On 9/4/2019 8:54 AM, Christoph Hellwig wrote:
On Tue, Sep 03, 2019 at 12:15:48PM -0700, Sagi Grimberg wrote:
The nvme_cleanup_cmd function should be called to avoid resource leakage
(it's the opposite to nvme_setup_cmd). Fix the error flow during command
submission and also fix the missing call in command completion.
Is it always called with nvme_complete_rq? Why not just put it there?
Yes, unless I am missing something we could call nvme_cleanup_cmd
at the beginning of nvme_complete_rq.
This will cause change is error flow in nvme_fc but I can check this.
Max, can you send one series for all the nvme_cleanup_cmd fixes and
cleanups and split that from the PI work? That might be a little
less confusing.
Yes I will. There is a connection between the patches but for now only
the nvme-pci supports T10 in the nvme subsystem, we can separate them.
There will be still a small gap in the error flow of the pci driver that
will call nvme_cleanup_cmd and do the t10 remap that he shouldn't (but
that's the behavior today)
