<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> <meta name="Generator" content="Microsoft Exchange Server">  <style></style> </head> <body> <font face="Calibri" size="2"><span style="font-size:10pt;"> <div style="padding-right:5pt;padding-left:5pt;"><font color="blue">[AMD Official Use Only - AMD Internal Distribution Only]<br> </font></div> <div style="margin-top:5pt;"><font face="Times New Roman" size="3"><span style="font-size:12pt;"><br> </span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;"><snipped></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;"> </span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > ></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > Thank you Mattias for the comments and question, please let me try</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > to explain the same below</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > ></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > >> We shouldn't have a separate CPU/cache hierarchy API instead?</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > ></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > Based on the intention to bring in CPU lcores which share same L3</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > (for better cache hits and less noisy neighbor) current API focuses</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > on using</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > ></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > Last Level Cache. But if the suggestion is `there are SoC where L2</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > cache are also shared, and the new API should be provisioned`, I am</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > also</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > ></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > > comfortable with the thought.</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > ></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> ></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > Rather than some AMD special case API hacked into <rte_lcore.h>, I</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > think we are better off with no DPDK API at all for this kind of functionality.</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> ></span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > A DPDK CPU/memory hierarchy topology API very much makes sense, but it</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> > should be reasonably generic and complete from the start.</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> </span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> Agreed. This one of those cases where the existing project hwloc which is part</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> of open-mpi is more complete and well supported. It supports multiple OS's</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> and can deal with more quirks.</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;"> </span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">Thank you Stephen for the inputs, last year when checked hwloc for distros there were anomalies for NUMA and Physical socket Identification on AMD EPYC Soc.</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">I will recheck the distros version of hwloc, if these work out fine I will re-work with hwloc libraries making it OS independent too.</span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;"> </span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> </span></font></div> <div><font face="Calibri" size="2"><span style="font-size:11pt;">> <a href="https://github.com/open-mpi/hwloc">https://github.com/open-mpi/hwloc</a></span></font></div> </span></font> </body> </html>