Wolfgang Denk wrote:
Dear Paul Gortmaker,
In message 4AC3C540.9050004@windriver.com you wrote:
If you still think it is best to maintain current behaviour and not stop after the 1st error, that is fine, I can do that, but I just wanted to be sure it was clear why I did it this way.
I have used the code many times (well, to be honest, not sooo many times, but several times) exactly that way: letting it run forever (or, for a long time), while manipulating the hardware (like using a hair dryer resp. cooling spray on it). In such a situation it is very useful when the code does _not_ terminate after the first error (even is this might have been the intention in earlier versions).
Definitely a valid use case. Hopefully one I never have to use personally, mind you.
So beause (1) it is the behaviour users might be used to, (2) I see use cases for this and (3) adding a new option will allow to have both beheaviours so anybody can chose what he wants, I think we should do as I suggested.
OK. I can do that. What about the CONFIG_ALT_MEMTEST then? Should it be changed to run continuously as well, so at least the two tests are consistent in their default behaviours?
Paul.
Best regards,
Wolfgang Denk