<html><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">Yeah, take a look at Grue-O0 i386:<div><br></div><div><span class="Apple-style-span" style="font-family: -webkit-monospace; font-size: 11px; ">LLC compile:<br>multisource/Applications/ClamAV/clamscan: 6.94% (4.4268 => 4.1196)<br>multisource/Applications/JM/lencod/lencod: 8.37% (5.3160 => 4.8711)<br>multisource/Applications/SPASS/SPASS: 6.11% (6.0424 => 5.6731)<br>multisource/Applications/sqlite3/sqlite3: 7.14% (3.3431 => 3.1043)<br>multisource/Benchmarks/MiBench/consumer-typeset/consumer-typeset: 7.29% (5.2923 => 4.9067)<br>multisource/Benchmarks/tramp3d-v4/tramp3d-v4: 5.36% (9.2806 => 8.7834)<br>external/Povray/povray: 7.55% (5.1237 => 4.7371)<br>external/SPEC/CFP2000/177.mesa/177.mesa: 7.65% (3.9692 => 3.6656)<br>external/SPEC/CINT2000/176.gcc/176.gcc: 7.18% (13.7394 => 12.7527)<br>external/SPEC/CINT2000/252.eon/252.eon: 6.20% (3.7286 => 3.4973)<br>external/SPEC/CINT2000/253.perlbmk/253.perlbmk: 6.74% (5.2106 => 4.8592)<br>external/SPEC/CINT2000/254.gap/254.gap: 7.38% (4.8105 => 4.4556)<br>external/SPEC/CINT2000/255.vortex/255.vortex: 7.29% (4.2463 => 3.9369)<br>external/SPEC/CINT2006/400.perlbench/400.perlbench: 6.96% (10.3442 => 9.6239)<br>external/SPEC/CINT2006/403.gcc/403.gcc: 6.88% (32.7735 => 30.5199)<br>external/SPEC/CINT2006/445.gobmk/445.gobmk: 7.32% (6.7599 => 6.2651)<br>external/SPEC/CINT2006/464.h264ref/464.h264ref: 8.21% (3.8471 => 3.5311)</span></div><div><font class="Apple-style-span" face="-webkit-monospace" size="3"><span class="Apple-style-span" style="font-size: 11px;"><br></span></font></div><div><font class="Apple-style-span" face="-webkit-monospace" size="3"><span class="Apple-style-span" style="font-size: 11px;">Very nice!</span></font></div><div><font class="Apple-style-span" face="-webkit-monospace" size="3"><span class="Apple-style-span" style="font-size: 11px;"><br></span></font></div><div><font class="Apple-style-span" face="-webkit-monospace" size="3"><span class="Apple-style-span" style="font-size: 11px;">Evan</span></font></div><div><font class="Apple-style-span" face="-webkit-monospace" size="3"><span class="Apple-style-span" style="font-size: 11px;"><br></span></font><div><div>On Jul 8, 2008, at 12:28 AM, Duncan Sands wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite"><div>Hi Dan,<br><br><blockquote type="cite">This drastically reduces the number of calls to malloc/free made<br></blockquote><blockquote type="cite">during the "Emit" phase of scheduling, as well as later phases<br></blockquote><blockquote type="cite">in CodeGen. Combined with other changes, this speeds up the<br></blockquote><blockquote type="cite">"instruction selection" phase of CodeGen by 10% in some cases.<br></blockquote><br>which platform are you measuring this speedup on? Darwin?<br><br>Thanks,<br><br>Duncan.<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></div></blockquote></div><br></div></body></html>