0000721: Random invalidSizeOffsetCombination error

(0002231)
malcom2073 (manager)
2012-09-28 02:27

I am currently unable to reproduce this issue. If you can reproduce, please provide steps required to do so, and pastebin the ENTIRE application console output.

(0002233)
malcom2073 (manager)
2012-09-28 02:32

Please provide more information so I can diagnose. Console pastebin of the entire application, and steps to reproduce reliably.

(0002235)
ToxicGumbo (reporter)
2012-09-28 03:51

Steps to repeat using build fa5aaf3-DEV (version first noticed):

- Run emstudio build from Ubuntu command line on 64bit machine
- Let the ECU interrogation complete
- When the "EMS Info" screen comes up, double-click on "VE Table" (first entry)
- On the pop-up table view, click "Save to Flash"
- Error pops up: "Command failed with error: invalidSizeOffsetCombination"

This also happens under Windows7/64 with your prebuilt EXE of the same hash. This is using a Prolific pl2303 serial adapter with a TA card and firmware 0.2.0-SNAPSHOT-104-g3f59fde-DEV. It should be noted that any changes to the RAM location will be properly stored and still exhibit the same error.

Full Ubuntu run from command line here, including clicking OK on the error and exiting:
http://pastebin.com/Jk9v6nCL [^]

(0002236)
ToxicGumbo (reporter)
2012-09-28 04:17

DeuceEFI called me out on the firmware being way out of date. I made an amateur mistake of having two firmware folders while on Windows and ended up loading an s19 file from the wrong folder. Apologies.

Tests fine with fa5aaf3-DEV and firmware a5ebb52-DEV.

Closing.

(0002238)
Fred (administrator)
2012-09-28 07:00

Jeff, what was the firmware version that you saw this with? Please don't assume that it was not OK. Ask me instead. Hash of old one, please?

(0002239)
ToxicGumbo (reporter)
2012-09-28 08:03

Abbreviated hash mentioned above. Full hash: 3f59fde9c4601df7be8e37a783e44357c5876ebe

(0002240)
Fred (administrator)
2012-09-28 12:23

Apologies! I didn't see it there. Yes, you were right, and here is the likely reason:

fred@cheetah:~/workspaces/eclipse/freeems-vanilla/src$ git log --oneline 3f59fde..master |grep API
f1d97fe Make it OK to have size AND offset zero for whole block burn. This is consistent with other location access API calls.

Notes
(0002231) malcom2073 (manager) 2012-09-28 02:27	I am currently unable to reproduce this issue. If you can reproduce, please provide steps required to do so, and pastebin the ENTIRE application console output.

(0002233) malcom2073 (manager) 2012-09-28 02:32	Please provide more information so I can diagnose. Console pastebin of the entire application, and steps to reproduce reliably.

(0002235) ToxicGumbo (reporter) 2012-09-28 03:51	Steps to repeat using build fa5aaf3-DEV (version first noticed): - Run emstudio build from Ubuntu command line on 64bit machine - Let the ECU interrogation complete - When the "EMS Info" screen comes up, double-click on "VE Table" (first entry) - On the pop-up table view, click "Save to Flash" - Error pops up: "Command failed with error: invalidSizeOffsetCombination" This also happens under Windows7/64 with your prebuilt EXE of the same hash. This is using a Prolific pl2303 serial adapter with a TA card and firmware 0.2.0-SNAPSHOT-104-g3f59fde-DEV. It should be noted that any changes to the RAM location will be properly stored and still exhibit the same error. Full Ubuntu run from command line here, including clicking OK on the error and exiting: http://pastebin.com/Jk9v6nCL [^]

(0002236) ToxicGumbo (reporter) 2012-09-28 04:17	DeuceEFI called me out on the firmware being way out of date. I made an amateur mistake of having two firmware folders while on Windows and ended up loading an s19 file from the wrong folder. Apologies. Tests fine with fa5aaf3-DEV and firmware a5ebb52-DEV. Closing.

(0002238) Fred (administrator) 2012-09-28 07:00	Jeff, what was the firmware version that you saw this with? Please don't assume that it was not OK. Ask me instead. Hash of old one, please?

(0002239) ToxicGumbo (reporter) 2012-09-28 08:03	Abbreviated hash mentioned above. Full hash: 3f59fde9c4601df7be8e37a783e44357c5876ebe

(0002240) Fred (administrator) 2012-09-28 12:23	Apologies! I didn't see it there. Yes, you were right, and here is the likely reason: fred@cheetah:~/workspaces/eclipse/freeems-vanilla/src$ git log --oneline 3f59fde..master \|grep API f1d97fe Make it OK to have size AND offset zero for whole block burn. This is consistent with other location access API calls.

Relationships