* [PATCH v2] Skip constant folding for fmin/max when either argument is sNaN [PR105414]
@ 2022-05-10 8:10 HAO CHEN GUI
2022-05-10 11:12 ` Richard Biener
0 siblings, 1 reply; 2+ messages in thread
From: HAO CHEN GUI @ 2022-05-10 8:10 UTC (permalink / raw)
To: gcc-patches
Cc: Segher Boessenkool, David, Peter Bergner, Kewen.Lin, Richard Biener
Hi,
This patch skips constant folding for fmin/max when either argument
is sNaN. According to C standard,
fmin(sNaN, sNaN)= qNaN, fmin(sNaN, NaN) = qNaN
So signaling NaN should be tested and skipped for fmin/max in match.pd.
The V2 patch splits the for loop and keeps MIN/MAX_EXPR unchanged.
Bootstrapped and tested on ppc64 Linux BE and LE with no regressions.
Is this okay for trunk? Any recommendations? Thanks a lot.
ChangeLog
2022-05-10 Haochen Gui <guihaoc@linux.ibm.com>
gcc/
PR target/105414
* match.pd (minmax): Skip constant folding for fmin/fmax when both
arguments are sNaN or one is sNaN and another is NaN.
gcc/testsuite/
PR target/105414
* gcc.dg/pr105414.c: New.
patch.diff
diff --git a/gcc/match.pd b/gcc/match.pd
index 6d691d302b3..6fb8806412a 100644
--- a/gcc/match.pd
+++ b/gcc/match.pd
@@ -3095,10 +3095,16 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
/* Simplifications of MIN_EXPR, MAX_EXPR, fmin() and fmax(). */
-(for minmax (min max FMIN_ALL FMAX_ALL)
+(for minmax (min max)
(simplify
(minmax @0 @0)
@0))
+/* For fmin() and fmax(), skip folding when both are sNaN. */
+(for minmax (FMIN_ALL FMAX_ALL)
+ (simplify
+ (minmax @0 @0)
+ (if (!tree_expr_maybe_signaling_nan_p (@0))
+ @0)))
/* min(max(x,y),y) -> y. */
(simplify
(min:c (max:c @0 @1) @1)
@@ -3198,12 +3204,13 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
(minmax @1 (convert @2)))))
(for minmax (FMIN_ALL FMAX_ALL)
- /* If either argument is NaN, return the other one. Avoid the
- transformation if we get (and honor) a signalling NaN. */
+ /* If either argument is NaN and other one is not sNaN, return the other
+ one. Avoid the transformation if we get (and honor) a signalling NaN. */
(simplify
(minmax:c @0 REAL_CST@1)
- (if (real_isnan (TREE_REAL_CST_PTR (@1))
- && (!HONOR_SNANS (@1) || !TREE_REAL_CST (@1).signalling))
+ (if (real_isnan (TREE_REAL_CST_PTR (@1))
+ && (!HONOR_SNANS (@1) || !TREE_REAL_CST (@1).signalling)
+ && !tree_expr_maybe_signaling_nan_p (@0))
@0)))
/* Convert fmin/fmax to MIN_EXPR/MAX_EXPR. C99 requires these
functions to return the numeric arg if the other one is NaN.
diff --git a/gcc/testsuite/gcc.dg/pr105414.c b/gcc/testsuite/gcc.dg/pr105414.c
new file mode 100644
index 00000000000..78772700acf
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/pr105414.c
@@ -0,0 +1,30 @@
+/* { dg-do run { target { *-*-linux* *-*-gnu* } } } */
+/* { dg-options "-O1 -fsignaling-nans -lm" } */
+/* { dg-add-options ieee } */
+/* { dg-require-effective-target issignaling } */
+
+
+#define _GNU_SOURCE
+#include <stdio.h>
+#include <math.h>
+
+int main()
+{
+ double a = __builtin_nans ("");
+
+ if (issignaling (fmin (a, a)))
+ __builtin_abort ();
+
+ if (issignaling (fmax (a, a)))
+ __builtin_abort ();
+
+ double b = __builtin_nan ("");
+
+ if (issignaling (fmin (a, b)))
+ __builtin_abort ();
+
+ if (issignaling (fmax (a, b)))
+ __builtin_abort ();
+
+ return 0;
+}
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH v2] Skip constant folding for fmin/max when either argument is sNaN [PR105414]
2022-05-10 8:10 [PATCH v2] Skip constant folding for fmin/max when either argument is sNaN [PR105414] HAO CHEN GUI
@ 2022-05-10 11:12 ` Richard Biener
0 siblings, 0 replies; 2+ messages in thread
From: Richard Biener @ 2022-05-10 11:12 UTC (permalink / raw)
To: HAO CHEN GUI
Cc: gcc-patches, Segher Boessenkool, David, Peter Bergner, Kewen.Lin
On Tue, May 10, 2022 at 10:10 AM HAO CHEN GUI <guihaoc@linux.ibm.com> wrote:
>
> Hi,
> This patch skips constant folding for fmin/max when either argument
> is sNaN. According to C standard,
> fmin(sNaN, sNaN)= qNaN, fmin(sNaN, NaN) = qNaN
> So signaling NaN should be tested and skipped for fmin/max in match.pd.
