Re: An example of bugs for Hot Standby - Mailing list pgsql-hackers

From Andres Freund
Subject Re: An example of bugs for Hot Standby
Date
Msg-id 201001201104.35671.andres@anarazel.de
Whole thread Raw
In response to Re: An example of bugs for Hot Standby  (Simon Riggs <simon@2ndQuadrant.com>)
Responses Re: An example of bugs for Hot Standby
List pgsql-hackers
On Wednesday 20 January 2010 10:52:24 Simon Riggs wrote:
> On Wed, 2010-01-20 at 10:45 +0100, Andres Freund wrote:
> > LWLockAcquire
> 
> I'm using spinlocks, not lwlocks.
CancelDBBackends which is used in SendRecoveryConflictWithBufferPin which in 
turn used by CheckStandbyTimeout triggered by SIGALRM acquires the lwlock.

Now that case is a bit less dangerous because you would have to interrupt 
yourself to trigger a deadlock there because the code sleeps soon after 
setting up the handler.
If ever two SIGALRM occur consecutive there is a problem. 

Andres


pgsql-hackers by date:

Previous
From: "Jehan-Guillaume (ioguix) de Rorthais"
Date:
Subject: Re: Bug ? different behaviour between 8.3 and 8.4 won IS NULL with sub arrays of nulls
Next
From: Pavel Stehule
Date:
Subject: Re: review: More frame options in window functions