On Tue, Mar 31, 2009 at 06:42:40PM -0700, arvind vasudev wrote: > What I am working on is a closed system, where a custom PCIe endpoint > needs to send a message to the driver, which has some data. This message > come in asynchronously. Since the data contained in this message is small > enough that it can be encapsulated in a TLP, the choice is to see if we > can do a send this message on directly to the CPU. I'd suggest sending two messages -- a memory write and then an MSI interrupt. > I guess the issue is that if the message comes into the Root Complex, > what does it do with it? Does it generate an ACPI general event, like it > does for hotplug, or does it do something else like generate an interrupt > to the CPU using what is configured in the device's configuration > space. If like you mentioned, the implementation is not defined in the > spec, so that would mean that if someone is using this feature, then they > will need to handle one of the two scenarios previously mentioned. The > question really was to see if someone is using it. I don't know of anyone using it. I don't think it's supposed to be used between a device and a root complex; more between two devices from the same vendor. -- Matthew Wilcox Intel Open Source Technology Centre "Bill, look, we understand that you're interested in selling us this operating system, but compare it to ours. We can't possibly take such a retrograde step." -- 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