Closed mkannwischer closed 8 months ago
OK, normally GitHub detects when I do a manual merge (had to fix the mupq
repo reference to the correct commit), but somehow it didn't work this time. So I'll close this as unmerged, but it was merged with c3dbd50.
Resolves https://github.com/mupq/pqm4/issues/283.
This adds the implementations from the NIST submission package of MQOM.
I eliminated the dynamic memory allocations that were present throughout the implementation. It was rather straightforward - but a lot of work. Right now it is using VLAs in a couple of places (which is fine for pqm4), but it should not be too hard to change those buffers to the worst case sizes.
The
cat5_gf251
parameter sets exceed the 4MB available on qemu'smps2-an386
so I did not include them because I cannot test them. All other parameter sets (includingcat5_gf251
) work on qemu.mqom_cat1_gf31_fast
andmqom_cat1_gf251_fast
are the only two parameter sets that fit on thenucleo-l4r5zi