[PATCH 1/2] powerpc/eeh: Fix crash in eeh_add_device_early() on Cell

[PATCH 1/2] powerpc/eeh: Fix crash in eeh_add_device_early() on Cell

[Michael Ellerman]

The recent change to the EEH probing causes a crash on Cell because
eeh_ops is NULL.

Check if EEH is enabled and if not bail out.

Fixes: ff57b454ddb9 ("powerpc/eeh: Do probe on pci_dn")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
 arch/powerpc/kernel/eeh.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/eeh.c b/arch/powerpc/kernel/eeh.c
index 76253eb146be..a4c62eb0ee48 100644
--- a/arch/powerpc/kernel/eeh.c
+++ b/arch/powerpc/kernel/eeh.c
@@ -1053,7 +1053,7 @@ void eeh_add_device_early(struct pci_dn *pdn)
 	struct pci_controller *phb;
 	struct eeh_dev *edev = pdn_to_eeh_dev(pdn);
 
-	if (!edev)
+	if (!edev || !eeh_enabled())
 		return;
 
 	/* USB Bus children of PCI devices will not have BUID's */
-- 
2.1.0

[PATCH 2/2] powerpc/cell: Fix iommu breakage caused by controller_ops change

[Michael Ellerman]

The recent patch to convert cell to use pci_controller_ops had a small
bug which broke machines using an iommu.

The set of phb->controller_ops was added after the check for name !=
"pci", meaning pcix/pcie PHBs weren't getting their ops set correctly.

Fixes: 9c1368fc50e7 ("powerpc/cell: Move controller ops from ppc_md to controller_ops")
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
 arch/powerpc/platforms/cell/setup.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/platforms/cell/setup.c b/arch/powerpc/platforms/cell/setup.c
index d1be268b1e6e..36cff28d0293 100644
--- a/arch/powerpc/platforms/cell/setup.c
+++ b/arch/powerpc/platforms/cell/setup.c
@@ -127,13 +127,13 @@ static int cell_setup_phb(struct pci_controller *phb)
 	if (rc)
 		return rc;
 
+	phb->controller_ops = cell_pci_controller_ops;
+
 	np = phb->dn;
 	model = of_get_property(np, "model", NULL);
 	if (model == NULL || strcmp(np->name, "pci"))
 		return 0;
 
-	phb->controller_ops = cell_pci_controller_ops;
-
 	/* Setup workarounds for spider */
 	if (strcmp(model, "Spider"))
 		return 0;
-- 
2.1.0

Re: [PATCH 2/2] powerpc/cell: Fix iommu breakage caused by controller_ops change

[Daniel Axtens]

[-- Attachment #1: Type: text/plain, Size: 574 bytes --]

On Tue, 2015-04-14 at 16:49 +1000, Michael Ellerman wrote:
> The recent patch to convert cell to use pci_controller_ops had a small
> bug which broke machines using an iommu.
> 
Oops.

> The set of phb->controller_ops was added after the check for name !=
> "pci", meaning pcix/pcie PHBs weren't getting their ops set correctly.
> 
> Fixes: 9c1368fc50e7 ("powerpc/cell: Move controller ops from ppc_md to controller_ops")
> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>


Thanks for that.

Acked-by: Daniel Axtens <dja@axtens.net>

Regards,
Daniel

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 860 bytes --]