On Fri, Oct 23, 2009 at 11:58 AM, Michael Creel <michael.cr...@uab.es>wrote:
> On Fri, Oct 23, 2009 at 12:35 PM, Jaroslav Hajek <high...@gmail.com>
> wrote:
> > On Fri, Oct 23, 2009 at 12:30 PM, Michael Creel <michael.cr...@uab.es>
> wrote:
> >> On Fri, Oct 23, 2009 at 12:28 PM, Jaroslav Hajek <high...@gmail.com>
> wrote:
> >>> On Fri, Oct 23, 2009 at 12:26 PM, Michael Creel <michael.cr...@uab.es>
> wrote:
> >>>> On Fri, Oct 23, 2009 at 12:19 PM, Jaroslav Hajek <high...@gmail.com>
> wrote:
> >>>>> On Fri, Oct 23, 2009 at 12:09 PM, Michael Creel <
> michael.cr...@uab.es> wrote:
> >>>>>> On Fri, Oct 23, 2009 at 12:00 PM, Jaroslav Hajek <high...@gmail.com>
> wrote:
> >>>>>>> On Fri, Oct 23, 2009 at 11:52 AM, Michael Creel <
> michael.cr...@uab.es> wrote:
> >>>>>>>> On Fri, Oct 23, 2009 at 11:23 AM, Jaroslav Hajek <
> high...@gmail.com> wrote:
> >>>>>>>>> On Fri, Oct 23, 2009 at 11:21 AM, Michael Creel <
> michael.cr...@uab.es> wrote:
> >>>>>>>>>> On Fri, Oct 23, 2009 at 10:57 AM, Jaroslav Hajek <
> high...@gmail.com> wrote:
> >>>>>>>>>>>
> >>>>>>>>>>> parcellfun works like cellfun, so this usage is incorrect. see
> help
> >>>>>>>>>>> for cellfun.
> >>>>>>>>>>> Also, for questions that concern exclusively OctaveForge
> functions,
> >>>>>>>>>>> please use the OctaveForge mailing list.
> >>>>>>>>>>>
> >>>>>>>>>>> octave:1> a = parcellfun (2, @sin, {1, 2})
> >>>>>>>>>>> parcellfun: 2/2 jobs done
> >>>>>>>>>>> a =
> >>>>>>>>>>>
> >>>>>>>>>>> 0.84147 0.90930
> >>>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>> Thanks. Sorry about the mail list confusion. I'm getting
> >>>>>>>>>> octave:1> a = parcellfun(2, @sin, {1,2})
> >>>>>>>>>> parcellfun: 2/2 jobs done
> >>>>>>>>>> warning: broken pipe -- some output may be lost
> >>>>>>>>>>
> >>>>>>>>>> At this point octave seems to hang indefinitely. I'm running 64
> bit
> >>>>>>>>>> Kubuntu 9.04. Octave is 3.3.50+ checked out yesterday.
> >>>>>>>>>>
> >>>>>>>>>> Michael
> >>>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> Do you use the SVN version of the general package? If not, do so.
> >>>>>>>>>
> >>>>>>>>> --
> >>>>>>>>
> >>>>>>>> I had been using the released general package, but I just made if
> from
> >>>>>>>> SVN, and I get the same result.
> >>>>>>>> Thanks, M.
> >>>>>>>>
> >>>>>>>
> >>>>>>> I'm using the same (3.3.50+ just built) on OpenSUSE 11.0 without
> >>>>>>> problems. Does __exit__ work as expected for you?
> >>>>>>> Otherwise, I'm afraid you'll need to investigate for yourself what
> is wrong.
> >>>>>>>
> >>>>>>> --
> >>>>>>> RNDr. Jaroslav Hajek
> >>>>>>> computing expert & GNU Octave developer
> >>>>>>> Aeronautical Research and Test Institute (VZLU)
> >>>>>>> Prague, Czech Republic
> >>>>>>> url: www.highegg.matfyz.cz
> >>>>>>>
> >>>>>>
> >>>>>> I'm re-sending this due to a confusing quoted-text misformatting
> problem.
