mohammed elsaeedy <mohammed.elsae...@kaust.edu.sa> wrote on 07/13/2010
05:11:20 AM:
>
>
> For the first proposal, I managed to get CopyTo to work, I was using
> localRail of type "ValRail" with a different size (smaller) of tempRail
> I think thats why it was blocked.
>
> but what about the second proposal?
The implementation of Rail.view() is commented out of the native
implementation of Rail, but the method wasn't commented out of Rail.x10
file. Looking at the comment in the X10 file, I suspect someone was about
to remove view entirely, removed it at the native implementation level,
but failed to remove it at the X10 API level.
--dave
>
> Thank you,
> Regards
> Mohammed @ KAUST
>
> On Tue, Jul 13, 2010 at 5:02 AM, mohammed elsaeedy <
> mohammed.elsae...@kaust.edu.sa> wrote:
>
> > Dear David,
> >
> > Thank you again for your reply, but I'm still facing some trouble
> >
> > 1) The first solution:
> >
> > Console.OUT.println("Before copying: "+ +here.id);
> > finish tempRail.copyTo(0,localRail,0,tempRail.length());
> > Console.OUT.println("finished copying: "+ +here.id);
> >
> > this is all done in the scope of place 1, so tempRail is the Rail of
the
> > Array at place 1, where as localRail is the Rail at place 0
> > so when I run this it stops indefinitely at the copyTo function, and
never
> > proceeds with execution.
> >
> > 2) the other proposal, which was using ValRail, the compiler gives me
the
> > following errors, which I really don't get
> >
> > *x10c++: In file included from Hello.cc:4:
> > Hello.inc: In member function ‘void Hello__closure__13::apply()’:
> > Hello.inc:290: error: ‘class x10::lang::Rail<double>’ has no
member
> > named ‘view’
> > x10c++: Non-zero return code: 1
> > 2 errors.
> > *
> >
> > at(matrix.dist()(0,1)) // PLACE 1
> > {
> > val temp= new Array[double]([0..3],((i):Point)=>0.0);
> >
> > val x1 =((matrix(0,1)) as Array[double](1));
> > val y =((vector(1)) as Array[double](1));
> >
> > Hello.myRun(x1.raw(),y.raw(),temp.raw());
> >
> >
> > val tempValRail= (temp.raw()as Rail[double]).view();
> >
> > at(matrix.dist()(0,0)) //PLACE 0
> > {
> > for(var i:int=0; i<2; i++)
> > {
> > Hello.result(i+2)+=tempValRail(i);
> > }
> > }
> > }
> >
> > Thanks a lot
> > Mohammed @ KAUST
> >
> > On Mon, Jul 12, 2010 at 10:32 PM, David P Grove <gro...@us.ibm.com>
wrote:
> >
> >>
> >>
> >> mohammed elsaeedy <mohammed.elsae...@kaust.edu.sa> wrote on 07/12/2010
> >> 03:43:36 PM:
> >> >
> >> > but my problem is when I want to sum up all the results in a global
> >> linear
> >> > array called result which is at place 0
> >> > when ever i want to add those elements that reside at place 1 to the
> >> values
> >> > of the result array, a bad place exception occurs.
> >> > so how can I pass an array (or the values of the array) that resides
at
> >> > place 1 to the result array that resides at place 0?
> >> >
> >>
> >> There isn't quite as much support for this in 2.0.4 as there should
be.
> >> I'm hoping to make improved support for this kind of operation part of
the
> >> Array API for 2.0.5 (was working on it this afternoon).
> >>
> >> In 2.0.4, what you need to do is the following series of steps.
> >>
> >> (1) call the raw() method on the array at place 1 to get the
> >> backing
> >> Rail.
> >> (2) either
> >> (a) use one of the Rail.copyTo/copyFrom methods to copy
the
> >> Rail from place 1 to place 0
> >> (b) convert the Rail to a ValRail, then capture the
ValRail
> >> in
> >> an "at" to copy it to place 0.
> >> (3) at place 0, take the copied values from the Rail/ValRail
and
> >> use
> >> them to update the Array.
> >>
> >> What I'm hoping to finish in time for 2.0.5 is to implement
> >> copyTo/copyFrom
> >> on Array so you can at least avoid having to know about the raw()
method
> >> to
> >> copy an Array's data from an Array in one place to another Array in a
> >> second place.
> >>
> >> --dave
> >>
> >>
>
------------------------------------------------------------------------------
> >> This SF.net email is sponsored by Sprint
> >> What will you do first with EVO, the first 4G phone?
> >> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> >> _______________________________________________
> >> X10-users mailing list
> >> X10-users@lists.sourceforge.net
> >> https://lists.sourceforge.net/lists/listinfo/x10-users
> >>
> >
> >
> >
> > --
> > Thank you for your concern.
> > Regards,
> > Mohammed El Sayed
> >
>
>
>
> --
> Thank you for your concern.
> Regards,
> Mohammed El Sayed
>
------------------------------------------------------------------------------
> This SF.net email is sponsored by Sprint
> What will you do first with EVO, the first 4G phone?
> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
> _______________________________________________
> X10-users mailing list
> X10-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/x10-users
------------------------------------------------------------------------------
This SF.net email is sponsored by Sprint
What will you do first with EVO, the first 4G phone?
Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first
_______________________________________________
X10-users mailing list
X10-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/x10-users
