Re: Threadsafe singletons

From:

"David Barrett-Lennard" <davidbl@iinet.net.au>

Newsgroups:

comp.lang.c++.moderated,comp.programming.threads

Date:

2 Aug 2006 23:19:10 -0400

Message-ID:

<1154573901.064377.170560@75g2000cwc.googlegroups.com>

Gene Bushuyev wrote:

"David Barrett-Lennard" <davidbl@iinet.net.au> wrote in message
news:1154525824.378002.232050@m73g2000cwd.googlegroups.com...
[...]

That suggests it is fairly easy to change the code in my OP to ensure
s_init is initialised. Eg

static volatile int s_dummy;
static struct InitMySingleton
{
InitMySingleton() { s_dummy = 0; MySingleton::GetInstance(); }
} s_init;

That won't fool a good compiler :-) Now you have s_dummy that is not used,
so
the program doesn't have an observable side effect in InitMySingleton.
Therefore, following as-if rule compiler can still throw the whole thing
away.

I was going by your brief summary from the C++ standard. Anyway,
according to James the issue seems rather academic. Do you agree?

I
suggest you play with VC8 and look at the assembly, it used to be pretty
smart
removing the dead code.

Used to be? :)

Cheers,
David Barrett-Lennard

[ See http://www.gotw.ca/resources/clcm.htm for info about ]
[ comp.lang.c++.moderated. First time posters: Do this! ]