[llvm] dc9664a - CodeGen: Strengthen definition of F{MIN|MAX}NUM_IEEE nodes (#85195)

[via llvm-commits]

Author: Matt Arsenault
Date: 2024-04-22T10:13:04+02:00
New Revision: dc9664a8adae17f2083fbcc8e96cfce606c56d57

URL: https://github.com/llvm/llvm-project/commit/dc9664a8adae17f2083fbcc8e96cfce606c56d57
DIFF: https://github.com/llvm/llvm-project/commit/dc9664a8adae17f2083fbcc8e96cfce606c56d57.diff

LOG: CodeGen: Strengthen definition of F{MIN|MAX}NUM_IEEE nodes (#85195)

Previously these were declared as having the 2008 behavior, with
underspecified signed zero handling. Currently, AMDGPU, PPC and
LoongArch mark these as legal. The AMDGPU and PPC instructions respect
the signed zero behavior. The LoongArch documentation doesn't state, but
I'm assuming it also does.

Added: 
    

Modified: 
    llvm/docs/GlobalISel/GenericOpcode.rst
    llvm/include/llvm/CodeGen/ISDOpcodes.h
    llvm/include/llvm/Target/GenericOpcodes.td

Removed: 
    


################################################################################
diff  --git a/llvm/docs/GlobalISel/GenericOpcode.rst b/llvm/docs/GlobalISel/GenericOpcode.rst
index a12627c01d205b..492d30280f4776 100644
--- a/llvm/docs/GlobalISel/GenericOpcode.rst
+++ b/llvm/docs/GlobalISel/GenericOpcode.rst
@@ -521,16 +521,32 @@ The return value of (FMAXNUM 0.0, -0.0) could be either 0.0 or -0.0.
 G_FMINNUM_IEEE
 ^^^^^^^^^^^^^^
 
-Perform floating-point minimum on two values, following the IEEE-754 2008
-definition. This 
diff ers from FMINNUM in the handling of signaling NaNs. If one
-input is a signaling NaN, returns a quiet NaN.
+Perform floating-point minimum on two values, following IEEE-754
+definitions. This 
diff ers from FMINNUM in the handling of signaling
+NaNs.
+
+If one input is a signaling NaN, returns a quiet NaN. This matches
+IEEE-754 2008's minnum/maxnum for signaling NaNs (which 
diff ers from
+2019).
+
+These treat -0 as ordered less than +0, matching the behavior of
+IEEE-754 2019's minimumNumber/maximumNumber (which was unspecified in
+2008).
 
 G_FMAXNUM_IEEE
 ^^^^^^^^^^^^^^
 
-Perform floating-point maximum on two values, following the IEEE-754 2008
-definition. This 
diff ers from FMAXNUM in the handling of signaling NaNs. If one
-input is a signaling NaN, returns a quiet NaN.
+Perform floating-point maximum on two values, following IEEE-754
+definitions. This 
diff ers from FMAXNUM in the handling of signaling
+NaNs.
+
+If one input is a signaling NaN, returns a quiet NaN. This matches
+IEEE-754 2008's minnum/maxnum for signaling NaNs (which 
diff ers from
+2019).
+
+These treat -0 as ordered less than +0, matching the behavior of
+IEEE-754 2019's minimumNumber/maximumNumber (which was unspecified in
+2008).
 
 G_FMINIMUM
 ^^^^^^^^^^

diff  --git a/llvm/include/llvm/CodeGen/ISDOpcodes.h b/llvm/include/llvm/CodeGen/ISDOpcodes.h
index 49d51a27e3c0f6..078a936b061a32 100644
--- a/llvm/include/llvm/CodeGen/ISDOpcodes.h
+++ b/llvm/include/llvm/CodeGen/ISDOpcodes.h
@@ -971,10 +971,16 @@ enum NodeType {
   FMINNUM,
   FMAXNUM,
 
-  /// FMINNUM_IEEE/FMAXNUM_IEEE - Perform floating-point minimum or maximum on
-  /// two values, following the IEEE-754 2008 definition. This 
diff ers from
-  /// FMINNUM/FMAXNUM in the handling of signaling NaNs. If one input is a
-  /// signaling NaN, returns a quiet NaN.
+  /// FMINNUM_IEEE/FMAXNUM_IEEE - Perform floating-point minimumNumber or
+  /// maximumNumber on two values, following IEEE-754 definitions. This 
diff ers
+  /// from FMINNUM/FMAXNUM in the handling of signaling NaNs, and signed zero.
+  ///
+  /// If one input is a signaling NaN, returns a quiet NaN. This matches
+  /// IEEE-754 2008's minnum/maxnum behavior for signaling NaNs (which 
diff ers
+  /// from 2019).
+  ///
+  /// These treat -0 as ordered less than +0, matching the behavior of IEEE-754
+  /// 2019's minimumNumber/maximumNumber.
   FMINNUM_IEEE,
   FMAXNUM_IEEE,
 

diff  --git a/llvm/include/llvm/Target/GenericOpcodes.td b/llvm/include/llvm/Target/GenericOpcodes.td
index e8cf8fcb647f45..8380d2738d164b 100644
--- a/llvm/include/llvm/Target/GenericOpcodes.td
+++ b/llvm/include/llvm/Target/GenericOpcodes.td
@@ -795,10 +795,17 @@ def G_FMAXNUM : GenericInstruction {
   let isCommutable = true;
 }
 
-// FMINNUM_IEEE/FMAXNUM_IEEE - Perform floating-point minimum or maximum on
-// two values, following the IEEE-754 2008 definition. This 
diff ers from
-// FMINNUM/FMAXNUM in the handling of signaling NaNs. If one input is a
-// signaling NaN, returns a quiet NaN.
+// FMINNUM_IEEE/FMAXNUM_IEEE - Perform floating-point minimumNumber or
+// maximumNumber on two values, following IEEE-754 definitions. This
+// 
diff ers from FMINNUM/FMAXNUM in the handling of signaling NaNs, and
+// signed zero.
+//
+// If one input is a signaling NaN, returns a quiet NaN. This matches
+// IEEE-754 2008's minnum/maxnum behavior for signaling NaNs (which
+// 
diff ers from 2019).
+//
+// These treat -0 as ordered less than +0, matching the behavior of
+// IEEE-754 2019's minimumNumber/maximumNumber.
 def G_FMINNUM_IEEE : GenericInstruction {
   let OutOperandList = (outs type0:$dst);
   let InOperandList = (ins type0:$src1, type0:$src2);