P08 Support Information

[kur4o]

I am not quiet sure but think earlier revisions used the v6 dlc registers, while the later ones use p01 dlc registers, but you mentioned at some point the same code was working on both pcms, so it might not be the case.

Can you log the bus with external tools, if there is some missed responses it will be logged.Try x1 too.

[Gampy]

[04:30:22:592] Sending upload request for kernel size 1171, loadaddress FFA800

Hmm, wrong address ... Should be FFB000.

I know it was B000 that was used with the first read.

Tried both I assume as Kernel only no Loader ??

-Enjoy

[antus]

The FFA800 address is kernel, the FFB000 address is loader, as from develop:

Code: Select all

                // P08
                case 9364970:
                case 12206029:
                case 12208773:
                    this.Description = "P08";
                    this.IsSupported = false;
                    this.LoaderRequired = true;
                    this.ValidationMethod = PcmType.P08;
                    this.HardwareType = PcmType.P08;
                    this.KernelFileName = "Kernel-P08.bin";
                    this.KernelBaseAddress = 0xFFA800;
                    this.LoaderFileName = "Loader-P08.bin";
                    this.LoaderBaseAddress = 0xFFB000;
                    this.ImageBaseAddress = 0x0;
                    this.ImageSize = 512 * 1024;
                    //this.RAMSize = 0x4DFF;
                    this.KeyAlgorithm = 13;
                    this.ChecksumSupport = false;
                    this.FlashCRCSupport = true;
                    this.FlashIDSupport = true;
                    this.KernelVersionSupport = true;
                    //this.KernelMaxBlockSize = 4096;
                    break;

[antus]

@gampy I still had the test build you first sent me laying around. The flash CRC is not implemented, but I ran it up and it read the P08 fine again. So whatever the difference is, maybe we can compare this kernel in a disassembler and the current one. If you have source for the test build, even better!

Having said that, it might not be anything obvious as very small changes to anything seemed to break this one too. It might be more luck than a real fix.

Tonights log of the first testing build from last week attached. If anyone wants to try the RE and compare angle and has a P08 the kernel can be grabbed out the log. Or if anyone has a P08 and tries the test build earlier in the thread and it does not work like mine, let us know and we can send this one to try.

[antus]

Ok looks like we have it working. A combination of different problems, mostly of my own making, but everything has to be perfect for this stuff to work. So loading the kernel, as is, and building it for FFB000 is working for me. The loader is not. I definitely took the road through points c, d and e, to get from a to b on this one. Doh! The loader is something I'll come back to.

[Gampy]

Antus,

Your PCM is newer then mine, yours is a 2000 Alero 2.4L, mine is a 98 S10 2.2L.

My numbers are,

Code: Select all

     Year .........................................: 1998
     Make ........................................: Chevrolet
     Model .......................................: S-10 Ext Cab 123" WB LS
     Engine ......................................: 2.2L L4 (LN2)
     Transmission ..............................: Automatic/Manual (M50, M30, MW2)
     Service Number ..........................: 16228016
     Hardware Number ........................: 16204850
     Flash Chip ID Code ......................: Intel 28F400B, 512kb (0x0089889D)
     Operating System Number (OsID)....: 9364970
     Calibration Number ......................: 9365003
     Serial Number .............................: CTXJSZ8324
     Vehicle Identification Number ........: 1GCCS1941WK140476

-Enjoy

[zack4200]

I made a couple more in-vehicle reads, one with Gampy's build 824.13 that I used originally, as well as CheckBuild56 from a few posts back. I halted the kernel after both but still ended up with a security light and no start. I'm not sure whether it's really beneficial to include the logs/bin from the older version, but I'll attach them just in case it is helpful.

[Gampy]

Thank you!

We do appreciate the tests very much!

You should delete the old test build 824.13 ... It is out of date!
CheckBuild 56 includes everything in that test build.

As far as the codes and no start ... I am told that is going to happen with some brain dead modules, they'll need to be restarted.
We do send a command to all modules on the bus to be quite in the beginning, and we also send the command at the end to tell them it's ok to start chattering again, some modules are just like some people, they don't listen or obey!

-Enjoy

[zack4200]

Thank you!

We do appreciate the tests very much!

You should delete the old test build 824.13 ... It is out of date!
will do, I mostly used that build again out of curiosity to see if manually halting the kernel after would change anything with the no start.
CheckBuild 56 includes everything in that test build.

As far as the codes and no start ... I am told that is going to happen with some brain dead modules, they'll need to be restarted.
We do send a command to all modules on the bus to be quite in the beginning, and we also send the command at the end to tell them it's ok to start chattering again, some modules are just like some people, they don't listen or obey!
ah okay good to know. I guess I really should stop being lazy and just pull out the PCM and bench read it. or better yet take a quick trip to the junkyard and get a couple spares.

-Enjoy

[Gampy]

Spares are the easiest way to go IMO ... That way you swap a PCM rather then Pull, Program, Install.
The problem with that technique is you lose the accumulated learned settings when you swap, it'll have the same last learned settings, of course depending on how the PCM is written.

I know it would be easier for this old man then to crawl in and out from under the dash, not a problem for you young kids though.

-Enjoy