[Gluster-devel] spurious failure with sparse-file-heal.t test

[Krishnan Parthasarathi]

> Seems like we are already inside pool->lock while doing statedump which
> does writes to files, so may be I shouldn't think too much :-/. I will
> take a look at your patch once.

Ah, now it makes sense. I thought you knew about pool->lock already.
I was wondering why you were even trying to avoid locks when these
data structures are designed to use explict mutexes to synchronize across
threads that access them concurrently.