[PATCH 18/19] [POWERPC] Fixup skipping of PowerMac PCI<->PCI bridge "closed" resources
Benjamin Herrenschmidt
benh at kernel.crashing.org
Thu Dec 20 14:55:04 EST 2007
Apple firmware has a strange way to "close" bridge resources by setting
them to some bogus values that overlap RAM (strangely, I haven't seen it
conflicting with DMA so far...). This explicitely closes them to avoid
problems. Previously, they would be closed as a consequence of failing
to be allocated, but this makes it more explicit, and thus the log
message is more explicit too.
Signed-off-by: Benjamin Herrenschmidt <benh at kernel.crashing.org>
---
arch/powerpc/kernel/pci-common.c | 22 ++++++++++++++++++++++
1 file changed, 22 insertions(+)
--- linux-merge.orig/arch/powerpc/kernel/pci-common.c 2007-12-19 15:47:23.000000000 +1100
+++ linux-merge/arch/powerpc/kernel/pci-common.c 2007-12-19 15:58:59.000000000 +1100
@@ -776,6 +776,7 @@ DECLARE_PCI_FIXUP_HEADER(PCI_ANY_ID, PCI
static void __devinit __pcibios_fixup_bus(struct pci_bus *bus)
{
+ struct pci_controller *hose = pci_bus_to_host(bus);
struct pci_dev *dev = bus->self;
pr_debug("PCI: Fixup bus %d (%s)\n", bus->number, dev ? pci_name(dev) : "PHB");
@@ -793,6 +794,27 @@ static void __devinit __pcibios_fixup_bu
if (!res->flags || bus->self->transparent)
continue;
+ /* On PowerMac, Apple leaves bridge windows open over
+ * an inaccessible region of memory space (0...fffff)
+ * which is somewhat bogus, but that's what they think
+ * means disabled...
+ *
+ * We clear those to force them to be reallocated later
+ *
+ * We detect such regions by the fact that the base is
+ * equal to the pci_mem_offset of the host bridge and
+ * their size is smaller than 1M.
+ */
+ if (res->start == hose->pci_mem_offset &&
+ res->end < 0x100000) {
+ printk(KERN_INFO
+ "PCI: Closing bogus Apple Firmware"
+ " region %d on bus 0x%02x\n",
+ i, bus->number);
+ res->flags = 0;
+ continue;
+ }
+
pr_debug("PCI:%s Bus rsrc %d %016llx-%016llx [%x] fixup...\n",
pci_name(dev), i,
(unsigned long long)res->start,\
More information about the Linuxppc-dev
mailing list