OSDev.org

PCMPEQD PCMPEQW PCMPEQB EMMS added
i think i added them all, and i dubt any of them will work. except EMMS, which implemented as ;

note for other emulator developers: mmx uses the reverse rm register order.

suprise: mmx works.

(hal of the opcodes are still broken, but i will fix them soon. )

paddsb and friends <- these opcodes are doing some completly different things compared to my excpectations.

(okay, so they clamping instead of just rolling through... jesus christ)

mmx shifts and bitwise operators was missing too.

The new, realistic develop schedule of the software:
2015: MMX and speed optimizations
2016: Sound Card and Network Card
2017: Botting Windows XP/2k, if possible to do it with zero documentation
2018: PAE, SSE and SSE2 (pentium4 level)
2019: 64 bit
2020: SMP
2021: Eliminating external bios and vgabios.

adding new features and new instructions, and bug-fixing attempts are discontinued due to the death of x86 architecture. new hardwares and host platforms will be probably added later aniwa, but as x86 is almost no more, there is no demand to emulate it in more serious ways.

Geri wrote:adding new features and new instructions, and bug-fixing attempts are discontinued due to the death of x86 architecture. new hardwares and host platforms will be probably added later aniwa, but as x86 is almost no more, there is no demand to emulate it in more serious ways.

That's a right decision, since x86 is only massively used in ATMs, that architecture is just a big and useless sack. x86 doesn't include x86_64, right?

Which architecture did both Microsoft and Sony use in their brand new top-of-the-line home entertainment systems?

Here's a hint: I bet they'd both run MS-DOS 6.11

i will not open an argument about this. we all know that arm won. this is why i started the emulator aniway, but basically there is no notable demand of x86 application emulation beyond the existing extreme-slow opensource solutions. i will not piss against the wind. i will not waste more time and work in this, i alreday have spent too much into it. it was a hard decision to make, but i will focus on my own platform instead. and yeah, the whole x86-console universe is only a few 10 million piece per year, not relevant. x86 ended.

Geri, if you indeed have some bright ideas about possible emulator speedup techniques - you could contribute the ideas (not even need the code) and make other extreme-slow opensource solutions faster. Do you ?

Stanislav

well, my extreme bright ideas about emulation optimization is not really worked in the practice aniway. and the tricks those are actually working, will needed in other non-x86 related projects of mine. therefore, i am not interested in any kind of opensource-related contribution.