All of lore.kernel.org
 help / color / mirror / Atom feed
* Lowspeed USB device enumerationg under 5.4.288, but not under 6.8.1
@ 2026-04-02 15:34 Fischer, Simon
  2026-04-02 16:10 ` Greg KH
  2026-05-06  6:58 ` Michal Pecio
  0 siblings, 2 replies; 10+ messages in thread
From: Fischer, Simon @ 2026-04-02 15:34 UTC (permalink / raw)
  To: regressions@lists.linux.dev

Dear list,

I've tried to dig deeper into the issues, tried different quirks and small patches, but was unable to solve this or find the root cause. As this seems to be a regression of 6.8 from 5.4, I hope this is the correct list?

I got setup with a low speed isochronous USB device (Optec filter wheel), that is connected to a Dell R6515 server via a Lindy 300m 2 Port Fibre Optic USB 3.0 Extender (which enumerates as a chain of USB3 super speed hubs).
The server is running Ubuntu 24.04 with a 6.8.1 RT kernel. When connecting the device, it fails to enumerate:

		[ 1843.938804] usb 2-1.2.1: new SuperSpeed USB device number 14 using xhci_hcd
		[ 1843.950893] xhci_hcd 0000:04:00.3: Waiting for status stage event
		[ 1843.950937] xhci_hcd 0000:04:00.3: Waiting for status stage event
		[ 1843.950981] xhci_hcd 0000:04:00.3: Waiting for status stage event
		[ 1843.950998] usb 2-1.2.1: New USB device found, idVendor=0424, idProduct=7016, bcdDevice= 6.26
		[ 1843.951006] usb 2-1.2.1: New USB device strings: Mfr=1, Product=2, SerialNumber=0
		[ 1843.951013] usb 2-1.2.1: Product: USB7016 Smart Hub
		[ 1843.951018] usb 2-1.2.1: Manufacturer: Microchip
		[ 1843.951365] xhci_hcd 0000:04:00.3: add ep 0x81, slot id 4, new drop flags = 0x0, new add flags = 0x8
		[ 1843.951375] xhci_hcd 0000:04:00.3: xhci_check_bandwidth called for udev 00000000f628da0a
		[ 1843.951387] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1843.952768] xhci_hcd 0000:04:00.3: Successful Endpoint Configure command
		[ 1843.952825] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1843.952865] xhci_hcd 0000:04:00.3: Stopped on No-op or Link TRB for slot 4 ep 2
		[ 1843.952884] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1843.969034] hub 2-1.2.1:1.0: USB hub found
		[ 1843.969108] hub 2-1.2.1:1.0: 3 ports detected
		[ 1843.969459] xhci_hcd 0000:04:00.3: xHCI version 110 needs hub TT think time and number of ports
		[ 1843.969466] xhci_hcd 0000:04:00.3: Set up configure endpoint for hub device.
		[ 1843.969474] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1843.969529] xhci_hcd 0000:04:00.3: Successful Endpoint Configure command
		[ 1844.178598] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1844.178634] xhci_hcd 0000:04:00.3: Slot 5 output ctx = 0x0x0000000179f5f000 (dma)
		[ 1844.178643] xhci_hcd 0000:04:00.3: Slot 5 input ctx = 0x0x00000001311ce000 (dma)
		[ 1844.178652] xhci_hcd 0000:04:00.3: Set slot id 5 dcbaa entry 00000000bb87fc6b to 0x179f5f000
		[ 1844.208850] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.224776] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.240774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.242544] xhci_hcd 0000:04:00.3: Set root hub portnum to 3
		[ 1844.242554] xhci_hcd 0000:04:00.3: Set fake root hub portnum to 1
		[ 1844.242560] xhci_hcd 0000:04:00.3: udev->tt = 0000000000000000
		[ 1844.242565] xhci_hcd 0000:04:00.3: udev->ttport = 0x0
		[ 1844.242573] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1844.250778] usb 2-1.2.1.3: Device not responding to setup address.
		[ 1844.250888] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1844.250919] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1844.250959] xhci_hcd 0000:04:00.3: Slot 5 output ctx = 0x0x0000000179f5f000 (dma)
		[ 1844.250968] xhci_hcd 0000:04:00.3: Slot 5 input ctx = 0x0x00000001311ce000 (dma)
		[ 1844.250976] xhci_hcd 0000:04:00.3: Set slot id 5 dcbaa entry 00000000bb87fc6b to 0x179f5f000
		[ 1844.256773] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.272775] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.288888] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.304774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.320774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.336774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.352774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.368774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.384774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.400774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.416774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.432774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.448774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.458535] xhci_hcd 0000:04:00.3: Set root hub portnum to 3
		[ 1844.458544] xhci_hcd 0000:04:00.3: Set fake root hub portnum to 1
		[ 1844.458550] xhci_hcd 0000:04:00.3: udev->tt = 0000000000000000
		[ 1844.458556] xhci_hcd 0000:04:00.3: udev->ttport = 0x0
		[ 1844.458563] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1844.464774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.466663] usb 2-1.2.1.3: Device not responding to setup address.
		[ 1844.466771] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1844.466801] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1844.466843] xhci_hcd 0000:04:00.3: Slot 5 output ctx = 0x0x0000000179f5f000 (dma)
		[ 1844.466851] xhci_hcd 0000:04:00.3: Slot 5 input ctx = 0x0x00000001311ce000 (dma)
		[ 1844.466860] xhci_hcd 0000:04:00.3: Set slot id 5 dcbaa entry 00000000bb87fc6b to 0x179f5f000
		[ 1844.480774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.496774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.512774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.528774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.544774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.560774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.576774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.592774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.608776] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.624777] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.640774] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.656775] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.672776] xhci_hcd 0000:04:00.3: ep 0x81 - asked for 2 bytes, 1 bytes untransferred
		[ 1844.674524] usb 2-1.2.1.3: device not accepting address 15, error -71
		[ 1844.675151] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1844.675178] xhci_hcd 0000:04:00.3: Bad real port.
		[ 1844.675188] xhci_hcd 0000:04:00.3: // Ding dong!

