cxl: Route eeh events to all slices for pci_channel_io_perm_failure state
authorVaibhav Jain <vaibhav@linux.vnet.ibm.com>
Thu, 23 Feb 2017 03:27:26 +0000 (08:57 +0530)
committerMichael Ellerman <mpe@ellerman.id.au>
Mon, 20 Mar 2017 09:19:36 +0000 (20:19 +1100)
commit07f5ab6002a4f0b633f3495157166f9f6180871b
tree06dabfd6c40e6e786269674611ce12d2cb13669c
parent97da3854c526d3a6ee05c849c96e48d21527606c
cxl: Route eeh events to all slices for pci_channel_io_perm_failure state

Fix a boundary condition where in some cases an eeh event with state ==
pci_channel_io_perm_failure wont be passed on to a driver attached to
the virtual PCI device associated with a slice. This will happen in case
the slice just before (n-1) doesn't have any vPHB bus associated with
it, that results in an early return from cxl_pci_error_detected()
callback.

With state == pci_channel_io_perm_failure, the adapter will be removed
irrespective of the return value of cxl_vphb_error_detected(). So we now
always return PCI_ERS_RESULT_DISCONNECTED for this case i.e even if
the AFU isn't using a vPHB (currently returns PCI_ERS_RESULT_NONE).

Fixes: e4f5fc001a6("cxl: Do not create vPHB if there are no AFU configuration records")
Signed-off-by: Vaibhav Jain <vaibhav@linux.vnet.ibm.com>
Reviewed-by: Matthew R. Ochs <mrochs@linux.vnet.ibm.com>
Reviewed-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Acked-by: Frederic Barrat <fbarrat@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
drivers/misc/cxl/pci.c