This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: Is gcc optimized for thread level parallelism?

From: Diego Novillo <dnovillo at redhat dot com>
To: gcc at gcc dot gnu dot org
Cc: Jim Wilson <wilson at specifix dot com>, x z <hzmonte at hotmail dot com>
Date: Thu, 20 Oct 2005 22:25:38 -0400
Subject: Re: Is gcc optimized for thread level parallelism?
References: <BAY102-F324780031FE48568D50F3CA0700@phx.gbl> <43582474.3000203@specifix.com>

On Thursday 20 October 2005 19:12, Jim Wilson wrote:

> See http://gcc.gnu.org/projects/gomp/
>
> This is still in early stages of implementation.  Don't expect anything
> to work yet.
>
Indeed.

The implementation is at a stage where the C front end can be tested with 
benchmarks (openmpbench_C and NAS-OMP-C both work fine).  Fortran is 
almost there (openmpbench_F works), and C++ should be ready for wider 
testing in a few weeks.

We plan to merge the branch into mainline for the 4.2 release.

Automatic parallelization passes may be added at a later stage by having a 
pass that emits the appropriate OpenMP IL instructions.

References:
- Is gcc optimized for thread level parallelism?
  - From: x z
- Re: Is gcc optimized for thread level parallelism?
  - From: Jim Wilson

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]