Changes between Initial Version and Version 1 of Ticket #61327, comment 4


Ignore:
Timestamp:
Jun 21, 2021, 6:50:08 PM (3 years ago)
Author:
kencu (Ken)
Comment:

Legend:

Unmodified
Added
Removed
Modified

  • Ticket #61327, comment 4

    initial v1  
    1 Here we go. Once we open up the openmp can of worms, we have to follow whereever it all leads. This means sorting out that anything that leaves flags behind in pkgconfig files, etc, needs to somehow always force an openmp compiler to be used with anything that might ever access those pkgconfig files, and we have no such mechanism in place at present.
     1Once we open up the openmp issue, we have to follow wherever it leads. The issue has been that software that leaves openmp flags behind in pkgconfig files, etc, needs to always force an openmp compiler to be used with anything that might access those pkgconfig files, and we have no mechanism in place at present for one port to force another to do that.
    22
    3 And because no Apple clang compilers can use the openmp flags, they all fail.
    4 
    5 And Ryan has made it clear he does not want every build to be forced to a macports clang compiler as that causes the drives on the buildbots to die prematurely due to all the installing and uninstalling of macports-clang-9.0 (or whichever is the head of the line at that time).
    6 
    7 And this is indeed how we came to disable openmp in virtually all builds, until Apple clangs supported openmp (which we all thought might happen before now, but has not happened).
    8 
    9 I would suggest we go back to disabling openmp for 99.99% of cases, unless people can sort this out for a specific lonely build, in a specific case, that can affect no other software.
     3And because no Apple clang compilers can use the openmp flags (when last I tested), they all fail.