Custom Query (1030 matches)

? (sqrt #c(-382810.37 0.0))
#C(-2.7044967E-5 618.71674)

The realpart of the result should be zero.

See ​http://clozure.com/pipermail/openmcl-devel/2013-January/013968.html

Given

(defun baz (A G H)
  (LABELS ((%F14 (x y z)
             (IF (OR (< -1 -2 (DOTIMES (I 0 (FLET ((%F11 (F11-1 F11-2
                                                                F11-3) 48)) -16))
                                (RETURN-FROM %F14 5)))
                     nil)
               nil
               (COUNT (DPB (COUNT A '(16)) (BYTE 11 0) (RETURN-FROM %F14
                                                         1))
                      #(2)))))
    (%F14 1 0 0)))

calling

(baz 1 2 3)

in the REPL on x8664 reports that (non-volatile) registers were clobbered in the call.

This worked correctly in earlier releases; it seems to have broken around the time that the x86 backend started trying to "elide" speculative push/pop pairs.

Hello CCL

First I want to thank you for your excellent Lisp. This ticket applies to the Mac IDE for 1.11-r16651 (DarwinX8664).

Having looked at the code for printing, I realize that there are some structural issues that make it more difficult to detect circularity when inspecting. However:

Inspecting a circular list works fine, but if the circles are within structs, CLOS objects, etc., then the inspector generates a stack overflow, and no recovery is possible.

I know you are busy, and perhaps the Mac IDE is not a priority. I have seen that there are efforts underway to build an IDE from Atom. I understand this.

So, if you plan to fix this inspection crash problem for the Mac IDE, I will be glad to wait. But if you are not planning to fix it, then please let me know {brooks.martin@ sympatico.ca} and I will become very motivated to find another way to work with CCL.

thanks, Martin