[PATCH] Fix wrong assumption in SRA

Richard Guenther richard.guenther@gmail.com
Tue May 27 11:59:00 GMT 2008


On Tue, May 27, 2008 at 12:43 PM, Eric Botcazou <ebotcazou@adacore.com> wrote:
>> Should it not reject these instead?
>
> That's not necessary for Ada (this works fine in our 4.1 and 4.3 based
> compiler so far) and would disable SRA for the testcase I've attached.
> I think that you have to be careful with integral types because of sign
> or zero extension, but this doesn't matter for aggregate types as long as
> their fields themselves are correctly represented.

Ok - the patch is ok then.

Thanks,
Richard.



More information about the Gcc-patches mailing list