> Because if we're going to merge something upstream, we might as well > merge the real fix (that fixes both mthca and mlx4) rather than merging > a workaround and then leaving the bug for another year or however long > it takes to hit it again.
Oh and of course the fact that the patch requires driver code to look at internal implementation details of struct completion is a big hint that this is the wrong approach too. - R. _______________________________________________ general mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
