Re: [PATCH V4 3/5] nvme-pci: avoid nvme_dev_disable to be invoked in nvme_timeout

Hi Keith Would you please take a look at this patch. I really need your suggestion on this. Sincerely Jianchao On 03/09/2018 10:01 AM, jianchao.wang wrote: > Hi Keith > > Can I have the honor of getting your comment on this patch? > > Thanks in advance > Jianchao > > On 03/08/2018 02:19 PM, J

Re: [PATCH V4 3/5] nvme-pci: avoid nvme_dev_disable to be invoked in nvme_timeout

Hi Keith Can I have the honor of getting your comment on this patch? Thanks in advance Jianchao On 03/08/2018 02:19 PM, Jianchao Wang wrote: > nvme_dev_disable will issue command on adminq to clear HMB and > delete io cq/sqs, maybe more in the future. When adminq no response, > it has to depends

[PATCH V4 3/5] nvme-pci: avoid nvme_dev_disable to be invoked in nvme_timeout

nvme_dev_disable will issue command on adminq to clear HMB and delete io cq/sqs, maybe more in the future. When adminq no response, it has to depends on timeout path. However, nvme_timeout has to invoke nvme_dev_disable before return, so that the DMA mappings could be released safely. This will int