Hi, Am 08.09.2014 um 16:25 schrieb WangChen: > 1. > I tested it on my computer. > If all name in X10_HOSTLIST are invalid then the program won’t start. > if one of them is right, then the program will running on it. > > 2. > Since you used finish construct, “the End” should always be printed at last. > I get the right output when compiling with native x10(x10c++) > But i get the same results as you do when compiling with managed > x10(x10c). (“the End” not printed at last)
Remember that println does buffer. To write directly (without buffering), call Console.OUT.flush() after every Console.OUT.println(...). This *should* fix ordering. Cheers, Marco > > 3. > To synchronize places you could use the barrier() method in the Team class. > > |*barrier <../../x10/util/Team.html#barrier()>*()| > Blocks until all team members have reached the barrier. > > > I’m new to X10 too and i’m not sure what i said was all right. > Hope i could help. > > > 在 2014年9月8日，下午9:29，Michael Dietrich > <michael.dietr...@s2010.tu-chemnitz.de > <mailto:michael.dietr...@s2010.tu-chemnitz.de>> 写道： > >> Hello, >> >> I'm sorry for my late answer. Finally I've written a test program like >> this >> >> >> import x10.io.Console; >> >> class HelloWholeWorld { >> public static def main(args:Rail[String]):void { >> finish for (p in Place.places()) { >> at (p) async Console.OUT.println(here+" says hello: "+ >> Runtime.getName()); >> } >> >> Console.OUT.println("The End"); >> >> } >> } >> >> The following occurs >> >> [console]$ export X10_NPLACES=2 >> [console]$ export X10_HOSTLIST="name1,name2" >> [console]$ ./HelloWholeWorld >> Place(0) says hello: 28901@name1 >> The End >> Place(1) says >> hello: 25821@name2 >> >> [console]$ >> >> >> If I enter invalid names I don't get any output. "The End" neither. So >> it seems the program doesn't even start if it gets wrong input. If >> this is correct I would be content with this. Is there a way to avoid >> all these spaces? >> >> One simple question since I'm unexperienced with X10 yet: How to >> synchronize the places, so that "The End" is always printed at last? >> >> Thank you very much for your help. >> >> bye >> >> >> Zitat von WangChen <wang...@163.com <mailto:wang...@163.com>>: >> >>> Hi, Michael >>> I’m reading the doc these days and i found an api that might be >>> useful to you. >>> In the Runtime class: >>> getName >>> >>> public static String getName() >>> Get a string that identifies the X10 execution environment of this >>> place. The string is undefined but should be helpful for debugging, >>> typically containing the pid and hostname. On java it is equivalent >>> to >>> java.lang.management.ManagementFactory.getRuntimeMXBean().getName(). >>> >>> >>> >>> >>> 在 2014年8月27日，下午11:33，Michael Dietrich >>> <michael.dietr...@s2010.tu-chemnitz.de >>> <mailto:michael.dietr...@s2010.tu-chemnitz.de>> 写道： >>> >>>> Hello, >>>> >>>> I am using X10 as a part of my bachelor's thesis about high >>>> performance programming languages (so maybe some more questions may be >>>> asked later :-) ). >>>> >>>> My first questions are about using several places: >>>> >>>> * If I set X10_NPLACES=n with having less than n hosts, the places >>>> will be distributed among the existing hosts, right? >>>> >>>> * Is there a possibility to find out which place is located on which >>>> host? In Chapel you have here.name. >>>> >>>> * Should X10_NTHREADS be equal to the amount of cores within a host? >>>> >>>> * Let there be X10_HOSTLIST="A,B,foo". A and B are the hosts I want to >>>> use while foo is a randomly chosen set of letters for simulating a >>>> mistyped or currently inactive computer. >>>> If I start a program with this parameter I don't get an error message >>>> though there is something wrong. If I do isDead() on every place every >>>> of them returns false. >>>> Did I maybe forget to set something so it doesn't distribute the >>>> places to other computers or does X10 ignore host names which cannot >>>> be reached? >>>> >>>> Thank you very much >>>> bye >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> Slashdot TV. >>>> Video for Nerds. Stuff that matters. >>>> http://tv.slashdot.org/ >>>> _______________________________________________ >>>> X10-users mailing list >>>> X10-users@lists.sourceforge.net >>>> https://lists.sourceforge.net/lists/listinfo/x10-users >> >> >> >> >> ------------------------------------------------------------------------------ >> Want excitement? >> Manually upgrade your production database. >> When you want reliability, choose Perforce >> Perforce version control. Predictably reliable. >> http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk >> _______________________________________________ >> X10-users mailing list >> X10-users@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/x10-users > > > > ------------------------------------------------------------------------------ > Want excitement? > Manually upgrade your production database. > When you want reliability, choose Perforce > Perforce version control. Predictably reliable. > http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk > > > > _______________________________________________ > X10-users mailing list > X10-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/x10-users > ------------------------------------------------------------------------------ Want excitement? Manually upgrade your production database. When you want reliability, choose Perforce Perforce version control. Predictably reliable. http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk _______________________________________________ X10-users mailing list X10-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/x10-users
