Merge pull request #1089 from Subv/neg_bits

Shaders: Corrected the 'abs' and 'neg' bit usage in the float arithmetic instructions.

src/video_core/engines/shader_bytecode.h

@@ -292,6 +292,10 @@ union Instruction {
         }
     } alu;
 
+    union {
+        BitField<48, 1, u64> negate_b;
+    } fmul;
+
     union {
         BitField<48, 1, u64> is_signed;
     } shift;