Skip to contentSkip to conversation

New: Intel Wireless card 3165 does not get detected but bluetooth works

Bug 201647 Bjorn Helgaas 3 participants 12 messages Nov 9, 2018 – Dec 11, 2018 1 message not in the archive PCI

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]/

#1 lore raw

[Resend because I forgot to send as plain text and linux-pci rejected it.]

The report below shows only one PCI-PCI bridge (00:13.0) but no devices below it. I wonder if that's related to whatever's wrong with this host bridge window:

acpi PNP0A08:00: host bridge window [mem 0x1000000010000 window]
(ignored, not CPU addressable)
pci 0000:00:13.0: PCI bridge to [bus 01]

Mertarg10, would you mind attaching the "sudo lspci -vv" output, too?

---------- Forwarded message ---------
From: <[email protected]>
Date: Fri, Nov 9, 2018 at 4:10 AM
Subject: [Bug 201647] New: Intel Wireless card 3165 does not get detected but bluetooth works
To: <[email protected]>

https://bugzilla.kernel.org/show_bug.cgi?id=201647

            Bug ID: 201647
           Summary: Intel Wireless card 3165 does not get detected but
                    bluetooth works
           Product: Drivers
           Version: 2.5
    Kernel Version: 4.19.1
          Hardware: Intel
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: high
          Priority: P1
         Component: PCI
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No

Created attachment 279385
--> https://bugzilla.kernel.org/attachment.cgi?id=279385&action=edit
Contains outputs of dmesg and lspci

This bug affects most of the devices with a Celeron N4000 and an Intel wifi 3165 Ac adapter.

When using Linux wifi is not working however, Bluetooth is working fine. Also, Bluetooth part of this chip is connected via btusb and the wifi part of this chip is connected via PCIe.

I have contacted the Linux wireless team of Intel, which suggested me to report this bug as a PCIe bug.

--
You are receiving this mail because:
You are watching the assignee of the bug.

#2 lore raw

[+cc Emmanuel, LKML]

Quoted 25 lines · Bjorn Helgaas

---------- Forwarded message ---------
From: <[email protected]>
Date: Fri, Nov 9, 2018 at 4:10 AM
Subject: [Bug 201647] New: Intel Wireless card 3165 does not get detected but bluetooth works

https://bugzilla.kernel.org/show_bug.cgi?id=201647

            Bug ID: 201647
           Summary: Intel Wireless card 3165 does not get detected but
                    bluetooth works
           Product: Drivers
           Version: 2.5
    Kernel Version: 4.19.1
          Hardware: Intel
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: high
          Priority: P1
         Component: PCI
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No
Quoted 6 lines

This bug affects most of the devices with a Celeron N4000 and an Intel wifi 3165 Ac adapter.

When using Linux wifi is not working however, Bluetooth is working fine. Also, Bluetooth part of this chip is connected via btusb and the wifi part of this chip is connected via PCIe.

Can you attach a screenshot of the Windows 10 device manager info for the wifi adapter to the bugzilla? If you can get a raw hex dump of its config space, that would be awesome.

Also attach a copy of your kernel .config file (typically in /boot/).

My only guess is that maybe the system keeps wifi completely powered down and uses hotplug to add it when needed. [1] mentions wifi being on pcibus 1 under Windows. Your lspci does show bridge 00:13.0 leading to bus 01, but Linux doesn't find any devices on bus 01.

Hotplug could be done via either acpiphp (ACPI mediated hotplug) or pciehp (native PCIe hotplug). Your dmesg shows you do have acpiphp.

I can't tell about pciehp (your .config will show that), but I think pciehp will only claim bridges where SltCap contains HotPlug+, and yours shows HotPlug-, so I don't think pciehp will do anything on your system.

Even if the system does use hotplug, I don't know what mechanism the OS would use to wake up the device, since we don't know it even exists. I guess there could be some magic switch accessible via USB. But if that were the case, I'm sure Emmanuel would know about it.

