Re: possible regression in usb-storage?

[Johann Klammer <klammerj@a1.net>]

On 9/24/26 4:28 PM, Alan Stern wrote:
> On Thu, Sep 24, 2026 at 07:18:20AM +0200, Johann Klammer wrote:
>>
>>
>> On 9/24/26 4:08 AM, Alan Stern wrote:
>>>
>>> Better still would be is you could also collect a usbmon trace under an
>>> earlier kernel that the drive worked with, so we can compare the working
>>> trace with a current non-working trace.
>>>
>>
>> this one is from a 4.9.210-1+deb9u1~deb8u1 (2020-06-09) i686 GNU/Linux
>> the device node shows up and can be mounted but the drive light is on
>> permanently and I get
>>> [31473.010372] usb 2-7: USB disconnect, device number 4
>>> [31473.016105] sd 4:0:0:0: [sdc] Synchronizing SCSI cache
>>> [31473.256060] sd 4:0:0:0: [sdc] Synchronize Cache(10) failed: Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK
> 
> When do you get this?  When you unplug the drive?

yes

> 
>> the log filtered by device 2:004
> 
> ...
> 
>>> ddba6300 2786664488 S Bi:2:004:2 -115 112 <
>>> dcd97860 2786664509 S Bo:2:004:1 -115 32 = 01000001 00000000 00000000 00000000 12000000 24000000 00000000 00000000
> 
> This trace shows that the uas driver was bound to the device, not
> usb-storage.  So it is not at all comparable to the earlier traces.
> 
Sorry. is a different box. will edit modprobe.conf and regenerate..