lkml.org 
[lkml]   [1996]   [May]   [5]   [last100]   RSS Feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
/
DateSat, 4 May 1996 00:33:24 +0100 (BST)
FromBryn Paul Arnold Jones <>
SubjectRe: Is clobber "memory" in include/asm-i386/system.h necessary?
On Thu, 2 May 1996, Linus Torvalds wrote:
> 
> Essentially, you can think of the memory clobbers as barriers, and think of
> "cli" and "sti" as
> 
> 	cli();
> 	barrier();
> 
> and
> 
> 	barrier();
> 	sti();
> 
> respectively.
> 

Hmm, the clobbers make gcc not optermize across the the cli()/sti()/..., 
but gcc can still optermize before, inside, and after (just not across) 
correct ?  So the real advantage is gcc can optermize (hmm, I think I'm 
spelling optermize wrong) throughout, and get rid of the 
silly/slow/unused/whatever, which wouldent happen with volitile (as well 
as suiteing you (Linus) ;).  Am I correct, or way off ?

> 		Linus
> 
Bryn
--
PGP key pass phrase forgotten,   \ Overload -- core meltdown sequence 
again :(                          |            initiated.
                                 / This space is intentionally left   
                                |  blank, apart from this text ;-)
                                 \____________________________________


\
 
 \ /
  Last update: 2005-03-22 13:37    [from the cache]
©2003-2008