Feature/openqxd

[chaoos]

• Adds interface openQxD <--> QUDA
• Adds capability for C* boundaries in 1,2 or 3 spatial directions
• Fixed CPU reordering
• Adds openQCD gamma matrix basis (and basis transformations)
• Adds order class for spinors
• Adds order class for clover field (U(1) clover field is transfered from openQxD)
• Adds order class for gauge field
• Changed all the comments in quda.h to C89 style comments
• Rebased on your current develop branch

[chaoos]

are there ever cases where OpenQ*D would need both recon-13/9 and recon-12/8 simultaneously?

Originally yes, but I saw that it makes stuff increasingly complicated. So I added a new Cmake flag QUDA_QCD_PLUS_QED=ON/OFF as you proposed to build Wilson ops with QUDA_RECONSTRUCT_NO/8/12 if set to OFF (the default) or QUDA_RECONSTRUCT_NO/9/13 if set to ON.

See 452f6e6

[weinbe2]

I'm happy with this PR from a visual standpoint. I still need to do build/run tests, in particular with QCD+QED, but that's more to make sure nothing unexpectedly broke, there are a few things I'd like to try beyond what's covered by our CI. This is awesome work, @chaoos .

Based on an offline convo, I'll take on updating the Wilson dslash tests to properly test recon-13 and -9 in a separate PR. This would be done by generating random U(3) instead of SU(3) fields. Since I've had to deal with this relatively recently with HISQ (as is relevant for long links) this should be relatively straightforward for me.

@maddyscientist , I believe you wanted to do some further review.

[maddyscientist]

Thank you @chaoos for all the work you have done and going through the review process to get this in shape to merge.

I have one small comment (regarding enabling strict prototyping for the C interface test), but beyond that, this is good to go. Approving.

[weinbe2]

Thank you @chaoos for all the work you have done and going through the review process to get this in shape to merge.

I have one small comment (regarding enabling strict prototyping for the C interface test), but beyond that, this is good to go. Approving.

I'll take point on addressing the strict prototypic for the C interface as part of updating the unit tests to explicitly check recon-9/13 when QCD+QED is enabled, I'm happy to merge this. Thanks @chaoos for all of your work (and patience when I was slow to review it)!