On Fri, May 25, 2012 at 01:25:29PM -0700, Kent Overstreet wrote: > This is for the new bio splitting code. When we split a bio, if the > split occured on a bvec boundry we reuse the bvec for the new bio. But > that means bio_free() can't free it, hence the explicit flag. I keep having the same complaints. The explanation isn't detailed enough. So, the necessity for this patch arises from the fact that the current bio_split doesn't use the usual bio allocation path to create split bio - it just supports fixed bvec allocation via bio_pair allocation which is allowable because of the severe restrictions the current bio_split() - but the new bio_split() wants to do it in general manner and thus wants the usual bvec allocation. Why do I (or anyone for that matter) have to go forward in the patch series to reconstruct the rationale? Why doesn't the patch description already explain this? Why isn't this still fixed when lack of proper patch description has already been pointed out multiple times? I'm gonna stop here on this series. * *Please* spend more effort on patch description and understanding and applying the reviews. If you don't like the opinions expressed in reviews, please argue. If reviews get ignored, why would anyone review your patches at all? * It would probably be better to split the series so that more experimental / constroversial stuff is in separate patch series. Thanks. -- tejun -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
