USB 3.0 devices negotiate to spd=HIGH for all ports

Sun Aug 28 18:06:08 UTC 2016

On 08/28/16 17:52, Adam Stylinski wrote:
> On Sun, Aug 28, 2016 at 10:18 AM, Adam Stylinski <kungfujesus06 at gmail.com>
> wrote:
>
>> Looking at my motherboard's manual, this is what my BIOS exposes:
>>
>> USB 2.0 Controller Use this item to enable or disable the use of USB 2.0
>> controller. USB 3.0 Controller Use this item to enable or disable the use
>> of USB 3.0 controller. Legacy USB Support Use this option to select legacy
>> support for USB devices. There are four con guration options: [Enabled],
>> [Disabled], [Auto] and [UEFI Setup Only]. The default value is [Enabled].
>> Please refer to below descriptions for the details of these four options:
>> [Enabled] - Enables support for legacy USB. [Disabled] - USB devices are
>> not allowed to use under legacy OS and UEFI setup when [Disabled] is
>> selected. If you have USB compatibility issue, it is recommended to select
>> [Disabled] to enter OS. [Auto] - Enables legacy support if USB devices are
>> connected. [UEFI Setup Only] - USB devices are allowed to use only under
>> UEFI setup and Windows / Linux OS. Legacy USB 3.0 Support Use this option
>> to enable or disable legacy support for USB 3.0 devices.
>>
>> I believe 3.0 controller is enabled, or I wouldn't be able to use those
>> ports at all.  Now the Legacy USB 3.0 option is a bit vaguely named, as it
>> could be taken to mean, use USB 3.0 ports in legacy mode.  It could also
>> mean, "allow USB 2.0 devices with 3.0 ports".  The default for that option
>> is disabled, but it may have been misconstrued and enabled at one point.
>> Getting to the BIOS unfortunately requires me installing a video card in
>> there that's not 15 years old, unfortunately, as it's one of those
>> sophisticated new BIOSes that work with a mouse and require fancier VGA
>> BIOSes than what this S3 Virge card provides.  It's definitely worth a
>> look, though.
>>
>> On Sat, Aug 27, 2016 at 2:05 PM, Adam Stylinski <kungfujesus06 at gmail.com>
>> wrote:
>>
>>> The only quirk specific code I can find in Linux's kernel for XHCI is
>>> this:
>>> http://lxr.free-electrons.com/source/drivers/usb/host/pci-quirks.c
>>>
>>> I believe I have one of the SBxxx AMD chipsets, though I can't recall
>>> which one.  I don't see anything in there referring to my PCI IDs, unless
>>> I'm looking at the wrong outputs in pciconf -lv.  I see one quirk in
>>> particular regarding power management but that seems to be more about
>>> stability rather than speed.  I'm sure I'm missing something.
>>>
>>> On Sat, Aug 27, 2016 at 1:43 PM, Hans Petter Selasky <hps at selasky.org>
>>> wrote:
>>>
>>>> On 08/27/16 18:41, Adam Stylinski wrote:
>>>>
>>>>> Hello,
>>>>>
>>>>> I definitely have a USB 3.0 controller that is being recognized by XHCI:
>>>>>
>>>>> [adam at nasbox ~]$ sudo usbconfig
>>>>> ugen0.1: <XHCI root HUB 0x1b6f> at usbus0, cfg=0 md=HOST spd=SUPER
>>>>> (5.0Gbps) pwr=SAVE (0mA)
>>>>> ugen2.1: <OHCI root HUB ATI> at usbus2, cfg=0 md=HOST spd=FULL (12Mbps)
>>>>> pwr=SAVE (0mA)
>>>>> ugen1.1: <XHCI root HUB 0x1b6f> at usbus1, cfg=0 md=HOST spd=SUPER
>>>>> (5.0Gbps) pwr=SAVE (0mA)
>>>>> ugen3.1: <EHCI root HUB ATI> at usbus3, cfg=0 md=HOST spd=HIGH (480Mbps)
>>>>> pwr=SAVE (0mA)
>>>>> ugen4.1: <OHCI root HUB ATI> at usbus4, cfg=0 md=HOST spd=FULL (12Mbps)
>>>>> pwr=SAVE (0mA)
