* [PATCH] match: Use uniform_integer_cst_p in bitwise_inverted_equal_p [PR115238]
@ 2024-05-27 4:13 Andrew Pinski
2024-05-27 6:32 ` Richard Biener
0 siblings, 1 reply; 2+ messages in thread
From: Andrew Pinski @ 2024-05-27 4:13 UTC (permalink / raw)
To: gcc-patches; +Cc: Andrew Pinski
I noticed while working on the `a ^ CST` patch, that bitwise_inverted_equal_p
would check INTEGER_CST directly and not handle vector csts that are uniform.
This moves over to using uniform_integer_cst_p instead of checking INTEGER_CST
directly.
Bootstrapped and tested on x86_64-linux-gnu with no regressions.
PR tree-optimization/115238
gcc/ChangeLog:
* generic-match-head.cc (bitwise_inverted_equal_p): Use
uniform_integer_cst_p instead of checking INTEGER_CST.
* gimple-match-head.cc (gimple_bitwise_inverted_equal_p): Likewise.
gcc/testsuite/ChangeLog:
* gcc.dg/tree-ssa/bitops-9.c: New test.
Signed-off-by: Andrew Pinski <quic_apinski@quicinc.com>
---
gcc/generic-match-head.cc | 6 ++++--
gcc/gimple-match-head.cc | 6 ++++--
gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c | 15 +++++++++++++++
3 files changed, 23 insertions(+), 4 deletions(-)
create mode 100644 gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c
diff --git a/gcc/generic-match-head.cc b/gcc/generic-match-head.cc
index 3709fe5456d..641d8e9b2de 100644
--- a/gcc/generic-match-head.cc
+++ b/gcc/generic-match-head.cc
@@ -146,8 +146,10 @@ bitwise_inverted_equal_p (tree expr1, tree expr2, bool &wascmp)
return false;
if (!tree_nop_conversion_p (TREE_TYPE (expr1), TREE_TYPE (expr2)))
return false;
- if (TREE_CODE (expr1) == INTEGER_CST && TREE_CODE (expr2) == INTEGER_CST)
- return wi::to_wide (expr1) == ~wi::to_wide (expr2);
+ tree cst1 = uniform_integer_cst_p (expr1);
+ tree cst2 = uniform_integer_cst_p (expr2);
+ if (cst1 && cst2)
+ return wi::to_wide (cst1) == ~wi::to_wide (cst2);
if (operand_equal_p (expr1, expr2, 0))
return false;
if (TREE_CODE (expr1) == BIT_NOT_EXPR
diff --git a/gcc/gimple-match-head.cc b/gcc/gimple-match-head.cc
index d5908f4e9a6..e26fa0860ee 100644
--- a/gcc/gimple-match-head.cc
+++ b/gcc/gimple-match-head.cc
@@ -295,8 +295,10 @@ gimple_bitwise_inverted_equal_p (tree expr1, tree expr2, bool &wascmp, tree (*va
return false;
if (!tree_nop_conversion_p (TREE_TYPE (expr1), TREE_TYPE (expr2)))
return false;
- if (TREE_CODE (expr1) == INTEGER_CST && TREE_CODE (expr2) == INTEGER_CST)
- return wi::to_wide (expr1) == ~wi::to_wide (expr2);
+ tree cst1 = uniform_integer_cst_p (expr1);
+ tree cst2 = uniform_integer_cst_p (expr2);
+ if (cst1 && cst2)
+ return wi::to_wide (cst1) == ~wi::to_wide (cst2);
if (operand_equal_p (expr1, expr2, 0))
return false;
diff --git a/gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c b/gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c
new file mode 100644
index 00000000000..a18b6bf3214
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c
@@ -0,0 +1,15 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -fdump-tree-optimized-raw" } */
+/* PR tree-optimization/115238 */
+
+
+#define vector8 __attribute__((vector_size(2*sizeof(int))))
+
+void f(int a, vector8 int *b)
+{
+ a = 1;
+ *b = a | ((~a) ^ *b);
+}
+/* { dg-final { scan-tree-dump-not "bit_xor_expr, " "optimized" } } */
+/* { dg-final { scan-tree-dump-times "bit_ior_expr, " 1 "optimized" } } */
+/* { dg-final { scan-tree-dump-times "bit_not_expr, " 1 "optimized" } } */
--
2.43.0
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH] match: Use uniform_integer_cst_p in bitwise_inverted_equal_p [PR115238]
2024-05-27 4:13 [PATCH] match: Use uniform_integer_cst_p in bitwise_inverted_equal_p [PR115238] Andrew Pinski
@ 2024-05-27 6:32 ` Richard Biener
0 siblings, 0 replies; 2+ messages in thread
From: Richard Biener @ 2024-05-27 6:32 UTC (permalink / raw)
To: Andrew Pinski; +Cc: gcc-patches
On Mon, May 27, 2024 at 6:14 AM Andrew Pinski <quic_apinski@quicinc.com> wrote:
>
> I noticed while working on the `a ^ CST` patch, that bitwise_inverted_equal_p
> would check INTEGER_CST directly and not handle vector csts that are uniform.
