Hi Matthijs, > -----Original Message----- > From: Matthijs Kooijman [mailto:matthijs@xxxxxxxx] > > Before, the DISCONNINT interrupt was enabled in > dwc2_enable_host_interrupts, but handled in dwc2_handle_common_intr, > while the RXFLVL interrupt was enabled in dwc2_enable_commont_interrupts > and handled in dwc_handle_hcd_intr. I guess you meant dwc2_hcd_intr, not dwc_handle_hcd_intr. > This moves the enabling of these interrupts around to match the spot > where they are handled. Actually, this is conceptually wrong. The DISCONNINT interrupt is a host-only interrupt, so it _should_ be enabled in dwc2_enable_host_interrupts(). There may be a reason why this interrupt is handled in dwc2_handle_common_intr() instead of in dwc2_hcd_intr(). It could be because, if the user does the disconnect by unplugging the USB cable from the host, then the core will immediately switch to device mode, so the dwc2_is_host_mode() test in dwc2_hcd_intr() will fail, and the DISCONNINT will not get cleared. Instead of this patch, maybe it would be better to move the handling of DISCONNINT to dwc2_hcd_intr(), but _before_ the call to dwc2_is_host_mode(). And put a comment there that explains this. And, the RXFLVL interrupt is a host+device mode interrupt, so it _should_ be enabled in dwc2_enable_common_interrupts. The only handler for it is in dwc2_hcd_intr() because there is no device- mode driver code yet. Once there is, there will either be two handlers, or else a common handler with different code depending on the mode. -- Paul -- 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
