Updated driver with new information related to ID and gamelist for tourvis.cpp [system11, Tafoid]
Fixed instances in source where "Tecmo" was shown as "Temco". This also covers MT#6293. (nw)
* Added several missing functions to rgbgen
* Fixed logical shift right in rgbgen
* Fixed sra that should be sra_imm in rdptpipe
* Added some simple SSE4.1 optimisations in rgbsse
* Re-organised rgbsse, rgbvmx and rgbgen to be in more logical order
* Fixed return on some modifying operators
* Made some more reference parameters const
* Removed inline qualifier from a number of methods as it's implied when body is present at declaration
* Mark some constructors explicit
Trying to evaluate __GLIBC__ will result in an error if is not defined,
if the preprocessor does not short-cut the evaluation.
Split the macros onto separate lines and define the result in a new
BX_USE_GLIBC_PTHREAD_SETNAME_NP macro to avoid duplication.
Fix potential compilation error by ensuring __GLIBC__ is only evaluated
when actually defined.
When __GLIBC__ is defined, we do not need any additional headers on BSD
platforms (hence why using #elif).
- Keyboard mapping adjusted to fit current MAME
- Fixed screen raw parameters
- Fixed disp.dex Test 20 (Video SYNC Gen. Add. Test) for color displays
3c505: The Ethernet frame check sequence is now appended to transmit data [Hans Ostermeyer]
input_common: fixed 0.172 regression for non-US keyboards [Hans Ostermeyer]
problem is caused by <SDL2/SDL_cpuinfo.h> puling in <altivec.h> which #defines vector, pixel and bool
however when the compiler is in Apple Altivec mode, these are context-sensitive keywords and don't need to be #defined
we never need the #defines in our own code because we use the GCC-specific __vector instead of vector
so we trick the header to think the compiler is in Apple Altivec mode to suppress the #defines
- Simulation code in nzeroteam confirms that COP sort DMA function uses a dword key
- Several priorities are still wrong (e.g. bikes invisible on character select screen)
