On 12:41 Mon 27 Jul , Hal Rosenstock wrote: > > The intent of the patch is to cause a NULL ptr fault rather than using > memory not currently owned. The latter can have unpredictable results.
This could be nice for debugging, but the case itself is illegal and should not happen in a main stream (unless there are bugs which should be fixed). BTW LASH checks this pointer for NULL in several places, so this patch will effectively hide a potential bugs. Sasha _______________________________________________ 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
