Date: Tue, 17 Jun 2014 08:44:55 -0700 From: Sean Bruno <sbruno@ignoranthack.me> To: Hans Petter Selasky <hps@selasky.org> Cc: freebsd-usb@freebsd.org Subject: Re: Lenovo T61, USB fails to power on after resume [solved?] Message-ID: <1403019895.39699.5.camel@bruno> In-Reply-To: <538E2E55.6000202@selasky.org> References: <1401807398.96874.3.camel@bruno> <538DEFD3.2010406@selasky.org> <1401813374.1114.0.camel@bruno> <538DFEA8.3090607@selasky.org> <1401820088.1120.9.camel@bruno> <538E292A.6090804@selasky.org> <1401826306.1282.2.camel@bruno> <538E2E55.6000202@selasky.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 2014-06-03 at 22:21 +0200, Hans Petter Selasky wrote:
> On 06/03/14 22:11, Sean Bruno wrote:
> > On Tue, 2014-06-03 at 21:59 +0200, Hans Petter Selasky wrote:
> >> On 06/03/14 20:28, Sean Bruno wrote:
> >>> On Tue, 2014-06-03 at 18:58 +0200, Hans Petter Selasky wrote:
> >>>> On 06/03/14 18:36, Sean Bruno wrote:
> >>>>> On Tue, 2014-06-03 at 17:54 +0200, Hans Petter Selasky wrote:
> >>>>>> On 06/03/14 16:56, Sean Bruno wrote:
> >>>>>>> Noted that on resume, the USB ports on my T61 don't seem to be active.
> >>>>>>>
> >>>>>>> How should I go about debugging this?
> >>>>>>>
> >>>>>>> sean
> >>>>>>
> >>>>>> Hi,
> >>>>>>
> >>>>>> The USB stack performs the same EHCI/OHCI/UHCI/XHCI reset which is does
> >>>>>> during power on, when it resumes. Ensure the ports are powered. +5V.
> >>>>>> Might be a BIOS/PCI/ACPI issue.
> >>>>>>
> >>>>>> --HPS
> >>>>>>
> >>>>>
> >>>>>
> >>>>> Is there something in the output of usbconfig that I can poke at to see
> >>>>> if the hardware *thinks* it is powered on?
> >>>>>
> >>>>> sean
> >>>>>
> >>>>>
> >>>>
> >>>> Yes, there is the port status.
> >>>>
> >>>> struct usb_port_status {
> >>>> uWord wPortStatus;
> >>>> #define UPS_CURRENT_CONNECT_STATUS 0x0001
> >>>> #define UPS_PORT_ENABLED 0x0002
> >>>> #define UPS_SUSPEND 0x0004
> >>>> #define UPS_OVERCURRENT_INDICATOR 0x0008
> >>>> #define UPS_RESET 0x0010
> >>>> #define UPS_PORT_L1 0x0020 /* USB 2.0 only */
> >>>> /* The link-state bits are valid for Super-Speed USB HUBs */
> >>>> #define UPS_PORT_LINK_STATE_GET(x) (((x) >> 5) & 0xF)
> >>>> #define UPS_PORT_LINK_STATE_SET(x) (((x) & 0xF) << 5)
> >>>> #define UPS_PORT_LS_U0 0x00
> >>>> #define UPS_PORT_LS_U1 0x01
> >>>> #define UPS_PORT_LS_U2 0x02
> >>>> #define UPS_PORT_LS_U3 0x03
> >>>> #define UPS_PORT_LS_SS_DIS 0x04
> >>>> #define UPS_PORT_LS_RX_DET 0x05
> >>>> #define UPS_PORT_LS_SS_INA 0x06
> >>>> #define UPS_PORT_LS_POLL 0x07
> >>>> #define UPS_PORT_LS_RECOVER 0x08
> >>>> #define UPS_PORT_LS_HOT_RST 0x09
> >>>> #define UPS_PORT_LS_COMP_MODE 0x0A
> >>>> #define UPS_PORT_LS_LOOPBACK 0x0B
> >>>> #define UPS_PORT_LS_RESUME 0x0F
> >>>> #define UPS_PORT_POWER 0x0100
> >>>> #define UPS_PORT_POWER_SS 0x0200 /* super-speed only */
> >>>> #define UPS_LOW_SPEED 0x0200
> >>>> #define UPS_HIGH_SPEED 0x0400
> >>>> #define UPS_OTHER_SPEED 0x0600 /* currently FreeBSD
> >>>> specific */
> >>>> #define UPS_PORT_TEST 0x0800
> >>>> #define UPS_PORT_INDICATOR 0x1000
> >>>> #define UPS_PORT_MODE_DEVICE 0x8000 /* currently FreeBSD
> >>>> specific */
> >>>> uWord wPortChange;
> >>>> #define UPS_C_CONNECT_STATUS 0x0001
> >>>> #define UPS_C_PORT_ENABLED 0x0002
> >>>> #define UPS_C_SUSPEND 0x0004
> >>>> #define UPS_C_OVERCURRENT_INDICATOR 0x0008
> >>>> #define UPS_C_PORT_RESET 0x0010
> >>>> #define UPS_C_PORT_L1 0x0020 /* USB 2.0 only */
> >>>> #define UPS_C_BH_PORT_RESET 0x0020 /* USB 3.0 only */
> >>>> #define UPS_C_PORT_LINK_STATE 0x0040
> >>>> #define UPS_C_PORT_CONFIG_ERROR 0x0080
> >>>> } __packed;
> >>>>
> >>>> It is probed regularly by the UHUB driver and the port status is printed
> >>>> in dmesg.
