From mboxrd@z Thu Jan 1 00:00:00 1970 From: peter.chen@freescale.com (Peter Chen) Date: Mon, 28 Jan 2013 11:32:32 +0800 Subject: [RESEND PATCH v5 3/7] usb: chipidea: add otg id switch and vbus connect/disconnect detect In-Reply-To: <87wqv1bnyd.fsf@ashishki-desk.ger.corp.intel.com> References: <1358733418-17969-1-git-send-email-peter.chen@freescale.com> <1358733418-17969-4-git-send-email-peter.chen@freescale.com> <87y5fimwma.fsf@ashishki-desk.ger.corp.intel.com> <20130125062827.GD29795@nchen-desktop> <87wqv1bnyd.fsf@ashishki-desk.ger.corp.intel.com> Message-ID: <20130128033203.GC2673@nchen-desktop> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Fri, Jan 25, 2013 at 11:40:10AM +0200, Alexander Shishkin wrote: > Peter Chen writes: > > > On Thu, Jan 24, 2013 at 05:25:17PM +0200, Alexander Shishkin wrote: > > What doesn't happen is the reset into device mode (unless you have > _REGS_SHARED set, which by the way seems a bit misleading) and we're > still doing nothing till the vbus interrupt comes. So the REGS_SHARED is > the real problem in this case. Yes, at imx it sets REGS_SHARD. The REGS_SHARED is needed at below conditions: - System boots up with usb device at OTG port using A-plug cable, at this situation, any gadget initialization should not do any registers operation, the related functions are udc_start and ci13xxx_start. - At udc_irq, we do not want any wakeup, id, and host interrupt to call internal code. > -- Best Regards, Peter Chen