Sebastian, On Mon, Jan 26, 2015 at 2:57 AM, Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx> wrote: > On 01/21/2015 06:06 PM, Bin Liu wrote: >> Hi Sebastian, > > Hi Bin, > >> With this patch, hubs stop working on TI AM335x EVMs when autosuspend >> is enabled. >> >> I booted the board, connected a hub to the USB1 host port, it got >> enumerated properly, then connected a thumb drive to the hub, but the >> drive was not enumerated, no any log from kernel. Removing the drive, >> no any kernel message either. Finally removed the hub, no disconnect >> for the hub. Now check MUSB1 DEVCTL register, it value is 0x5D. >> >> Reverted this patch, the issue disappeared. Or disable usbcore >> autosuspend, the issue did not happen. >> >> I tested 7+ hubs, all have the same issue. >> >> I tested on git://git.ti.com/ti-linux-kernel/ti-linux-kernel.git, tag >> ti2014.10.01. have not tested with mainline kernel yet. >> >> Have you validated this test case? > > No, I had no USB-hub attached. It would be nice if you could figure out > which part is responsible for the missing detection because otherwise > suspend/resume is broken. I think I found the cause. Plugging a device behind a hub, this is related to runtime PM, so the finish_resume_work() should also be added in musb_runtime_resume(), also musb_host_resume_root_hub() should not be moved as did in your patch, this call is used to trigger the bus resume. A patch will be coming soon - tomorrow. Regards, -Bin. > >> Thanks, >> -Bin. > > Sebastian > -- 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
