Re: [Unicon-group] Unexpected garbage collections

2017-04-11 Thread Jeffery, Clint (jeffe...@uidaho.edu)
Don: the compiler is as clever as you make it, but it cannot optimize away the trapped variables that are implicit side effects of various VM instructions that it generates. It needs the cooperation of the VM in the form of alternative instructions for known "read-only" uses of certain

Re: [Unicon-group] Unexpected garbage collections

2017-04-11 Thread Don Ward
So the problem is the underlying instruction set available to the compiler: got it. Another thought on David’s problem. Is this an Icon program or a Unicon program? If the latter, would using an array instead of a table get around the trapped variable allocation? If the look-ups really are

[Unicon-group] benchmarks

2017-04-11 Thread Jeffery, Clint (jeffe...@uidaho.edu)
Thanks to everyone who sent me benchmarks. Things I've noticed: 1. Unicon 13 is generally faster than Unicon 12. Range is 5% to 400%, typical might be 70-80%. 2. We improved our calculation of CPU clock speed on Linux; need to develop one on MacOS. 3. Some Concurrent benchmarks run slower

Re: [Unicon-group] Has anyone tried to compile Unicon with MSYS2?

2017-04-11 Thread Jafar Al-Gharaibeh
Jovan, It's been since I built on Windows 7 after upgrading to Windows 10, but the process should be the same and I successfully do builds on Windows 10 on a regular basis. config64.sh is present, where did you get the sources from? svn checkout? --Jafar On Tue, Apr 11, 2017 at 1:45 PM,

Re: [Unicon-group] Unexpected garbage collections

2017-04-11 Thread David Gamey
Don, I'm not at all concerned about a little time/memory trade-off space. I seem to be caught in a bit of a catch-22 on this.  To use a list or array, I'll need to convert  the string to integer and then perform the look up. The only other l mechanisms I can think of are using ord, tables (with

Re: [Unicon-group] Unexpected garbage collections

2017-04-11 Thread Don Ward
> On 11 Apr 2017, at 18:26, Robert Heckendorn wrote: > > What kind of optimizations does the Unicon compiler have? This is a tough question (to give a short and comprehensive answer to) and I’m not sure I’m the best person to attempt it, but I’ll give it a try.

[Unicon-group] Has anyone tried to compile Unicon with MSYS2?

2017-04-11 Thread Jovan Trujillo
Hi all, I tried to compile Unicon on my Windows 7 machine using GCC and MSYS2. The make command: make WUnicon64 returns and error saying it can't find config64.sh in config/win32/gcc I tried copying config.sh to configh64.sh without much more success. Is this file missing or do I generate

Re: [Unicon-group] benchmarks

2017-04-11 Thread Bruce & Breeanna Rennie
Good morning Clinton, On the following machine Times reported below reflect averages over three executions. Expect 20 minutes for suite to run to completion. Word Size Main Memory C Compiler clockOS 32 bit 972 MB gcc 4.4.7 1.0 GHz UNIX CPU 4x Intel(R) Atom(TM) CPU N570 @

Re: [Unicon-group] Has anyone tried to compile Unicon with MSYS2?

2017-04-11 Thread Sergey Logichev
Hi Jovan!I quite successful compiled Unicon 13.0 on my Win7/Win10 machines a month ago from svn sources.I used msys\1.0 from mingw and gcc from mingw-w64 (or mingw). Suppose, your problem in some inconsistency in unicon\config directory. Try to download current svn copy of unicon. All should be