From: Toru Kisuki <tkisuki@tachyum.com>
Date: Mon, 19 Jun 2023 17:51:09 +0000 (-0600)
Subject: Do not allow "x + 0.0" to "x" optimization with -fsignaling-nans
X-Git-Tag: basepoints/gcc-15~8187
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=827b2a279fc6ad5bb76e4d2c2eb3432955b5e11c;p=thirdparty%2Fgcc.git

Do not allow "x + 0.0" to "x" optimization with -fsignaling-nans

gcc/
	PR rtl-optimization/110305
	* simplify-rtx.cc (simplify_context::simplify_binary_operation_1):
	Handle HONOR_SNANS for x + 0.0.
---

diff --git a/gcc/simplify-rtx.cc b/gcc/simplify-rtx.cc
index 9c68d3606723..1b581447de68 100644
--- a/gcc/simplify-rtx.cc
+++ b/gcc/simplify-rtx.cc
@@ -2831,7 +2831,8 @@ simplify_context::simplify_binary_operation_1 (rtx_code code,
 	 when x is NaN, infinite, or finite and nonzero.  They aren't
 	 when x is -0 and the rounding mode is not towards -infinity,
 	 since (-0) + 0 is then 0.  */
-      if (!HONOR_SIGNED_ZEROS (mode) && trueop1 == CONST0_RTX (mode))
+      if (!HONOR_SIGNED_ZEROS (mode) && !HONOR_SNANS (mode)
+	  && trueop1 == CONST0_RTX (mode))
 	return op0;
 
       /* ((-a) + b) -> (b - a) and similarly for (a + (-b)).  These