Re: Unanswered questions

From: Paul H. Hargrove (PHHargrove_at_lbl_dot_gov)
Date: Sun Apr 26 2009 - 20:27:55 PDT

  • Next message: Andreev Nikita: "Re[2]: Unanswered questions"
    Andreev Nikita wrote:
    > Hi, guys.
    > I sent you two letters at 17 and 18 of April. But haven't received any
    > reply.
    > Anything wrong with them? A can provide any clarification if needed. I
    > really need to know answers to these questions.
    > All the best,
    > Nikita.
    I believe you are referring to the two messages I quote below.  If that 
    is the case, then I believe you have received no reply because nobody 
    here knows an answer.  I, for instance, know very little about GASP 
    other than by reading the documentation (which you appear to have done a 
    better job of than I).  I also do not keep track of who is using our 
    software for what projects.
    In partial response to the question of UPC rising in popularity: I can 
    say that both IBM and Cray appear to be taking UPC seriously when they 
    are designing and marketing their machines.
    With respect to line and column numbers w/ GASP, I would guess that the 
    --line-comm-stats flag to ppwrun may be required.  I didn't reply 
    earlier because that is just a guess.  If you are using GASP w/o PPW, 
    then I have no guess at all.
    Perhaps others could step up and provide some answers?
    > *From:* Andreev Nikita (/
    > *Date:* Thu Apr 16 2009 - 19:12:00 PDT
    > Hello.
    > Thanks Paul for clear previous answers.
    > I want to know if there is any evidence of rising PGAS popularity in
    > HPC field. Could you provide list of big clusters and projects where
    > PGAS languages is being used (as much as possible).
    > I would really appreciate that.
    > Thanks in advance,
    > Nikita.
    > *From:* Nikita Andreev (/
    > *Date:* Sat Apr 18 2009 - 04:13:56 PDT
    > Paul,
    > >> 2. In GASP split-phase barrier consists of two events: GASP_UPC_NOTIFY 
    > >> and GASP_UPC_WAIT. How does upc_fence and ordinary barrier appears in 
    > >> trace? There are GASP_UPC_BARRIER and GASP_UPC_FENCE but it turns out 
    > >> that same GASP_UPC_NOTIFY and GASP_UPC_WAIT are being used.
    > >
    > > This I am not sure about.  I would expect that the fence should NOT be 
    > > generating the NOTIFY and WAIT GASP events.  If you have a simple example 
    > > that you are sure is not seeing the GASP events implicit final barrier, 
    > > please share it and we can take a look at it.
    > It seems to me that in current GASP implementation in Berkeley UPC 
    > translator fence as long as ordinary barrier generate NOTIFY and WAIT 
    > events. Application and trace it generates are in enclosure. You can see 
    > many ordinary barriers and fence in code but you won't see neither 
    > GASP_UPC_BARRIER nor GASP_UPC_FENCE in trace file.
    > And another question. Does GASP in Berkeley UPC translator support filename, 
    > linenum and colnum args of gasp_event_notify event? As you can see in trace 
    > file I provided linenum and colnum are always 0 and filenum equals to some 
    > odd values.
    > All regards,
    > Nikita. 
    Paul H. Hargrove                          PHHargrove_at_lbl_dot_gov
    Future Technologies Group                 Tel: +1-510-495-2352
    HPC Research Department                   Fax: +1-510-486-6900
    Lawrence Berkeley National Laboratory     

  • Next message: Andreev Nikita: "Re[2]: Unanswered questions"