<html><head><meta http-equiv="Content-Type" content="text/html charset=utf-8"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class="">Great!<div class="">I thought I had to look at our pipeline at O0 to make sure optimized regalloc was supported (<a href="https://bugs.llvm.org/show_bug.cgi?id=33022" class="">https://bugs.llvm.org/show_bug.cgi?id=33022</a> in mind). Glad I was wrong, it saves me some time.</div><div class=""><br class=""><div><blockquote type="cite" class=""><div class="">On May 22, 2017, at 12:51 AM, Kristof Beyls <<a href="mailto:kristof.beyls@arm.com" class="">kristof.beyls@arm.com</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><div class=""><br class=""><blockquote type="cite" class="">On 22 May 2017, at 09:09, Diana Picus <<a href="mailto:diana.picus@linaro.org" class="">diana.picus@linaro.org</a>> wrote:<br class=""><br class="">Hi Quentin,<br class=""><br class="">I actually did a run with -mllvm -optimize-regalloc -mllvm<br class="">-regalloc=greedy over the weekend and the test does pass with that.<br class="">Haven't measured the compile time though.<br class=""><br class="">Cheers,<br class="">Diana<br class=""></blockquote><br class="">I also did my usual benchmarking run with the same options as Diana did above:<br class="">- Comparing against -O0 without globalisel: 2.5% performance drop, 0.8% code size improvement.<br class=""></div></div></blockquote><div><br class=""></div><div>That’s compared to 9.5% performance drop and 2.8% code size regression, without that regalloc scheme, right?</div><br class=""><blockquote type="cite" class=""><div class=""><div class="">- Comparing against -O0 without globalisel but with the above regalloc options: 5.6% performance drop, 1% code size drop.<br class=""><br class="">In summary, the measurements indicate some good improvements.<br class="">I also haven't measure the impact on compile time.<br class=""></div></div></blockquote><div><br class=""></div><div>Do you have a mean to make this measurement?</div><div>Ahmed did a bunch of compile time measurements on our side and I wanted to see if I need to put him on the hook again :).</div><br class=""><blockquote type="cite" class=""><div class=""><div class=""><br class="">Doing a few quick spot checks on the generated code, I still see some constants being created in the entry BB and stored on the stack.<br class=""></div></div></blockquote><div><br class=""></div><div>Feel free to file PRs if you think that shouldn’t happen.</div><div><br class=""></div><div>@Eric, how does it look on your side?</div><div><br class=""></div><div>Cheers,</div><div>-Quentin</div><br class=""><blockquote type="cite" class=""><div class=""><div class="">I haven't tried to investigate if the entry BB seems like a good place to materialize those remaining constants.<br class=""><br class="">Thanks,<br class=""><br class="">Kristof<br class=""></div></div></blockquote></div><br class=""></div></body></html>