[1] https://www.chinamobilemag.de/forum/hardware/3779-teclast-f5-linux-erkennt-kein-wlan-intel-3165-in-ubuntu-18-04-1.html?start=10

#3 lore raw
Quoted 50 lines

[+cc Emmanuel, LKML]

Quoted 25 lines · Bjorn Helgaas

---------- Forwarded message ---------
From: <[email protected]>
Date: Fri, Nov 9, 2018 at 4:10 AM
Subject: [Bug 201647] New: Intel Wireless card 3165 does not get detected but bluetooth works

https://bugzilla.kernel.org/show_bug.cgi?id=201647

            Bug ID: 201647
           Summary: Intel Wireless card 3165 does not get detected but
                    bluetooth works
           Product: Drivers
           Version: 2.5
    Kernel Version: 4.19.1
          Hardware: Intel
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: high
          Priority: P1
         Component: PCI
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No
Quoted 6 lines

This bug affects most of the devices with a Celeron N4000 and an Intel wifi 3165 Ac adapter.

When using Linux wifi is not working however, Bluetooth is working fine. Also, Bluetooth part of this chip is connected via btusb and the wifi part of this chip is connected via PCIe.

Can you attach a screenshot of the Windows 10 device manager info for the wifi adapter to the bugzilla? If you can get a raw hex dump of its config space, that would be awesome.

Also attach a copy of your kernel .config file (typically in /boot/).

My only guess is that maybe the system keeps wifi completely powered down and uses hotplug to add it when needed. [1] mentions wifi being on pcibus 1 under Windows. Your lspci does show bridge 00:13.0 leading to bus 01, but Linux doesn't find any devices on bus 01.

Hotplug could be done via either acpiphp (ACPI mediated hotplug) or pciehp (native PCIe hotplug). Your dmesg shows you do have acpiphp.

Quoted 9 lines

I can't tell about pciehp (your .config will show that), but I think pciehp will only claim bridges where SltCap contains HotPlug+, and yours shows HotPlug- , so I don't think pciehp will do anything on your system.

Even if the system does use hotplug, I don't know what mechanism the OS would use to wake up the device, since we don't know it even exists. I guess there could be some magic switch accessible via USB.
But if that were the case, I'm sure Emmanuel would know about it.

Hm... Don't be so sure... :)
I don't think we have anything as fancy as this.
I guess you can try to dig into the BIOS settings?
I have heard of such a switch that would make the device disappear.

#4 lore raw
Quoted 56 lines · Emmanuel
Quoted 47 lines · Bjorn Helgaas
Quoted 25 lines

---------- Forwarded message ---------
From: <[email protected]>
Date: Fri, Nov 9, 2018 at 4:10 AM
Subject: [Bug 201647] New: Intel Wireless card 3165 does not get detected but bluetooth works

https://bugzilla.kernel.org/show_bug.cgi?id=201647

            Bug ID: 201647
           Summary: Intel Wireless card 3165 does not get detected but
                    bluetooth works
           Product: Drivers
           Version: 2.5
    Kernel Version: 4.19.1
          Hardware: Intel
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: high
          Priority: P1
         Component: PCI
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No
Quoted 6 lines

This bug affects most of the devices with a Celeron N4000 and an Intel wifi 3165 Ac adapter.

When using Linux wifi is not working however, Bluetooth is working fine. Also, Bluetooth part of this chip is connected via btusb and the wifi part of this chip is connected via PCIe.

Can you attach a screenshot of the Windows 10 device manager info for the wifi adapter to the bugzilla? If you can get a raw hex dump of its config space, that would be awesome.

Also attach a copy of your kernel .config file (typically in /boot/).

My only guess is that maybe the system keeps wifi completely powered down and uses hotplug to add it when needed. [1] mentions wifi being on pcibus 1 under Windows. Your lspci does show bridge 00:13.0 leading to bus 01, but Linux doesn't find any devices on bus 01.

Hotplug could be done via either acpiphp (ACPI mediated hotplug) or pciehp (native PCIe hotplug). Your dmesg shows you do have acpiphp.

Quoted 8 lines

I can't tell about pciehp (your .config will show that), but I think pciehp will only claim bridges where SltCap contains HotPlug+, and yours shows HotPlug- , so I don't think pciehp will do anything on your system.

Even if the system does use hotplug, I don't know what mechanism the OS would use to wake up the device, since we don't know it even exists. I guess there could be some magic switch accessible via USB.
But if that were the case, I'm sure Emmanuel would know about it.

