data-ref: Improve non-loop disambiguation [PR106019]

When dr_may_alias_p is called without a loop context, it tries
to use the tree-affine interface to calculate the difference
between the two addresses and use that difference to check whether
the gap between the accesses is known at compile time.  However, as the
example in the PR shows, this doesn't expand SSA_NAMEs and so can easily
be defeated by things like reassociation.

One fix would have been to use aff_combination_expand to expand the
SSA_NAMEs, but we'd then need some way of maintaining the associated
cache.  This patch instead reuses the innermost_loop_behavior fields
(which exist even when no loop context is provided).

It might still be useful to do the aff_combination_expand thing too,
if an example turns out to need it.

Tested on aarch64-linux-gnu and x86_64-linux-gnu.  OK to install?

Richard


gcc/
	PR tree-optimization/106019
	* tree-data-ref.cc (dr_may_alias_p): Try using the
	innermost_loop_behavior to disambiguate non-loop queries.

gcc/testsuite/
	PR tree-optimization/106019
	* gcc.dg/vect/bb-slp-pr106019.c: New test.
---
 gcc/testsuite/gcc.dg/vect/bb-slp-pr106019.c | 15 +++++++++++++++
 gcc/tree-data-ref.cc                        | 19 +++++++++++++++++++
 2 files changed, 34 insertions(+)
 create mode 100644 gcc/testsuite/gcc.dg/vect/bb-slp-pr106019.c

> Am 21.06.2022 um 17:16 schrieb Richard Sandiford via Gcc-patches <gcc-patches@gcc.gnu.org>:
> 
> When dr_may_alias_p is called without a loop context, it tries
> to use the tree-affine interface to calculate the difference
> between the two addresses and use that difference to check whether
> the gap between the accesses is known at compile time.  However, as the
> example in the PR shows, this doesn't expand SSA_NAMEs and so can easily
> be defeated by things like reassociation.
> 
> One fix would have been to use aff_combination_expand to expand the
> SSA_NAMEs, but we'd then need some way of maintaining the associated
> cache.  This patch instead reuses the innermost_loop_behavior fields
> (which exist even when no loop context is provided).
> 
> It might still be useful to do the aff_combination_expand thing too,
> if an example turns out to need it.
> 
> Tested on aarch64-linux-gnu and x86_64-linux-gnu.  OK to install?
> 

Ok.

Thanks,
Richard 

> Richard
> 
> 
> gcc/
>    PR tree-optimization/106019
>    * tree-data-ref.cc (dr_may_alias_p): Try using the
>    innermost_loop_behavior to disambiguate non-loop queries.
> 
> gcc/testsuite/
>    PR tree-optimization/106019
>    * gcc.dg/vect/bb-slp-pr106019.c: New test.
> ---
> gcc/testsuite/gcc.dg/vect/bb-slp-pr106019.c | 15 +++++++++++++++
> gcc/tree-data-ref.cc                        | 19 +++++++++++++++++++
> 2 files changed, 34 insertions(+)
> create mode 100644 gcc/testsuite/gcc.dg/vect/bb-slp-pr106019.c
> 
> diff --git a/gcc/testsuite/gcc.dg/vect/bb-slp-pr106019.c b/gcc/testsuite/gcc.dg/vect/bb-slp-pr106019.c
> new file mode 100644
> index 00000000000..218d7cca33d
> --- /dev/null
> +++ b/gcc/testsuite/gcc.dg/vect/bb-slp-pr106019.c
> @@ -0,0 +1,15 @@
> +/* { dg-do compile } */
> +
> +void f(double *p, long i)
> +{
> +    p[i+0] += 1;
> +    p[i+1] += 1;
> +}
> +void g(double *p, long i)
> +{
> +    double *q = p + i;
> +    q[0] += 1;
> +    q[1] += 1;
> +}
> +
> +/* { dg-final { scan-tree-dump-not "can't determine dependence" slp2 } } */
> diff --git a/gcc/tree-data-ref.cc b/gcc/tree-data-ref.cc
> index 8b7edf2124a..90242948c27 100644
> --- a/gcc/tree-data-ref.cc
> +++ b/gcc/tree-data-ref.cc
> @@ -2968,6 +2968,25 @@ dr_may_alias_p (const struct data_reference *a, const struct data_reference *b,
>      disambiguation.  */
>   if (!loop_nest)
>     {
> +      tree tree_size_a = TYPE_SIZE_UNIT (TREE_TYPE (DR_REF (a)));
> +      tree tree_size_b = TYPE_SIZE_UNIT (TREE_TYPE (DR_REF (b)));
> +
> +      if (DR_BASE_ADDRESS (a)
> +      && DR_BASE_ADDRESS (b)
> +      && operand_equal_p (DR_BASE_ADDRESS (a), DR_BASE_ADDRESS (b))
> +      && operand_equal_p (DR_OFFSET (a), DR_OFFSET (b))
> +      && poly_int_tree_p (tree_size_a)
> +      && poly_int_tree_p (tree_size_b)
> +      && !ranges_maybe_overlap_p (wi::to_widest (DR_INIT (a)),
> +                      wi::to_widest (tree_size_a),
> +                      wi::to_widest (DR_INIT (b)),
> +                      wi::to_widest (tree_size_b)))
> +    {
> +      gcc_assert (integer_zerop (DR_STEP (a))
> +              && integer_zerop (DR_STEP (b)));
> +      return false;
> +    }
> +
>       aff_tree off1, off2;
>       poly_widest_int size1, size2;
>       get_inner_reference_aff (DR_REF (a), &off1, &size1);
> -- 
> 2.25.1
>