Re: parallel translation

["Jeffrey R. Carter" <spam.jrcarter.not@acm.nospam.org>]

Samuel Tardieu wrote:
>>>>>> "Jeffrey" == Jeffrey R Carter <spam.jrcarter.not@acm.nospam.org> writes:
> 
> Jeffrey> If P1 depends on the body of P2 (perhaps by calling an
> Jeffrey> inlined subprogram with -gnatn), then you have a problem
> Jeffrey> similar to what you described. I'm not sure how GNAT handles
> Jeffrey> this, but perhaps the task compiling P1 can wait until
> Jeffrey> appropriate versions of p2.ali and p2.o are created.
> 
> As you wrote, GNAT is source based. If P1 depends on the body of P2,
> it will use its source file (p2.adb in this case), and this is totally
> independant of the compilation of P2 which may well happen in
> parallel or after the compilation of P1.
> 
> Any unit can be compiled without depending on any other unit
> compilation. Moreover, the compiler never reads any ali file. gnatmake
> does (to determine what needs to be compiled), and gnatbind does as
> well (to determine the elaboration order).

Thanks for the clarification.

-- 
Jeff Carter
"Go and boil your bottoms."
Monty Python & the Holy Grail
01