On 04/12/2012 01:21 PM, Rodrigo Severo wrote: > Jeff, I believe that Alex is exactly saying that "stat /a" is failing with an > input/output error. In such a case how could a "stat /a" be suficient? And not > being suficient, what would be the fix? The fact that the stat() is returning an error means that we're getting *into* stat(), and that's the important thing because that's where self-heal would occur* if it were working in this case. The problem is that it doesn't seem to be, and I'm still looking into that. * Actually it's in lookup, which precedes stat, but that's an implementation detail which shouldn't matter to anyone not actually hacking the code.
