forums.ps2dev.org

fowl416 wrote:I could not agree with you more.

This was fixed in r1514. Give details if you're still having problems.
http://svn.ps2dev.org/log.php?repname=p ... %2FSDL_ttf

Looks like the toolchain script didn't install the headers in /usr/local/pspdev/psp/sdk/include properly. Make sure you're running it with permission to write to /usr/local/pspdev.

The fix in 1489 is definitely right for clock, but I think time and gettimeofday should keep __psp_set_errno. You could check the return value of time(123) and gettimeofday(123, 456) to see. I can't test it at the moment.

__psp_set_errno should probably be removed from clock() and maybe time() in newlib-psp/newlib/libc/sys/psp/libcglue.c.

danzel wrote:0x74: Connection Timed out
0x6f: Connection Refused

Those are ETIMEDOUT and ECONNREFUSED in sys/errno.h.

Try adding -G0 to your own program too.

Sounds to me like your newlib patches are very outdated. Get the psptoolchain dir from subversion (svn://svn.pspdev.org/psp/trunk/psptoolchain) and use that.

I checked; there's no activity on this line at all when the battery is not present.

Why do you think it is a bidirectional protocol? Because there's at least some communication initiated by the PSP, and it's used to get status information from the battery. I could build a better circuit and figure out which data comes from where, but unless we think this is more than just battery ...

I propose an environment variable be used to preserve aspect ratio I propose we make it even more flexible: SDL_VIDEO_WINDOW_POS=x1,y1,x2,y2. For example fullscreen is 0,0,479,271, 4:3 aspect ratio would be around 59,0,421,271, and an unscaled 320x240 is 80,16,400,256. Could specify a fullscreen ro...

The IPL is only accesible by the sceNand driver, so it's relatively protected from a casual bricking. They do reflash it from 1.x -> 2.x, so it could still get messed up for entirely innocent behavior, although that's unlikely since the IPL is so small. ... which *only* initializes flash, DDR, GPIO...

This isn't quite true - some do and some don't. For example, the SVGA driver definately does wait for a vblank, and the DirectX5 driver will depending on the video driver settings. The SDL docs say that SDL_Flip() should set up a flip but return immediately, so having the flip wait for vblank would...

I wonder if the IPL would even be involved with flash recovery, seeing as how the IPL is in flash too.

There's some stuff in the existing code that could be cleaned up a bit. For example GuUpdateRects should only update the rects it's passed, not the entire screen, and there's a bug I'd like to track down where the scaled screen is sometimes shifted right a few pixels, and sometimes doesn't show up a...

dankydoo wrote:I see that SIO is the most unintrusive, what exactly do we need for a USB implementation?

We'd need to figure out how to write a USB driver on the PSP, or turn on the USB storage driver and use files on the memory stick to communicate with the PC.

It's the same linker that Linux uses, but you generally don't see this problem when linking against shared libraries. When doing static linking as on the PSP, the linker goes in order and only includes objects found in each library if it needs them to solve an earlier dependency. You can repeat libr...

nem: When does IPL run? When power is applied or when the power switch is pressed?

OK, I've cleaned up the errno handling in r1223. It should be more or less correct now, except that the actual errno values still need to get fixed up.

From http://forums.ps2dev.org/viewtopic.php?t=3435:

The most important thing to note is your module must start in kernel mode, but you app should only be in user mode

It looks to me like you're running your main thread in kernel mode, and the sign-extension on the pc is just a side-effect of that.

It was broken back in r1096 when the newlib patch as regenerated. It should be fixed (for good) in r1218.

Just follow README.PSP and it will get installed into /usr/local/pspdev/psp/lib. Make sure you're running bash to avoid the problems above.

Well of course the best place to start is the first error: LDFLAGS="-L`psp-config --pspsdk-path`/lib -lc -lpspuser -lpsputility" ./configure --host psp --prefix=`psp-config --psp-prefix` configure: error: expected an absolute directory name for --prefix: psp-config ...

Yeah, you'll need to also rebuild newlib. Use the latest toolchain from oopo's site (currently psptoolchain-20051021.tgz) or get it from svn.

Both psptoolchain-20050906.tgz and psptoolchain-20051011.tgz are out of date; get it from Subversion instead:

OK, rev 1172 has these changes (move dirent inside DIR, zero dirent before calling sceIoDread). Don't free the dirent returned by readdir() anymore. Let me know if it gives you any trouble.

The toolchain script now copies the built docs to /usr/local/pspdev/psp/sdk/doc, so if you have doxygen installed when you build the toolchain, you should be able to point your browser to [url]file:///usr/local/pspdev/psp/sdk/doc/html/index.html[/url] when it's done.

Seems to work just fine here, can you provide an example?

Sure, paste that (and config.log) on rafb.net/paste

No idea, it works here. Maybe try removing and reinstalling everything from a clean slate (http://wiki.ps2dev.org/psp:programming_faq)

I finally got around to testing this. Nice work! I only had some minor issues for which I've committed fixes: accept ascii 127 as backspace pass argp in load_start_module, otherwise the first exec wouldn't get any argv[0] passed. Only tested it on some simple small programs so far, but I'll start us...