>>>>> ugen5.1: <EHCI root HUB ATI> at usbus5, cfg=0 md=HOST spd=HIGH (480Mbps)
>>>>> pwr=SAVE (0mA)
>>>>> ugen7.1: <OHCI root HUB ATI> at usbus7, cfg=0 md=HOST spd=FULL (12Mbps)
>>>>> pwr=SAVE (0mA)
>>>>> ugen6.1: <OHCI root HUB ATI> at usbus6, cfg=0 md=HOST spd=FULL (12Mbps)
>>>>> pwr=SAVE (0mA)
>>>>> ugen8.1: <EHCI root HUB ATI> at usbus8, cfg=0 md=HOST spd=HIGH (480Mbps)
>>>>> pwr=SAVE (0mA)
>>>>> ugen0.2: <Extreme SanDisk> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps)
>>>>> pwr=ON (200mA)
>>>>>
>>>>>
>>>>> xhci0 at pci0:3:0:0: class=0x0c0330 card=0x70231849 chip=0x70231b6f
>>>>> rev=0x01
>>>>> hdr=0x00
>>>>>     vendor     = 'Etron Technology, Inc.'
>>>>>     device     = 'EJ168 USB 3.0 Host Controller'
>>>>>     class      = serial bus
>>>>>     subclass   = USB
>>>>> xhci1 at pci0:4:0:0: class=0x0c0330 card=0x70231849 chip=0x70231b6f
>>>>> rev=0x01
>>>>> hdr=0x00
>>>>>     vendor     = 'Etron Technology, Inc.'
>>>>>     device     = 'EJ168 USB 3.0 Host Controller'
>>>>>     class      = serial bus
>>>>>     subclass   = USB
>>>>>
>>>>> However, when I connect a USB 3.0 device into a USB 3.0 port, it only
>>>>> gives
>>>>> me spd=HIGH.  This is a Sandisk CZ80, but I've other devices in the past
>>>>> and have had the same behavior.  Plugging a device in with
>>>>> hw.usb.xhci.debug=1 doesn't appear to print too many useful things, at
>>>>> least not to an end user.  I won't yet paste the full output, but it
>>>>> looks
>>>>> something like this:
>>>>>
>>>>> xhci_transfer_insert: qh_pos = 2
>>>>> xhci_check_transfer: slot=1 epno=3 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_check_transfer: TD is last
>>>>> xhci_check_transfer: slot=1 epno=3 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_device_generic_enter:
>>>>> xhci_setup_generic_chain_sub: NTRB=1
>>>>> xhci_setup_generic_chain_sub: LINK=0x241ff800
>>>>> xhci_setup_generic_chain: first=0xfffff800241ff980
>>>>> last=0xfffff800241ff980
>>>>> xhci_device_generic_start:
>>>>> xhci_transfer_insert: qh_pos = 2
>>>>> xhci_check_transfer: slot=1 epno=4 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_check_transfer: TD is last
>>>>> xhci_check_transfer: slot=1 epno=4 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_device_generic_enter:
>>>>> xhci_setup_generic_chain_sub: NTRB=1
>>>>> xhci_setup_generic_chain_sub: LINK=0x55fc6a300
>>>>> xhci_setup_generic_chain: first=0xfffff8055fc6a480
>>>>> last=0xfffff8055fc6a480
>>>>> xhci_device_generic_start:
>>>>> xhci_transfer_insert: qh_pos = 0
>>>>> xhci_check_transfer: slot=1 epno=3 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_check_transfer: TD is last
>>>>> xhci_check_transfer: slot=1 epno=3 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_device_generic_enter:
>>>>> xhci_setup_generic_chain_sub: NTRB=1
>>>>> xhci_setup_generic_chain_sub: LINK=0x3d958d000
>>>>> xhci_setup_generic_chain: first=0xfffff803d958d180
>>>>> last=0xfffff803d958d180
>>>>> xhci_device_generic_start:
>>>>> xhci_transfer_insert: qh_pos = 1
>>>>> xhci_check_transfer: slot=1 epno=3 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_check_transfer: TD is last
>>>>> xhci_check_transfer: slot=1 epno=3 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_device_generic_enter:
>>>>> xhci_setup_generic_chain_sub: NTRB=1
>>>>> xhci_setup_generic_chain_sub: LINK=0x7d676f000
>>>>> xhci_setup_generic_chain: first=0xfffff807d676f180
>>>>> last=0xfffff807d676f180
