Thread: Re: [COMMITTERS] pgsql: Forbid setval() during recovery.
heikki@postgresql.org (Heikki Linnakangas) writes: > Forbid setval() during recovery. This prevents the PANIC reported by > Erik Rijkers. Patch by Andres Freund. ISTM this is the wrong fix. The real bug is that setval() doesn't check XactReadOnly. Now XactReadOnly is only a "soft" read only mode, but I don't see an argument for either setval() or nextval() to be allowed when that's set. Any objections to changing that? regards, tom lane
On Saturday 20 February 2010 17:34:50 Tom Lane wrote: > heikki@postgresql.org (Heikki Linnakangas) writes: > > Forbid setval() during recovery. This prevents the PANIC reported by > > Erik Rijkers. Patch by Andres Freund. > > ISTM this is the wrong fix. The real bug is that setval() doesn't > check XactReadOnly. Now XactReadOnly is only a "soft" read only > mode, but I don't see an argument for either setval() or nextval() > to be allowed when that's set. > Any objections to changing that? Not from my side - sounds more sensible... Andres