Restrict the two sources of vect_recog_cond_expr_convert_pattern to be of the same type when convert is extension.
Commit Message
> > +(match (cond_expr_convert_p @0 @2 @3 @6)
> > + (cond (simple_comparison@6 @0 @1) (convert@4 @2) (convert@5 @3))
> > + (if (types_match (TREE_TYPE (@2), TREE_TYPE (@3))
>
> But in principle @2 or @3 could safely differ in sign, you'd then need to ensure
> to insert sign conversions to @2/@3 to the signedness of @4/@5.
>
It turns out differ in sign is not suitable for extension(but ok for truncation),
because it's zero_extend vs sign_extend.
The patch add types_match check when convert is extension.
Bootstrapped and regtested on x86_64-pc-linux-gnu{-m32,}.
And native Bootstrapped and regtested on CLX.
Ok for trunk?
gcc/ChangeLog:
PR tree-optimization/104551
PR tree-optimization/103771
* match.pd (cond_expr_convert_p): Add types_match check when
convert is extension.
* tree-vect-patterns.cc
(gimple_cond_expr_convert_p): Adjust comments.
(vect_recog_cond_expr_convert_pattern): Ditto.
gcc/testsuite/ChangeLog:
* gcc.target/i386/pr104551.c: New test.
---
gcc/match.pd | 8 +++++---
gcc/testsuite/gcc.target/i386/pr104551.c | 24 ++++++++++++++++++++++++
gcc/tree-vect-patterns.cc | 6 ++++--
3 files changed, 33 insertions(+), 5 deletions(-)
create mode 100644 gcc/testsuite/gcc.target/i386/pr104551.c
Comments
On Wed, Feb 16, 2022 at 05:03:09PM +0800, liuhongt via Gcc-patches wrote:
> > > +(match (cond_expr_convert_p @0 @2 @3 @6)
> > > + (cond (simple_comparison@6 @0 @1) (convert@4 @2) (convert@5 @3))
> > > + (if (types_match (TREE_TYPE (@2), TREE_TYPE (@3))
> >
> > But in principle @2 or @3 could safely differ in sign, you'd then need to ensure
> > to insert sign conversions to @2/@3 to the signedness of @4/@5.
> >
> It turns out differ in sign is not suitable for extension(but ok for truncation),
> because it's zero_extend vs sign_extend.
>
> The patch add types_match check when convert is extension.
>
> Bootstrapped and regtested on x86_64-pc-linux-gnu{-m32,}.
> And native Bootstrapped and regtested on CLX.
>
> Ok for trunk?
>
> gcc/ChangeLog:
>
> PR tree-optimization/104551
> PR tree-optimization/103771
> * match.pd (cond_expr_convert_p): Add types_match check when
> convert is extension.
> * tree-vect-patterns.cc
> (gimple_cond_expr_convert_p): Adjust comments.
> (vect_recog_cond_expr_convert_pattern): Ditto.
>
> gcc/testsuite/ChangeLog:
>
> * gcc.target/i386/pr104551.c: New test.
> ---
> gcc/match.pd | 8 +++++---
> gcc/testsuite/gcc.target/i386/pr104551.c | 24 ++++++++++++++++++++++++
> gcc/tree-vect-patterns.cc | 6 ++++--
> 3 files changed, 33 insertions(+), 5 deletions(-)
> create mode 100644 gcc/testsuite/gcc.target/i386/pr104551.c
>
> diff --git a/gcc/match.pd b/gcc/match.pd
> index 05a10ab6bfd..8e80b9f1576 100644
> --- a/gcc/match.pd
> +++ b/gcc/match.pd
> @@ -7692,11 +7692,13 @@ and,
> (if (INTEGRAL_TYPE_P (type)
> && INTEGRAL_TYPE_P (TREE_TYPE (@2))
> && INTEGRAL_TYPE_P (TREE_TYPE (@0))
> - && INTEGRAL_TYPE_P (TREE_TYPE (@3))
> && TYPE_PRECISION (type) != TYPE_PRECISION (TREE_TYPE (@0))
> && TYPE_PRECISION (TREE_TYPE (@0))
> == TYPE_PRECISION (TREE_TYPE (@2))
> - && TYPE_PRECISION (TREE_TYPE (@0))
> - == TYPE_PRECISION (TREE_TYPE (@3))
> + && (types_match (TREE_TYPE (@2), TREE_TYPE (@3))
> + || ((TYPE_PRECISION (TREE_TYPE (@0))
> + == TYPE_PRECISION (TREE_TYPE (@3)))
> + && INTEGRAL_TYPE_P (TREE_TYPE (@3))
> + && TYPE_PRECISION (TREE_TYPE (@3)) > TYPE_PRECISION (type)))
> && single_use (@4)
> && single_use (@5))))
I find this quite unreadable, it looks like if @2 and @3 are treated
differently. I think keeping the old 3 lines and just adding
&& (TYPE_PRECISION (TREE_TYPE (@0)) >= TYPE_PRECISION (type)
|| (TYPE_UNSIGNED (TREE_TYPE (@2))
== TYPE_UNSIGNED (TREE_TYPE (@3))))
after it ideally with a comment why would be better.
Note, if the precision of @0 and type is the same, I think signedness can
still differ, no?
Jakub
On Wed, Feb 16, 2022 at 10:17 PM Jakub Jelinek via Gcc-patches
<gcc-patches@gcc.gnu.org> wrote:
>
> On Wed, Feb 16, 2022 at 05:03:09PM +0800, liuhongt via Gcc-patches wrote:
> > > > +(match (cond_expr_convert_p @0 @2 @3 @6)
> > > > + (cond (simple_comparison@6 @0 @1) (convert@4 @2) (convert@5 @3))
> > > > + (if (types_match (TREE_TYPE (@2), TREE_TYPE (@3))
> > >
> > > But in principle @2 or @3 could safely differ in sign, you'd then need to ensure
> > > to insert sign conversions to @2/@3 to the signedness of @4/@5.
> > >
> > It turns out differ in sign is not suitable for extension(but ok for truncation),
> > because it's zero_extend vs sign_extend.
> >
> > The patch add types_match check when convert is extension.
> >
> > Bootstrapped and regtested on x86_64-pc-linux-gnu{-m32,}.
> > And native Bootstrapped and regtested on CLX.
> >
> > Ok for trunk?
> >
> > gcc/ChangeLog:
> >
> > PR tree-optimization/104551
> > PR tree-optimization/103771
> > * match.pd (cond_expr_convert_p): Add types_match check when
> > convert is extension.
> > * tree-vect-patterns.cc
> > (gimple_cond_expr_convert_p): Adjust comments.
> > (vect_recog_cond_expr_convert_pattern): Ditto.
> >
> > gcc/testsuite/ChangeLog:
> >
> > * gcc.target/i386/pr104551.c: New test.
> > ---
> > gcc/match.pd | 8 +++++---
> > gcc/testsuite/gcc.target/i386/pr104551.c | 24 ++++++++++++++++++++++++
> > gcc/tree-vect-patterns.cc | 6 ++++--
> > 3 files changed, 33 insertions(+), 5 deletions(-)
> > create mode 100644 gcc/testsuite/gcc.target/i386/pr104551.c
> >
> > diff --git a/gcc/match.pd b/gcc/match.pd
> > index 05a10ab6bfd..8e80b9f1576 100644
> > --- a/gcc/match.pd
> > +++ b/gcc/match.pd
> > @@ -7692,11 +7692,13 @@ and,
> > (if (INTEGRAL_TYPE_P (type)
> > && INTEGRAL_TYPE_P (TREE_TYPE (@2))
> > && INTEGRAL_TYPE_P (TREE_TYPE (@0))
> > - && INTEGRAL_TYPE_P (TREE_TYPE (@3))
> > && TYPE_PRECISION (type) != TYPE_PRECISION (TREE_TYPE (@0))
> > && TYPE_PRECISION (TREE_TYPE (@0))
> > == TYPE_PRECISION (TREE_TYPE (@2))
> > - && TYPE_PRECISION (TREE_TYPE (@0))
> > - == TYPE_PRECISION (TREE_TYPE (@3))
> > + && (types_match (TREE_TYPE (@2), TREE_TYPE (@3))
> > + || ((TYPE_PRECISION (TREE_TYPE (@0))
> > + == TYPE_PRECISION (TREE_TYPE (@3)))
> > + && INTEGRAL_TYPE_P (TREE_TYPE (@3))
> > + && TYPE_PRECISION (TREE_TYPE (@3)) > TYPE_PRECISION (type)))
> > && single_use (@4)
> > && single_use (@5))))
>
> I find this quite unreadable, it looks like if @2 and @3 are treated
> differently. I think keeping the old 3 lines and just adding
> && (TYPE_PRECISION (TREE_TYPE (@0)) >= TYPE_PRECISION (type)
> || (TYPE_UNSIGNED (TREE_TYPE (@2))
> == TYPE_UNSIGNED (TREE_TYPE (@3))))
Yes, good idea.