> This moves over to using uniform_integer_cst_p instead of checking INTEGER_CST
> directly.
>
> Bootstrapped and tested on x86_64-linux-gnu with no regressions.
OK
> PR tree-optimization/115238
>
> gcc/ChangeLog:
>
> * generic-match-head.cc (bitwise_inverted_equal_p): Use
> uniform_integer_cst_p instead of checking INTEGER_CST.
> * gimple-match-head.cc (gimple_bitwise_inverted_equal_p): Likewise.
>
> gcc/testsuite/ChangeLog:
>
> * gcc.dg/tree-ssa/bitops-9.c: New test.
>
> Signed-off-by: Andrew Pinski <quic_apinski@quicinc.com>
> ---
> gcc/generic-match-head.cc | 6 ++++--
> gcc/gimple-match-head.cc | 6 ++++--
> gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c | 15 +++++++++++++++
> 3 files changed, 23 insertions(+), 4 deletions(-)
> create mode 100644 gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c
>
> diff --git a/gcc/generic-match-head.cc b/gcc/generic-match-head.cc
> index 3709fe5456d..641d8e9b2de 100644
> --- a/gcc/generic-match-head.cc
> +++ b/gcc/generic-match-head.cc
> @@ -146,8 +146,10 @@ bitwise_inverted_equal_p (tree expr1, tree expr2, bool &wascmp)
> return false;
> if (!tree_nop_conversion_p (TREE_TYPE (expr1), TREE_TYPE (expr2)))
> return false;
> - if (TREE_CODE (expr1) == INTEGER_CST && TREE_CODE (expr2) == INTEGER_CST)
> - return wi::to_wide (expr1) == ~wi::to_wide (expr2);
> + tree cst1 = uniform_integer_cst_p (expr1);
> + tree cst2 = uniform_integer_cst_p (expr2);
> + if (cst1 && cst2)
> + return wi::to_wide (cst1) == ~wi::to_wide (cst2);
> if (operand_equal_p (expr1, expr2, 0))
> return false;
> if (TREE_CODE (expr1) == BIT_NOT_EXPR
> diff --git a/gcc/gimple-match-head.cc b/gcc/gimple-match-head.cc
> index d5908f4e9a6..e26fa0860ee 100644
> --- a/gcc/gimple-match-head.cc
> +++ b/gcc/gimple-match-head.cc
> @@ -295,8 +295,10 @@ gimple_bitwise_inverted_equal_p (tree expr1, tree expr2, bool &wascmp, tree (*va
> return false;
> if (!tree_nop_conversion_p (TREE_TYPE (expr1), TREE_TYPE (expr2)))
> return false;
> - if (TREE_CODE (expr1) == INTEGER_CST && TREE_CODE (expr2) == INTEGER_CST)
> - return wi::to_wide (expr1) == ~wi::to_wide (expr2);
> + tree cst1 = uniform_integer_cst_p (expr1);
> + tree cst2 = uniform_integer_cst_p (expr2);
> + if (cst1 && cst2)
> + return wi::to_wide (cst1) == ~wi::to_wide (cst2);
> if (operand_equal_p (expr1, expr2, 0))
> return false;
>
> diff --git a/gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c b/gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c
> new file mode 100644
> index 00000000000..a18b6bf3214
> --- /dev/null
> +++ b/gcc/testsuite/gcc.dg/tree-ssa/bitops-9.c
> @@ -0,0 +1,15 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -fdump-tree-optimized-raw" } */
> +/* PR tree-optimization/115238 */
> +
> +
> +#define vector8 __attribute__((vector_size(2*sizeof(int))))
> +
> +void f(int a, vector8 int *b)
> +{
> + a = 1;
> + *b = a | ((~a) ^ *b);
> +}
> +/* { dg-final { scan-tree-dump-not "bit_xor_expr, " "optimized" } } */
> +/* { dg-final { scan-tree-dump-times "bit_ior_expr, " 1 "optimized" } } */
> +/* { dg-final { scan-tree-dump-times "bit_not_expr, " 1 "optimized" } } */
> --
> 2.43.0
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2024-05-27 6:32 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-05-27 4:13 [PATCH] match: Use uniform_integer_cst_p in bitwise_inverted_equal_p [PR115238] Andrew Pinski
2024-05-27 6:32 ` Richard Biener
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).