Folks - I'm wondering what this list thinks of making a change to the mmc driver to make use the the High Priority Interrupt (HPI) feature that was added in the eMMC 4.41 spec to give priority to reads over writes. Part of this could be writing a custom i/o scheduler that just resorts the requests on the queue so that reads are always favored. (And obviously taking into account where a read is dependent on a previous write.) But the secondary ideas is to use HPI to interrupt a write in progress as long as the incoming read isn't depending on the write. Then the write would have to finished later, I guess. So my question for this list is, is this even feasible ?? Only the i/o scheduler knows if of incoming reads. The driver itself just knows about the current request. Does the i/o scheduler know what request is in progress ?? (CFQ scheduler for example.)
Charles Johnson Ultra-Mobility Group UMSE Software Architecture Intel Corporation [email protected] -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
