Jon, please would you put together a more-detailed patch header before I send this upstream? I have already seen this analysis, but please include it with the patch so other people can review it too. It needs to mention: 1) all the code reordering in the patch 2) the fact clear_region() is no longer protected by the lock and why these changes are all OK. Alasdair -- agk@xxxxxxxxxx -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
