From mboxrd@z Thu Jan 1 00:00:00 1970 From: jgunthorpe@obsidianresearch.com (Jason Gunthorpe) Date: Wed, 8 Jul 2015 17:16:12 -0600 Subject: [PATCH v4 3/5] tee: generic TEE subsystem In-Reply-To: <20150708223325.GA5843@kroah.com> References: <1436350592-7732-1-git-send-email-jens.wiklander@linaro.org> <1436350592-7732-4-git-send-email-jens.wiklander@linaro.org> <20150708171026.GA11740@obsidianresearch.com> <20150708211129.GA29824@kroah.com> <20150708222649.GA20068@obsidianresearch.com> <20150708223325.GA5843@kroah.com> Message-ID: <20150708231612.GB20068@obsidianresearch.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Wed, Jul 08, 2015 at 03:33:25PM -0700, Greg Kroah-Hartman wrote: > > The basic issue is that cdev_del doesn't seem to be synchronizing. > > > > The use after free race is then something like: > > > > struct tpm_chip { > > struct device dev; > > struct cdev cdev; > > Oops, right there's your problem. You can't have two reference counted > objects trying to manage the memory of a single structure. No matter > what you do, it's going to be a pain to deal with this, so don't :) Sure, generally, yes, but that isn't done for no reason, it is to make open straightforward: static int tpm_open(struct inode *inode, struct file *file) { struct tpm_chip *chip = container_of(inode->i_cdev, struct tpm_chip, cdev); We need to recover the tpm_chip associated with the char device node, in a way that is holding a kref on it, without racing with cdev_del/etc This scheme does mean that if we have a struct file we have a kref on the cdev, and if we have cdev then we have a kref on the tpm_chip, which is really easy to use properly. > > Ie we need cdev to hold a ref on tpm_chip->dev until cdev_put is > > called. > > No, separate them, make the cdev a pointer and all should be fine. Okay, cdev_alloc takes care of the cdev lifetime. Do you have a simple solution to replace container_of as well? What would you think about something like: cdev_alloc(&chip->dev.kref) ? Jason