[x86] Correct ix86_rtx_cost for multi-word multiplication.
Commit Message
This is the i386 backend specific piece of my revised patch for
PR middle-end/98865, where Richard Biener has suggested that I perform
the desired transformation during RTL expansion where the backend can
control whether it is profitable to convert a multiplication into a
bit-wise AND and a negation. This works well for x86_64, but alas
exposes a latent bug with -m32, where a DImode multiplication incorrectly
appears to be cheaper than negdi2+anddi3(!?). The fix to ix86_rtx_costs
is to report that a DImode (multi-word) multiplication actually requires
three SImode multiplications and two SImode additions. This also corrects
the cost of TImode multiplication on TARGET_64BIT.
This patch has been tested on x86_64-pc-linux-gnu with make bootstrap
and make -k check, both with and without --target_board=unix{-m32}, with
no new failures. This change avoids the need for a !ia32 target selector
for the upcoming new test case gcc.target/i386/pr98865.c.
Ok for mainline?
2022-05-17 Roger Sayle <roger@nextmovesoftware.com>
gcc/ChangeLog
* config/i386/i386.cc (ix86_rtx_costs) [MULT]: When mode size
is wider than word_mode, a multiplication costs three word_mode
multiplications and two word_mode additions.
Thanks in advance,
Roger
--
Comments
On Tue, May 17, 2022 at 6:44 PM Roger Sayle <roger@nextmovesoftware.com> wrote:
>
>
> This is the i386 backend specific piece of my revised patch for
> PR middle-end/98865, where Richard Biener has suggested that I perform
> the desired transformation during RTL expansion where the backend can
> control whether it is profitable to convert a multiplication into a
> bit-wise AND and a negation. This works well for x86_64, but alas
> exposes a latent bug with -m32, where a DImode multiplication incorrectly
> appears to be cheaper than negdi2+anddi3(!?). The fix to ix86_rtx_costs
> is to report that a DImode (multi-word) multiplication actually requires
> three SImode multiplications and two SImode additions. This also corrects
> the cost of TImode multiplication on TARGET_64BIT.
>
> This patch has been tested on x86_64-pc-linux-gnu with make bootstrap
> and make -k check, both with and without --target_board=unix{-m32}, with
> no new failures. This change avoids the need for a !ia32 target selector
> for the upcoming new test case gcc.target/i386/pr98865.c.
> Ok for mainline?
>
>
> 2022-05-17 Roger Sayle <roger@nextmovesoftware.com>
>
> gcc/ChangeLog
> * config/i386/i386.cc (ix86_rtx_costs) [MULT]: When mode size
> is wider than word_mode, a multiplication costs three word_mode
> multiplications and two word_mode additions.
LGTM.
Thanks,
Uros.
>
> Thanks in advance,
> Roger
> --
>
@@ -20634,7 +20634,17 @@ ix86_rtx_costs (rtx x, machine_mode mode, int outer_code_i, int opno,
op0 = XEXP (op0, 0), mode = GET_MODE (op0);
}
- *total = (cost->mult_init[MODE_INDEX (mode)]
+ int mult_init;
+ // Double word multiplication requires 3 mults and 2 adds.
+ if (GET_MODE_SIZE (mode) > UNITS_PER_WORD)
+ {
+ mult_init = 3 * cost->mult_init[MODE_INDEX (word_mode)]
+ + 2 * cost->add;
+ nbits *= 3;
+ }
+ else mult_init = cost->mult_init[MODE_INDEX (mode)];
+
+ *total = (mult_init
+ nbits * cost->mult_bit
+ rtx_cost (op0, mode, outer_code, opno, speed)
+ rtx_cost (op1, mode, outer_code, opno, speed));