Hi Wenbo,
From: Wenbo Wang <[email protected]>[v3] Do request irq in nvme_init_queue() to handle request irq failures There is one problem with the original patch. Since init queue happens before request irq, online_queue might be left increased if request irq fails. This version merges request irq into nvme_init_queue() because: 1. It solves the online_queue counting issue if request irq fails. 2. nvme_init_queue() is always followed by request irq, this change simplifies code. This version also solves another possible race condition by moving adminq free irq before iounmap. [v2] Remove duplicate init code in nvme_alloc_queue() During reset process, the nvme_dev->bar (ioremapped) may change, so nvmeq->q_db shall be also updated by nvme_init_queue(). [v1] Currently nvmeq irq is enabled before queue init, so a spurious interrupt triggered nvme_process_cq may access nvmeq->q_db just before it is updated, this could cause kernel panic.
So the patch changes should not appear in the change-log itself, its meant for the reviewers. The change log should document the fix.
Signed-off-by: Wenbo Wang <[email protected]> Reviewed-by: Wenwei Tao <[email protected]> CC: [email protected] ---
Here (under the '---' separator) the patch vX changes should be documented in the form: Changes from v1: - A2 - B2 - C2 Changes from v0: - A1 - B1 - C1 This way git-am won't include it in the commit log message. Other then that the patch itself looks fine, Reviewed-by: Sagi Grimberg <[email protected]>
