forums.ps2dev.org

Urm be careful though (I haven't checked where the buffer is defined) but there is always the risk that increasing the size could make existing code fail due to higher memory requirements (well when recompiled against a new libc). If it is a static buffer (I guess likely) then you could blow away al...

Odd that, never seen that before. You aren't already running the psplink prx as a plugin or something strange are you as that error code kind of implies the library is already there. Then again I have never tried it with CF Enabler, people have said they have had psplinkusb running on a PSP-3000 bef...

The exception handler in the sdk is abit old now, it doesn't really work in most scenarios, it only worked before because of the fact that kernel ELFs were loaded into user memory regions. You should use psplink unless you have a good reason not to, otherwise you might need to use psplink's exceptio...

Nice, certainly plenty of stuff I never worked out when implementing usbhostfs, surprised it works ;) At least I can see why the padding was there (I kinda guessed it might be something like extra fields but I didn't need to spend any extra time finding out). Now we will see if any of those people w...

Ahum, looks like the is a bug in psplink's exception handler :) It never restores the lo/hi registers so stepping over the instructions trashes the values and you get crap back. Try checking out the latest psplinkusb and updating your psplink.prx file.

The simplest way is to put the number into something like windows calc and then change to hex, you will get something like 0xFFFFFFFF8002013A, you only care about the lower 32bits. Or when you print it out of your program use unsigned hex formatting with printf (i.e. %x). Or to be technically correc...

psp-gdb shouldn't do anything to your application, there is no concept of a "debugging" version of a process. And about the only way to post-mortem debug is wait for the crash to occur in psplink and use its internal debug features to see what went wrong. Of course if it is not dying in gd...

For a start you really should look at what the error code is telling you. If you convert it to hex and look it up in pspkerror (or using the error command in pspsh) you find it is 0x8002013a which just so happens to be SCE_KERNEL_ERROR_LIBRARY_NOT_YET_LINKED. So for some reason you have specified th...

Your problem is you don't understand how c++ works, in theory you can export static/global c++ functions no problem, if you understood what the export builder does and what c++ does. Alas you haven't bothered to try and understand.

If it is a hardware bug it might not be fixable, after all the RSX isn't their's and I doubt NVidia are _that_ bothered about security, not something they usually need to care about.

I am sorry but the likelyhood that any decent games would come out for the linux side of things is small to minimal, and due to the shit poor processor in the thing (and the general laziness of most "porters") then you would be far better off spending the $400 on a cheapo PC than using a P...

Simple you have two PSPs if you care that much :)

No it doesn't. And who gives a fuck anyway

What the hell are you talking about? As far as I know other-os still works on the latest firmwares. And sure the card slots are gone but that is why you create an other-os partition, stick on a boot loader (a one time use of a usb key or card reader) and I think that even will load your code over th...

Well you do need to but something around zlib which handles the zip format itself but yes not hard :)

hmm... I will look into it then. I think I see it, the patch changed the loader code to always use program headers, think I will just add a check for loading an ELF and in that situation always use the section headers :)

Update: Try it now, hopefully fixed it :)

Well the real problem is the program header is garbage. No idea what you used to build it, most likely as the ELF is not loaded at the "standard" address you have a custom linker script? Or is this "third party code"? I am sure I am paranoid, but I would swear some of the changes...

Seems to work on my ELF files I have built using the SDK, turn on the debugging option and see if there is something specific it is failing on.

Well that error code isn't in the ones which came from puzzle bobble's debug symbols so there is really little way to tell what it means :)

Nice work :)

The fact that it has type info means it is not externed propery as a C export and so it can't find the mangled name in the libraries

While FMCB is all nice 'n' legal (well depends which country you are in I suppose) the forums are basically awash with people using it to play warez. Oh well, things never change :)

Have you ever actually programmed in C?

It shoudn't cause a problem, but it depends if you are using all kernel memory up as well cause at that point the usb driver can go screwy, and there is not a massive amount of free ram on a modern psp firmware.

That said I have never seen it do so :)

Sorry it took so long, got around to committing the patch in SVN 2457, works nicely :)

As the modules need to load into user memory (no way around that, and you wouldn't have space in kernel mode anyway) then you will just hit the problem that games will probably have consumed all memory anyway. Tbh you are screwed unless you implement a simple kernel mode version and talk to the kern...

Why would you want to :)

I guess try and geting psplink to work, that would at least rule out whether it is a general problem or your code. Tbh there are a few machines that for what ever reason just do not work with usbhostfs but it is rare.

Sigh... You do know you need to be running usbhostfs_pc on the PC side otherwise it has nothing to communicate with.

http://svn.ps2dev.org/filedetails.php?r ... rev=0&sc=0

An example of using the ansynchronous pipes with usbhostfs to allow a simple connection between PSP and PC.