While doing EEH recovery, the PCI devices of the problematic PE
should be removed and then added to the system again. During the
so-called hotplug event, the PCI devices of the problematic PE
will be probed through early/late phase. We would delay EEH probe
on late point for PowerNV platform since the PCI device isn't
available in early phase.
Signed-off-by: Gavin Shan <sha...@linux.vnet.ibm.com>
---
arch/powerpc/platforms/pseries/eeh.c | 16 +++++++++++++++-
1 files changed, 15 insertions(+), 1 deletions(-)
diff --git a/arch/powerpc/platforms/pseries/eeh.c
b/arch/powerpc/platforms/pseries/eeh.c
index ffe34c4..c65e14c 100644
--- a/arch/powerpc/platforms/pseries/eeh.c
+++ b/arch/powerpc/platforms/pseries/eeh.c
@@ -758,6 +758,14 @@ static void eeh_add_device_early(struct device_node *dn)
{
struct pci_controller *phb;
+ /*
+ * If we're doing EEH probe based on PCI device, we
+ * would delay the probe until late stage because
+ * the PCI device isn't available this moment.
+ */
+ if (!eeh_probe_mode_devtree())
+ return;
+
if (!of_node_to_eeh_dev(dn))
return;
phb = of_node_to_eeh_dev(dn)->phb;
@@ -766,7 +774,6 @@ static void eeh_add_device_early(struct device_node *dn)
if (NULL == phb || 0 == phb->buid)
return;
- /* FIXME: hotplug support on POWERNV */
eeh_ops->of_probe(dn, NULL);
}
@@ -817,6 +824,13 @@ static void eeh_add_device_late(struct pci_dev *dev)
edev->pdev = dev;
dev->dev.archdata.edev = edev;
+ /*
+ * We have to do the EEH probe here because the PCI device
+ * hasn't been created yet in the early stage.
+ */
+ if (eeh_probe_mode_dev())
+ eeh_ops->dev_probe(dev, NULL);
+
eeh_addr_cache_insert_dev(dev);
}
--
1.7.5.4
_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev
