SCIENTIFIC-LINUX-USERS Archives

February 2007

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Connie Sieh <[log in to unmask]>
Reply To:
Connie Sieh <[log in to unmask]>
Date:
Fri, 16 Feb 2007 14:54:45 -0600
Content-Type:
TEXT/PLAIN
Parts/Attachments:
TEXT/PLAIN (47 lines)
On Fri, 16 Feb 2007, Ioannis Vranos wrote:

> Connie Sieh wrote:
>> OpenMP is a standard with precompiler directives for multithreading. 
>>> From what I recall
>>> from my previous W*ndows programming days it was for Fortran and C++. I 
>>> haven't checked if
>>> it is provided for other languages since then, but anyway. :-) In any 
>>> case, it is about
>>> the programmer providing specific multithreading guarantees about the 
>>> execution of
>>> specific code.
>>> 
>>> 
>>> BTW, I read in a Fedora web page that they compile its binaries with 
>>> specific flags. Are
>>> the flags used for compiling SL the same as the flags used for EL?
>> 
>> Fedora is NOT Enterprise.  We do not know what flags are used to build 
>> Enterprise as the Upstream Vendor uses a inhouse build system.
>> 
>> Note that in the future (Fedora 7) that the build system to build Fedora 7 
>> will be the same build system used for the Fedora Extras as Fedora Core and 
>> Fedora Extras are merging and they are working on a more open build system. 
>> Since the Fedora Extras build system is opensource then maybe in the future 
>> one may expect that this info maybe more available.
>
>
>
> How about using the flags used for Fedora Core at least? I can't find where I 
> read about them though.

I did not think that the flags for Fedora Core have been made public. 
Do you have a reference to these flags.

-connie
>
>
> However OpenMP was not available to GCC sometime ago (that is when VS 8 .NET 
> was under development and I checked MINGW32/GCC at that time), no compile 
> problem of course since unknown #pragmas are ignored. So if GCC doesn't 
> provide OpenMP yet, it may compile, I think the SMP of SL 5 may suffer, and 
> perhaps bugs/incompatibilities may be introduced (for example gcc-specific or 
> other multithreading code may make wrong assumptions about OpenMP based code, 
> or interoperate with one another etc).
>

ATOM RSS1 RSS2