Is there a way to explicitly ensure that my driver traces are flushed to the disk and safe.
This is in a case where I suspect my next line of code will crash the system (with no memory dump)
but I want to keep my traces up to this point.
thanks
No. The answer is to hook up a kernel debugger. The kernel debugger will get every byte of your debug prints.
Thanks but since I’ll be sending the instrumented driver to test sights that cannot set up a debugger,
I can specify once a second flush of traces, and since this is not a time critical code path
I will capture the state to trace file, delay for a second to get a flush and do the suspect register read.