[dm-devel] Re: panic in dm_pg_init_complete+0x10

[Alasdair G Kergon]

On Tue, Mar 08, 2005 at 04:23:59PM -0500, goggin, edward wrote:
> Furthermore, I strongly suspect the panic is related to having
> multiple pg_init requests outstanding for the same multipath.

It was not the intention to allow concurrent pg_init requests
of any sort - but they are indeed not locked out correctly.

> (1) It is possible that an outstanding pg_init request and a new one

I don't see a need to allow that.  Only one priority group is
in use at once.  Even switch_group has to wait without being able
to cancel already-submitted pg_init I/O.
Not mea

Alasdair
-- 
agk at redhat.com