PDA

View Full Version : Please read before posting bug reports



Blacky
April 23rd, 2008, 07:38 PM
Earlier this week a problem in the V2.5.16(0414) firmware was uncovered. Actually it effects all V2.05.xx firmware versions. It had to do with the C compiler producing incorrect code and causing CPU registers to be overwritten.

The results of overwriting CPU register values will cause as many different and seemingly unrelated bugs as you can imagine. A new firmware version will be released as soon as I get confirmation from the compiler manufacturer of the workaround I intend to use.

Please don't post anymore bugs related to any firmware earlier than 2.05.17 (which will be the next release). There is every possibility that any bug you encounter is caused by the problem described above.

Once the new firmware is available, please retry any functions/operations that have failed with earlier versions. Be sure to reformat all file systems (internal and SD card) before trying again with the new firmware.

Regards
Paul

Garry
April 24th, 2008, 03:39 AM
Whoa ... bugs like that suck ... especially as it can be hard or next to impossible to locate unless you can take a look at the asm code created ...

Blacky
April 24th, 2008, 08:29 AM
Whoa ... bugs like that suck ... especially as it can be hard or next to impossible to locate unless you can take a look at the asm code created ...
That's how I found it - yeah I was a bit annoyed. :cussing:

The assembler for a case statement (with case variable in R0) should have been:

A840 MOV R4,[R0]
06F4F5FF ADD R4,#0FFF5H
2D26 JMPR cc_Z,?C0077 But it was incorrectly compiled as:

A8E0 MOV R14,[R0]
F04E MOV R4,R14
06F4F5FF ADD R4,#0FFF5H
2D26 JMPR cc_Z,?C0077 Turns out the compiler was saving the value of the case variable in R14 to be re-used in statements within the same case block. Unfortunately the compiler "forgot" that it was using R14 for something else.

Regards
Paul

nqcv8
April 24th, 2008, 07:43 PM
Earlier this week a problem in the V2.5.16(0414) firmware was uncovered. Actually it effects all V2.05.xx firmware versions. It had to do with the C compiler producing incorrect code and causing CPU registers to be overwritten.

The results of overwriting CPU register values will cause as many different and seemingly unrelated bugs as you can imagine. A new firmware version will be released as soon as I get confirmation from the compiler manufacturer of the workaround I intend to use.

Please don't post anymore bugs related to any firmware earlier than 2.05.17 (which will be the next release). There is every possibility that any bug you encounter is caused by the problem described above.

Once the new firmware is available, please retry any functions/operations that have failed with earlier versions. Be sure to reformat all file systems (internal and SD card) before trying again with the new firmware.

Regards
Paul


Paul

any idea when the new version will be released?...I am using 2.05.16(414) and its causing me some grief... mainly takes several attempts the see the controller when scanning ..and gives errors when trying to log. Seems to work a little better if you start logging without the WB02 connected to the V2 and plugging it in once you have the scan started. Often have Reboot laptop start and stop car and reset V2 to get any sort of logging working.

Don t know if this is related or not.

Cheers

Sean

Blacky
April 24th, 2008, 10:07 PM
Within 24-48 hours.
Regards
Paul

nqcv8
April 24th, 2008, 10:18 PM
Within 24-48 hours.
Regards
Paul

:cheers: Paul

Thanks for your quick reply :)

joecar
April 25th, 2008, 01:40 AM
Funny thing is, I did a cal-only flash and tested the DVT the days before you posted this thread Paul, and it all went fine (on my 2001 LS1).

:good:

JoshH
April 25th, 2008, 07:26 AM
Would this cause problems with the ECM/TCM or is it a FlashScan problem? The reason I ask is because the other day I tried to do some BB logging. I got an error, and ever since then, I've had a CEL that comes on as soon as I start my truck. I can clear it, but if I try to check the DTCs to see why it's on, there is nothing that shows up.

Blacky
April 25th, 2008, 08:18 AM
Funny thing is, I did a cal-only flash and tested the DVT the days before you posted this thread Paul, and it all went fine (on my 2001 LS1).

:good:

It is difficult to predict what can go wrong because of this fault. The three main problems seem to be:
1. Screen displays get messed up.
2. File system gets corrupted and files don't save properly or just disappear.
3. USB comms can lock up.

Regards
Paul

Blacky
April 25th, 2008, 08:22 AM
Would this cause problems with the ECM/TCM or is it a FlashScan problem? The reason I ask is because the other day I tried to do some BB logging. I got an error, and ever since then, I've had a CEL that comes on as soon as I start my truck. I can clear it, but if I try to check the DTCs to see why it's on, there is nothing that shows up.

BBL can not change anything in your ECM/TCM. Have you tried removed power from the ECM to allow it to reset completely?