> >>>>>>
> >>>>>> If it type __exit__ from the octave prompt, I get dropped out to the
> >>>>>> shell. That's the expected behavior, correct? A little more
> >>>>>> experimentation shows that it works sometimes:
> >>>>>>
> >>>>>> octave:1> parcellfun(1, @sin, {1,2})
> >>>>>> parcellfun: 2/2 jobs done
> >>>>>> octave:2> [a] = parcellfun(1, @sin, {1,2})
> >>>>>> parcellfun: 2/2 jobs done
> >>>>>> a =
> >>>>>>
> >>>>>> 0.84147 0.90930
> >>>>>>
> >>>>>> octave:3> [a] = parcellfun(2, @sin, {1,2,3})
> >>>>>> parcellfun: 3/3 jobs done
> >>>>>> a =
> >>>>>>
> >>>>>> 0.84147 0.90930 0.14112
> >>>>>>
> >>>>>> octave:4> [a] = parcellfun(2, @sin, {1,2})
> >>>>>> parcellfun: 2/2 jobs done
> >>>>>> warning: broken pipe -- some output may be lost
> >>>>>>
> >>>>>> M.
> >>>>>>
> >>>>>
> >>>>> Never happens for me. I just tried at least 20 times, even 100 times
> >>>>> with a for loop. So I can hardly help you.
> >>>>> One place that may generate this is the block at lines 254-258; if
> >>>>> closing pipes by subprocesses fails. But I don't know why that should
> >>>>> happen for you and not for me.
> >>>>>
> >>>>> --
> >>>>> RNDr. Jaroslav Hajek
> >>>>> computing expert & GNU Octave developer
> >>>>> Aeronautical Research and Test Institute (VZLU)
> >>>>> Prague, Czech Republic
> >>>>> url: www.highegg.matfyz.cz
> >>>>>
> >>>>
> >>>> OK, thanks for clearing up the syntax for me. I don't have the
> >>>> knowledge to investigate further.
> >>>> Thanks, Michael
> >>>>
> >>>
> >>> I could make you some test patches that may help with the issue, but
> >>> I'm short of time right now. If you can wait until November and remind
> >>> me then, I'll try to help you more.
> >>>
> >>> regards
> >>>
> >>> --
> >>> RNDr. Jaroslav Hajek
> >>> computing expert & GNU Octave developer
> >>> Aeronautical Research and Test Institute (VZLU)
> >>> Prague, Czech Republic
> >>> url: www.highegg.matfyz.cz
> >>>
> >>
> >> It's no hurry for me, I was just investigating it as a way of
> >> parallelizing some tasks. I'd be happy to test it again in the future.
> >> Cheers, M.
> >>
> >
> > Well, unless it actually makes problems, I'd say ignore the warning
> > for the time being. Chances are that it won't even manifest on your
> > real problem, because it looks like a race condition dependent on
> > speed of execution of some code. But I'll be glad if eventually the
> > problem is discovered so that parcellfun/pararrayfun are improved.
> >
> > regards
> >
> > --
> > RNDr. Jaroslav Hajek
> > computing expert & GNU Octave developer
> > Aeronautical Research and Test Institute (VZLU)
> > Prague, Czech Republic
> > url: www.highegg.matfyz.cz
> >
>
> Well, I have tried with a more time consuming example, and it occurs
> there, too. Octave locks up when this happens, so it's not just a
> warning, it's pretty much a show stopper. This does look like a very
> promising feature, though.
> M.
>
Hi Michael,
any updates? Want to try some debugging patches?
--
RNDr. Jaroslav Hajek
computing expert & GNU Octave developer
Aeronautical Research and Test Institute (VZLU)
Prague, Czech Republic
url: www.highegg.matfyz.cz
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Octave-dev mailing list
Octave-dev@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/octave-dev
