[1.44][build] Improving support for VC10 and Adding support for Intel C++ 12

I'm working through updating msvc.jam to better support VC10, primarily so that intel-win.jam works more cleanly for Intel C++ 12. It looks like the last substantial revision to msvc.jam was for VC8, although it still works well for VC9 and VC10. The changes I've made work well; my next post is to Intel about "internal error: 0_1374", which seems to be a catch-all. My overall goal is to better support the new target architectures and optimization features available in VC10 and Intel C++ 12. I'm new here, so I wanted to make sure that 1) I'm not duplicating work; 2) I'm following the appropriate procedure. BTW, I posted to the Boost Build list a couple days ago and there has been no traffic at all. Any guidance is appreciated.

On Sun, Nov 21, 2010 at 9:36 AM, Jeff Benshetler <jeff.benshetler@gmail.com> wrote:
I'm working through updating msvc.jam to better support VC10, primarily so that intel-win.jam works more cleanly for Intel C++ 12. It looks like the last substantial revision to msvc.jam was for VC8, although it still works well for VC9 and VC10. The changes I've made work well; my next post is to Intel about "internal error: 0_1374", which seems to be a catch-all. My overall goal is to better support the new target architectures and optimization features available in VC10 and Intel C++ 12.
I'm new here, so I wanted to make sure that 1) I'm not duplicating work; 2) I'm following the appropriate procedure. BTW, I posted to the Boost Build list a couple days ago and there has been no traffic at all. Any guidance is appreciated.
Take a look at http://beta.boost.org/development/tests/trunk/developer/summary.html and http://beta.boost.org/development/tests/release/developer/summary.html. No one is testing the Intel compiler on Windows for either trunk or release branches. Without at least one platform champion to daily run tests and raise a red flag when config changes need to be made, it is hard to support a compiler. Intel used to have a someone assigned to that task, but no longer. So getting a daily trunk test running would be a logical first step. --Beman
participants (2)
-
Beman Dawes
-
Jeff Benshetler