	[this repeats multiple times until giving up]

		[ 1864.698065] usb 2-1.2.1.3.4: device not accepting address 25, error -71
		[ 1864.698674] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1864.698711] xhci_hcd 0000:04:00.3: Bad real port.
		[ 1864.698718] usb 2-1.2.1.3-port4: unable to enumerate USB device
		[ 1864.699443] xhci_hcd 0000:04:00.3: Cancel URB 00000000c8fb420b, dev 1.2.1.3, ep 0x81, starting at offset 0x16e8fb050
		[ 1864.699455] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1864.699474] xhci_hcd 0000:04:00.3: Stopped on Transfer TRB for slot 5 ep 2
		[ 1864.699481] xhci_hcd 0000:04:00.3: Removing canceled TD starting at 0x16e8fb050 (dma) in stream 0 URB 00000000c8fb420b
		[ 1864.699489] xhci_hcd 0000:04:00.3: Set TR Deq ptr 0x16e8fb060, cycle 1

		[ 1864.699495] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1864.699500] xhci_hcd 0000:04:00.3: xhci_giveback_invalidated_tds: Keep cancelled URB 00000000c8fb420b TD as cancel_status is 2
		[ 1864.699510] xhci_hcd 0000:04:00.3: Successful Set TR Deq Ptr cmd, deq = @16e8fb060
		[ 1864.699517] xhci_hcd 0000:04:00.3: xhci_handle_cmd_set_deq: Giveback cancelled URB 00000000c8fb420b TD
		[ 1864.699522] xhci_hcd 0000:04:00.3: Giveback URB 00000000c8fb420b, len = 0, expected = 2, status = -115
		[ 1864.711179] xhci_hcd 0000:04:00.3: Cancel URB 00000000301149a6, dev 1.2.1, ep 0x81, starting at offset 0x125370300
		[ 1864.711194] xhci_hcd 0000:04:00.3: // Ding dong!
		[ 1864.711211] xhci_hcd 0000:04:00.3: Stopped on Transfer TRB for slot 4 ep 2
		[ 1864.711218] xhci_hcd 0000:04:00.3: Removing canceled TD starting at 0x125370300 (dma) in stream 0 URB 00000000301149a6
		[ 1864.711226] xhci_hcd 0000:04:00.3: Set TR Deq ptr 0x125370310, cycle 1

		[ 1864.711232] xhci_hcd 0000:04:00.3: // Ding dong!


When booting into a 5.4.288 PREEMPT_RT kernel we were using before with, everything works fine (device enumerates and works).
When exchanging the USB Extender for a Lindy 200m Fibre Optic USB 2.0 Extender, everything works fine with the 6.8.1 kernel.
I cannot inject a USB 2.0 hub before the extender, because it is a super speed device and not compatible.

Any suggestions how to fix or further investigate the issue would be greatly appreciated!

