You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@apr.apache.org by Blair Zajac <bl...@orcaware.com> on 2002/04/29 23:55:42 UTC

RedHat comments on APR using /usr/include/asm/atomic.h

There are some interesting comments from a RedHat developer on why
APR should not include /usr/include/asm/atomic.h on Linux platforms
at line 160 of apr/include/apr_atomic.h

http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=63643

Maybe someone more familiar with the atomic code should take a
look at it.

Best,
Blair

-- 
Blair Zajac <bl...@orcaware.com>
Web and OS performance plots - http://www.orcaware.com/orca/

Re: RedHat comments on APR using /usr/include/asm/atomic.h

Posted by Justin Erenkrantz <je...@apache.org>.
On Mon, Apr 29, 2002 at 02:55:42PM -0700, Blair Zajac wrote:
> There are some interesting comments from a RedHat developer on why
> APR should not include /usr/include/asm/atomic.h on Linux platforms
> at line 160 of apr/include/apr_atomic.h
> 
> http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=63643
> 
> Maybe someone more familiar with the atomic code should take a
> look at it.

I submitted a patch to use our own assembly primatives on Linux
based on FreeBSD's atomic code.  I know Thom spent some time
trying to get it to work on the other platforms (i.e. cleaning up
the #defines on the other platforms).

However, this has all been sidetracked by the recent push to
remove atomics entirely.

I'm on the fence about removing it  - I don't really care what
happens, but people need to step up and get it fixed on Linux
if it remains.  -- justin

Re: RedHat comments on APR using /usr/include/asm/atomic.h

Posted by Pier Fumagalli <pi...@betaversion.org>.
"Blair Zajac" <bl...@orcaware.com> wrote:

> There are some interesting comments from a RedHat developer on why
> APR should not include /usr/include/asm/atomic.h on Linux platforms
> at line 160 of apr/include/apr_atomic.h
> 
> http://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=63643
> 
> Maybe someone more familiar with the atomic code should take a
> look at it.

So, at the end, did we decide to keep atomic or to drop them? I have some
code relying on those for the WebApp module and would like to know whether I
have to patch it and use intra-process locks or just leave it as is :)

Thanks for hints :)

    Pier

--
I think that it's extremely foolish to name a server after the current U.S.
President.                                                 B.W. Fitzpatrick