Flag Day - GCC4 and F77

Contributed by weerd on 2010-05-27 from the bigger-binaries dept.

Recently, there has been a lot of work going into the tree to move some architectures to GCC4 (the latest non-GPL3 version). Those of you following source-changes@ and/or ports-changes@ will have seen the large amount of commits to prepare for this new compiler. Just today, Mark Kettenis (kettenis@) committed the bits to switch the amd64 and sparc64 architectures over to GCC4.

For the people following -current, please be advised that this is a flag day, you will need to follow the steps in the faq to upgrade your machine. There was another one about config(8) needing an update very recently and soon more architectures will follow amd64 and sparc64, so keep a close eye on that page if you're on an architecture that hasn't moved yet.

In addtion to the move to GCC4, FORTRAN f77 is moving out of the src tree into the ports tree. Some of the reasons were stated by Marc Espie (espie@) on the ports@ mailing list.

Update: Marco Peereboom requests more people help test the move to GCC4. Interested people should check out his website where he collects results and which describes the testing procedure. If you want to send him your results, send them to marco@openbsd.org.

(Comments are closed)

Comments

By Marc Espie (espie) espie@nerim.net on 2010-05-27 08:28

Note that it is really, really important people start playing with that when they can.

Specifically, the sparc64 and amd64 architectures have been selected because they benefit a bit more from gcc4, but there are other architectures in the queue... they're *all* depending on test results from amd64/sparc64 proving that the change is okay...

and this includes selecting your favorite apps and giving them a run.

I don't have much illusion about that. Actual developers are still catching bugs better than most of our user, but still, one can dream...
Comments
1. By Scot Bontrager (cloudkucooland) on 2010-05-27 17:42 http://www.indievisible.org/
  
  > Note that it is really, really important people start playing with that when they can.
  >
  > Specifically, the sparc64 and amd64 architectures have been selected because they benefit a bit more from gcc4, but there are other architectures in the queue... they're *all* depending on test results from amd64/sparc64 proving that the change is okay...
  >
  > and this includes selecting your favorite apps and giving them a run.
  >
  > I don't have much illusion about that. Actual developers are still catching bugs better than most of our user, but still, one can dream...
  
  Been running a gcc4/amd64 build for 24 hours now with no issues. Happened to put a new drive in yesterday and it seems slow, but I think that's the drive and not the build. I have seen systat reporting livelocks on my em interface. I'll watch and report.
By jirib (jirib) jirib@mailinator.com on 2010-05-27 08:48 pcc

what about pcc compiler?
Comments
1. By Medina Clavijo (ventejuy) on 2010-05-27 11:05 http://ventejuy.es
  
  > what about pcc compiler?
  
  +1 pcc
2. By phessler (phessler) on 2010-05-27 13:44 http://theapt.org
  
  > what about pcc compiler?
  
  pcc doesn't do c++, which was one of the primary reasons for this upgrade.
3. By Marc Espie (espie) on 2010-05-27 15:00
  
  > what about pcc compiler?
  
  pcc might happen when people spend enough time making it work.
  
  You have little ideas how many hours were sunk in the gcc4 update. And that's a comparatively small amount of time compared to what would be needed to get pcc into shape.
  
  Yeah, pcc is all cool and that, but unless a lot of actual people with madz programming skills jump in and start coding, it won't happen anytime in the near future.
  Comments
  1. By Ted Walther (TedWalther) on 2010-05-30 22:25 http://reactor-core.org/
    
    > Yeah, pcc is all cool and that, but unless a lot of actual people with madz programming skills jump in and start coding, it won't happen anytime in the near future.
    
    Ragge must have been watching this thread; the past few days he's really gotten busy with bug-fixes and optimizations.
By Anonymous Coward (zdw) zdw@artisancomputer.com on 2010-05-27 16:09 http://artisancomputer.com

If the goal is performance and license compatibility, LLVM is another good alternative to GCC:

http://llvm.org/

Actively worked on by a larger community than pcc.
Comments
1. By Matthew Szudzik (mszudzik) on 2010-05-28 20:27
  > If the goal is performance and license compatibility, LLVM is another good alternative to GCC:
  >
  > http://llvm.org/
  >
  > Actively worked on by a larger community than pcc.
  The relative merits of LLVM, GCC, and PCC have been discussed repeatedly. For example, see
  http://undeadly.org/cgi?action=article&sid=20090623180311&pid=48 http://undeadly.org/cgi?action=article&sid=20080629133049
  Comments
  1. By Medina Clavijo (ventejuy) on 2010-05-30 11:57 http://ventejuy.es
    
    > > If the goal is performance and license compatibility, LLVM is another good alternative to GCC:
    > >
    > > http://llvm.org/
    > >
    > > Actively worked on by a larger community than pcc.
    >
    >
    > The relative merits of LLVM, GCC, and PCC have been discussed repeatedly. For example, see
    >
    > http://undeadly.org/cgi?action=article&sid=20090623180311&pid=48
    > http://undeadly.org/cgi?action=article&sid=20080629133049
    
    Very good articles. I think PCC KISS phisolophy is a very valuable thing.
    
    Comments
    
    By Peter Ljung (peter_ljung) on 2010-06-01 14:47 http://www.lounge.se
    
    I suppose one attractive aspect of LLVM/Clang is future (LLVM 2.8) support for ARMv7 architecture. GCC has not included ARMv7 until 4.3 release as I understand.
    
    LLVM 2.8 Release Notes
    
    GCC 4.3 Release Series
    
    Would be great to see OpenBSD on platforms like Beagle Board, Gumstix and future ARM cortex based laptops and servers some day ...
    
    Comments
    
    By Nima Hoda (Nima) on 2010-06-01 21:18
    
    > I suppose one attractive aspect of LLVM/Clang is future (LLVM 2.8) support for ARMv7 architecture. GCC has not included ARMv7 until 4.3 release as I understand.
    >
    > Would be great to see OpenBSD on platforms like Beagle Board, Gumstix and future ARM cortex based laptops and servers some day ...
    
    If I understand correctly, ARMv7 maintains compatibility with ARMv5, so I don't think specific support is necessary.
    
    There has been work done on making a beagleboard port (mostly by Dale Rahn, I believe):
    
    http://www.openbsd.org/cgi-bin/cvsweb/src/sys/arch/arm/armv7/
    http://www.openbsd.org/cgi-bin/cvsweb/src/sys/arch/beagle/
    
    From beagle/conf/GENERIC:
    # CPU options
    options CPU_ARMv7 # Support the ARMv7
    makeoptions CPUFLAGS="-mcpu=armv5" # dont have gcc v7 support yet.
    
    -Nima

Latest Articles

Thu, Apr 18
- 05:05 Coming soon to a -current system near you: parallel raw IP input (0)
Wed, Apr 17
- 05:33 In -current, default write format for tar(1) changed to "pax" (10)
Wed, Apr 10
- 18:50 OpenSMTPD 7.5.0p0 Released (2)
Tue, Apr 09
- 04:49 20 years since "and we're just starting": undeadly.org turns 20 (2024-04-09) (13)
Fri, Apr 05
- 06:16 OpenBSD 7.5 released (3)
Thu, Mar 28
- 18:18 LibreSSL 3.8.4 and 3.9.1 released (0)
Tue, Mar 12
- 06:53 OpenSSH 9.7/9.7p1 released! (0)
Mon, Mar 11
- 11:28 Game of Trees 0.97 released (0)
Sun, Mar 10
- 09:06 LibreSSL versions 3.8.3 and 3.9.0 released (0)

Credits

Copyright © 2004-2008 Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to April 2nd 2004 as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]