>>>>> xhci_device_generic_start:
>>>>> xhci_transfer_insert: qh_pos = 0
>>>>> xhci_check_transfer: slot=1 epno=4 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>> xhci_check_transfer: TD is last
>>>>> xhci_check_transfer: slot=1 epno=4 stream=256 remainder=0 status=1
>>>>> xhci_check_transfer: stream_id=0
>>>>>
>>>>> What can I do to end this frustrating issue of USB 2.0 only-ness?
>>>>>
>>>>
>>>> Try to google the PCI IDs of your device and see if Linux has any quirks
>>>> for your card. Sometimes these cards need special register writes to enable
>>>> USB 3.0.
>>>>
>>>> --HPS
>>>>
>>>>
>>>
>>
> Alright, so I went through the crazy hurdles to actually get to the BIOS,
> and the USB legacy support for USB 3.0 was enabled.  I disabled it, and it
> has no apparently effect, other than the fact that setting up the device
> takes a lot longer now:
>
> usbd_setup_device_desc: getting device descriptor at addr 1 failed,
> USB_ERR_TIMEOUT
> usbd_setup_device_desc: getting device descriptor at addr 1 failed,
> USB_ERR_TIMEOUT
> usbd_setup_device_desc: getting device descriptor at addr 1 failed,
> USB_ERR_TIMEOUT
> usbd_setup_device_desc: getting device descriptor at addr 1 failed,
> USB_ERR_TIMEOUT
> usbd_setup_device_desc: getting device descriptor at addr 1 failed,
> USB_ERR_TIMEOUT
> ugen0.2: <Unknown> at usbus0 (disconnected)
> uhub_reattach_port: could not allocate new device
> ugen0.2: <SanDisk> at usbus0
> umass0: <SanDisk Extreme, class 0/0, rev 2.10/0.10, addr 1> on usbus0
> umass0:  SCSI over Bulk-Only; quirks = 0x0100
> umass0:14:0:-1: Attached to scbus14
> da23 at umass-sim0 bus 0 scbus14 target 0 lun 0
> da23: <SanDisk Extreme 0001> Removable Direct Access SPC-4 SCSI device
> da23: Serial Number [PROBABLY NOT IMPORTANT ENOUGH TO SCRUB, BUT DOING SO
> ANYWAY]
> da23: 40.000MB/s transfers
> da23: 59840MB (122552320 512 byte sectors)
> da23: quirks=0x2<NO_6_BYTE>
>
> [adam at nasbox ~]$ sudo usbconfig
> ugen0.1: <XHCI root HUB 0x1b6f> at usbus0, cfg=0 md=HOST spd=SUPER
> (5.0Gbps) pwr=SAVE (0mA)
> ugen2.1: <OHCI root HUB ATI> at usbus2, cfg=0 md=HOST spd=FULL (12Mbps)
> pwr=SAVE (0mA)
> ugen1.1: <XHCI root HUB 0x1b6f> at usbus1, cfg=0 md=HOST spd=SUPER
> (5.0Gbps) pwr=SAVE (0mA)
> ugen3.1: <EHCI root HUB ATI> at usbus3, cfg=0 md=HOST spd=HIGH (480Mbps)
> pwr=SAVE (0mA)
> ugen4.1: <OHCI root HUB ATI> at usbus4, cfg=0 md=HOST spd=FULL (12Mbps)
> pwr=SAVE (0mA)
> ugen6.1: <OHCI root HUB ATI> at usbus6, cfg=0 md=HOST spd=FULL (12Mbps)
> pwr=SAVE (0mA)
> ugen5.1: <EHCI root HUB ATI> at usbus5, cfg=0 md=HOST spd=HIGH (480Mbps)
> pwr=SAVE (0mA)
> ugen7.1: <OHCI root HUB ATI> at usbus7, cfg=0 md=HOST spd=FULL (12Mbps)
> pwr=SAVE (0mA)
> ugen8.1: <EHCI root HUB ATI> at usbus8, cfg=0 md=HOST spd=HIGH (480Mbps)
> pwr=SAVE (0mA)
> ugen0.2: <Extreme SanDisk> at usbus0, cfg=0 md=HOST spd=HIGH (480Mbps)
> pwr=ON (200mA)
>
> Which steps are next?  I apologize to the list for top-posting, it's been a
> while since I've been to a mailing list and I decided to do it through
> Gmail for whatever reason instead of mutt.
>

Hi,

Typically USB 3.0 replaces all the USB 2.0 ones, so you should only have 
USB 3.0 enabled.

BTW: Can you apply and try these patches. Do they make any difference? 
Especially in the case of the USB_ERR_TIMEOUT's.

https://svnweb.freebsd.org/changeset/base/304597
https://svnweb.freebsd.org/changeset/base/304629

--HPS