releases/4.17.12/nvme-pci-fix-aer-reset-handling.patch

   1 From foo@baz Sat Jul 28 10:14:30 CEST 2018
   2 From: Keith Busch <keith.busch@intel.com>
   3 Date: Thu, 24 May 2018 16:16:04 -0600
   4 Subject: nvme-pci: Fix AER reset handling
   5
   6 From: Keith Busch <keith.busch@intel.com>
   7
   8 [ Upstream commit 72cd4cc28e234ed7189ee508ed65ab60c80a97c8 ]
   9
  10 The nvme timeout handling doesn't do anything if the pci channel is
  11 offline, which is the case when recovering from PCI error event, so it
  12 was a bad idea to sync the controller reset in this state. This patch
  13 flushes the reset work in the error_resume callback instead when the
  14 channel is back to online. This keeps AER handling serialized and
  15 can recover from timeouts.
  16
  17 Link: https://bugzilla.kernel.org/show_bug.cgi?id=199757
  18 Fixes: cc1d5e749a2e ("nvme/pci: Sync controller reset for AER slot_reset")
  19 Reported-by: Alex Gagniuc <mr.nuke.me@gmail.com>
  20 Tested-by: Alex Gagniuc <mr.nuke.me@gmail.com>
  21 Signed-off-by: Keith Busch <keith.busch@intel.com>
  22 Signed-off-by: Christoph Hellwig <hch@lst.de>
  23 Signed-off-by: Sasha Levin <alexander.levin@microsoft.com>
  24 Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
  25 ---
  26  drivers/nvme/host/pci.c |   14 +++++---------
  27  1 file changed, 5 insertions(+), 9 deletions(-)
  28
  29 --- a/drivers/nvme/host/pci.c
  30 +++ b/drivers/nvme/host/pci.c
  31 @@ -2679,19 +2679,15 @@ static pci_ers_result_t nvme_slot_reset(
  32
  33         dev_info(dev->ctrl.device, "restart after slot reset\n");
  34         pci_restore_state(pdev);
  35 -       nvme_reset_ctrl_sync(&dev->ctrl);
  36 -
  37 -       switch (dev->ctrl.state) {
  38 -       case NVME_CTRL_LIVE:
  39 -       case NVME_CTRL_ADMIN_ONLY:
  40 -               return PCI_ERS_RESULT_RECOVERED;
  41 -       default:
  42 -               return PCI_ERS_RESULT_DISCONNECT;
  43 -       }
  44 +       nvme_reset_ctrl(&dev->ctrl);
  45 +       return PCI_ERS_RESULT_RECOVERED;
  46  }
  47
  48  static void nvme_error_resume(struct pci_dev *pdev)
  49  {
  50 +       struct nvme_dev *dev = pci_get_drvdata(pdev);
  51 +
  52 +       flush_work(&dev->ctrl.reset_work);
  53         pci_cleanup_aer_uncorrect_error_status(pdev);
  54  }
  55