Environment in EEPROM on MPC5200 based board
Hi all,
I'm using Lite5200 Board and tried to configure U-Boot to have environment in EEPROM. I use Hard I2C driver an everything works fine when the environment is in flash.
When setting CFG_ENV_IS_IN_EEPROM U-Boot don't start (hang-up before any messages are printed on screen).
If anyone know something about this problem, any help is greatly appreciated.
Thanks, Sebastian
Hallo, Herr Schau!
Quoted 3 lines · In message <[email protected]> you wrote:
I'm using Lite5200 Board and tried to configure U-Boot to have environment in EEPROM. I use Hard I2C driver an everything works fine
Why don't you keep the environment in flash as it is no?
In my opinion using the environment in EEPROM is just asking for trouble. It has serious disadvantages:
* Loooooong boot times, which are also not deterministic (just setting the board to a new console baud rate can double the boot time, etc.)
* Less reliable -- in flash, you can have redundand flash sectors; so far I never heard of a single case of data loss in such a configuration. In EEPROM, you may experience the infamous "I2C Edge Conditions" problem (see doc/I2C_Edge_Conditions) - I have seen cases where a failing power supply caused slow voltage rise times; when powering on such a board, the CPU would start to run, then the watchdog would resert it because the voltages were still too low. This happened 2...5 times for each power on. In the end the CPU would run fine, but in 2 out of 3 boot cycles the environment in EEPROM was lost because of the Edge Conditions problem.
I _strongly_ recommend NOT to put the envrionment in EEPROM (nor any other critical data).
Quoted 2 lines
When setting
CFG_ENV_IS_IN_EEPROMU-Boot don't start (hang-up before any messages are printed on screen).
Maybe you have some debugging enabled? This will try to use printf() to the console when the "badrate" variable is being read to initialize the console --> deadlock.
Quoted 2 lines
If anyone know something about this problem, any help is greatly appreciated.
Just don't do it.
Best regards,
Wolfgang Denk
Signature · 4 lines
Software Engineering: Embedded and Realtime Systems, Embedded Linux Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de If a group of N persons implements a COBOL compiler, there will be N-1 passes. Someone in the group has to be the manager. - T. Cheatham