On Oct 20, 2011, at 10:02, Wolfgang Denk wrote:
Dear "Moffett, Kyle D", In message 8B4AC84D-1F22-4326-B75A-FB3CC39A5CF7@boeing.com you wrote:
Would you accept a patch which makes it possible for a board to not implement a "reset" command at all?
There are a few places in common/cmd_bootm.c which are converted to use panic("...") instead of printf("...")+do_reset().
This is not acceptable, as changes behaviour: panic() will halt the system, not reset it.
That is obviously wrong, as a 5 second glance at panic() in the file lib/vsprintf.c would tell you. For 540 of the 567 board configs in the include/configs/ directory, panic will directly call do_reset(); only 27 of the configs (less than 5%) set CONFIG_PANIC_HANG.
The only change with the patch for boards without CONFIG_PANIC_HANG is that panic() has an extra udelay() to ensure that the serial console messages go out before the reset.
For the boards that *do* set CONFIG_PANIC_HANG, none of the fatal errors in common/cmd_bootm.c should cause U-Boot to reset, they are all valid panic() conditions, such as GZIP overwrite errors and fatal image format issues. In those cases this is also a bugfix.
Patch to follow shortly.
Cheers, Kyle Moffett
-- Curious about my work on the Debian powerpcspe port? I'm keeping a blog here: http://pureperl.blogspot.com/