On Tue, 23 Mar 2010, Mindaugas Kavaliauskas wrote:
Hi,
> >2010-03-23 15:19 UTC+0100 Przemyslaw Czerpak (druzus/at/priv.onet.pl)
> > * harbour/include/hbatomic.h
> > + added HB_SPINLOCK_TRY(l)
> > + added support for recursive spin locks
> I'm interested why sl parameter in hb_spinlock_try_r()
> (hbatomic.h#486) is used in brackets. It's a function, it's not the
> macro define. Is it just copy-paste typo, or has some meaning?
It's not a typo. This code is correct but the redundant parenthesis
in (sl) are results of workarounds for BCC. This compiler was not
able to inline this functions:
static _HB_INLINE_ int hb_spinlock_try_r( struct hb_spinlock_r * sl )
{
HB_SPINLOCK_T * l = &sl->lock;
if( *l != HB_SPINLOCK_INIT )
{
if( sl->thid == HB_THREAD_SELF() )
{
sl->count++;
return 1;
}
}
else if( HB_SPINLOCK_TRY( l ) )
{
sl->thid = HB_THREAD_SELF();
sl->count = 1;
return 1;
}
return 0;
}
and was generating warning about it. The same was with
hb_spinlock_acquire_r(). Finally I converted the second
one to macro and introduced additional variable 'r' to
pacify the warnings. Anyhow maybe I'll change it to macro
to in the future so I left parenthesis encapsulating 'sl'
parameter.
BTW what pragma allows to disable this BCC warnings?
I do not like that I have to introduce such hacks so
I'd prefer to simply disable warnings in BCC builds.
best regards,
Przemek
_______________________________________________
Harbour mailing list (attachment size limit: 40KB)
Harbour@harbour-project.org
http://lists.harbour-project.org/mailman/listinfo/harbour
