MC-Server on Raspberry Pi

[xoft]

Just for fun I ran a profiled version of MCServer on the RasPi. Quite unexpectedly, The function eating up the most time was the lighting thread's PrepareSkyLight() function with 17 % of the run time. Who would've thought?
Here's the list of top time-eaters. Notice how the generators are actually doing rather well.

Code:

% cumulative self self total time seconds seconds calls s/call s/call name 17.84 10.22 10.22 398 0.03 0.03 cLightingThread::PrepareSkyLight 12.50 17.38 7.16 882 0.01 0.02 deflate_slow 10.77 23.55 6.17 1217115 0.00 0.00 cCaveTunnel::ProcessChunk 10.10 29.34 5.79 15968 0.00 0.00 fill_window 8.38 34.14 4.80 1345982 0.00 0.00 longest_match 8.25 38.87 4.73 422 0.01 0.04 cLightingThread::LightChunk 5.85 42.22 3.35 370359 0.00 0.00 cHeiGenBiomal::GetHeightAt 5.69 45.48 3.26 8901 0.00 0.00 cLightingThread::CalcLightStep 4.08 47.82 2.34 484 0.00 0.00 cChunk::CreateBlockEntities 2.34 49.16 1.34 478 0.00 0.02 cStructGenTrees::GenStructures 2.09 50.36 1.20 1775634 0.00 0.00 cNoise::CubicNoise2D 1.75 51.36 1.00 6123 0.00 0.00 adler32 1.66 52.31 0.95 726731 0.00 0.00 cBioGenVoronoi::VoronoiBiome 1.43 53.13 0.82 426000 0.00 0.00 cNoise::CubicNoise3D 1.20 53.82 0.69 1814 0.00 0.00 cIniFile::FindValue 0.98 54.38 0.56 34071 0.00 0.00 cStructGenRavines::cRavine::ProcessChunk 0.59 54.72 0.34 484 0.00 0.01 cStructGenWormNestCaves::GenStructures 0.54 55.03 0.31 866 0.00 0.00 compress_block

I guess the ARM architecture is good for math programming, but not so good for memory access. PrepareSkyLight jumps all over more than a 600 KiB block of memory, reading it and writing to it byte-wise. Generators usually take a few numeric values and chew them over and over until they get a final number out of it, and that's where a many-register platform wins.

I'd like to run a few more tests on this - to see if there's any kind of RAM cache somewhere in there, if it's better to access the memory consecutively or it is the same as accessing it randomly.