Project

General

Profile

Support #151

Support #150: Align -optimize flags with current compiler releases

Align -optimize flags with current release of GNU compilers

Added by Phil Miller over 6 years ago. Updated about 1 year ago.

Status:
New
Priority:
Normal
Category:
-
Target version:
-
Start date:
Due date:
% Done:

0%


Description

We provide a generic wrapper for various backend compilers in charmc, and a flag -optimize that's meant to provide good optimization settings for any given backend without breaking things. Every new compiler release brings new optimization settings along with it. We haven't necessarily kept up with these releases and the options they bring to the table.

We should make sure that the flags we pass GNU compilers under -optimize are appropriate for the most recent few releases (probably 4.4-4.8, given how long they stay in service), in terms of performance, reliability, and compatibility.

History

#1 Updated by Eric Bohm over 6 years ago

  • Assignee set to Ramprasad Venkataraman

#2 Updated by Eric Bohm almost 6 years ago

  • Assignee changed from Ramprasad Venkataraman to Michael Robson

#3 Updated by Eric Bohm almost 3 years ago

  • Tracker changed from Bug to Support

#4 Updated by Sam White about 1 year ago

The GCC documentation doesn't specify very many optimization options that aren't enabled by default at -O3 and which don't relax semantics. I did find a couple:

-fipa-pta

Perform interprocedural pointer analysis and interprocedural modification and reference analysis. This option can cause excessive memory and compile-time usage on large compilation units. It is not enabled by default at any optimization level.

-flto

Enable link-time optimization. Already have an open issue for that here: https://charm.cs.illinois.edu/redmine/issues/1817

Also available in: Atom PDF