On Wed, Oct 10, 2012 at 8:39 AM, Jakub Jelinek <ja...@redhat.com> wrote:
> Hi!
>
> On 32-bit HWI warning there is a warning about signed (INTVAL) vs. unsigned
> ({,op_}precision) comparison (on 64-bit HWI that doesn't trigger, as
> INTVAL is 64-bit signed long, while precision is 32-bit unsigned).
> This patch fixes the warning and makes sure we don't try to optimize
> shifts by negative count.
>
> Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
Ok.
Thanks,
Richard.
> 2012-10-10 Jakub Jelinek <ja...@redhat.com>
>
> PR middle-end/54862
> * simplify-rtx.c (simplify_truncation): Compare UINTVAL instead of
> INTVAL of second argument with precision resp. op_precision.
>
> --- gcc/simplify-rtx.c.jj 2012-10-08 21:37:33.000000000 +0200
> +++ gcc/simplify-rtx.c 2012-10-09 13:48:24.269969321 +0200
> @@ -668,7 +668,7 @@ simplify_truncation (enum machine_mode m
> && CONST_INT_P (XEXP (op, 1))
> && GET_CODE (XEXP (op, 0)) == SIGN_EXTEND
> && GET_MODE (XEXP (XEXP (op, 0), 0)) == mode
> - && INTVAL (XEXP (op, 1)) < precision)
> + && UINTVAL (XEXP (op, 1)) < precision)
> return simplify_gen_binary (ASHIFTRT, mode,
> XEXP (XEXP (op, 0), 0), XEXP (op, 1));
>
> @@ -680,7 +680,7 @@ simplify_truncation (enum machine_mode m
> && CONST_INT_P (XEXP (op, 1))
> && GET_CODE (XEXP (op, 0)) == ZERO_EXTEND
> && GET_MODE (XEXP (XEXP (op, 0), 0)) == mode
> - && INTVAL (XEXP (op, 1)) < precision)
> + && UINTVAL (XEXP (op, 1)) < precision)
> return simplify_gen_binary (LSHIFTRT, mode,
> XEXP (XEXP (op, 0), 0), XEXP (op, 1));
>
> @@ -692,7 +692,7 @@ simplify_truncation (enum machine_mode m
> && (GET_CODE (XEXP (op, 0)) == ZERO_EXTEND
> || GET_CODE (XEXP (op, 0)) == SIGN_EXTEND)
> && GET_MODE (XEXP (XEXP (op, 0), 0)) == mode
> - && INTVAL (XEXP (op, 1)) < precision)
> + && UINTVAL (XEXP (op, 1)) < precision)
> return simplify_gen_binary (ASHIFT, mode,
> XEXP (XEXP (op, 0), 0), XEXP (op, 1));
>
> @@ -705,8 +705,7 @@ simplify_truncation (enum machine_mode m
> && 2 * precision <= op_precision
> && CONST_INT_P (XEXP (op, 1))
> && (INTVAL (XEXP (op, 1)) & (precision - 1)) == 0
> - && INTVAL (XEXP (op, 1)) >= 0
> - && INTVAL (XEXP (op, 1)) < op_precision)
> + && UINTVAL (XEXP (op, 1)) < op_precision)
> {
> int byte = subreg_lowpart_offset (mode, op_mode);
> int shifted_bytes = INTVAL (XEXP (op, 1)) / BITS_PER_UNIT;
>
> Jakub
