Re: An example of bugs for Hot Standby

Andres Freund <andres@anarazel.de>

From: Andres Freund <andres@anarazel.de>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-hackers@postgresql.org, Simon Riggs <simon@2ndquadrant.com>, Hiroyuki Yamada <yamada@kokolink.net>
Date: 2010-01-20T09:45:00Z
Lists: pgsql-hackers
On Wednesday 20 January 2010 06:30:28 Tom Lane wrote:
> Andres Freund <andres@anarazel.de> writes:
> > Is there any supported platform with sizeof(sig_atomic_t) <4 - I would
> > doubt so?
> 
> Er ... what?  I believe there are live platforms with sig_atomic_t = char.
> If we're assuming more that's a must-fix.
The reason I have asked is that the code is doing things like:
/*
 * Used by backends when they receive a request to check for buffer pin waits.
 */
int
GetStartupBufferPinWaitBufId(void)
{
	int bufid;

	/* use volatile pointer to prevent code rearrangement */
	volatile PROC_HDR *procglobal = ProcGlobal;

	SpinLockAcquire(ProcStructLock);

	bufid = procglobal->startupBufferPinWaitBufId;

	SpinLockRelease(ProcStructLock);

	return bufid;
}

or similar things with LWLockAcquire in a signal handler which strikes me as a 
not that good idea. As at least on x86 reading an integer is atomic the above 
spinlock is pointless. My cross arch experience is barely existing, so...

Andres