ParaFROST 3.5.0

Added constant forall variants.
Improved simplifier API.
Replaced grid-stride loops with for_parallel macros.
Renamed GPU printers for better readability.
Renamed CNF info. Variables for better readability.
Added missing API members to vector classes.
Allowed wrapup API to force solver state.
Fixed the calculation of the elected size during allocation.
Fixed incremental solving allocation: Replaced all broken incremental vectors by a dynamic search space allocation for better locality and reliability.
Fixed marking variables being eliminated or frozen.
Fixed the checking of empty CNF in incremental solving.
Fixed incremental solving when MDM is enabled.
Fixed the allocation size of vectors defined in Variables structure.
Various micro-optimizations and minor typo/bug fixes.

ParaFROST 3.4.5

Now compiling only uisng GPU native architecture.

ParaFROST 3.4.4

Fixed many issues during compilation under latest CUDA version.

ParaFROST 3.4.3

Fixed a bug in phase-1 of VE.

ParaFROST 3.4.2

Fixed a bug in counting kernels and another bug when freeing the cacher memory.

ParaFROST 3.4.1

Minor fixes (eg. fixed bug occurs with gcc 13)

ParaFROST 3.4.0

Optimized and cleaned GPU code.

ParaFROST 3.3.7

Resetting memory after 'vars' allocation to fix a bug when reallocating that structure.

ParaFROST 3.3.6

Fixed MDM in incremental solving and assumptions mapping.

ParaFROST 3.3.5

Fixed a bug with incremental solving.