<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="">Here the main problem seem to be that we don’t have a fast way to check for “is this in a certain set”.<div class=""><br class=""></div><div class="">In my opinion here the problem is that we don’t have a good (read fast) way for checking that a certain condition applies to an llvm Value.</div><div class="">I believe this bit is trying to address that.</div><div class=""><br class=""></div><div class="">Maybe there is room for improvement in LLVM set-like data structures?</div><div class="">Maybe there is a way to use a BitVector instead and generate from the Value/Instruction a unique linear ID that can be used with that?</div><div class=""><br class=""></div><div class="">Marcello</div><div class=""><div><blockquote type="cite" class=""><div class="">On 15 Sep 2015, at 14:49, Pete Cooper via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" class="">llvm-dev@lists.llvm.org</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><meta http-equiv="Content-Type" content="text/html charset=utf-8" class=""><div style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class=""><br class=""><div class=""><blockquote type="cite" class=""><div class="">On Sep 15, 2015, at 2:16 PM, Owen Anderson <<a href="mailto:resistor@mac.com" class="">resistor@mac.com</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><meta http-equiv="Content-Type" content="text/html charset=utf-8" class=""><div style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class=""><br class=""><div class=""><blockquote type="cite" class=""><div class="">On Sep 14, 2015, at 5:02 PM, Mehdi Amini via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" class="">llvm-dev@lists.llvm.org</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><blockquote type="cite" style="font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class=""><br class="Apple-interchange-newline">On Sep 14, 2015, at 2:58 PM, Pete Cooper via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" class="">llvm-dev@lists.llvm.org</a>> wrote:<br class=""><br class=""><br class=""><blockquote type="cite" class="">On Sep 14, 2015, at 2:49 PM, Matt Arsenault via llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" class="">llvm-dev@lists.llvm.org</a>> wrote:<br class=""><br class="">On 09/14/2015 02:47 PM, escha via llvm-dev wrote:<br class=""><blockquote type="cite" class="">I would assume that it’s just considered to be garbage. I feel like any sort of per-pass side data like this should come with absolute minimal contracts, to avoid introducing any more inter-pass complexity.<br class=""></blockquote>I would think this would need to be a verifier error if it were ever non-0<br class=""></blockquote>+1<br class=""><br class="">Otherwise every pass which ever needs this bit would have to first zero it out just to be safe, adding an extra walk over the whole functions.<br class=""><br class="">Of course otherwise the pass modifying it will have to zero it, which could also be a walk over the whole function. So either way you have lots iterate over, which is why i’m weary of this approach tbh.<br class=""></blockquote><br style="font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class=""><span style="font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">Every pass which ever uses this internally would have to set it to zero when it is done, adding an extra walk over the whole functions as you noticed. This goes against “you don’t pay for what you don’t use”, so definitively -1 for this. Better to cleanup before use.</span><br style="font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class=""><span style="font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px; float: none; display: inline !important;" class="">I agree that the approach does not scale/generalize well, and we should try to find an alternative if possible. Now *if* it is the only way to improve performance significantly, we might have to weight the tradeoff.</span><br style="font-family: Helvetica; font-size: 12px; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px;" class=""></div></blockquote></div><br class=""><div class="">Does anyone have any concrete alternative suggestions to achieve the speedup demonstrated here?</div></div></div></blockquote>Honestly, not an alternative to speedup ADCE. This appears to be the fastest way to improve that particular pass.</div><div class=""><br class=""></div><div class="">My worry here comes from prior experience. I have done this exactly change years ago on a non-LLVM compiler. Adding one bit to each instruction was great, we used it in a few passes, then we found we needed a second bit, then a third. That quickly became unmanageable, and if we add this bit then we put ourselves on a repeat of that.</div><div class=""><br class=""></div><div class="">I think we should weigh up what this change gets us in overall compile time. A 30% speedup in ADCE is great, and if a few other passes would benefit from this and get similar speedups, then even better. However, if thats only worth say 1% of overall compile time, then i’d argue its not worth it. </div><div class=""><br class=""></div><div class="">Pete<br class=""><blockquote type="cite" class=""><div class=""><div style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class=""><div class=""><br class=""></div><div class="">—Owen</div></div></div></blockquote></div><br class=""></div>_______________________________________________<br class="">LLVM Developers mailing list<br class=""><a href="mailto:llvm-dev@lists.llvm.org" class="">llvm-dev@lists.llvm.org</a><br class="">http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev<br class=""></div></blockquote></div><br class=""></div></body></html>