may be Gautum can help? Sun On Sat, Jan 15, 2011 at 9:12 AM, Mike Murphy <mmur...@nvidia.com> wrote: > I don't have time to try and build and test the change, but looking at the > source diffs they look okay to me. I assume your basic smoke testing > included generating multiple targets successfully? > > -----Original Message----- > From: Sun Chan [mailto:sun.c...@gmail.com] > Sent: Thursday, January 13, 2011 10:31 PM > To: David Coakley > Cc: open64-devel > Subject: Re: [Open64-devel] patch for statically linked target scheduling > info (repost) > > I am hoping Mike would have time, he is better qualified to do that. > May be some other people also? > Sun > > On Fri, Jan 14, 2011 at 2:28 PM, David Coakley <dcoak...@gmail.com> wrote: >> I am still looking for a gatekeeper to review my changes to statically >> link the target scheduling info. >> >> I would like to thank Gang Yu for testing the patch and reporting some >> problems exposed by the SL target. >> >> An updated patch is attached. Compared to the original patch, the changes >> are: >> >> In targ_info/generate/si_gen.cxx: >> >> - Rewrite the incorrect "res_id_set_cmp" comparison function. >> - Improve the error handling in TOP_SCHED_INFO_MAP::Output_Data. >> >> In targ_info/Makefile.gbase: >> >> - Fix errors in the PROC_SI_OBJS lists. >> - Add missing dependencies for correct make -j builds. >> >> Also, I corrected some typos in the targ_info/proc/SL/*.cxx files. >> >> -David Coakley / AMD Open Source Compiler Engineering >> >> On Wed, Jan 5, 2011 at 1:46 PM, David Coakley <dcoak...@gmail.com> wrote: >>> Hi all, >>> >>> I have completed the implementation for my proposal to remove the >>> scheduling info DSO files and include that data directly into the >>> backend. Here is the thread with the original proposal for reference: >>> >>> http://sourceforge.net/mailarchive/forum.php?thread_name=AANLkTik3EfLKqAyXbLnhrBAKMY1VfCiWYd9ssTnzx2oY%40mail.gmail.com&forum_name=open64-devel >>> >>> I'm now looking for someone to review/approve the patch. If there's >>> anyone familiar with the targ_info subsystem and si_gen.cxx, that's >>> where the majority of the changes are. >>> >>> A few notes: >>> >>> 1) Along with some basic smoke testing, I tested that the same code is >>> generated for x86-64 when building some non-trivial programs. I do >>> not expect any changes in the output of the compiler as a result of >>> this patch. >>> 2) I also checked the on-disk footprint and compilation time, and saw >>> no net change in either one. >>> 3) In osprey/common/targ_info/Makefile.gbase, I removed rules related >>> to files that do not exist in the source tree, e.g. pentium.so. If >>> someone can contribute a pentium_si.cxx, then I'm happy to add the >>> appropriate rules. Also, the use of the same filenames in the MIPS/SL >>> proc subdirectories looks sloppy -- do we really need all the sl* >>> files in the MIPS directory or r10000_si.cxx in the SL directory? >>> >>> -David Coakley / AMD Open Source Compiler Engineering >>> >> >> ------------------------------------------------------------------------------ >> Protect Your Site and Customers from Malware Attacks >> Learn about various malware tactics and how to avoid them. Understand >> malware threats, the impact they can have on your business, and how you >> can protect your company and customers by using code signing. >> http://p.sf.net/sfu/oracle-sfdevnl >> _______________________________________________ >> Open64-devel mailing list >> Open64-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/open64-devel >> >> > > ------------------------------------------------------------------------------ > Protect Your Site and Customers from Malware Attacks > Learn about various malware tactics and how to avoid them. Understand > malware threats, the impact they can have on your business, and how you > can protect your company and customers by using code signing. > http://p.sf.net/sfu/oracle-sfdevnl > _______________________________________________ > Open64-devel mailing list > Open64-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/open64-devel > ----------------------------------------------------------------------------------- > This email message is for the sole use of the intended recipient(s) and may > contain > confidential information. Any unauthorized review, use, disclosure or > distribution > is prohibited. If you are not the intended recipient, please contact the > sender by > reply email and destroy all copies of the original message. > ----------------------------------------------------------------------------------- >
------------------------------------------------------------------------------ Protect Your Site and Customers from Malware Attacks Learn about various malware tactics and how to avoid them. Understand malware threats, the impact they can have on your business, and how you can protect your company and customers by using code signing. http://p.sf.net/sfu/oracle-sfdevnl _______________________________________________ Open64-devel mailing list Open64-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/open64-devel
