On 03/04/2015 12:56 AM, Chris Wilson wrote: > On Tue, Mar 03, 2015 at 04:27:47PM -0800, Jesse Barnes wrote: >> My "rant du jour" still is. mlock() is a good solution for some things, >> but for the simple task of testing kernel swap out code, just running >> that code is the most straightforward thing to do, rather than trying to >> contort into it from userspace. > > Once upon a time, there was a minor request for MADV_POPULATE and > MADV_INVALIDATE (basically wrappers around get_pages() and put_pages()) > which would give you the fuzzy feeling of testing those paths without > actually testing the shrinker. But no use cases. Yeah, that sounds cool, but otoh adding stable ABI just for debug is a little scary... Jesse _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
