4
edits
Changes
Jump to navigation
Jump to search
→Floating-Point Behavior
Line 806:
Line 806:
| rsq(-0)
| rsq(-0)
| +inf
| +inf
−|
+| no -0 so differs from IEEE where rsq(-0) = -inf
|-
|-
| rsq(-2)
| rsq(-2)
Line 835:
Line 835:
| max(0, NaN)
| max(0, NaN)
| NaN
| NaN
−| max and min violate IEEE but match GLSL spec
+| max violates IEEE but match GLSL spec
|-
|-
| max(NaN, 0)
| max(NaN, 0)
Line 856:
Line 856:
| min(0, NaN)
| min(0, NaN)
| NaN
| NaN
−|
+| min violates IEEE but match GLSL spec
|-
|-
| min(NaN, 0)
| min(NaN, 0)