On Fri, 2007-06-15 at 11:59 -0400, Zhu Han wrote: > Hi, Kieran, > > I'm just wonder why you try to acquire the lock and increase the > hooks_usecount each time when you use the hook routine. Is there any > generic ways to synchronze the code path using hook routines and > netfront_accelerator_unloaded, considering you can synchronize the > tx/rx data path easily.
The lock protects the use_count variable. The use_count variable prevents the plugin module unloading while it is being used. I couldn't just use the lock to prevent the module unloading as the hook function (i) might block (and holding a spin_lock would be rather antisocial) (ii) might call back into netfront and try to take the lock again, which would deadlock. The data path hooks do not block, and are already protected by locks, so these are also taken when trying to unload the plugin module. For this reason it's not necessary to use the hooks_usecount on the data path. I think that RCU would only work in this situation if the hook functions didn't block, and wouldn't affect the data path locking overhead as it wouldn't be necessary there. Kieran - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
