Messages in this thread | | | Date | Wed, 12 Dec 2012 16:15:18 -0800 | From | Greg KH <> | Subject | Re: [PATCH 2/2] uio: do not expose inode to uio open/release hooks |
| |
On Thu, Dec 13, 2012 at 01:08:54AM +0100, Hans J. Koch wrote: > On Wed, Dec 12, 2012 at 07:08:18AM -0800, Greg KH wrote: > > On Wed, Dec 12, 2012 at 09:56:16AM +0100, Benedikt Spranger wrote: > > > Am Wed, 12 Dec 2012 09:50:54 +0100 > > > schrieb "Hans J. Koch" <hjk@hansjkoch.de>: > > > > > > > On Tue, Dec 11, 2012 at 08:46:48PM -0800, Greg KH wrote: > > > > > Yes, but what does that have to do with this in-kernel, internal api? > > > > > > > > Ah, OK. You're right, the commit message is confusing. > > > > > > > > Bene, it's enough to say we drop the inode parameter because nobody > > > > ever needed it. > > > I am fine with that. > > > > > > > I cannot see why this also helps with the other problem. > > > It would help, because we can defer calling the release hook until the > > > last mmap user is gone. In this case the inode pointer may not be valid > > > anymore. > > > > Which, again, is the same for any in-kernel driver with these types of > > callbacks. > > Is that a general mmap problem that wants to be fixed?
Not that I know of, but I haven't messed around in this area of the kernel in a long time, sorry.
greg k-h
| |