<div dir="ltr">Here're the numbers with max_app = 10 on both branches for btree and fib:<br><br>--- master ---<br><br>btree<br>Compile to livesets<br>runtime: 6m39s, gctime: 28.0s, systime: 2.7s.<br>Axioms: 0, Defs: 0, Disk: 0, Orcl: 0, Prims: 709909286; Total: 709909286<br><br>607 oracles<br><br>Repeat compilation with oracle<br>runtime: 29m39s, gctime: 2m13s, systime: 10.5s.<br>Axioms: 0, Defs: 0, Disk: 0, Orcl: 0, Prims: 3239510671; Total: 3239510671<br><br>fib<br>Compile to livesets<br>runtime: 6m36s, gctime: 26.1s, systime: 2.7s.<br>Axioms: 0, Defs: 0, Disk: 0, Orcl: 0, Prims: 704732789; Total: 704732789<br><br>597 oracles<br><br>Repeat compilation with oracle<br>runtime: 25m19s, gctime: 1m01s, systime: 5.5s.<br>Axioms: 0, Defs: 0, Disk: 0, Orcl: 0, Prims: 2618410506; Total: 2618410506<br><br>--- clos_stubs ---<br><br>btree<br>Compile to livesets<br>runtime: 4m01s, gctime: 6.1s, systime: 0.55600s.<br>Axioms: 0, Defs: 0, Disk: 0, Orcl: 0, Prims: 446369571; Total: 446369571<br><br>551 oracles<br><br>Repeat compilation with oracle<br>runtime: 19m24s, gctime: 1m52s, systime: 9.7s.<br>Axioms: 0, Defs: 0, Disk: 0, Orcl: 0, Prims: 2070001412; Total: 2070001412<br><br>fib<br>Compile to livesets<br>runtime: 3m58s, gctime: 6.3s, systime: 0.64800s.<br>Axioms: 0, Defs: 0, Disk: 0, Orcl: 0, Prims: 441178076; Total: 441178076<br><br>541 oracles<br><br>Repeat compilation with oracle<br>runtime: 16m29s, gctime: 1m33s, systime: 7.8s.<br>Axioms: 0, Defs: 0, Disk: 0, Orcl: 0, Prims: 1588238417; Total: 1588238417<br><br></div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Mar 7, 2017 at 11:31 AM, Yong Kiam <span dir="ltr"><<a href="mailto:tanyongkiam@gmail.com" target="_blank">tanyongkiam@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">I'll try and leave two of the benchmarks with max_app set to 10 before I go to lectures (still rebuilding on master).<br></div><div class="HOEnZb"><div class="h5"><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Mar 7, 2017 at 11:24 AM, Magnus Myreen <span dir="ltr"><<a href="mailto:magnus.myreen@gmail.com" target="_blank">magnus.myreen@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">What are we measuring? I'm keen to know how high max_app can be. I suspect we need it to be 10 in order to avoid generating bad applications when compiling the compiler. -- Magnus<div class="m_3056712115378551128HOEnZb"><div class="m_3056712115378551128h5"><br><div class="gmail_quote"><div dir="ltr">On Tue, 7 Mar 2017 at 16:28, Scott Owens <<a href="mailto:S.A.Owens@kent.ac.uk" target="_blank">S.A.Owens@kent.ac.uk</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hold off for now, I just thought of another easy improvement.<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
Scott<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> On 2017/03/07, at 15:12, Yong Kiam <<a href="mailto:tanyongkiam@gmail.com" class="m_3056712115378551128m_5782559481199725747gmail_msg" target="_blank">tanyongkiam@gmail.com</a>> wrote:<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> I can get you a before/after test on master and clos_stub on one of the benchmarks compiled in the logic.<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> (will take awhile though)<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> On Tue, Mar 7, 2017 at 10:04 AM, Scott Owens <<a href="mailto:S.A.Owens@kent.ac.uk" class="m_3056712115378551128m_5782559481199725747gmail_msg" target="_blank">S.A.Owens@kent.ac.uk</a>> wrote:<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> I should clarify that this is on the clos_stub branch, which I haven’t merged back to master.<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> Scott<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> > On 2017/03/07, at 15:00, Scott Owens <<a href="mailto:S.A.Owens@kent.ac.uk" class="m_3056712115378551128m_5782559481199725747gmail_msg" target="_blank">S.A.Owens@kent.ac.uk</a>> wrote:<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> ><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> > I’ve just finished the first improvement on reducing the number of stubs. I’d like to see where we are before trying to get further improvements. Can you re-try this, or even better, tell me what you were doing to get these numbers. Can we get anything meaningful without running the bootstrap (and the attendant 2 day wait)?<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> ><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> > Scott<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> ><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> On 2017/03/05, at 08:21, Magnus Myreen <<a href="mailto:magnus.myreen@gmail.com" class="m_3056712115378551128m_5782559481199725747gmail_msg" target="_blank">magnus.myreen@gmail.com</a>> wrote:<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> Hi all,<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> We should get rid of max_app. Reasons:<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> 2. The current compilation strategy produces too much stub code in<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> clos_to_bvl. If we really must have max_app (I argue below that<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> we don't), then it should be something like 10 or 15. With the<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> current compilation strategy, the binary produced for max_app=9 is<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> 176 kb larger than the binary for max_app=3. I wanted to try<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> compiling with max_app=10, but that was taking more than 5<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> minutes. Compiling with max_app=3 takes 7 seconds for me. I<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> believe the slow down in compilation speed with differing max_app<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> values is caused by the rapid increase in code size (number of<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> stubs in clos_to_bvl) as max_app gets larger. My measurements<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> are below while compiling factorial (and the basis library):<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=1: 10.810s (* very little room for opts in basis *)<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=2: 8.724s<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=3: 7.872s (* allows optimisations in basis code *)<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=4: 8.935s<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=5: 11.608s<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=6: 21.771s<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=7: 42.932s<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=8: 1m24.872s<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> >> max_app=9: 3m9.745s<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> ><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> > ______________________________<wbr>_________________<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> > Developers mailing list<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> > <a href="mailto:Developers@cakeml.org" class="m_3056712115378551128m_5782559481199725747gmail_msg" target="_blank">Developers@cakeml.org</a><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> > <a href="https://lists.cakeml.org/listinfo/developers" rel="noreferrer" class="m_3056712115378551128m_5782559481199725747gmail_msg" target="_blank">https://lists.cakeml.org/listi<wbr>nfo/developers</a><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> ______________________________<wbr>_________________<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> Developers mailing list<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> <a href="mailto:Developers@cakeml.org" class="m_3056712115378551128m_5782559481199725747gmail_msg" target="_blank">Developers@cakeml.org</a><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
> <a href="https://lists.cakeml.org/listinfo/developers" rel="noreferrer" class="m_3056712115378551128m_5782559481199725747gmail_msg" target="_blank">https://lists.cakeml.org/listi<wbr>nfo/developers</a><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
><br class="m_3056712115378551128m_5782559481199725747gmail_msg">
<br class="m_3056712115378551128m_5782559481199725747gmail_msg">
</blockquote></div>
</div></div></blockquote></div><br></div>
</div></div></blockquote></div><br></div>