Re: About barriers

From: Nikita Andreev (lestat_at_kemsu.ru)
Date: Sat Apr 18 2009 - 04:13:56 PDT

  • Next message: Benjamin Byington: "Storing Pointers to Shared in Non-UPC Environment"
    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. 
    
    
    



  • Next message: Benjamin Byington: "Storing Pointers to Shared in Non-UPC Environment"