<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=""><br class=""><div><blockquote type="cite" class=""><div class="">On May 30, 2016, at 12:02 AM, Sanjoy Das <<a href="mailto:sanjoy@playingwithpointers.com" class="">sanjoy@playingwithpointers.com</a>> wrote:</div><br class="Apple-interchange-newline"><div class=""><span style="font-family: Helvetica; font-size: 12px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: 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 have made some minor comments inline, but I still stand by my earlier comment that we should do something like this as a last resort. As an initial step we should at least evaluate how far we can we can get on relevant workloads without teaching SCEV about floating point values at all.</span><br style="font-family: Helvetica; font-size: 12px; font-style: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: 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="">Are there conceivable use cases for this infrastructure beyond vectorizing a small subcategory of loops of this form?</div><div class=""><br class=""></div><div class="">float x = init;<br class=""> for (int i=0;i<N;i++){<br class=""> A[i] = x;<br class=""> x += fp_inc; // Loop invariant variable or constant<br class=""> }</div><div class=""><br class=""></div><div class="">Can the vectorizer handle loops of this form without querying SCEV?</div><div class=""><br class=""></div><div class="">SCEV expressions have an inherent width. They are not infinite precision. This is the main challenge of working with SCEV expressions, as Sanjoy is well aware. What happens when incrementing a floating-point SCEV expression by a smaller amount than ULP. Eventually that will happen in a floating-point recurrence. Do we have to prove that floating-point recurrences behave a certain way before we can legally convert them to SCEV expressions?</div><div class=""><br class=""></div><div class="">Honestly, I’m not an expert in floating-point semantics, and I wouldn’t feel comfortable adding this to SCEV without buy-in from someone who is.</div><div class=""><br class=""></div><div class="">Andy</div></body></html>