Kernel threads and printf locking question

[John Baldwin]

On Sunday 11 January 2004 03:42 pm, Daan Vreeken [PA4DAN] wrote:
> Hi,
>
> I am writing a kernel module in which I have created a kernel thread with
> kthread_create. For debugging I have added a couple of printf's in the
> thread routine. Running the module on a single-proc system works fine, but
> when running it on one of my SMP machines the console freezes after a
> couple of seconds.
>
> I have been looking through the archives and through some other drivers (
> for example randomdev.c uses a kthread and printf's from it ). I found a
> version of randomdev.c with google that lock's Giant while calling printf,
> but randomdev.c in my src-tree has no locking there. (I'm using
> 5.1-RELEASE) Right now I have changed all printf's in my module into a
> macro-function that checks if ( curproc == our thread ) and locks Giant if
> true, then printf's. (since some of the routines get called both from the
> thread and from other places.)
>
> What is the current state of printf-locking when called from a kernel
> thread? Is locking Giant the only possible way to avoid problems on SMP
> systems, or is there a more fine-grained printf lock available? (on
> 5.1-RELEASE or -current)

printf really probably does require Giant right now, it just happens to be 
used very often after bootup when SMP is running.

-- 
John Baldwin <jhb at FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve"  =  http://www.FreeBSD.org