nmk/nmk16.cpp: upgraded interrupts trigger system based on PROM contents

[sergiopolog]

• Added a new system for triggering interrupts based on PROM contents: set_interrupt_timing() and nmk16_scanline.
• Bootleg sets, Afega and Comad games, and other sets that PROMs are still undumped (or not exist), on which the new interrupt triggering system cannot be applied, remain using the old set_hacky_interrupt_timing() and nmk16_hacky_scanline handlers.
• Added info about screen resolutions and timings.
• Removed "hacky" denomination to set_screen_xxxxxx() functions as the screen sizes are confirmed for each available resolution
• PROM dumps from hachamfp are added to hachamf and other clones, and marked as BAD_DUMP until they are properly dumped
• PROM dumps from bjtwin are added to nouryoku and clones, and marked as BAD_DUMP until they are properly dumped
• strahl and clones could benefit from the new interrupt system, but PROMs for them are undumped. They still use the old system until they are properly dumped
• ssmissin and airattck "Comad" games could also benefit from the new interrupt system, but the V-Timing PROM on them is a 82S147 (512x8bit) in which only half of space is used and A5 address line is tied to ground. The nmk16_scanline code could be tweaked to support them, but more research is needed to apply that. They still use the old system.
• Merged powerins.cpp into nmk16.cpp driver to make use the new interrupt system, too. New set_screen_midres() function is added to be used by powerins sets.
• Sprite DMA process might be triggered using PROM contents aswell, but that needs more testing and it could be a future enhancement.

[sergiopolog]

I marked the PR as draft because powerinsb and powerinsc bootleg sets are not working properly after powerins.cpp driver merge into nmk16.cpp. Needs further review.

Additionally: Is the flag MACHINE_SUPPORTS_SAVE present in powerins sets really needed? isn't that the default behaviour?

[Hammy1986]

Powerins bootleg C never worked correctly anyway. It's the electronic devices italy bootleg.
Strahl and Mecha fighter proms will be dumped sometime , along with a bootleg set of mecha fighter with raiden sound.

The reason the old power instinct driver was based on the bootlegs as it took many years for the original dumps to turn up.

[rb6502]

MACHINE_SUPPORTS_SAVE isn't a default choice by the core, but it should be everyone's default choice when writing/updating a driver :-)

[cuavas]

MACHINE_SUPPORTS_SAVE is supposed to be applied to systems that people have audited for proper save state support (including all the devices they use, etc.).

[Osso13]

I marked the PR as draft because powerinsb and powerinsc bootleg sets are not working properly after powerins.cpp driver merge into nmk16.cpp. Needs further review.

As far as I can see they crash because they don't have the ROM_REGION "vtiming" that the set_interrupt_timing(config); call in their machine_config expects

[sergiopolog]

I marked the PR as draft because powerinsb and powerinsc bootleg sets are not working properly after powerins.cpp driver merge into nmk16.cpp. Needs further review.

As far as I can see they crash because they don't have the ROM_REGION "vtiming" that the set_interrupt_timing(config); call in their machine_config expects

Thanks @Osso13 for pointing that. That was a copy/paste error and I wrongly included set_interrupt_timing() at maching_config for powerinsb and powerinsc, while they have their own interrupt config (via screen_vblank) as they don't have timing PROMs.

Fixed now and the PR promoted to Ready

[sergiopolog]

Is something else needed to be checked or included on this PR? Games were fairly tested with new interrupts system with no noticeable issues. Do you expect me to add or test something else?

Additional improvements such applying the new system to Comad games, also in ddealer, and dump Strahl PROMs could be included on a future PR