> Eric, > > how about change default min_size to 0 ? For io_size I can see that as io space access are pretty much legacy at this point. For mem_size I haven't seen a device that doesn't need something and 2M is small. > #define DEFAULT_HOTPLUG_IO_SIZE 0 > #define DEFAULT_HOTPLUG_MEM_SIZE 0 > /* pci=hpmemsize=nnM,hpiosize=nn can override this */ > unsigned long pci_hotplug_io_size = DEFAULT_HOTPLUG_IO_SIZE; > unsigned long pci_hotplug_mem_size = DEFAULT_HOTPLUG_MEM_SIZE; > > and later to use dmi or other quirks way increase those value. The immediate problem is that after the simulated hotplug we have shifted the resource assignments around on the qlogic adapter and the adapter no longer works. As for the root cause I think you hit the nail on the head with your first suggestion that we need in kernel understanding of the root bridge. If we are assigning resources that don't actually work something is messed up in resource routing from the cpu to the device. Eric -- 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
