pcmhacking.net

Latest log ... There is not much I like more then a good challenge and am always up to them, however, I am a bit confused (being nice)

NSFW wrote:but if you can figure out why the read requests had to be sent twice you might only need half as much patience.

that it wasn't mentioned that the Allpro has the EXACT same issue thus I beat my head thinking it was an issue with the Sparkfun board only when it is most likely a software issue.

On the good side of that ... I have learned a BUNCH, so I guess that's 'the price of an education'!
Don't get my words wrong, I'm super chuffed!

Now, armed with that new information maybe it can get resolved.

So the Sparkfun board works as good as the Allpro just a helluva lot slower!
Comparing logs (test kernel, read full) between the two 'tools' are basically the same except for speed.
All the same 'errors', 'no data', 'no responses', 'Timeouts', etc..., etc...
I have been thinking it was all because of the Sparkfun board.

BTW: That read was done with NO delays in kernel ReadMode35() ... The Sparkfun board can handle no more then ~5-7 delay with ElmPause(...).

It took me a while to realize that the AllPro was having the same problem. When I was testing the C kernel I was so exited to see it work that I didn't realize it was working slowly. Sorry about that!

NSFW wrote:When I was testing the C kernel I was so exited to see it work that I didn't realize it was working slowly.

Well yea, why wouldn't ya, it's something to accomplish a goal one is working on, if it wasn't then why would ya be doing it!

NSFW wrote:Sorry about that!

But excitement is no excuse for failure ... Fix it!
Yes, I'm kidding, I have no doubt it'll be fixed in good time, your work speaks for itself.

Maybe I'll suddenly get smart enough to fix it ...
One can dream!

I don't think the previous transaction is fully completed in some way, missed/lost completion code or something, but what do I know about this ... very little!

-Enjoy

The elm protocol requires a poll, and is synchronous. It sucks, and makes this harder than it should be, but its what you have to support for the cheap interfaces that were not designed for this. It certainly complicates the time sensitive parts for half the supported interfaces.

I don't know you (or nsfw) ... but I feel confident saying this.

If it was easy, you wouldn't be doing it!

Neither one of ya!

I know I wouldn't be, but I am nowhere near y'alls level in this stuff, never will be.
Doesn't stop me from trying.

-Enjoy

Thats right! I wrote ls1flash which reads the pcm, and its pcm kernel. I got write working but it wasnt stable due to the lack of my microsoft skills (whole thing not and object oriented app and runs in the ui thread so touch the window with the mouse and brick your pcm!). Nsfw did the open source subaru flash tools then bought a chev and wanted to make an equivalent for gm. We met here and the rest is history (...in the making?).

I helped out with open source Subaru stuff, but most of that was actually established before I got a Subaru (it was half the reason I chose to buy a Subaru). The Subaru flash tools were actually done by a company called Tactrix - it's not open source, but their interface is $150ish and their software saves raw binaries, which is really important. Most of the Subaru community revolves around an open source editor / datalogger called RomRaider where a bunch of people got involved with DIY tuning, dissasembly, hacking, companion utilties, etc. It's where I learned tuning, and that led to writing some utilities, doing some reverse engineering, modifying code, etc. So I was involved, but the flash tools were already there by the time I showed up.

But then I bought an 02 Corvette... and the tools weren't quite there yet.

I know Windows stuff, but needed help with embedded stuff... Antus knows embedded stuff, but needed help with Windows stuff...

The tools are almost ready...

Quit jabbering with me and GET TO WORK ...


[edit]
Whoa, wait a minute here ... What am I saying, if you quit jabbering with me and go to work I'll learn less! BAH HUMBUG!
[/edit]

Hi all,

Forget it, sorry, my bad ... Has something to do with my twisted PcmH, not the kernel, it's not y'alls issue, I'm apparently not getting the kernel loaded correctly.

[delete]
Mucking with micro-kernel trying to educate myself, this is what I see ... Please do correct me if I'm wrong!

If ReadMessage() returns 0, it should echo a toolPresent '8CFEF03F' + added diagnostic bytes, and continue.
else it should echo the command sent prefixed with '6CF010AA' + length, completion code and read state.
unless the command was '36E0', then it should echo '6CF01036E080' and continue.

Never see anything periodically.
Send '36E0' and it echos '6CF0107F36E02277'.
What am I missing, Where is that coming from?
[/delete]

Thanks

-Enjoy

Hi all,

How is the Kernels base address figured?

Thank you