Eric BEGOT <eric_begot@yahoo.fr> writes: > Well i'm writing a little driver and I don't know if i have to use > semaphores or spinlocks to synchronize it. I know that spinlocks are > prefered on SMP environment becoz there's no reschedule so no process > switch. > I guess that if the blocking time will be long, i have to use a > semaphore. But on my driver, this time depends on the user operations > and can vary from seconds to minutes. Well in fact, even a second is > long for a processor. But you see what i mean i'm sure :) > By reading the linux source code, i found a lot of spinlocks and few > (less than spinlocks :) ) semaphores. So spinlocks are really the > prefered synchronization method ? The kernel comes with a good general overview of locking methods in Documentation/DocBook/kernel-locking.tmpl, Rusty Russell's "Unreliable Guide To Locking". You can use "make psdocs" to generate that document in PostScript. It's also on the web in a few places, including this one: http://www.kernel.org/pub/linux/kernel/people/rusty/kernel-locking/ -- --Ed L Cashin | PGP public key: ecashin@uga.edu | http://noserose.net/e/pgp/ -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