Quoted 4 lines

Hm... Don't be so sure... :)
I don't think we have anything as fancy as this.
I guess you can try to dig into the BIOS settings?
I have heard of such a switch that would make the device disappear.

It's worth looking, but I don't understand how a BIOS switch would solve this problem. I assume that with the same BIOS settings, Windows works and Linux fails.

Maybe there would be a clue in an acpidump from affected machines, e.g., maybe we'd see some kind of ACPI hotplug notification. That seems like a long shot because we do have acpiphp in the kernel, and it *should* be handling such notifications, but it could always be broken.

The Windows device manager info (requested above) would be interesting.

Bjorn

#5 lore raw
Quoted 51 lines
Quoted 38 lines
Quoted 38 lines
Quoted 20 lines

https://bugzilla.kernel.org/show_bug.cgi?id=201647

            Bug ID: 201647
           Summary: Intel Wireless card 3165 does not get detected but
                    bluetooth works
           Product: Drivers
           Version: 2.5
    Kernel Version: 4.19.1
          Hardware: Intel
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: high
          Priority: P1
         Component: PCI
          Assignee: [email protected]
          Reporter: [email protected]
        Regression: No
Quoted 6 lines

This bug affects most of the devices with a Celeron N4000 and an Intel wifi 3165 Ac adapter.

When using Linux wifi is not working however, Bluetooth is working fine. Also, Bluetooth part of this chip is connected via btusb and the wifi part of this chip is connected via PCIe.

Can you attach a screenshot of the Windows 10 device manager info for the wifi adapter to the bugzilla? If you can get a raw hex dump of its config space, that would be awesome.

Also attach a copy of your kernel .config file (typically in /boot/).

My only guess is that maybe the system keeps wifi completely powered down and uses hotplug to add it when needed. [1] mentions wifi being on pcibus 1 under Windows. Your lspci does show bridge 00:13.0 leading to bus 01, but Linux doesn't find any devices on bus 01.

Quoted 6 lines
Quoted 6 lines

Hotplug could be done via either acpiphp (ACPI mediated hotplug) or pciehp (native PCIe hotplug). Your dmesg shows you do have acpiphp.

I can't tell about pciehp (your .config will show that), but I think pciehp will only claim bridges where SltCap contains HotPlug+, and yours shows HotPlug- , so I don't think pciehp will do anything on your

system.

Quoted 6 lines
Quoted 5 lines

Even if the system does use hotplug, I don't know what mechanism the OS would use to wake up the device, since we don't know it even exists. I guess there could be some magic switch accessible via USB. But if that were the case, I'm sure Emmanuel would know about it.

Quoted 8 lines
Quoted 4 lines

Hm... Don't be so sure... :)
I don't think we have anything as fancy as this.
I guess you can try to dig into the BIOS settings?
I have heard of such a switch that would make the device disappear.

It's worth looking, but I don't understand how a BIOS switch would solve this problem. I assume that with the same BIOS settings, Windows works and Linux fails.

I guess I had a typo there... I have *not* heard of such a switch.

Quoted 8 lines

Maybe there would be a clue in an acpidump from affected machines, e.g., maybe we'd see some kind of ACPI hotplug notification. That seems like a long shot because we do have acpiphp in the kernel, and it *should* be handling such notifications, but it could always be broken.

The Windows device manager info (requested above) would be interesting.

Indeed.
FWIW: I saw another problem like this with a 9650 device.

PS: PCI folks don't use bugzilla's anymore? It's all over the mailing list?

#6 lore raw
Quoted 3 lines · Emmanuel
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line

https://bugzilla.kernel.org/show_bug.cgi?id=201647

The Windows device manager info (requested above) would be interesting.

I *thought* I had requested it, but I can't find that. Mert, thanks for the screenshot you attached. Would you mind also capturing the other tabs from device manager? I'm really looking for a hex dump of the config space, which should be at least 256 bytes. I *think* that's available somewhere in device manager. If not, AIDA64
(https://www.aida64.com/downloads) should be able to get it.

Quoted 1 line

FWIW: I saw another problem like this with a 9650 device.

Do you have a pointer to any info about this 9650 issue? Maybe that would have a clue.

Quoted 1 line

PS: PCI folks don't use bugzilla's anymore? It's all over the mailing list?

