<html><head></head><body><div class="ydpa928af02yahoo-style-wrap" style="font-family: Helvetica Neue, Helvetica, Arial, sans-serif; font-size: 13px;"><div></div>
<div dir="ltr" data-setdir="false"><div><div dir="ltr" style="font-family: Helvetica Neue, Helvetica, Arial, sans-serif;">Hi, Sorry for spamming, but I noticed that my previous two emails are not being received by my yahoo email and the email is being treated as an attachement (lists.llvm.org archives says An HTML attachment was scrubbed...)</div><div dir="ltr" style="font-family: Helvetica Neue, Helvetica, Arial, sans-serif;"><br clear="none"></div><div dir="ltr" style="font-family: Helvetica Neue, Helvetica, Arial, sans-serif;">I posted this patch: https://reviews.llvm.org/D70688 [PGO] Add Value Profiling for Loop Trip Count (WIP).</div><div style="font-family: Helvetica Neue, Helvetica, Arial, sans-serif;">I'm looking for reviewers please. Thanks.</div></div></div>
</div><div id="yahoo_quoted_5509749505" class="yahoo_quoted">
<div style="font-family:'Helvetica Neue', Helvetica, Arial, sans-serif;font-size:13px;color:#26282a;">
<div>
On Wednesday, November 27, 2019, 01:42:30 p.m. EST, Wael Yehia <wmyehia2001@yahoo.com> wrote:
</div>
<div><br></div>
<div><br></div>
<div><div id="yiv1782621956"><div><div class="yiv1782621956ydp838bdb1ayahoo-style-wrap" style="font-family:Helvetica Neue, Helvetica, Arial, sans-serif;font-size:13px;"><div></div>
<div dir="ltr"><div><div dir="ltr">Hi, Sorry for spamming, but I noticed that my previous two emails are not being received by my yahoo email and the email is being treated as an attachement (lists.llvm.org archives says <span>An HTML attachment was scrubbed...)</span></div><div dir="ltr"><br clear="none"></div><div dir="ltr"> I posted this patch: https://reviews.llvm.org/D70688 [PGO] Add Value Profiling for Loop Trip Count (WIP).</div><div>I'm looking for reviewers please. Thanks.</div></div><br clear="none"></div><div><br clear="none"></div>
</div><div class="yiv1782621956yahoo_quoted" id="yiv1782621956yahoo_quoted_5768378425">
<div style="font-family:'Helvetica Neue', Helvetica, Arial, sans-serif;font-size:13px;color:#26282a;">
<div class="yiv1782621956yqt6965248776" id="yiv1782621956yqt45908"><div>
On Thursday, November 21, 2019, 06:32:21 a.m. EST, Bardia Mahjour via llvm-dev <llvm-dev@lists.llvm.org> wrote:
</div>
<div><br clear="none"></div>
<div><br clear="none"></div>
<div><div id="yiv1782621956"><div><p><font size="2">-----------------</font><br clear="none"><font size="2">Wed, Nov 20, 2019:</font><br clear="none"><font size="2">-----------------</font><br clear="none"><br clear="none"><font size="2">Loop Trip Count Value Profiling - Wael Yehia</font><br clear="none"><font size="2"> - Proposal is to associate value profiles to each loop using the </font><br clear="none"><font size="2"> llvm.loop metadata. </font><br clear="none"><font size="2"> - Suggestion for consideration: maybe value profiling should be done</font><br clear="none"><font size="2"> more generally on branches (as an extension to the presented proposal).</font><br clear="none"><br clear="none"><font size="2">LoopDeletion</font><br clear="none"><font size="2"> - Motivation: some passes cause/create empty loops, then instead of </font><br clear="none"><font size="2"> running loop deletion as a pass for all loops we may want to only</font><br clear="none"><font size="2"> run it on loop nests that contain empty ones.</font><br clear="none"><font size="2"> - Could we refactor it to create a loop-deletion utility?</font><br clear="none"><font size="2"> - Loop deletion has some special hand shake with the old pass manager. </font><br clear="none"><font size="2"> Need to detect if new pass manager is in use (via checking updater </font><br clear="none"><font size="2"> pointer).