Show description

Kern_err Used to report error conditions. For example, device drivers will often use this loglevel to report errors returned by the hardware (as opposed to a hardware failure, which is a bit more serious). ) kern_notice Situations that are normal, but still worthy of note. For example, many kernel routines report security related issues at this log level. kern_info Informational messages. Many device drivers print messages at this loglevel when logging information about the hardware they find and the initialization they perform.

It doesn’t protect much. It simply doesn’t allow one process to initialize the device while another process is reading or writing the structure (or vice versa). This may seem important, but really it isn’t that important since the very next read or write after the device has been reinitialized will fail anyway. up procedure. This increments the semaphore counter and releases any process(es) waiting on the semaphore so they can grab the resource. This is a rather trivial discussion of semaphores.

Printk and the C printf functions is that "<1>" prefix to the format string. This prefix specifies the severity or loglevel of the message. " in your source code. This was done because you’ll probably use (some of) these constants often and their names are unlikely to conflict with names in your code, so saving some typing is worthwhile. The comments appearing next to each constant pretty much state the severity level associated with the prefix. , those that immediately precede a kernel crash (such as an assertion failure right before a NULL reference).