Precise data-flow among regions is a hard problem especially if loops with function/method calls and dynamic data structures are involved. You may approximate it but you will lose quite a bit of information. If you are fine with access to arrays and dynamic data structures being approximated as accesses to single variables, scalar data-flow among regions should be good enough. If you want more precise information, you may want to look at literature on array data-flow, automatic task parallelism detection and the like.
-----Original Message----- From: Li Shengmei [mailto:lis...@gmail.com] Sent: Thursday, April 14, 2011 2:22 PM To: '"C. Bergström"' Cc: open64-devel@lists.sourceforge.net Subject: Re: [Open64-devel] About region analysis in open64 I have not detailed design. I think the region definition in open64 may be a reference. the existing research from HP (not open64), Intel (Roy Ju) and SGI -- Do you mean "a region based compilation infrastructure" written by Liu Yang et al.? I am reading that. I think the region related codes are from that paper, right? Thanks for your information about LLVM. I will have a look. :) May > -----Original Message----- > From: "C. Bergström" [mailto:cbergst...@pathscale.com] > Sent: Thursday, April 14, 2011 4:33 PM > To: Li Shengmei > Cc: 'Sun Chan'; open64-devel@lists.sourceforge.net > Subject: Re: [Open64-devel] About region analysis in open64 > > Li Shengmei wrote: > > Both if possible. > > I want to analyze the data flow among regions. > > > How do you plan to define a region? Have you looked at the existing > research from HP (not open64), Intel (Roy Ju) and SGI from the past? > LLVM also has a small amount of region based analysis and not sure if > that's of any interest. > > Thanks > > ./C ------------------------------------------------------------------------------ Benefiting from Server Virtualization: Beyond Initial Workload Consolidation -- Increasing the use of server virtualization is a top priority.Virtualization can reduce costs, simplify management, and improve application availability and disaster protection. Learn more about boosting the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev _______________________________________________ Open64-devel mailing list Open64-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/open64-devel ------------------------------------------------------------------------------ Benefiting from Server Virtualization: Beyond Initial Workload Consolidation -- Increasing the use of server virtualization is a top priority.Virtualization can reduce costs, simplify management, and improve application availability and disaster protection. Learn more about boosting the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev _______________________________________________ Open64-devel mailing list Open64-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/open64-devel
