[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [virtio-dev] Re: [PATCH v1] virtio-mmio: Specify wait needed in driver during reset
* Michael S. Tsirkin <firstname.lastname@example.org> [2021-08-16 10:37:18]: > > What about the case where device reset fails, which may require driver to > > initiate reset again or give up on device? Won't the read of features be forever > > deferred in that case? > > > > - vatsa > > That might be worth addressing but I feel your proposal does not > address it sufficiently. Ok let me take a stab at addressing it better in my next version. Basically here's what I am thinking. Pls let me know if there is anything else I need to consider: * Introduce a status bit for device to indicate "reset in progress". This could be adopted by other transports also (for ex PCI). * For v3 MMIO devices, driver, after initiating reset, should wait until status = 0 or device indicates "failure". Should the spec advise what a driver should do in case of failure after reset? Not sure if the outcome would be different if driver were to issue reset again. Again this advice should be common to both MMIO and PCI I think. - vatsa --- Qualcomm Innovation Center, Inc. is submitting the attached "feedback" as a non-member to the virtio-dev mailing list for consideration and inclusion.