Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 10, 2024

• docs: updated README

  Martinetto33

  committed
  22be82e

  Copy full SHA for 22be82e

Commits on Oct 24, 2024

• chore: deleted comment

  Martinetto33

  committed
  d06613a

  Copy full SHA for d06613a

• feat: I managed to incorporate the reduction

  Martinetto33

  committed
  3c7afff

  Copy full SHA for 3c7afff

Commits on Oct 23, 2024

• chore: removed debug functions and prints

  Martinetto33

  committed
  22d584e

  Copy full SHA for 22d584e

• the project is finished. Debug functions still available in the code, but some refactoring will be done

  Martinetto33

  committed
  ff4c9bf

  Copy full SHA for ff4c9bf

• fix: output produced in debug file corresponds with jdah's project for the first chunk

  Martinetto33

  committed
  5c9e8d1

  Copy full SHA for 5c9e8d1

Commits on Oct 21, 2024

• fix: changed type of b and h in cuda worldgen data from long to long long to reflect the s64 in jdah's code

  Martinetto33

  committed
  c77fcd8

  Copy full SHA for c77fcd8

• checkpoint before adding file logging code

  Martinetto33

  committed
  3017f9f

  Copy full SHA for 3017f9f

Commits on Oct 20, 2024

• checkpoint: I might have found a race condition between blocks when computing the vector v

  Martinetto33

  committed
  2e8f45f

  Copy full SHA for 2e8f45f

Commits on Oct 19, 2024

• partial fix: I avoided using some uninitialised memory in the device code; the world now looks MUCH better, but there are still some inconsistencies

  Martinetto33

  committed
  9a2ffb8

  Copy full SHA for 9a2ffb8

Commits on Oct 18, 2024

• feat: the world generation started to work, however the chunks are not 'natural' yet

  Martinetto33

  committed
  f256572

  Copy full SHA for f256572

Commits on Oct 17, 2024

• I think the double free error was fixed; I'm definitely on the right track because I saw something that almost made sense in the window; however, assertion that the biome id is in bounds fails

  Martinetto33

  committed
  41a81ac

  Copy full SHA for 41a81ac

Commits on Oct 16, 2024

• test: compute worldgen data gpu seems to give the correct results; removed free in worlgen.c to avoid double free problem; corrected a bug (I wrote & instead of %)

  Martinetto33

  committed
  673f1d8

  Copy full SHA for 673f1d8

• checkpoint before rewriting cuda part

  Martinetto33

  committed
  9f7aba0

  Copy full SHA for 9f7aba0

Commits on Oct 15, 2024

• fix: probably solved an important cast problem

  Martinetto33

  committed
  3311f5f

  Copy full SHA for 3311f5f

Commits on Oct 14, 2024

• checkpoint - I may have found the problem: compiling with different debug settings and different compilers, changes to the Makefile will be tried

  Martinetto33

  committed
  879add5

  Copy full SHA for 879add5

• debug: added debug prints in compute code

  Martinetto33

  committed
  3933736

  Copy full SHA for 3933736

Commits on Oct 13, 2024

• feat: initialisation occurs only once now

  Martinetto33

  committed
  7525f41

  Copy full SHA for 7525f41

• fix: solved some out of bounds accesses; need to initialise only once

  Martinetto33

  committed
  5f43483

  Copy full SHA for 5f43483

• fix: now passing correctly an array of pointers to kernel

  Martinetto33

  committed
  808e76c

  Copy full SHA for 808e76c

Commits on Oct 12, 2024

• error: polymorphism possibly not working in CUDA (?)

  Martinetto33

  committed
  1ed7d7c

  Copy full SHA for 1ed7d7c

Commits on Oct 8, 2024

• fix: removed matrix rapresenting biome table from cuda code; now dealing with illegal memcpy in device code

  Martinetto33

  committed
  26dc39c

  Copy full SHA for 26dc39c

• refactor: changed Makefile to compile c files with clang and .cu files with nvcc; linking done through nvcc; I can't believe it actually compiled

  Martinetto33

  committed
  32a0841

  Copy full SHA for 32a0841

• feat: made call to kernel and fixed some errors regarding device code accessing host variables; everything should be ready for integration

  Martinetto33

  committed
  3855558

  Copy full SHA for 3855558

Commits on Oct 7, 2024

• feat: almost completed the blocks generation kernel too (only decorations are missing); missing kernel call in function generateBlocks, but I'll do it tomorrow

  Martinetto33

  committed
  235590a

  Copy full SHA for 235590a

• feat: added kernel call

  Martinetto33

  committed
  4e9d9eb

  Copy full SHA for 4e9d9eb

• feat: added worldgen data smoothing; removed useless int pointer heightmap from struct CUDA_RESULT; it will have to be processed once the integration is done

  Martinetto33

  committed
  cdcbc29

  Copy full SHA for cdcbc29

• feat: almost completed heightmap generation

  Martinetto33

  committed
  428caf5

  Copy full SHA for 428caf5

Commits on Oct 6, 2024

• fix: forgot to add back worldgen.cu

  Martinetto33

  committed
  9c7131e

  Copy full SHA for 9c7131e

• fix: renamed all typedefs, functions, defines, symbols in order to avoid linking problems while resolving extern symbols

  Martinetto33

  committed
  1353f3e

  Copy full SHA for 1353f3e

• fix: corrected refactor of name Noise made acidentally with CLion refactoring

  Martinetto33

  committed
  1946b5a

  Copy full SHA for 1946b5a

• cuda additions compile, but the program crashes misteriously when trying to call compute functions in Noise structs

  Martinetto33

  committed
  b8e0b47

  Copy full SHA for b8e0b47

Commits on Oct 5, 2024

• chore: translated noise library to CUDA compatible code

  Martinetto33

  committed
  ade7a77

  Copy full SHA for ade7a77

• feat: I was able to include a C noise library in CUDA code!

  Martinetto33

  committed
  edf85a0

  Copy full SHA for edf85a0

• fix: inlined function hpc_gettime() in hpc.h and written documentation to justify this choice. No there are no errors at link time and the executable works as expected.

  Martinetto33

  committed
  630db32

  Copy full SHA for 630db32

Pagination

Previous

Next