Re: [HACKERS] Re: [PATCHES] patches for 6.2.1p6 - Mailing list pgsql-hackers

From dg@illustra.com (David Gould)
Subject Re: [HACKERS] Re: [PATCHES] patches for 6.2.1p6
Date
Msg-id 9804011956.AA08443@hawk.illustra.com
Whole thread Raw
In response to Re: [HACKERS] Re: [PATCHES] patches for 6.2.1p6  (Tom Ivar Helbekkmo <tih@Hamartun.Priv.NO>)
Responses Re: [HACKERS] Re: [PATCHES] patches for 6.2.1p6
List pgsql-hackers
> David Gould wrote:
>
> >Seriously, if you want to, please create a function to emulate the following:
> >
> > /*
> >  * tas(lock)
> >  *
> >  * Access to platform specific test_and_set functionality. Given pointer to
> >  * lock attempts to acquire the lock atomically.
> >  *
> >  * Returns 0 for success, nonzero for failure.
> >  */
> > typedef slock_t unsigned char;        /* or whatever works on the platform */
> >
> > int tas(slock_t *lock)
> > {
> >     slock_t    tmp;
> >
> >     /* atomic, interlocked */
> >     tmp = *lock;
> >     *lock = -1;             /* any nonzero will do here */
> >
> >     return (tmp != 0);
> > }
> >
> > Given this, I can fold the VAX right into the grand scheme, just like a
> > normal computer (;-)).
>
> Hmpf!  The true worth of a computer is a function of its weight!  :-)
>
> Sorry this took a while, but anyway, this should do it for the VAX (in
> fact, it's more or less the version of the code that I figured I'd use
> until Bruce asked me to bum it down maximally for performance, only
> now with the return values from tas() swapped).  I include the macros

What do you mean "now with the return values from tas() swapped"? I think
your code looks ok, but just want to be sure we are following the same
grand plan...

> that would fit the current (6.3) locking scheme:
>
> typedef unsigned char slock_t;
>
> int tas(slock_t *lock) {
>     register ret;
>
> asm("    movl $1, r0
>     bbssi $0,(%1),1f
>     clrl r0
> 1:    movl r0,%0"
>     : "=r"(ret)    /* return value, in register */
>     : "r"(lock)    /* argument, 'lock pointer', in register */
>     : "r0");    /* inline code uses this register */
>
>     return ret;
> }
>
> #define    S_LOCK(addr)        do { while (tas(addr)) ; } while (0)
> #define    S_UNLOCK(addr)        (*(addr) = 0)
> #define    S_INIT_LOCK(addr)    (*(addr) = 0)
>
> -tih
> --

Thanks, this is just what I was looking for. I will fold it in to my changes.
I have gotten a little snowed under with other tasks, but I expect to finalize
my patch next week and will post it.

-dg

David Gould            dg@illustra.com           510.628.3783 or 510.305.9468
Informix Software  (No, really)         300 Lakeside Drive  Oakland, CA 94612
 - Linux. Not because it is free. Because it is better.


pgsql-hackers by date:

Previous
From: "Thomas G. Lockhart"
Date:
Subject: Re: [HACKERS] Re: [PORTS] Port Bug Report: int2 negative numbers not parsed correctly
Next
From: Paul Raines
Date:
Subject: Re: Let's talk up 6.3