[prev in list] [next in list] [prev in thread] [next in thread] 

List:       pcc-list
Subject:    Re: powerpc: simmod() can produce incorrect code
From:       Stefan Kempf <stefan () nefkom ! net>
Date:       2008-06-07 8:41:24
Message-ID: 20080607084124.GA19357 () murdock ! my-fqdn ! de
[Download RAW message or body]

Gregory McGarry wrote:
> Would this be okay?
>
> Index: local.c
> ===================================================================
> RCS file: /cvsroot/pcc/arch/powerpc/local.c,v
> retrieving revision 1.11
> diff -u -r1.11 local.c
> --- local.c     11 May 2008 16:55:03 -0000      1.11
> +++ local.c     7 Jun 2008 00:48:52 -0000
> @@ -1267,6 +1267,9 @@
>
>         assert(p->n_op == MOD);
>
> +       if (!ISUNSIGNED(p->n_type))
> +               return;
> +
>  #define ISPOW2(n) ((n) && (((n)&((n)-1)) == 0))
>
>         // if the right is a constant power of two, then replace with AND

Looks right.
[prev in list] [next in list] [prev in thread] [next in thread]