November 2010 - cmucl-imp - mailman3.common-lisp.net

[cmucl-imp] format-related warning
by Didier Verna 11 Nov '10

11 Nov '10

Hello, I just discovered a small bug in a format string of mine thanks to CCL. It is illustrated below: the ~:; control sequence is meaningless in this context. (defun test () (format t "~:[[~A]~:;~A~]" t "foo") (format t "~:[[~A]~:;~A~]" nil "bar") (values)) This function still works correctly, but perhaps a warning would be nice to have, as CCL does. -- Resistance is futile. You will be jazzimilated. Scientific site: http://www.lrde.epita.fr/~didier Music (Jazz) site: http://www.didierverna.com

2 1

[cmucl-imp] Working set
by Helmut Eller 07 Nov '10

07 Nov '10

Is there an easy way to determine the working set size (resident size) that is needed to evaluate some form? TIME prints various things but not what I need. The Linux top command shows a column with the resident size; essentially I'd like to see the maximum of the those values sampled while executing the form. Does the GC have this information already? Would it be difficult/expensive to use getrusage in GC hooks to collect the information? Helmut

2 3

Re: [cmucl-imp] Test
by Raymond Toy 02 Nov '10

02 Nov '10

On 11/1/10 6:37 PM, Martin Cracauer wrote: > Raymond Toy wrote on Mon, Nov 01, 2010 at 06:33:34PM -0400: >> On 11/1/10 6:25 PM, Martin Cracauer wrote: >>> Raymond Toy wrote on Sun, Oct 31, 2010 at 09:34:25AM -0400: >>>> Testing list after move. >>> Looks like replying through Gmane is broken. Can somebody confirm >>> this? I'm not using it. >> It's been broken ever since we closed the lists. >> >> I think I brought this up back then and you said gmane was subscribed. >> I didn't pursue it any further. > I think something just got hosed there, even I cannot post to the list > anymore :-) > Post to which list? Sending messages using gmane? Or sending messages to cmucl-imp and having to archived by gmane? The latter is working for me. Oh, wait. I see that your messages aren't getting archived by gmane after you moved the list. But my messages are. Don't know what's up with that. :-( Ray

2 1

Re: [cmucl-imp] Test
by Raymond Toy 01 Nov '10

01 Nov '10

On 11/1/10 6:25 PM, Martin Cracauer wrote: > Raymond Toy wrote on Sun, Oct 31, 2010 at 09:34:25AM -0400: >> Testing list after move. > Looks like replying through Gmane is broken. Can somebody confirm > this? I'm not using it. It's been broken ever since we closed the lists. I think I brought this up back then and you said gmane was subscribed. I didn't pursue it any further. Ray

1 0

[cmucl-imp] 20b patch 000 available
by Raymond Toy 01 Nov '10

01 Nov '10

Well, that didn't last long. I've discovered a bug in REALPART and IMAGPART for x86/sse2 builds. Here is an example of the bug: (defun zot-r (z) (declare (type (simple-array (complex double-float) (*)) z)) (realpart (aref z 0))) (compile 'zot-r) (let ((z (make-array 1 :element-type '(complex double-float) :initial-element #c(42d0 -42d0)))) (zot-r z)) This returns 0d0 instead of 42d0. If you replace realpart with imagpart, you also get 0d0 instead of -42d0. This has been fixed in 2010-11 snapshot. For 20b, you can find a patch for this at <http://common-lisp.net/project/cmucl/downloads/release/20b/patches/cmucl-20…> As usual, download this patch, untar it somewhere. Load cmucl-20b-patch-000.asd, and (asdf:oos 'asdf:load-op :cmucl-20b-patch-000). (Alternatively, you can untar it somewhere where asdf can find the patch-000.asd file.) Sorry for the trouble, Ray

1 0