Are you checking DTCs using FlahScan as a hand held device or as a pass through to a laptop?

If hand held, after you check the DTCs, can you save the trace file, copy the trace file back to the PC using EFILive_Explorer and email it to me.

If pass through, can you generate an error report (menu: File->Generate error report), save it and email it to me.

Email to paul@efilive.com
Include a link to this thread so I can correlate the two.

Regards
Paul

rdaniels
April 25th, 2008, 08:59 AM
Paul, I am also seeing a strange problem. I did a calibration update to my ECM. The truck will no longer remote start. V2 shows no DTC either direct or through the scan tool. My stand alone code reader will not connect?
Rock

JoshH
April 25th, 2008, 12:56 PM
BBL can not change anything in your ECM/TCM. Have you tried removed power from the ECM to allow it to reset completely?

Are you checking DTCs using FlahScan as a hand held device or as a pass through to a laptop?

If hand held, after you check the DTCs, can you save the trace file, copy the trace file back to the PC using EFILive_Explorer and email it to me.

If pass through, can you generate an error report (menu: File->Generate error report), save it and email it to me.

Email to paul@efilive.com
Include a link to this thread so I can correlate the two.

Regards
Paul
I tried reading DTCs with both methods and neither one will show any codes. I'll try to get a trace file and an error report, but I'm not sure I know how to make a trace file or copy it back to EFILive_Explorer.

GMPX
April 26th, 2008, 10:46 AM
Paul, I am also seeing a strange problem. I did a calibration update to my ECM. The truck will no longer remote start. V2 shows no DTC either direct or through the scan tool. My stand alone code reader will not connect?
Rock

What vehicle type is this happening on? Diesel or gas?

Cheers,
Ross

rdaniels
April 26th, 2008, 06:58 PM
What vehicle type is this happening on? Diesel or gas?

Cheers,
Ross

LMM Diesel

azzhauler
June 3rd, 2008, 03:13 AM
If this is just a firmware issue, it wouldn't affect me if I haven't updated any firmware correct? I have only extracted the files and run the scanner/tuner, and uploaded calibrations to the car(unless it does an automatic firmware upgrade). I have noticed on 2 occasions in the past few days where I uploaded a cal to my car and after the guages did not reset and would not reset with key on/off until I reuploaded the file.

Blacky
June 3rd, 2008, 08:09 AM
If this is just a firmware issue, it wouldn't affect me if I haven't updated any firmware correct? I have only extracted the files and run the scanner/tuner, and uploaded calibrations to the car(unless it does an automatic firmware upgrade). I have noticed on 2 occasions in the past few days where I uploaded a cal to my car and after the guages did not reset and would not reset with key on/off until I reuploaded the file.

This thread was explaining a boot block problem. If your boot block is NOT version 2.05.06 then this problem will not affect you.
Regards
Paul

knagel
June 22nd, 2008, 03:26 PM
what was the solution to joshh mystery codes? I had the same thing happen to me yesterday. Got done with a pull and lost power and set a CEL. checked with scan tool and flashscan and no codes. Didn't read this post til now so didn't generate a bug report or save anything on flashscan. I just pressed clear codes on flashscan and light went away.

kev

Blacky
June 22nd, 2008, 04:36 PM
The LBZ/LMM controllers are bosch controllers and they reply to the FlashScan scan tool a little differently than the Delphi controllers.

It ended up being a timing issue. FlashScan was just not waiting long enough for the Bosch controller to send the codes.

A new version that fixes the problem is available here:
http://forum.efilive.com/showthread.php?t=8384
Just posted an hour ago.

Regards
Paul

knagel
June 23rd, 2008, 01:00 AM
I used auto update on 8.1.1 and it worked but on 7.5.4 it said there were no updates. My scan tool and tune tool still say build 62??? Do I need to install off web site and if so, do I need to make "old" folders as to not overwrite files?

Kevin

Blacky
June 23rd, 2008, 09:33 AM
I used auto update on 8.1.1 and it worked but on 7.5.4 it said there were no updates. My scan tool and tune tool still say build 62???

Yes, you need to install V7.5.4 from here:
http://forum.efilive.com/showthread.php?t=8384
V7.5.3 has not been released as an auto update install yet.


Do I need to install off web site and if so, do I need to make "old" folders as to not overwrite files?
Kevin

Yes, the V7.5.4 (build 63) install is the last version where you will need to do that. Once this version (build 63) is installed, all future installs will not overwrite your config files.

Installing build 63 will also reset the registry entries for both the Scan and Tune tool. That will cause window sizes and placements and property settings to get reset to install defaults. You will need to go into the properties window in both scan and tune and re-apply any settings that you require.

Regards
Paul