[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Openexr-devel] Thread safety
From: |
Brendan Bolles |
Subject: |
Re: [Openexr-devel] Thread safety |
Date: |
Wed, 3 Apr 2013 11:40:59 -0700 |
On Apr 3, 2013, at 9:17 AM, Christopher Horvath wrote:
> I think it should be a config-level decision. People who (god-forbid) need to
> maintain different thread-capable installations could then have their
> packaging system deal with "exr_omp", "exr_posix", "exr_cxx11", "exr_tbb".
Right, it seems like we should be able to support many different threading
libraries by adding new IlmThread implementations and just choosing which to
compile. We currently have IlmThreadPosix.cpp and IlmThreadWin32.cpp, how
about IlmThreadTBB.cpp and IlmThreadStd.cpp? Either way we need a single
abstraction to connect to any threading architecture, which we already have in
IlmThread.h.
OMP would not appear to be compatible with making an IlmThread implementation,
but I guess there's nothing stopping someone from sprinkling OMP pragmas
throughout the code. Is there harm from using OMP on top of IlmThread?
I agree that C++11 std::threads should eventually become the default when
available.
Brendan
- [Openexr-devel] Thread safety, Larry Gritz, 2013/04/02
- Re: [Openexr-devel] Thread safety, Peter Hillman, 2013/04/02
- Re: [Openexr-devel] Thread safety, Larry Gritz, 2013/04/02
- Re: [Openexr-devel] Thread safety, Peter Hillman, 2013/04/02
- Re: [Openexr-devel] Thread safety, Christopher Horvath, 2013/04/03
- Re: [Openexr-devel] Thread safety, Piotr Stanczyk, 2013/04/03
- Re: [Openexr-devel] Thread safety, Christopher Horvath, 2013/04/03
- Re: [Openexr-devel] Thread safety, Piotr Stanczyk, 2013/04/03
- Re: [Openexr-devel] Thread safety, Bob Friesenhahn, 2013/04/03
- Re: [Openexr-devel] Thread safety,
Brendan Bolles <=
- Re: [Openexr-devel] Thread safety, Paul Miller, 2013/04/04
Re: [Openexr-devel] Thread safety, jcupitt, 2013/04/03