrated Rate Matching Hub
> Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
> Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
> Bus 002 Device 004: ID 8087:07dc Intel Corp. Bluetooth wireless interface
> Bus 002 Device 003: ID 0bda:c822 Realtek Semiconductor Corp. Bluetooth Radio
> Bus 002 Device 002: ID 04f2:b3b2 Chicony Electronics Co., Ltd TOSHIBA
> Web Camera - FHD
> Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
> 
> I have no devices plugged into a USB port.

Okay, now I get the picture.  The Intel PCIe card contains an 
EHCI USB host controller plus a couple of on-board USB Bluetooth 
devices and an on-board USB webcam, in addition to the PCIe wifi 
device.

Which means you're looking at the problem all wrong.  It isn't a 
USB problem at all; it's a PCI problem.  Namely, why doesn't the 
system detect the USB host controller on the PCIe board?

I have added the PCI maintainer and mailing list to the CC.  
Maybe they can help shed some light.

The original Suse Bugzilla report:

	https://bugzilla.suse.com/show_bug.cgi?id=1186889

shows the Realtek board at PCI address 0000:03:00.0, but there's 
no mention of a USB host controller on that board.  The only host 
controller on the system is the one at address 0000:00:14.0, 
which is xHCI and is directly on the motherboard.

Furthermore, there's no trace of any mention of an EHCI USB host 
controller in the system log.  So maybe the board has to be told 
somehow to turn that controller on before it will show up, and 
the rtw_8822ce driver isn't giving the appropriate order.

Can the bug reporter get information from Windows about the USB 
host controllers, and in particular, the one on the RTL8822 
board?

Alan Stern

Re: Strange problem with USB device

[Ibrahim Erturk]

Hi,

I've already attached logs and a snapshot from the device manager on
the windows side into the bug report. Hope this helps.

Regards,
Ibrahim ERTURK

On Wed, Jun 9, 2021 at 5:12 AM Alan Stern <stern@rowland.harvard.edu> wrote:
>
> On Tue, Jun 08, 2021 at 03:56:11PM -0500, Larry Finger wrote:
> > On 6/8/21 1:53 PM, Alan Stern wrote:
> > > I don't get it.  If this is a PCIe device, why should it appear
> > > on a USB bus?  Wouldn't you expect it to show up as a PCI device
> > > on a PCI bus instead?
> > >
> >
> > I do not know the internal details, but Realtek packages a PCIe wifi
> > device and a bluetooth USB device in the same package. Intel does the
> > same thing on my Wireless 7260.
> >
> > My lsusb shows:
> > Bus 003 Device 002: ID 8087:8000 Intel Corp. Integrated Rate Matching Hub
> > Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
> > Bus 001 Device 002: ID 8087:8008 Intel Corp. Integrated Rate Matching Hub
> > Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
> > Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
> > Bus 002 Device 004: ID 8087:07dc Intel Corp. Bluetooth wireless interface
> > Bus 002 Device 003: ID 0bda:c822 Realtek Semiconductor Corp. Bluetooth Radio
> > Bus 002 Device 002: ID 04f2:b3b2 Chicony Electronics Co., Ltd TOSHIBA
> > Web Camera - FHD
> > Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
> >
> > I have no devices plugged into a USB port.
>
> Okay, now I get the picture.  The Intel PCIe card contains an
> EHCI USB host controller plus a couple of on-board USB Bluetooth
> devices and an on-board USB webcam, in addition to the PCIe wifi
> device.
>
> Which means you're looking at the problem all wrong.  It isn't a
> USB problem at all; it's a PCI problem.  Namely, why doesn't the
> system detect the USB host controller on the PCIe board?
>
> I have added the PCI maintainer and mailing list to the CC.
> Maybe they can help shed some light.
>
> The original Suse Bugzilla report:
>
>         https://bugzilla.suse.com/show_bug.cgi?id=1186889
>
> shows the Realtek board at PCI address 0000:03:00.0, but there's
> no mention of a USB host controller on that board.  The only host
> controller on the system is the one at address 0000:00:14.0,
> which is xHCI and is directly on the motherboard.
>
> Furthermore, there's no trace of any mention of an EHCI USB host
> controller in the system log.  So maybe the board has to be told
> somehow to turn that controller on before it will show up, and
> the rtw_8822ce driver isn't giving the appropriate order.
>
> Can the bug reporter get information from Windows about the USB
> host controllers, and in particular, the one on the RTL8822
> board?
>
> Alan Stern

Re: Strange problem with USB device

[Alan Stern]

On Fri, Jun 11, 2021 at 03:08:16AM +0300, Ibrahim Erturk wrote:
> Hi,
> 
> I've already attached logs and a snapshot from the device manager on
> the windows side into the bug report. Hope this helps.

Yes, it does help.  Although the information in those reports is somewhat 
disorganized, it clearly shows there is only one USB host controller in the 
system, and that is the one Linux detects.  So my impression that we weren't 
finding the host controller was wrong.

Back to my earlier guess: The Realtek board has to be told to do something in 
order to make the Bluetooth device start working, such as turning on a power 
source.  (And perhaps that is what the RealTek people were talking about when 
they suggested the problem could be in the rtw8822 power-up sequence.)  Whatever 
it is, the rtw8822 driver isn't doing it.

This means it's still a PCI problem.

Alan Stern

PS: Larry, the discrepancy between Windows reporting an Intel USB hub and Linux 
reporting two Linux Foundation hubs isn't real -- or at least, it's what should 
be expected.  I can explain in more detail if you're curious, but you don't need 
to worry about it.