bug-mit-scheme
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Bug-mit-scheme] [bug #27887] i386 assembly hook for generic multiplicat


From: Taylor R. Campbell
Subject: [Bug-mit-scheme] [bug #27887] i386 assembly hook for generic multiplication multiplies fixnums incorrectly
Date: Fri, 30 Oct 2009 21:24:56 +0000
User-agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X; en-US) AppleWebKit/525.18 (KHTML, like Gecko, Safari/525.20) OmniWeb/v622.6.1.0.111015

URL:
  <http://savannah.gnu.org/bugs/?27887>

                 Summary: i386 assembly hook for generic multiplication
multiplies fixnums incorrectly
                 Project: MIT/GNU Scheme
            Submitted by: riastradh
            Submitted on: Fri 30 Oct 2009 21:24:55 GMT
                Category: microcode
                Severity: 2 - Minor
                Priority: 5 - Normal
              Item Group: Incorrect behavior
                  Status: None
                 Privacy: Public
             Assigned to: riastradh
         Originator Name: 
        Originator Email: 
             Open/Closed: Open
         Discussion Lock: Any
                Keywords: 

    _______________________________________________________

Details:

((fluid-let ((compiler:open-code-primitives? #f))
   (compile-procedure
    (eval `(LAMBDA (X Y) (,(make-primitive-procedure '&*) X Y))
          (nearest-repl/environment))))
 1
 2)
;Value: 128

This is a reminder to me to commit the fix (shift EAX right by 6 before
IMUL), once I have cleaned up a large number of other changes that it
shouldn't get swept into.

(This doesn't really matter at all -- nobody ever inhibits open-coding of
primitives, so that the fixnum case is always open-coded and the assembly
hook's fixnum case is never used.  So this bug has lain dormant for fifteen
years.)




    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/bugs/?27887>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/





reply via email to

[Prev in Thread] Current Thread [Next in Thread]