On Fri, Nov 9, 2018 at 6:05 PM Wilco Dijkstra <wilco.dijks...@arm.com> wrote:
>
> Richard Biener wrote:
> >Marc Glisse wrote:
> >> Let's try with C = DBL_MIN and x = 婊BL_MAX. I don't believe it involves
> >> signed zeros or infinities, just an underflow. First, the result depends on
> >> the rounding mode. And in the default round-to-nearest, both divisions give
> >> 0, and thus compare the same with 0, but we replace that with a sign test
> >> on
> >> x, where they clearly give opposite answers.
> >>
> >> What would be the proper flag to test to check if we care about underflow?
> >
> > We have none specific so this makes it flag_unsafe_math_optimizations.
>
> Right I have added the unsafe math check again like in the previous version:
>
>
> The patch implements some of the optimizations discussed in
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71026.
>
> Simplify (C / x >= 0.0) into x >= 0.0 with -funsafe-math-optimizations
> (since C / x can underflow to zero if x is huge, it's not safe otherwise).
> If C is negative the comparison is reversed.
>
>
> Simplify (x * C1) > C2 into x > (C2 / C1) with -funsafe-math-optimizations.
> If C1 is negative the comparison is reversed.
>
> OK for commit?
OK.
Thanks,
Richard.
> ChangeLog
> 2018-11-09 Wilco Dijkstra <wdijk...@arm.com>
> Jackson Woodruff <jackson.woodr...@arm.com>
>
> gcc/
> PR 71026/tree-optimization
> * match.pd: Simplify floating point comparisons.
>
> gcc/testsuite/
> PR 71026/tree-optimization
> * gcc.dg/div-cmp-1.c: New test.
> * gcc.dg/div-cmp-2.c: New test.
>
> --
> diff --git a/gcc/match.pd b/gcc/match.pd
> index
> 94fbab841f5e36bd33fda849a686fd80886ee1ff..f6c76510f95be2485e5bacd07edab336705cbd25
> 100644
> --- a/gcc/match.pd
> +++ b/gcc/match.pd
> @@ -405,6 +405,21 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
> (rdiv @0 (negate @1))
> (rdiv (negate @0) @1))
>
> +(if (flag_unsafe_math_optimizations)
> + /* Simplify (C / x op 0.0) to x op 0.0 for C != 0, C != Inf/Nan.
> + Since C / x may underflow to zero, do this only for unsafe math. */
> + (for op (lt le gt ge)
> + neg_op (gt ge lt le)
> + (simplify
> + (op (rdiv REAL_CST@0 @1) real_zerop@2)
> + (if (!HONOR_SIGNED_ZEROS (@1) && !HONOR_INFINITIES (@1))
> + (switch
> + (if (real_less (&dconst0, TREE_REAL_CST_PTR (@0)))
> + (op @1 @2))
> + /* For C < 0, use the inverted operator. */
> + (if (real_less (TREE_REAL_CST_PTR (@0), &dconst0))
> + (neg_op @1 @2)))))))
> +
> /* Optimize (X & (-A)) / A where A is a power of 2, to X >> log2(A) */
> (for div (trunc_div ceil_div floor_div round_div exact_div)
> (simplify
> @@ -4049,6 +4064,22 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
> (rdiv @2 @1))
> (rdiv (op @0 @2) @1)))
>
> + (for cmp (lt le gt ge)
> + neg_cmp (gt ge lt le)
> + /* Simplify (x * C1) cmp C2 -> x cmp (C2 / C1), where C1 != 0. */
> + (simplify
> + (cmp (mult @0 REAL_CST@1) REAL_CST@2)
> + (with
> + { tree tem = const_binop (RDIV_EXPR, type, @2, @1); }
> + (if (tem
> + && !(REAL_VALUE_ISINF (TREE_REAL_CST (tem))
> + || (real_zerop (tem) && !real_zerop (@1))))
> + (switch
> + (if (real_less (&dconst0, TREE_REAL_CST_PTR (@1)))
> + (cmp @0 { tem; }))
> + (if (real_less (TREE_REAL_CST_PTR (@1), &dconst0))
> + (neg_cmp @0 { tem; })))))))
> +
> /* Simplify sqrt(x) * sqrt(y) -> sqrt(x*y). */
> (for root (SQRT CBRT)
> (simplify
> diff --git a/gcc/testsuite/gcc.dg/div-cmp-1.c
> b/gcc/testsuite/gcc.dg/div-cmp-1.c
> new file mode 100644
> index
> 0000000000000000000000000000000000000000..cd1a5cd3d6fee5a10e9859ca99b344fa3fdb7f5f
> --- /dev/null
> +++ b/gcc/testsuite/gcc.dg/div-cmp-1.c
> @@ -0,0 +1,29 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -funsafe-math-optimizations -fdump-tree-optimized-raw"
> } */
> +
> +int
> +cmp_mul_1 (float x)
> +{
> + return x * 3 <= 100;
> +}
> +
> +int
> +cmp_mul_2 (float x)
> +{
> + return x * -5 > 100;
> +}
> +
> +int
> +div_cmp_1 (float x, float y)
> +{
> + return x / 3 <= y;
> +}
> +
> +int
> +div_cmp_2 (float x, float y)
> +{
> + return x / 3 <= 1;
> +}
> +
> +/* { dg-final { scan-tree-dump-times "mult_expr" 1 "optimized" } } */
> +/* { dg-final { scan-tree-dump-not "rdiv_expr" "optimized" } } */
> diff --git a/gcc/testsuite/gcc.dg/div-cmp-2.c
> b/gcc/testsuite/gcc.dg/div-cmp-2.c
> new file mode 100644
> index
> 0000000000000000000000000000000000000000..f4ac42a196a804747d0b578e0aa2131671c8d3cf
> --- /dev/null
> +++ b/gcc/testsuite/gcc.dg/div-cmp-2.c
> @@ -0,0 +1,29 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -funsafe-math-optimizations -ffinite-math-only
> -fdump-tree-optimized-raw" } */
> +
> +int
> +cmp_1 (float x)
> +{
> + return 5 / x >= 0;
> +}
> +
> +int
> +cmp_2 (float x)
> +{
> + return 1 / x <= 0;
> +}
> +
> +int
> +cmp_3 (float x)
> +{
> + return -2 / x >= 0;
> +}
> +
> +int
> +cmp_4 (float x)
> +{
> + return -5 / x <= 0;
> +}
> +
> +/* { dg-final { scan-tree-dump-not "rdiv_expr" "optimized" } } */
> +