Personally, I use bugzilla as a convenient place to archive artifacts like logs, screenshots, etc. But it doesn't seem as convenient for conversations as the mailing list, and it doesn't have the visibility that the list does.

Bjorn

#7 lore raw

+ the person who reported the problem with 9560

Quoted 14 lines · Emmanuel
Quoted 3 lines
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line

https://bugzilla.kernel.org/show_bug.cgi?id=201647

The Windows device manager info (requested above) would be

interesting.

I *thought* I had requested it, but I can't find that. Mert, thanks for the screenshot you attached. Would you mind also capturing the other tabs from device manager? I'm really looking for a hex dump of the config space, which should be at least 256 bytes. I *think* that's available somewhere in device manager. If not, AIDA64
(https://www.aida64.com/downloads) should be able to get it.

Quoted 4 lines
Quoted 1 line

FWIW: I saw another problem like this with a 9650 device.

Do you have a pointer to any info about this 9650 issue? Maybe that would have a clue.

I added him here so that you can ask whatever you like :)
The story is the same. I am attaching a picture I got from the reporter. I can confirm that this device ID / subdevice ID is supported by the linux driver.

Quoted 7 lines
Quoted 2 lines

PS: PCI folks don't use bugzilla's anymore? It's all over the mailing list?

Personally, I use bugzilla as a convenient place to archive artifacts like logs, screenshots, etc. But it doesn't seem as convenient for conversations as the mailing list, and it doesn't have the visibility that the list does.

#8 lore raw
Quoted 11 lines · Emmanuel
Quoted 7 lines · Emmanuel
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

https://bugzilla.kernel.org/show_bug.cgi?id=201647

FWIW: I saw another problem like this with a 9650 device.

Do you have a pointer to any info about this 9650 issue? Maybe that would have a clue.

I added him here so that you can ask whatever you like :)
The story is the same. I am attaching a picture I got from the reporter. I can confirm that this device ID / subdevice ID is supported by the linux driver.

Was there a resolution, i.e., if somebody figured out how to make the 9650 wifi work, maybe a similar solution would work for the 3165?

#9 lore raw
Quoted 10 lines · Emmanuel
Quoted 8 lines
Quoted 7 lines · Emmanuel
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

https://bugzilla.kernel.org/show_bug.cgi?id=201647

FWIW: I saw another problem like this with a 9650 device.

Do you have a pointer to any info about this 9650 issue? Maybe that would have a clue.

Quoted 7 lines
Quoted 3 lines

I added him here so that you can ask whatever you like :) The story is the same. I am attaching a picture I got from the reporter.
I can confirm that this device ID / subdevice ID is supported by the linux

driver.

Was there a resolution, i.e., if somebody figured out how to make the 9650 wifi work, maybe a similar solution would work for the 3165?

Unfortunately not, I am stuck there at the exact same point. No Wifi device shows up in lspci...

#10 lore raw
Quoted 12 lines · Emmanuel
Quoted 12 lines · Emmanuel
Quoted 11 lines
Quoted 7 lines · Emmanuel
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

https://bugzilla.kernel.org/show_bug.cgi?id=201647

FWIW: I saw another problem like this with a 9650 device.

Do you have a pointer to any info about this 9650 issue? Maybe that would have a clue.

I added him here so that you can ask whatever you like :) The story is the same. I am attaching a picture I got from the reporter.
I can confirm that this device ID / subdevice ID is supported by the linux

Quoted 7 lines
Quoted 4 lines

driver.

Was there a resolution, i.e., if somebody figured out how to make the 9650 wifi work, maybe a similar solution would work for the 3165?

Unfortunately not, I am stuck there at the exact same point. No Wifi device shows up in lspci...

Is there a public spec? My patience for reverse engineering things that are documented somewhere that I'm not allowed to see is pretty limited. Presumably *you* should be able to see the spec, since this is an Intel device and you have an @intel.com address?

From what we can tell so far, I don't *think* this is a PCI issue. If the device doesn't respond to a PCI config access, the PCI core can't really be expected to do anything about it.

Bjorn

#11 lore raw
Quoted 22 lines · Emmanuel
Quoted 2 lines
Quoted 2 lines · Emmanuel
Quoted 1 line
Quoted 1 line

On Tue, Dec 11, 2018 at 06:35:20AM +0000, Grumbach, Emmanuel

Quoted 17 lines · wrote:
Quoted 16 lines
Quoted 11 lines
Quoted 6 lines
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

https://bugzilla.kernel.org/show_bug.cgi?id=201647

FWIW: I saw another problem like this with a 9650 device.

