forums.ps2dev.org

Are you sure this is how it works? (This is a naive question, don't look for any hidden thing.) And if it's indeed the way it works, do we know whether the 1st stage bootloader code can be changed by a firmware? If so, then validation of the OtherOS part would be needed anyway, though it'd probably ...

That's a very good point. So let's hope they will keep it as is for us early adopters :)

So they don't want to pay to implement virtualization of new components, and they would keep support for all of old hardware in upcoming releases? In all companies I have worked for, no matter their sizes, that wouldn't be called cost effective. And you know, we'd better be sure than being sorry whe...

If that's a software cost issue only, then the Other Boot support will be removed from all of the upcoming firmwares, no matter whether you install it on a Slim PS3 or not.

So much for paranoia: Jim just posted a patch that works with FW 2.50.

EDIT: Here is the link.

A few things:
- abs is not in math.h/libm, it's part of libc
- try to move -lm at end of link command-line

1. I posted the link to IBM result in my previous post (it's close to the end of the article)
2. If you want to search SPEC 2000 results use the official site: http://www.spec.org/cgi-bin/osgresults?conf=cpu2000

In order, no (or poor can't remember :D) branch prediction. IBM has posted SPECint 2k results of 423 and fp of 387 ( ref ). That basically means the PPU has a similar performance of a PIII 800 MHz for integers and about 10% better than the PIII for FP. So if you don't use the Cell "properly&quo...

Stupid question just in case: do you use -O2 or -O3 when you compile?

And again: the Cell PPU is a very poor processor, using it alone will only be giving you very deceiving results.

What d-range meant is that your for loops write 1100 x 800 words per frame and that doing it this way on the PPU is far from optimal. The "standard" way to write big amounts of data on the Cell is by using DMA with multiple SPU's. However even at 50 fps, we are talking of 1100 x 800 x 4 x ...

I'm sorry for this noobish question but I have only worked on little endian systems, are there any preprocessor commands that will have a Big Endian system revert to Little Endian? Or must I be byte swapping all my data from right->left to left->right? It's a property of the processor not of the la...

Now I just need to figure out why my app reports I'm out of memory. Perhaps I should just have it run as an otheros instead of trying to share resources within Ubuntu. Funny thing is, I figured it would be a straight port to PS3/UBuntu since I'm using SDL framework and the port from x86/Windows to ...

Your program works on little-endian machines, but the PS3 is big-endian.

That's a bit extreme :) Try something like this: apt-get install linux-source-2.6.12 (replacing the version number with the kernel you have). But if you really don't feel at home using Ubuntu, I am not sure other distros will make it easier for you. I am not trying to look negative, I just warn you ;)

that's it, and I've been looking the source for one day, I have found ps3fb.c on the net, but... How should I install it ? No... You have to install the kernel source package (sorry don't know how to do that for Ubuntu, I only use rpm based distros). I am afraid that if you don't know what I mean, ...

It seems you only have the kernel headers installed, not the source.

phaielip wrote:I thought that's why I haven't applied the " Glaurung's kernel patch", but here is another problem : I can't find the "ps3fb.c" file on my system to apply the patch, and I don't even hnow were it is supposed to be ?

It's in the kernel sources (first hit with google :-) )

unsolo wrote:Its so great to se the one ppc only distribution not following the rules...

Is it ydl 6 that doesn't follow the rules or your hacked Gentoo distrib? :)

spu_yuv2argb_scaler_handle spu_yuv2argb_scaler spu_yuv2argb_scaler.eo make[1]: spu_yuv2argb_scaler_handle: Command not found make[1]: *** [spu_yuv2argb_scaler.eo] Error 127 make[1]: Leaving directory `/u...

Ref: ydl ml Yellow Dog Enthusiasts, YDL v6.0 was today delivered to the beta test team comprised of TSS staff, individuals from chip vendors and Power OEMs, and community members. YDL v6.0 is a complete rebuild against the CentOS foundation with select Fedora 7 components. The final shipping version...

Well I don't know how I can convince you. I work for a company that does GPUs, so I have some inside view, which of course I can't share unless you sign an NDA ;) I can't really convince you either for the same reasons even though I have the PS3 OpenGL driver code sitting right in front of me. :) H...

Well I don't know how I can convince you. I work for a company that does GPUs, so I have some inside view, which of course I can't share unless you sign an NDA ;)

aegiswings wrote:OpenGL *is* all done in hardware these days. A prize to anyone who can point out one non-trivially part of OpenGL that isn't implemented in hardware?

OpenGL shading language compilation? What's my prize? :)

Long gone are the days when the hardware is simply used to accelerate parts of the rendering pipeline. Actually what NVIDIA hardware provides is very close to a full OpenGL interface (I don't know about ATI) and it's become impossible to interrupt that pipeline and replaces parts of it with softwar...

ps2devman wrote:If a coder makes the effort to code at low level, he gets in no time something incredibly efficient. But of course, non coders will ask again and again something allowing immediate port of existing sources, requiring high level apis.

Yeah real coders only talk to baremetal in assembly language :)

I don't think Sony is all evil. But it's a big company with many dev teams and marketing teams, all with different targets and understanding.
For instance, I am ready to bet that the Sony guys who are working on the PS3 kernel don't even know all what is in a firmware upgrade...

I found on the Wiki of Second Life that some of the required libraries are only available in binary format, so I guess either it's impossible to get it to work on PS3, or you would have to remove some features.

d-range wrote:There's nothing wrong with templates for performance-critical code

I agree, except that deciphering error messages with templates can be a real pain.

Also always declare copy constructors as private methods to avoid using them by accident :D

Did I say I hate C++ ? ;)

The code bloat argument is irrelevant if you know what you're doing. That's the problem with C++, unless you know what you're doing, you can quickly make underperforming code. C++ adds virtually no runtime overhead, as long as you don't throw in virtual functions everywhere and start deriving every...

If anyone out there has a 1Gbit network card already in their PC, please try this out and see if the results come out good. Let's do some optimistic math: let's say your 1 Gbit network can deliver 100 MB/s. A full HD picture is 9MB so at most you will get 11-12 FPS if you use it directly. So basica...