[RFC PATCH] ipmi/of: Don't use unavailable interfaces

[Benjamin Herrenschmidt <benh@kernel.crashing.org>]

If an IPMI controller is used by the firmware and as such marked with
a reserved status, we shouldn't use it.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
---

Not tested, found by inspection as we are investigating how to plumb
IPMI on one of our future platforms. Our FW uses the device-tree for
its own internal use and tags stuff that it reserves for itself using
the "status" property. This is likely to be the case of our BT interface.

diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c
index 5d66568..4fc8931 100644
--- a/drivers/char/ipmi/ipmi_si_intf.c
+++ b/drivers/char/ipmi/ipmi_si_intf.c
@@ -2658,6 +2658,9 @@ static int ipmi_probe(struct platform_device *dev)
 	if (!match)
 		return -EINVAL;
 
+	if (!of_device_is_available(np))
+		return -EINVAL;
+
 	ret = of_address_to_resource(np, 0, &resource);
 	if (ret) {
 		dev_warn(&dev->dev, PFX "invalid address from OF\n");