>
> The V2 patch splits the for loop and keeps MIN/MAX_EXPR unchanged.
>
> Bootstrapped and tested on ppc64 Linux BE and LE with no regressions.
> Is this okay for trunk? Any recommendations? Thanks a lot.
OK.
Richard.
> ChangeLog
> 2022-05-10 Haochen Gui <guihaoc@linux.ibm.com>
>
> gcc/
> PR target/105414
> * match.pd (minmax): Skip constant folding for fmin/fmax when both
> arguments are sNaN or one is sNaN and another is NaN.
>
> gcc/testsuite/
> PR target/105414
> * gcc.dg/pr105414.c: New.
>
>
> patch.diff
> diff --git a/gcc/match.pd b/gcc/match.pd
> index 6d691d302b3..6fb8806412a 100644
> --- a/gcc/match.pd
> +++ b/gcc/match.pd
> @@ -3095,10 +3095,16 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
>
> /* Simplifications of MIN_EXPR, MAX_EXPR, fmin() and fmax(). */
>
> -(for minmax (min max FMIN_ALL FMAX_ALL)
> +(for minmax (min max)
> (simplify
> (minmax @0 @0)
> @0))
> +/* For fmin() and fmax(), skip folding when both are sNaN. */
> +(for minmax (FMIN_ALL FMAX_ALL)
> + (simplify
> + (minmax @0 @0)
> + (if (!tree_expr_maybe_signaling_nan_p (@0))
> + @0)))
> /* min(max(x,y),y) -> y. */
> (simplify
> (min:c (max:c @0 @1) @1)
> @@ -3198,12 +3204,13 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
> (minmax @1 (convert @2)))))
>
> (for minmax (FMIN_ALL FMAX_ALL)
> - /* If either argument is NaN, return the other one. Avoid the
> - transformation if we get (and honor) a signalling NaN. */
> + /* If either argument is NaN and other one is not sNaN, return the other
> + one. Avoid the transformation if we get (and honor) a signalling NaN. */
> (simplify
> (minmax:c @0 REAL_CST@1)
> - (if (real_isnan (TREE_REAL_CST_PTR (@1))
> - && (!HONOR_SNANS (@1) || !TREE_REAL_CST (@1).signalling))
> + (if (real_isnan (TREE_REAL_CST_PTR (@1))
> + && (!HONOR_SNANS (@1) || !TREE_REAL_CST (@1).signalling)
> + && !tree_expr_maybe_signaling_nan_p (@0))
> @0)))
> /* Convert fmin/fmax to MIN_EXPR/MAX_EXPR. C99 requires these
> functions to return the numeric arg if the other one is NaN.
> diff --git a/gcc/testsuite/gcc.dg/pr105414.c b/gcc/testsuite/gcc.dg/pr105414.c
> new file mode 100644
> index 00000000000..78772700acf
> --- /dev/null
> +++ b/gcc/testsuite/gcc.dg/pr105414.c
> @@ -0,0 +1,30 @@
> +/* { dg-do run { target { *-*-linux* *-*-gnu* } } } */
> +/* { dg-options "-O1 -fsignaling-nans -lm" } */
> +/* { dg-add-options ieee } */
> +/* { dg-require-effective-target issignaling } */
> +
> +
> +#define _GNU_SOURCE
> +#include <stdio.h>
> +#include <math.h>
> +
> +int main()
> +{
> + double a = __builtin_nans ("");
> +
> + if (issignaling (fmin (a, a)))
> + __builtin_abort ();
> +
> + if (issignaling (fmax (a, a)))
> + __builtin_abort ();
> +
> + double b = __builtin_nan ("");
> +
> + if (issignaling (fmin (a, b)))
> + __builtin_abort ();
> +
> + if (issignaling (fmax (a, b)))
> + __builtin_abort ();
> +
> + return 0;
> +}
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2022-05-10 11:12 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-05-10 8:10 [PATCH v2] Skip constant folding for fmin/max when either argument is sNaN [PR105414] HAO CHEN GUI
2022-05-10 11:12 ` 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).