Kind regards
Simon Fischer


Developer
Department E5 - CoDaC


Max Planck Institute for Plasmaphysics
Wendelsteinstrasse 1
17491 Greifswald, Germany

Phone: +49(0)3834 88 1215


^ permalink raw reply	[flat|nested] 10+ messages in thread
* RE: Lowspeed USB device enumerationg under 5.4.288, but not under 6.8.1
@ 2026-05-06  7:19 Fischer, Simon
  2026-05-06  8:34 ` Michal Pecio
  0 siblings, 1 reply; 10+ messages in thread
From: Fischer, Simon @ 2026-05-06  7:19 UTC (permalink / raw)
  To: Michal Pecio; +Cc: regressions@lists.linux.dev



> On Wed, 6 May 2026 08:58, Pecio, Michal wrote: 
> > On Thu, 2 Apr 2026 15:34:35 +0000, Fischer, Simon wrote:
> > Dear list,
> >
> > I've tried to dig deeper into the issues, tried different quirks and
> > small patches, but was unable to solve this or find the root cause.
> > As this seems to be a regression of 6.8 from 5.4, I hope this is the
> > correct list?
> >
> > I got setup with a low speed isochronous USB device (Optec filter
> > wheel), that is connected to a Dell R6515 server via a Lindy 300m 2
> > Port Fibre Optic USB 3.0 Extender (which enumerates as a chain of
> > USB3 super speed hubs). The server is running Ubuntu 24.04 with a
> > 6.8.1 RT kernel. When connecting the device, it fails to enumerate:
> 
> Hi,
> 
> Are you still fighting with this?
> 
> I'm afraid I have no better suggestions than others: bisect it.
> To avoid bisecting all commits, you could manually bisect major
> releases and then test USB commits added by the "bad" release,
> as the problem likely came from USB subsystem.
> 
> By "low speed", do you mean the 1Mbps data rate defined by USB 1.x?
> That's odd, such devices aren't allowed to use isochronous endpoints
> so it would be big luck if they work at all with any SW and HW.
> 
> Would you mind posting its 'lsusb -v', from the working system?
> 
> Your log shows a failure to enumerate some SuperSpeed device, while
> any USB1/USB2 device should be showing on another bus (probably 1),
> unless your range extender somehow changes "apparent" speed.
> 
> Also, if the problem reproduces on current kernels, you could try
> sending logs (from a current kernel) to linux-usb in order to reach
> more USB developers.
> 
> Regards,
> Michal

Hi Michal,

thanks for coming back. I had as of yet postponing an update for workload reasons, but mostly because my Outlook is very bad at following the rules of this mailing list :-)

I have been able to reproduce the problem with a current (6.19) kernel and independent of PREEMPT_RT status. But I have also been able to identify that the difference between my Ubuntu provided 5.4 and 6.8 kernels that leads to the problem, is actually the CONFIG_HZ build config parameter.

Ubuntu's 6.8 kernel is by default built with CONFIG_HZ=1000, while the old 5.4 kernel was built with CONFIG_HZ=250. If I rebuild Ubuntu's 5.4 with 1000Hz, I also cannot talk to the devices, and when I compile 6.19 or 6.8 with 250Hz, they work (currently running a rebuilt Ubuntu 6.8.1 with only that one parameter updated, I sometimes still see errors, but an automatic retry of the driver is then successful).

So, a regression this is not :-)

Since I cannot imagine in what way the system interrupt frequency would directly affect this part in such a reproducible and significant manner, I would think it more likely that the USB core or xhci drivers actually use that setting value somewhere themselves and maybe wrongly. So this should probably realy go to some other list. If you know which, I would greatly appreciate you forwarding my information  or telling me where to take it.

Thanks a lot & best regards
Simon

> 
> > [...]
> >
> >
> > When booting into a 5.4.288 PREEMPT_RT kernel we were using before
> > with, everything works fine (device enumerates and works). When
> > exchanging the USB Extender for a Lindy 200m Fibre Optic USB 2.0
> > Extender, everything works fine with the 6.8.1 kernel. I cannot
> > inject a USB 2.0 hub before the extender, because it is a super speed
> > device and not compatible.
> >
> > Any suggestions how to fix or further investigate the issue would be
> > greatly appreciated!
> >
> > Kind regards
> > Simon Fischer
> >
> >
> > Developer
> > Department E5 - CoDaC
> >
> >
> > Max Planck Institute for Plasmaphysics
> > Wendelsteinstrasse 1
> > 17491 Greifswald, Germany
> >
> > Phone: +49(0)3834 88 1215
> >

^ permalink raw reply	[flat|nested] 10+ messages in thread
end of thread, other threads:[~2026-05-06  8:34 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-04-02 15:34 Lowspeed USB device enumerationg under 5.4.288, but not under 6.8.1 Fischer, Simon
2026-04-02 16:10 ` Greg KH
2026-04-03 16:34   ` Simon Fischer
2026-04-07  6:33   ` AW: " Fischer, Simon
2026-04-07  6:49     ` Greg KH
2026-04-07  7:25     ` Thorsten Leemhuis
2026-04-07  7:28   ` Fischer, Simon
2026-05-06  6:58 ` Michal Pecio
  -- strict thread matches above, loose matches on Subject: below --
2026-05-06  7:19 Fischer, Simon
2026-05-06  8:34 ` Michal Pecio

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.