<div dir="ltr"><div><div><br></div>Please ensure that you have more than just Carlo's feedback from the OMP side on this. If Intel isn't in agreement then you're leaving out the most important stakeholder. (I'd also ask that Hal weigh in since he has quite a bit of insight into the politics for this)<br><br></div><div>Pushing forward quickly (steamroller) with something just because the thread "stalled" doesn't make any sense. Please work with those who have experience and expertise.<br></div><div><br></div>I'll leave technical feedback for later after rereading the meat of proposal.<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Sat, Apr 23, 2016 at 9:50 AM, Carlo Bertolli via cfe-dev <span dir="ltr"><<a href="mailto:cfe-dev@lists.llvm.org" target="_blank">cfe-dev@lists.llvm.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><p><span class="">Hi Chandler<br><br>About your points:<br><br></span><span class="">> 3) Initially, leave the OpenMP offloading stuff targeted at OpenMP. Then, as it evolves, consider moving it to be another runtime in the broad project if and when it makes sense.<br>> 4) As both OpenMP and SE evolve and are used some in the project, evaluate whether there is a common core that makes sense to extract. If so, do it and rebase them appropriately.<br><br></span><span class="">I agree with you.<br><br>I had some discussion within my group and there seems to be some agreement that OpenMP offloading, if properly extended, could be used as a plugin for SE. As time progresses and we learn more about SE we will be able to re-evaluate this.<br><br>Thanks<br><br>-- Carlo<br><br><img src="cid:1__=8FBBF50DDF9AE5EE8f9e8a93df938690918c8FB@" alt="Inactive hide details for Chandler Carruth via cfe-dev ---04/22/2016 06:02:24 PM---I feel like this thread got a bit stalled. I" height="16" width="16" border="0"><font color="#424282">Chandler Carruth via cfe-dev ---04/22/2016 06:02:24 PM---I feel like this thread got a bit stalled. I'd like to pick it up and try to suggest a path forward.</font><br><br><font color="#5F5F5F" size="2">From: </font><font size="2">Chandler Carruth via cfe-dev <<a href="mailto:cfe-dev@lists.llvm.org" target="_blank">cfe-dev@lists.llvm.org</a>></font><br><font color="#5F5F5F" size="2">To: </font><font size="2">Mehdi Amini <<a href="mailto:mehdi.amini@apple.com" target="_blank">mehdi.amini@apple.com</a>>, Jason Henline <<a href="mailto:jhen@google.com" target="_blank">jhen@google.com</a>>, Hal Finkel <<a href="mailto:hfinkel@anl.gov" target="_blank">hfinkel@anl.gov</a>></font><br><font color="#5F5F5F" size="2">Cc: </font><font size="2">llvm-dev <<a href="mailto:llvm-dev@lists.llvm.org" target="_blank">llvm-dev@lists.llvm.org</a>>, cfe-dev <<a href="mailto:cfe-dev@lists.llvm.org" target="_blank">cfe-dev@lists.llvm.org</a>>, "<a href="mailto:openmp-dev@lists.llvm.org" target="_blank">openmp-dev@lists.llvm.org</a>" <<a href="mailto:openmp-dev@lists.llvm.org" target="_blank">openmp-dev@lists.llvm.org</a>></font><br><font color="#5F5F5F" size="2">Date: </font><font size="2">04/22/2016 06:02 PM</font><br><font color="#5F5F5F" size="2">Subject: </font><font size="2">Re: [cfe-dev] [Openmp-dev] [llvm-dev] RFC: Proposing an LLVM subproject for parallelism runtime and support libraries</font><br><font color="#5F5F5F" size="2">Sent by: </font><font size="2">"cfe-dev" <<a href="mailto:cfe-dev-bounces@lists.llvm.org" target="_blank">cfe-dev-bounces@lists.llvm.org</a>></font><br></span></p><hr style="color:#8091a5" align="left" size="2" noshade width="100%"><br><br><br><div><div class="h5"><font size="4">I feel like this thread got a bit stalled. I'd like to pick it up and try to suggest a path forward.</font><br><br><font size="4">I don't hear any real objections to the overall idea of having an LLVM subproject for parallelism runtimes and support libraries. I think we should get that created.</font><br><br><font size="4">I don't actually see any real objections to StreamExecutor being one of the runtimes. There are some interesting questions however:</font><br><font size="4">- Is there common code in the OpenMP runtime that could be unified with this?</font><br><font size="4">- Could OpenMP end up using SE or some common shared library between them as a basis for offloading?</font><br><font size="4">- Would instead it make more sense to have the OpenMP offload library be a plugin for StreamExecutor?</font><br><br><font size="4">I don't know the answer to any of these really, but I also don't think that they should prevent us from making progress here. And I think if anything, they'll become easier to answer if we do.</font><br><br><font size="4">So my suggestion would be:</font><br><font size="4">1) Create the broader scoped LLVM subproject, including writing up its charter, scope, plans, etc.</font><br><br><font size="4">2) Add stream executor to it</font><br><br><font size="4">3) Initially, leave the OpenMP offloading stuff targeted at OpenMP. Then, as it evolves, consider moving it to be another runtime in the broad project if and when it makes sense.</font><br><br><font size="4">4) As both OpenMP and SE evolve and are used some in the project, evaluate whether there is a common core that makes sense to extract. If so, do it and rebase them appropriately.</font><br><br><br></div></div><font size="4">Does this make sense? Are there objections to moving forward here?</font><tt>_______________________________________________<br>cfe-dev mailing list<br><a href="mailto:cfe-dev@lists.llvm.org" target="_blank">cfe-dev@lists.llvm.org</a><br></tt><tt><a href="http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev" target="_blank">http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev</a></tt><tt><br></tt><br><br><br>
<p></p></div>
<br>_______________________________________________<br>
cfe-dev mailing list<br>
<a href="mailto:cfe-dev@lists.llvm.org">cfe-dev@lists.llvm.org</a><br>
<a href="http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev" rel="noreferrer" target="_blank">http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-dev</a><br>
<br></blockquote></div><br></div>