<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">Oh, and it did pass SPEC2000 and SPEC2006, so it can't be <i>too</i> wrong, can it?<div><br class="webkit-block-placeholder"></div><div>--Owen</div><div><br><div><div>On Jan 9, 2008, at 10:44 PM, Owen Anderson wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite">You really need to address these to Evan. I asked him what to do for LiveIntervalAnalysis, and this is what he told me :-)<br><br>--Owen<br><br>On Jan 9, 2008, at 10:08 PM, Chris Lattner wrote:<br><br><blockquote type="cite"><blockquote type="cite">URL: <a href="http://llvm.org/viewvc/llvm-project?rev=45815&view=rev">http://llvm.org/viewvc/llvm-project?rev=45815&view=rev</a><br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">Log:<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">Don't use LiveVariables::VarInfo::DefInst.<br></blockquote></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">Cool, question though:<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite"><blockquote type="cite">+++ llvm/trunk/lib/CodeGen/LiveIntervalAnalysis.cpp Wed Jan 9<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">21:12:54 2008<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">@@ -309,7 +309,14 @@<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite"> // are actually two values in the live interval. Because of<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">this we<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite"> // need to take the LiveRegion that defines this register and<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">split it<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite"> // into two values.<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">- unsigned DefIndex =<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">getDefIndex(getInstructionIndex(vi.DefInst));<br></blockquote></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">In the old code, when it introduces multiple definitions of a vreg,<br></blockquote><blockquote type="cite">did it clear DefInst? If so, this code can only be triggered for<br></blockquote><blockquote type="cite">vregs with a single def.<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite"><blockquote type="cite"><br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">+ MachineRegisterInfo& MRI = mbb->getParent()->getRegInfo();<br></blockquote></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">I think that MRI should become and ivar in liveintervalanalysis.<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite"><blockquote type="cite"><br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">+ unsigned lowIndex = ~0U;<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">+ for (MachineRegisterInfo::def_iterator DI =<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">MRI.def_begin(interval.reg),<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">+ DE = MRI.def_end(); DI != DE; ++DI)<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">+ if (getInstructionIndex(&*DI) < lowIndex)<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">+ lowIndex = getInstructionIndex(&*DI);<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">+<br></blockquote></blockquote><blockquote type="cite"><blockquote type="cite">+ unsigned DefIndex = getDefIndex(lowIndex);<br></blockquote></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">It would be really nice if you could drop the loop. It's unclear why<br></blockquote><blockquote type="cite">it's right anyway, the numbering is not necessarily contiguous across<br></blockquote><blockquote type="cite">code...<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">-Chris<br></blockquote><blockquote type="cite"><br></blockquote><blockquote type="cite">_______________________________________________<br></blockquote><blockquote type="cite">llvm-commits mailing list<br></blockquote><blockquote type="cite"><a href="mailto:llvm-commits@cs.uiuc.edu">llvm-commits@cs.uiuc.edu</a><br></blockquote><blockquote type="cite"><a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits">http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits</a><br></blockquote><br>_______________________________________________<br>llvm-commits mailing list<br><a href="mailto:llvm-commits@cs.uiuc.edu">llvm-commits@cs.uiuc.edu</a><br>http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits<br></blockquote></div><br></div></body></html>