Partial page discard requests are ignored and the documentation on why this is correct behaviour sucks. A readahead patch looked like a "regression" to a random IO storage benchmark because posix_fadvise() was used incorrectly to force IO requests to go to disk. In reality, the benchmark sucked but it was non-obvious why. Patch 1 updates the kernel comment in case someone "fixes" either readahead or fadvise for inappropriate reasons. Patch 2 updates the relevant man page on the rough off chance that application developers do not read kernel source comments. -- 2.1.2 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>