> >>>>
> >>>> Turn on like this:
> >>>>
> >>>> sysctl hw.usb.uhub.debug=16
> >>>>
> >>>> By resetting the root HUB, you can write new power on bits:
> >>>>
> >>>> usbconfig -d X.1 set_config 255
> >>>> usbconfig -d X.1 set_config 0
> >>>>
> >>>> --HPS
> >>>
> >>> Well, that's problematic. The USB tree looks like this normally:
> >>>
> >>> ugen0.1: <UHCI root HUB Intel> at usbus0, cfg=0 md=HOST spd=FULL
> >>> (12Mbps) pwr=SAVE (0mA)
> >>> ugen1.1: <UHCI root HUB Intel> at usbus1, cfg=0 md=HOST spd=FULL
> >>> (12Mbps) pwr=SAVE (0mA)
> >>> ugen2.1: <EHCI root HUB Intel> at usbus2, cfg=0 md=HOST spd=HIGH
> >>> (480Mbps) pwr=SAVE (0mA)
> >>> ugen3.1: <UHCI root HUB Intel> at usbus3, cfg=0 md=HOST spd=FULL
> >>> (12Mbps) pwr=SAVE (0mA)
> >>> ugen4.1: <UHCI root HUB Intel> at usbus4, cfg=0 md=HOST spd=FULL
> >>> (12Mbps) pwr=SAVE (0mA)
> >>> ugen5.1: <UHCI root HUB Intel> at usbus5, cfg=0 md=HOST spd=FULL
> >>> (12Mbps) pwr=SAVE (0mA)
> >>> ugen6.1: <EHCI root HUB Intel> at usbus6, cfg=0 md=HOST spd=HIGH
> >>> (480Mbps) pwr=SAVE (0mA)
> >>> ugen0.2: <Biometric Coprocessor STMicroelectronics> at usbus0, cfg=0
> >>> md=HOST spd=FULL (12Mbps) pwr=ON (100mA)
> >>>
> >>>
> >>> But, on resume ... sometimes ... ugen0.1 is just flatout gone (along
> >>> with the ugen0.2 device, obviously). This only seems to happen with
> >>> various USB device plugged in (tried about 4 different make/model usb
> >>> sticks and ext drives).
> >>>
> >>> So, resetting doesn't work as the device is literally gone. Thoughts?
> >>>
> >>> sean
> >>>
> >>
> >> Setting hw.usb.debug=15 should give you some hints. Are you sure the PCI
> >> device is still there after resume?
> >>
> >> --HPS
> >>
> >
> > I did a pre/post resume pciconf -lv and I see no difference between the
> > two.
> >
> > On initial resume, ugen0.1 appears to be there in usbconfig output, but
> > trying to set_config on it, causes it to dissapear and causes usbconfig
> > to hang:
> >
> > root@bruno:/home/sbruno # usbconfig -d 0.1 set_config 255
> > load: 0.36 cmd: usbconfig 1540 [UGONE] 4.19r 0.00u 0.00s 0% 2064k
> > load: 0.36 cmd: usbconfig 1540 [UGONE] 5.00r 0.00u 0.00s 0% 2064k
> > load: 0.36 cmd: usbconfig 1540 [UGONE] 5.16r 0.00u 0.00s 0% 2064k
> > load: 0.36 cmd: usbconfig 1540 [UGONE] 5.35r 0.00u 0.00s 0% 2064k
> > load: 0.36 cmd: usbconfig 1540 [UGONE] 5.51r 0.00u 0.00s 0% 2064k
> >
> >
>
> Hi,
>
> What kind of USB devices are attached? Are any of these USB devices
> preventing detach? Could you check the USB threads using kgdb?
>
> --HPS
>
Btw, this appears to be solvable with a patch to acpi_ibm(4) from trasz:
http://lists.freebsd.org/pipermail/freebsd-current/2014-June/050721.html
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1403019895.39699.5.camel>