Victor Nieto

The d_subdiv16 didn't make any difference, but I was able to reduce the crashes by disabling the main menu drawing. And still have some crashes when the game tries to...

Maybe it's better to make the visplanes unlimited, I already did a fully working implementation (there is a branch) for it. I remember it was minimally slower, and the main...

No work has been done for the 68k or PPC FastDoom ports. I really want to port FastDoom to other architectures, but that will happen after the 1.0 release is...

Thanks for the ideas @Awesoken2 ! Pretty interesting optimization using LEA instruction, I'll try it for sure. As for register stalls, the 486 has 1 cycle penalty when writing to...

The 386SX (and some 386DXs) suffers from huge memory bottleneck, specially if the system doesn't have any cache. Reading and writing to the CR2/CR3 register is faster than doing from...

Is it possible to use the ESP register as a additional register if the interrupts are disabled? I've been trying to use it but I get random lockups.

> While reading my trusty "i486 Microprocessor Programmer's Reference Manual", page 844 (G-8), I noticed this stinker under "PREFIX OPCODES": > > "On either processor, all prefix opcodes, including OFh,...

@Awesoken2 I've been able to use your idea to render columns using interleaved code, the code is in this branch: https://github.com/viti95/FastDoom/tree/ken-silverman-optimizations For now only is implemented for high detail and...

Some benchmarks (only for R_DrawColumn high detail): UMC Green 486 (50MHz): Old code: 42.732 fps New code (LEA+SHR): 43.307 fps Intel 486DX2 (66MHz): Old code: 37.651 fps New code (LEA+SHR):...

Yes I repeat the benchmarks multiple times, so the result is pretty much stable. And yeah benchmarks are slow on Doom, as every frame has to be rendered to get...