> after it ideally with a comment why would be better.
> Note, if the precision of @0 and type is the same, I think signedness can
> still differ, no?
We have TYPE_PRECISION (type) != TYPE_PRECISION (TREE_TYPE (@0)).
>
> Jakub
>
@@ -7692,11 +7692,13 @@ and,
(if (INTEGRAL_TYPE_P (type)
&& INTEGRAL_TYPE_P (TREE_TYPE (@2))
&& INTEGRAL_TYPE_P (TREE_TYPE (@0))
- && INTEGRAL_TYPE_P (TREE_TYPE (@3))
&& TYPE_PRECISION (type) != TYPE_PRECISION (TREE_TYPE (@0))
&& TYPE_PRECISION (TREE_TYPE (@0))
== TYPE_PRECISION (TREE_TYPE (@2))
- && TYPE_PRECISION (TREE_TYPE (@0))
- == TYPE_PRECISION (TREE_TYPE (@3))
+ && (types_match (TREE_TYPE (@2), TREE_TYPE (@3))
+ || ((TYPE_PRECISION (TREE_TYPE (@0))
+ == TYPE_PRECISION (TREE_TYPE (@3)))
+ && INTEGRAL_TYPE_P (TREE_TYPE (@3))
+ && TYPE_PRECISION (TREE_TYPE (@3)) > TYPE_PRECISION (type)))
&& single_use (@4)
&& single_use (@5))))
new file mode 100644
@@ -0,0 +1,24 @@
+/* { dg-do run } */
+/* { dg-options "-O3 -mavx2" } */
+/* { dg-require-effective-target avx2 } */
+
+unsigned int
+__attribute__((noipa))
+test(unsigned int a, unsigned char p[16]) {
+ unsigned int res = 0;
+ for (unsigned b = 0; b < a; b += 1)
+ res = p[b] ? p[b] : (char) b;
+ return res;
+}
+
+int main ()
+{
+ unsigned int a = 16U;
+ unsigned char p[16];
+ for (int i = 0; i != 16; i++)
+ p[i] = (unsigned char)128;
+ unsigned int res = test (a, p);
+ if (res != 128)
+ __builtin_abort ();
+ return 0;
+}
@@ -929,8 +929,10 @@ vect_reassociating_reduction_p (vec_info *vinfo,
with conditions:
1) @1, @2, c, d, a, b are all integral type.
2) There's single_use for both @1 and @2.
- 3) a, c and d have same precision.
+ 3) a, c have same precision.
4) c and @1 have different precision.
+ 5) c, d are the same type or they can differ in sign when convert is
+ truncation.
record a and c and d and @3. */
@@ -952,7 +954,7 @@ extern bool gimple_cond_expr_convert_p (tree, tree*, tree (*)(tree));
TYPE_PRECISION (TYPE_E) != TYPE_PRECISION (TYPE_CD);
TYPE_PRECISION (TYPE_AB) == TYPE_PRECISION (TYPE_CD);
single_use of op_true and op_false.
- TYPE_AB could differ in sign.
+ TYPE_AB could differ in sign when (TYPE_E) A is a truncation.
Input: