]> git.proxmox.com Git - mirror_ubuntu-zesty-kernel.git/commit
fm10k: reinitialize queuing scheme after calling init_hw
authorJacob Keller <jacob.e.keller@intel.com>
Fri, 16 Oct 2015 17:56:59 +0000 (10:56 -0700)
committerTim Gardner <tim.gardner@canonical.com>
Mon, 29 Feb 2016 15:57:23 +0000 (08:57 -0700)
commitc0283501adbfd6113edf973eb3644802025fb526
tree056ea900211b79b9f999feb8afdab44cf8ae3df9
parent244cd8778846e052eeb2feccb9d064816e06cafa
fm10k: reinitialize queuing scheme after calling init_hw

BugLink: http://bugs.launchpad.net/bugs/1536475
The init_hw function may fail, and in the case of VFs, it might change
the number of maximum queues available. Thus, for every flow which
checks init_hw, we need to ensure that we clear the queue scheme before,
and initialize it after. The fm10k_io_slot_reset path will end up
triggering a reset so fm10k_reinit needs this change. The
fm10k_io_error_detected and fm10k_io_resume also need to properly clear
and reinitialize the queue scheme.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
(cherry picked from commit 875328e4bce696e85edcda3c4b0ec80fd525e3a3)
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
drivers/net/ethernet/intel/fm10k/fm10k_pci.c