According to a comment in bdevsw_add() in xnu/bsd/kern/bsd_stubs.c, a fix was applied to address Radar bug 2842228. I don't know the history behind the fix, but it seems to have completely broken the function. The change appears to be trying to skip the first item in the bdevsw table by using an initial index value of 1. However the items are being examined via the devsw pointer, not by indexing bdevsw. This means that the index doesn't correspond to the item being examined. The result is a bit of a mess. When I call bdevsw_add() with an index of -1, the for() loop returns immediately because the first entry in bdevsw is empty. Since index is 1 though, devsw then gets set to bdevsw[1], which isn't empty. The second memcmp() fails and the function returns -1. If someone could kindly confirm for me that this is wrong, I'll post a bug to Radar. Thanks, Sam _______________________________________________ darwin-kernel mailing list | darwin-kernel@lists.apple.com Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-kernel Do not post admin requests to the list. They will be ignored.