I just put up kmod.c.v6 on http://athena.res.cmu.edu/kmod/
Changes from the last version:
passing of slot to run_modprobe (Thanks to Edouard Parmelan)
Complete switch-over to using signals to activate kmod. Before everyone
starts commenting on how hackish that is, let me explain my reasoning. A
signal can be delivered to a process at any time. This eliminates the
race condition of a wake_up to kmod being issued while kmod is running.
We ensure that every signal results in processing of the slot list.
Anyway, I still can't figure out why signal handlers aren't being called
when signals are being issued to kmod, though it seems necessary for a
signal handler to be installed, for the signal to interrupt kmod. <shrug>
Anyway - this seems to be running okay for me. Can people (preferably on
Alphas and Sparcs as well as intel hardware) test this out? I'd like to
send it to Linus soon...
Greg Zornetzer - gaz+@andrew.cmu.edu
"Light shines brightest in the darkest night"
http://www.contrib.andrew.cmu.edu/~gaz
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu