liveness analysis with stack slots

[Maurizio Vitale]

On Jul 6, 2007, at 1:56 PM, Andrew Haley wrote:

> Maurizio Vitale writes:
>> Andrew Haley <aph-gcc@littlepinkcloud.COM> writes:
>>
>>> Maurizio Vitale writes:
>>>>
>>>>
>>>> Can somebody shed some light on this and hint at why GCC (and the
>>>> Intel compiler) decide that they cannot remove that code?
>>>
>>> It would be easy for us to do so if you provided a test case.   
>>> Without
>>> one it's impossible.  It doesn't look to me like the expression is
>>> totally dead, but without the source I can't tell.
>>
>> The entire source is large and I don't think it would help much other
>> than allowing people to run experiments on their own.
>
> Well, yes.  Those of us reading might be able to figure out for
> ourselves what is happening!
>
> I don't want the entire source, just a snippet that demonstrates the
> problem.  Is it impossible to do that?
>

Impossible not, but rather problematic because it is all metatemplate  
programming.
I'm working on reducing it, but would still need boost::mpl and  
boost::proto.
How much source would be too much for this list?

In the mean time I'm pursuing a path where code that GCC doesn't  
understand as
dead is not generated in the first place, thanks to template magic.
This would cure the symptom, but not explain GCC's reasoning,

Best regards,

	Maurizio
>>