On Tue, Oct 31, 2017 at 06:07:39PM +0100, Stefan Agner wrote: > Hi Peter, > > I am currently trying to come up with a upstreamable solution for USB > device/host switch for our NXP i.MX 7 CoM (Colibri iMX7). My colleague > Sanchayan already posted on the mailing list in a earlier attempt for > our Colibri VFxx modules (NXP Vybrid): > http://www.spinics.net/lists/linux-usb/msg140810.html > > The situation is very similar: The device is not OTG capable, but just > allows to use a Micro-USB connector as Host when using a OTG cable. VBUS > is enabled in hardware depending on ID-Pin. We have a single 3.3V input > to the module "usbc_det". Back then your advice was to wait for the DCD > framework, however, reading up on the mailing list discussion it seems > that this did not make it mainline in a form it would be easily usable > with the Chipidea IP: > https://patchwork.kernel.org/patch/9172269/ > > Downstream we currently use a custom "state machine" which directly > calls ci_role_stop/start and usb_gadget_vbus_(dis)connect: > http://git.toradex.com/cgit/linux-toradex.git/commit/drivers/usb/chipidea/core.c?h=toradex_4.1-2.0.x-imx-next&id=1f43d373946ee34ce0cfeb667a9aa48d6995bfea > > I tried to use the Chipidea OTG FSM and extcon. The device tree bindings > state: > > - extcon: phandles to external connector devices. First phandle should point to > > external connector, which provide "USB" cable events, the second should point > > to external connector device, which provide "USB-HOST" cable events. If one > > of the external connector devices is not required, empty <0> phandle should > > be specified. > > It sounds like any of the two extcon are optional, also in > drivers/usb/chipidea/core.c a comment notes: "Each one of them is not > mandatory" > > On Colibri iMX7 the usbc_det is also connected to USB_OTG1_VBUS, and it > leads to a proper B Session Valid in OTGSC registers. > > So I tried specifying ID only (since VBUS is switched in hardware and > can be detected through B Session Valid), but this seems not to be stuck > in the OTG FSM (with an additional debug print in ci_otg_fsm_event): > [ 264.327871] ci_otg_fsm_irq > [ 264.333153] ci_hdrc ci_hdrc.0: ci_otg_fsm_event a_wait_vrise > [ 264.347984] ci_cable_notifier, event 0 > [ 264.354329] ci_otg_fsm_irq > [ 265.360930] ci_hdrc ci_hdrc.0: remove, state 4 > [ 265.367862] usb usb1: USB disconnect, device number 1 > [ 265.380912] ci_hdrc ci_hdrc.0: USB bus 1 deregistered > > It seems to me that the CI OTG FSM is too complex for what I am trying > to achieve. Any advice how I should go about it? > Stefan, if you only search for solution for dual-role switch through ID pin, you do not need OTG FSM. Just keep your vbus and ID value are correct during every manual operation. -- Best Regards, Peter Chen -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
