<html><head></head><body><div style="color:#000; background-color:#fff; font-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, sans-serif;font-size:10px"><div id="yui_3_16_0_ym19_1_1466850166033_31632"><span></span></div> <div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><font size="3">Thanks for clarifying.</font></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><font size="3" id="yui_3_16_0_ym19_1_1466850166033_31978">The example I was using is the SpecInt-2006 astar benchmark with rivers.cfg & rivers.bin file. </font></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><font size="3"><br></font></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><font size="3" id="yui_3_16_0_ym19_1_1466850166033_31979">As the loopmicroopbuffersize is 24 & 16 for m1 & kryo, so was expecting the effect on loop unroll pass, as this factor is considered as partial unrolling factor while unrolling. </font></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><font size="3"><br></font></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><font size="3" id="yui_3_16_0_ym19_1_1466850166033_31980">Thanks,</font></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><font size="3">Pankaj</font></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><font size="3"><br></font></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><br></div><div class="qtdSeparateBR" id="yui_3_16_0_ym19_1_1466850166033_31784"><br><br></div><div class="yahoo_quoted" id="yui_3_16_0_ym19_1_1466850166033_31777" style="display: block;"> <div style="font-family: HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, sans-serif; font-size: 10px;" id="yui_3_16_0_ym19_1_1466850166033_31776"> <div style="font-family: HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, sans-serif; font-size: 16px;" id="yui_3_16_0_ym19_1_1466850166033_31775"> <div dir="ltr"><font size="2" face="Arial"> On Saturday, June 25, 2016 4:18 PM, Renato Golin <renato.golin@linaro.org> wrote:<br></font></div> <br><br> <div class="y_msg_container" id="yui_3_16_0_ym19_1_1466850166033_31774">On 25 June 2016 at 10:50, Pankaj Gode via cfe-dev<br clear="none"><<a shape="rect" ymailto="mailto:cfe-dev@lists.llvm.org" href="mailto:cfe-dev@lists.llvm.org" id="yui_3_16_0_ym19_1_1466850166033_31783">cfe-dev@lists.llvm.org</a>> wrote:<br clear="none">> While using clang driver to create binary for aarch64 subtarget,<br clear="none">> do we need to give any other flag apart from '--target=aarch64-linux-gnu<br clear="none">> -mcpu=<> -mtune=<>' for using a AArch64's subtarget scheduling model?<br clear="none"><br clear="none">Nope, that should be enough.<div class="yqt1398412378" id="yqtfd12023"><br clear="none"><br clear="none"><br clear="none">> I see same executable (in size & contents) generated for subtargets of<br clear="none">> aarch64 such as cortex-a57, exynos-m1 & kryo even after replacing the cpu<br clear="none">> names. Should I include some other flag ?</div><br clear="none"><br clear="none">Both M1 and Kryo have different scheduling parameters, but not<br clear="none">radically so. It's possible that your code is not hitting any of those<br clear="none">differences.<br clear="none"><br clear="none">Perhaps if you could provide an example of what the code is and how<br clear="none">you expected it to be in the assembly output (maybe comparing with<br clear="none">GCC's output), we'd know what is not being done.<br clear="none"><br clear="none">cheers,<br clear="none">-renato<div class="yqt1398412378" id="yqtfd20928"><br clear="none"></div><br><br></div> </div> </div> </div></div></body></html>