Psp Emulation And Universl Remote

[rike]

well i heard some ware that some one was working on psp emulation on pandora what ever hapened to that also i heard that thar is soft ware to make the pandora a universal remote is that true ps sory about the spelling

 

[kingoddball]

PSP = JayFoxRox is working on that.

Universal Remote. Is it via Wifi? Pandora does not have InfraRed (unless I am mistaken)

 

[trooper]

tacoman, I know you are new here, But just a polite bit of advice, Please make sure in future
when you create a topic, That you post it in the correct forum. Oh, And please go easy on
the submit button too, I mean three duplicate topics ?. Easy boy. :rolleyes:

Thanks.

Trooper

 

[rike]

ya sory a bout that

 

[WizardStan]

Universal Remote. Is it via Wifi? Pandora does not have InfraRed (unless I am mistaken)

It does not have infrared, but it'd be pretty easy to hack one onto the EXT port. Also cheap USB->infrared adapters exist.

 

[gruso]

JayFoxRox keeps all his PSP emulator updates here: http://www.gp32x.de/board/index.php?/blog/37-pandora-psp/

Long story short: It's on hold for now, maybe he'll get back into it when he has hardware. If you hold your breath for a release you will turn purple and DIE.

 

[rike]

i gess i'll just hafe to play my boring real psp for now

 

[JayFoxRox]

JayFoxRox keeps all his PSP emulator updates here: http://www.gp32x.de/board/index.php?/blog/37-pandora-psp/

Long story short: It's on hold for now, maybe he'll get back into it when he has hardware. If you hold your breath for a release you will turn purple and DIE.

Right now I'm doing research which will help me to write the emulator later, however, the emulator itself is not changing much at the moment. However, there is one stable version which runs basic homebrew games.
And IrDA (http://de.wikipedia.org/wiki/Infrared_Data_Association) is also working in an older version, it was trivial to implement (meaning it won't be a problem to have it in the newer version too again). So if your devices support IrDA you can just stick to the PSP emulator. I couldn't test it yet though because I don't have a IR transmitter (I will probably get a USB one once I know which are supported under linux and which can do more than just IrDA) and as I also wasn't able to find a library I just decided to route all IR functions to a file handle. A possible implementation would be to use this as input for another application which implements the IrDA signal on its own and then uses the parallel port with an IR diode directly attached to it (I doubt that this is fast enough though in which case the second application would have to use some sort of library for example).
Raw IR signals are not working at all yet though (because its depends on timing which is not emulated yet).

 

[Kaze0110]

What firmware libraries is it able to run homebrew games on? 1.5, or upwards of 3.0+ ?

 

[bennollan]

What firmware libraries is it able to run homebrew games on? 1.5, or upwards of 3.0+ ?

With the official sony firmwares homebrew would work on 1.50 and 2.xx but there was homebrew enabled firmware that went all the way up to 5.00

 

[Drak]

Psp emulation will never be possible (at a playable state) on the pandora. There might be some proof of concept later on, but other than that if that's one of the only reasons you want a pandora you can forget it :rolleyes:.

 

[willrandship]

I hope that Rolling eyes thing means you're sarcastic. Otherwise I will not pity your flaming corpse.

 

[Alpha2]

If the Pandora can emulate a PS1 or even early DC emulation it's not too far off to think it could do early PSP games eventually. Only reason any of the later games get close to PS2 quality is because they're running at an incredibly lower resolution and sometimes at the higher clock speed setting.

 

[WizardStan]

Psp emulation will never be possible (at a playable state) on the pandora. There might be some proof of concept later on, but other than that if that's one of the only reasons you want a pandora you can forget it :rolleyes: .

You seem to know a lot more about PSP emulation than JayFoxRox who has already completed the proof of concept which is capable of running some homebrew games right now as he said just a couple posts above yours. I suggest you tell JayFoxRox that you are more knowledgeable than he, and make sure he knows he's wasting his time. :rolleyes:

 

[ari64]

Psp emulation will never be possible (at a playable state) on the pandora. There might be some proof of concept later on, but other than that if that's one of the only reasons you want a pandora you can forget it :rolleyes: .

You seem to know a lot more about PSP emulation than JayFoxRox who has already completed the proof of concept which is capable of running some homebrew games right now as he said just a couple posts above yours. I suggest you tell JayFoxRox that you are more knowledgeable than he, and make sure he knows he's wasting his time. :rolleyes:

People are already complaining about N64 emulation being slow, so unless Fox comes up with a way of making MIPS emulation much faster, I suspect people will find PSP emulation slow.

