> -----Original Message----- > From: gregkh@xxxxxxxxxxxxxxxxxxx [mailto:gregkh@xxxxxxxxxxxxxxxxxxx] > Sent: Thursday, January 09, 2014 12:15 PM > To: Dan Carpenter > Cc: KY Srinivasan; olaf@xxxxxxxxx; jasowang@xxxxxxxxxx; linux- > kernel@xxxxxxxxxxxxxxx; apw@xxxxxxxxxxxxx; devel@xxxxxxxxxxxxxxxxxxxxxx > Subject: Re: [PATCH 1/1] Drivers: hv: Implement the file copy service > > On Thu, Jan 09, 2014 at 11:09:21PM +0300, Dan Carpenter wrote: > > We've had this discussion before where you urge me to trust the host... > > > > Problem: This code is racy. > > Solution: The host will only send one message at a time. > > > > Now I have to audit the user space code on the host and I don't feel > > like doing that so you win. > > > > I wish we had a better way to do IPC. If kdbus were ready, that might > > have worked for this, and it's a better solution because both sender and > > reciever code will be written in a less trusting way. > > kdbus is almost ready, it might make 3.15, depending on the result of > work that is happening at linux.conf.au and FOSDEM. > > If it would be a better solution for this, that's even more reason to > get kdbus merged soon, no need to add something that doesn't really > work. > > But, how will kdbus help with this? It's a userspace <-> userspace > message transmission bus, would you want the kernel to be a receiver or > sender here? Greg, The transport between the kernel and the user in this driver is a regular char device and Vmbus is the transport between the host and the guest. As you observe, I am not sure how kdbus would be useful. K. Y > > thanks, > > greg k-h _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
