This is the mail archive of the gcc-patches@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: [patch] Lno branch merge part 12 -- induction variable optimizations


Zdenek Dvorak wrote:

Hello,



5) build_int_cst is rewritten to decide whether to sign extend or to
zero extend depending on the signedness of the type of the constant,
and to mask out bits outside of the precision, which should be what
we want most of the time.  Few places where this is incorrect should
use build_int_csts instead.






Like Joseph, I think (5) should be completely separated from this patch.

The patch to stor-layout.c is unncessary, and should not be included.




the patch to stor-layout is necessary for 5), since otherwise
build_int_cst is called before the bitsize of the type is set up.




Let me put it a different way: I think that you can find a way to avoid touching large parts of the rest of the compiler, and still do what you need to do. I'd rather not be considering whether or not the semantics of build_int_cst ought to change for all callers.



I will make the function with the chaned semantics local to ivopts for now, although later I would like to come back to the issue.

Perfect.

Thanks,

--
Mark Mitchell
CodeSourcery, LLC
(916) 791-8304
mark@codesourcery.com


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