 

[WizardStan]

People are already complaining about N64 emulation being slow, so unless Fox comes up with a way of making MIPS emulation much faster, I suspect people will find PSP emulation slow.

PSP is a lot more HLEable than N64.

 

[Exophase]

PSP is a lot more HLEable than N64.

I disagree. Or at least, when we're talking about the fully RCP HLE'd N64 emulation approach, that doesn't actually work for all games. I think that as far as graphics go this is at least as high a level of abstraction than what you get on PSP, and it's graphics that eat up most of the performance.

PSP's GPU does have a full transformation pipeline, but if games decide to circumvent it and use the VFPU instead (ie, they want something like programmable vertex shaders) then a whole lot more needs to be emulated. At the very least, games will probably be using VFPU for matrix transforms.

What JayFoxRox does have going for him, that Ari64's N64 emulation doesn't, is that it isn't counting cycles. This means somewhat faster CPU emulation, but what it really means is that you're emulating at exactly the rate Pandora can handle instead of some fixed number of instructions per second. Where this number is less than what a real PSP would have done the game may compensate - you may have choppier gameplay while still having proper overall game and music speed. Or you may have an all around slower game. Or you may have a game that doesn't work altogether. The latter part is subtle - while most game engines are not designed to not rely on a certain minimum number of clock cycles beyond something very small all it takes to break a game is some initialization routine or something that violates this. We may find that overclocking the Pandora actually improves compatibility.

JayFoxRox is also pretty confident that enough PSP games worth playing don't push the CPU very hard. This isn't that hard to believe; a lot of DS games don't push the CPU very hard, and that CPU is much weaker. The baseline for many games is 222MHz, since that's what the CPU was initially capped at. I doubt that PSP games at 222MHz can use anywhere close to 200 MIPS or wherever you'd place a theoretical number w/o memory stalls, because it has weak 2-way set associative L2 cache and slow main memory.

Compared to N64 emulation JayFoxRox has a slight advantage in not having to emulate 64-bit registers or double precision FPU operations. And a massive disadvantage in having to emulate VFPU operations.

 

[Nova]

Same guy?

 

[ari64]

What JayFoxRox does have going for him, that Ari64's N64 emulation doesn't, is that it isn't counting cycles. This means somewhat faster CPU emulation, but what it really means is that you're emulating at exactly the rate Pandora can handle instead of some fixed number of instructions per second. Where this number is less than what a real PSP would have done the game may compensate - you may have choppier gameplay while still having proper overall game and music speed. Or you may have an all around slower game. Or you may have a game that doesn't work altogether. The latter part is subtle - while most game engines are not designed to not rely on a certain minimum number of clock cycles beyond something very small all it takes to break a game is some initialization routine or something that violates this. We may find that overclocking the Pandora actually improves compatibility.

If it's not counting cycles, how does it deal with games which busy wait for interrupts? The cycle count is 1 or 2 extra cycles per branch. I doubt this accounts for more than a few percent of the time. My mips emulator spends most of its time waiting on the L2 cache and main memory. The memory subsystem on the OMAP3530 is really slow. Running the dynarec on an Intel Atom N270 @ 800mhz is about 30% faster than a OMAP3530 at the same clock speed.

 

[Exophase]

If it's not counting cycles, how does it deal with games which busy wait for interrupts? The cycle count is 1 or 2 extra cycles per branch. I doubt this accounts for more than a few percent of the time. My mips emulator spends most of its time waiting on the L2 cache and main memory. The memory subsystem on the OMAP3530 is really slow. Running the dynarec on an Intel Atom N270 @ 800mhz is about 30% faster than a OMAP3530 at the same clock speed.

PSP emulators w/o cycle counting, and other emulators that take the same approach, emulate interrupts by chaining them to interrupts on the target platform. If interrupts are emulated with original PC et al then there will still have to be checks for interrupts in the emitted code. I'm not sure JayFoxRox even intends to emulate interrupts in such a way at all, but I'll have to defer that explanation to him.

Is 30% just for the recompiler, or the entire emulation? Can you actually quantify "most of its time"? Have you checked the performance counters?

This is pretty bad news, given that I expect the generated x86 code to take more cycles internally... and Cortex-A8's L2 cache should be pretty low latency as far as these things go. I hear that at some Pandora revision there a boost in main RAM performance. OMAP3530 revs have been bumping Cortex-A8 revs, and I heard a long time ago that Cortex-A8 originally didn't perform critical word first on L2 line loads. Maybe the L3 interconnect has been improved too.