This is the mail archive of the gcc-help@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: how do I promise const to the optimizer?


On Wednesday 25 April 2007 17:17, Ian Lance Taylor wrote:
> Marco Manfredini <mldb@gmx.net> writes:

> Unfortunately a const pointer in C/C++ doesn't mean "the value this
> points to can not change."  It only means "the value this points to
> can not be changed via this pointer."  So there is no way to express
> what you want in the standard language.  Nor does gcc provide an
> extension along these lines.

I hoped there was an __attribute__((const)) for function parameters somewhere 
in the hide.

>
> gcc does provide the extension of annotating fun with __attribute__
> ((pure)).  If fun is indeed a pure function, then gcc might be perform
> the optimization you want.  I haven't tried.  Of course, if fun is not
> pure, that won't help.

pure doesn't help, yes. It isn't even available for my x86_64 target. What I 
thought of was some clever hack which tricks the compiler into thinking that 
a completely new pointer had appeared. I've tried: 

const val* __attribute__((malloc)) summon_pointer(const val *p) { return p; }

but with no luck. 

If the task is to create a new pointer for which the compiler can't see the 
relation to the original, then __asm__ abuse would be my next idea, but I 
don't have a clue how to twiddle the constraints for this case.

Marco


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]