0.9.11.22

  * fixed type derivation of float boundaries from numbers outside the
    appropriate float range

NEWS

@@ -16,6 +16,8 @@ changes in sbcl-0.9.12 relative to sbcl-0.9.11:
   * fixed bug: Tests for the (VECTOR T) type gave the wrong answer
     when given a vector displaced to an adjustable array.  (reported
     by Utz-Uwe Haus)
+  * bug fix: derivation of float boundaries from numbers outside the
+    appropriate float range (reported by John Wiseman)
   * improvements to DOCUMENTATION for TYPE and STRUCTURE doc-types:
     allow condition class objects as arguments to DOCUMENTATION and
     (SETF DOCUMENTATION); only find and set documentation for

src/compiler/float-tran.lisp

@@ -267,24 +267,30 @@
 ;;; defined range. Quite useful if we want to convert some type of
 ;;; bounded integer into a float.
 (macrolet
-    ((frob (fun type)
+    ((frob (fun type most-negative most-positive)
        (let ((aux-name (symbolicate fun "-DERIVE-TYPE-AUX")))
          `(progn
-           (defun ,aux-name (num)
-             ;; When converting a number to a float, the limits are
-             ;; the same.
-             (let* ((lo (bound-func (lambda (x)
-                                      (coerce x ',type))
-                                    (numeric-type-low num)))
-                    (hi (bound-func (lambda (x)
-                                      (coerce x ',type))
-                                    (numeric-type-high num))))
-               (specifier-type `(,',type ,(or lo '*) ,(or hi '*)))))
-
-           (defoptimizer (,fun derive-type) ((num))
-             (one-arg-derive-type num #',aux-name #',fun))))))
-  (frob %single-float single-float)
-  (frob %double-float double-float))
+            (defun ,aux-name (num)
+              ;; When converting a number to a float, the limits are
+              ;; the same.
+              (let* ((lo (bound-func (lambda (x)
+                                       (if (< x ,most-negative)
+                                           ,most-negative
+                                           (coerce x ',type)))
+                                     (numeric-type-low num)))
+                     (hi (bound-func (lambda (x)
+                                       (if (< ,most-positive x )
+                                           ,most-positive
+                                           (coerce x ',type)))
+                                     (numeric-type-high num))))
+                (specifier-type `(,',type ,(or lo '*) ,(or hi '*)))))
+
+            (defoptimizer (,fun derive-type) ((num))
+              (one-arg-derive-type num #',aux-name #',fun))))))
+  (frob %single-float single-float
+        most-negative-single-float most-positive-single-float)
+  (frob %double-float double-float
+        most-negative-double-float most-positive-double-float))
 ) ; PROGN
 
 ;;;; float contagion

tests/compiler.pure.lisp

@@ -2036,8 +2036,7 @@
 ;; Reported by John Wiseman, sbcl-devel
 ;; Subject: [Sbcl-devel] float type derivation bug?
 ;; Date: Tue, 4 Apr 2006 15:28:15 -0700
-(with-test (:name (:type-derivation :float-bounds)
-                  :fails-on :sbcl)
+(with-test (:name (:type-derivation :float-bounds))
   (compile nil '(lambda (bits)
                  (let* ((s (if (= (ash bits -31) 0) 1 -1))
                         (e (logand (ash bits -23) #xff))

version.lisp-expr

@@ -17,4 +17,4 @@
 ;;; checkins which aren't released. (And occasionally for internal
 ;;; versions, especially for internal versions off the main CVS
 ;;; branch, it gets hairier, e.g. "0.pre7.14.flaky4.13".)
-"0.9.11.21"
+"0.9.11.22"