[graphite] Fix PR39335: problem in canonicalize_loop_ivs

Sebastian Pop sebpop@gmail.com
Tue Mar 3 03:44:00 GMT 2009


On Mon, Mar 2, 2009 at 18:48, Zdenek Dvorak <rakdver@kam.mff.cuni.cz> wrote:
>>  tree
>> -canonicalize_loop_ivs (struct loop *loop, htab_t reduction_list, tree nit)
>> +canonicalize_loop_ivs (struct loop *loop, htab_t reduction_list, tree *nit)
>
> Update the function comment, explaining that *NIT is set to the value
> with that the the control variable is compared in the exit test (or
> perhaps the number of iterations in the type of the control variable,
> whichever describes your intended semantics better).

I added the following text to the comment of that function:

When the IV type precision has to be larger than *NIT type precision,
*NIT is converted to the larger type, the conversion code is inserted
before the loop, and *NIT is updated to the new definition.

Sebastian



More information about the Gcc-patches mailing list