</font><br clear="none"><br clear="none"><font size="2">Loop Analysis</font><br clear="none"><font size="2"> - It appears that invalidating the LoopInfo analysis in a loop pass </font><br clear="none"><font size="2"> does not cause the analysis to be rerun on next query.</font><br clear="none"><font size="2"> - This may be because it is not a well-defined concept for the loop</font><br clear="none"><font size="2"> pass manager to be invalidating the structure of the loop in the</font><br clear="none"><font size="2"> middle of iterating over it.</font><br clear="none"><br clear="none"><font size="2">Order of subloops and basic blocks in LoopInfo</font><br clear="none"><font size="2"> - LoopInfoBase::analyze tries to put loops and basic blocks in </font><br clear="none"><font size="2"> program order, however update functions don't preserve this order.</font><br clear="none"><font size="2"> - Might be a bug, but fixing it requires a lot of effort and has </font><br clear="none"><font size="2"> compile-time impact. It may be better to have the consumers that </font><br clear="none"><font size="2"> care about the order sort the blocks.</font><br clear="none"><br clear="none"><font size="2">Loop Unswitching:</font><br clear="none"><font size="2"> - when old unswitching partially unswitches it simplifies the </font><br clear="none"><font size="2"> branches, but the new unswitching doesn't do that as much. </font><br clear="none"><font size="2"> This causes SCEV to not be able to generate proper expressions</font><br clear="none"><font size="2"> in some cases. Philip has a patch to improve SCEV to handle</font><br clear="none"><font size="2"> these more complex cases.</font><br clear="none"><br clear="none"><font size="2">- Status Updates</font><br clear="none"><font size="2"> - DomTreeUpdater investigation (Kit)</font><br clear="none"><font size="2"> - no update and the problem seems to have been fixed!</font><br clear="none"><font size="2"> - Data Dependence Graph (Bardia)</font><br clear="none"><font size="2"> - New patch to do topological sort will be posted soon.</font><br clear="none"><font size="2"> - Loop Nest Analysis (Ettore)</font><br clear="none"><font size="2"> - Updated patch to be posted soon.</font><br clear="none"><br clear="none"><br clear="none"><font size="2">-----------------------------------</font><br clear="none"><font size="2">Tentative Agenda for Dec 4, 2019</font><br clear="none"><font size="2">-----------------------------------</font><br clear="none"><br clear="none"><font size="2">Scalar Evolution Education (Philip Reames)</font><br clear="none"><br clear="none"><font size="2">- Status Updates</font><br clear="none"><font size="2"> - Data Dependence Graph (Bardia)</font><br clear="none"><font size="2"> - Loop Nest Analysis (Ettore)</font><br clear="none"><font size="2"> - Follow up on multi-dimensional array indexing RFC (Siddharth)</font><br clear="none"><font size="2"> - Impact of Loop Rotation on existing passes (Min-Yih)</font><br clear="none"><font size="2"> - Bugzilla bugs update</font><br clear="none"><br clear="none"><br clear="none"><font size="2">Bardia Mahjour<br clear="none">Compiler Optimizations<br clear="none">IBM Toronto Software Lab</font><br clear="none">
</p></div></div>_______________________________________________<br clear="none">LLVM Developers mailing list<br clear="none"><a rel="nofollow" shape="rect" ymailto="mailto:llvm-dev@lists.llvm.org" target="_blank" href="mailto:llvm-dev@lists.llvm.org">llvm-dev@lists.llvm.org</a><br clear="none"><a rel="nofollow" shape="rect" target="_blank" href="https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev">https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev</a><br clear="none"></div></div>
</div>
</div></div></div></div>
</div>
</div></body></html>