In addition to the normal float, double, and long-double FFTW builds, this patch includes AVX2-enabled float and double libs in /usr/lib/haswell.
Resolves T8235.
JoshStrobl |
Triage Team |
In addition to the normal float, double, and long-double FFTW builds, this patch includes AVX2-enabled float and double libs in /usr/lib/haswell.
Resolves T8235.
Run several applications that use FFTW, ensuring that they link against the Haswell libs (gmic, gromacs, hugin, octave).
Automatic diff as part of commit; lint not applicable. |
Automatic diff as part of commit; unit tests not applicable. |
@joebonrichie Mentioned earlier in the thread that it might be worth adding --enable-fma`along with AVX2. Any luck with that?
Oh gosh, sorry @serebit. I did look though the changes, but it looks like I missed that one. Will rely om ctrl + f in the future :I