I think Daves theory sounds good, but I'd like to get some evidence for it. We already have fairly good trace points in the writeback code, but for this they aren't quite enough yet. Curt posted some patches to pass a why argument down the writeback stack so that we known the high-level caller in all tracepoints. I'll put that into my tree to check where these calls come from. -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
