--- In [email protected], Thomas Treyer <[EMAIL PROTECTED]> wrote: > > Hi Geert, > > I tried the I2C driver again. Now I have no problem any more, no more > crashes can be observed. All I did was to make a new fimage, and now > I get from I2C_WRITE the error EI2CBUSNFREE, which is correct, > because up to now I have no I2C slaves connected to the foxboard. > > I am sorry for the confusion, and I cannot explain what changed > between my old and new fimage. At least I cannot tell anything about > relevant changes, all I did was to add one directory- and one file- > command to the post file. > > I am pretty sure now, that the kernel crash problem has nothing to do > with your code, sorry for that.
No problem, Thomas. Pardon me for not coming back on your earlier question, but my previous answer was a kind of "silly". Of course, I tried my code without hardware attached (I told you I didn't, but that was the "silly" answer). Whenever there's no communication possible any more, the code "kicks out", like you just have experienced yourself. I think that's all there is to do. It's always such that the caller of the function *must* check the return value of that function and act appropriate. Best rgds, --Geert
