A new possible crash has been found in x64 port

A new possible crash has been found in x64 port. This problem looks innocent, but can potentially cause random unexpected crash: https://sourceforge.net/p/wp2latex/wp2latex/ci/a9e3f4f4f2f0a593a45964d167a87293e959d36b/

32 bit port is here for more than 10 years, but 64 port is still relatively new.

A new bugfix release 2.0.3 is available.

Anyway, ussing assembler is critical to push CPU load as low as possible. It is much more better when CamView with eunning camera consumes 5% of CPU and not 50%.

Benchmark - Flip in 1D direction C x64 [us]:

        1        2        4        8       16       24       32       64 
1D    91.46   174.57   28.736   30.705   32.118   87.131   31.319   32.119 
2D   192.09   364.31   77.789   64.341    67.99   184.43   66.332   68.571 
3D   330.03   626.03   133.49   110.41   116.88   322.98   114.47   117.32 

Benchmark - Flip in 1D direction ASM x64 [us]:

        1        2        4        8       16       24       32       64 
1D   47.178   180.51   8.1272   8.3669   7.2521   25.019    6.305   7.3133 
2D   98.361   367.76   25.707   16.817   15.766   51.013   13.618   15.425 
3D   169.89   634.45   44.597   28.897   27.684     87.6   23.669    26.14