> because it is locked down by BIOS to chipset, readback should be 0xfed98004. > > and pci_size will return 32k for 0xfed98000. A device with a read-only BAR doesn't conform to the PCI spec. We can't determine how much space the device consumes. It's just an accident that BIOS put it at an address that we happen to interpret as 32K. We have no idea if the device consumes 4K, 8K, 16K, or 32K. > so try get one update bios that could return resource for root bus. > > or you could try to live with booting with pci=nocrs before your get new BIOS. A BIOS update is not a useful answer. The point is that we need better reporting of the situation so we don't have to spend all this time debugging this issue again. We need to figure out how to identify this as a non spec-compliant device and continue as best we can. We probably need to report it to the user because if there's a device of unknown size consuming address space, we're liable to cause a conflict by placing another device on top of it. -- To unsubscribe from this list: send the line "unsubscribe linux-pci" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