Do you have a pointer to any info about this 9650 issue? Maybe that would have a clue.

I added him here so that you can ask whatever you like :) The story is the same. I am attaching a picture I got from the reporter. I can confirm that this device ID / subdevice ID is supported by the linux

driver.

Was there a resolution, i.e., if somebody figured out how to make the
9650 wifi work, maybe a similar solution would work for the 3165?

Quoted 7 lines
Quoted 2 lines

Unfortunately not, I am stuck there at the exact same point. No Wifi device shows up in lspci...

Is there a public spec? My patience for reverse engineering things that are documented somewhere that I'm not allowed to see is pretty limited. Presumably *you* should be able to see the spec, since this is an Intel device and you have an @intel.com address?

Ok, I start to understand. So no, no public spec. I am an Intel employee,
I used to maintain the WiFi driver (iwlwifi). I don't know much about PCI though. I am definitely not asking you to reverse engineer our device :)
I can try to ask internally what could be going on.
The reason I haven't done so so far and reached out to you is because I am not aware about switches that we need to toggle to make the device magically show up. So I thought it'd be related to the PCI core, which take me to the next point...

Quoted 4 lines

From what we can tell so far, I don't *think* this is a PCI issue. If the device doesn't respond to a PCI config access, the PCI core can't really be expected to do anything about it.

Ok, thanks. So I understand from you that you didn't see any weird log from the PCI core saying: "I can see this device, but I can't enumerate it because XYZ". I tried to look for such a thing but couldn't.

From this point, I guess I can't do much either. I'll try to get info internally.

#12 lore raw
Quoted 40 lines · Emmanuel
Quoted 22 lines · Emmanuel
Quoted 2 lines
Quoted 2 lines · Emmanuel
Quoted 1 line
Quoted 1 line

On Tue, Dec 11, 2018 at 06:35:20AM +0000, Grumbach, Emmanuel

Quoted 17 lines · wrote:
Quoted 16 lines
Quoted 11 lines
Quoted 6 lines
Quoted 3 lines
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

https://bugzilla.kernel.org/show_bug.cgi?id=201647

FWIW: I saw another problem like this with a 9650 device.

Do you have a pointer to any info about this 9650 issue? Maybe that would have a clue.

I added him here so that you can ask whatever you like :) The story is the same. I am attaching a picture I got from the reporter. I can confirm that this device ID / subdevice ID is supported by the linux

driver.

Was there a resolution, i.e., if somebody figured out how to make the
9650 wifi work, maybe a similar solution would work for the 3165?

Quoted 8 lines
Quoted 2 lines

Unfortunately not, I am stuck there at the exact same point. No Wifi device shows up in lspci...

Is there a public spec? My patience for reverse engineering things that are documented somewhere that I'm not allowed to see is pretty limited. Presumably *you* should be able to see the spec, since this is an Intel device and you have an @intel.com address?

Ok, I start to understand. So no, no public spec. I am an Intel employee, I used to maintain the WiFi driver (iwlwifi). I don't know much about PCI though. I am definitely not asking you to reverse engineer our device :) I can try to ask internally what could be going on. The reason I haven't done so so far and reached out to you is because I am not aware about switches that we need to toggle to make the device magically show up. So I thought it'd be related to the PCI core, which take me to the next point...

Quoted 8 lines
Quoted 3 lines

From what we can tell so far, I don't *think* this is a PCI issue. If the device doesn't respond to a PCI config access, the PCI core can't really be expected to do anything about it.

Ok, thanks. So I understand from you that you didn't see any weird log from the PCI core saying: "I can see this device, but I can't enumerate it because XYZ". I tried to look for such a thing but couldn't.

Right; we should be doing a config read to the device, but apparently we don't get a response so we think it's not there. It's still conceivable there's some hotplug or power management wrinkle that we aren't handling correctly, but those things are used on many systems, so it would have to be a fairly esoteric issue.

Quoted 2 lines

From this point, I guess I can't do much either. I'll try to get info internally.

Great, thanks!

Bjorn