Re: [PATCH 1/6] livepatch v5: avoid race when checking for state of the patch

[Petr Mladek <pmladek@suse.cz>]

On Wed 2014-12-10 09:25:06, Josh Poimboeuf wrote:
> On Wed, Dec 10, 2014 at 11:11:47AM +0100, Petr Mladek wrote:
> > On Tue 2014-12-09 12:32:49, Josh Poimboeuf wrote:
> > > On Tue, Dec 09, 2014 at 07:05:02PM +0100, Petr Mladek wrote:
> > > > klp_patch_enable() and klp_patch_disable() should check the current state
> > > > of the patch under the klp_lock. Otherwise, it might detect that the operation
> > > > is valid but the situation might change before it takes the lock.
> > > 
> > > Hi Petr,
> > > 
> > > Thanks for the patches.
> > > 
> > > I don't think this patch is necessary.  klp_is_enabled() doesn't check
> > > the state of the patch.  It checks the initialization state of the core
> > > module (klp_root_kobj), which can only be set in klp_init().  It's not
> > > protected by the lock, so I don't see the point of this patch.
> > 
> > Ah, I have misread the name and expected that it checked whether
> > the patch was enabled or disabled. The original code is OK then.
> > 
> > Well, Jiri Kosina pointed out that the check did not make much sense.
> > klp_is_enabled() could not be called if the livepatch module is not
> > loaded. And the later check for klp_patch_is_registered() is enough
> > to check whether the klp_enable_patch()/klp_disable_patch() calls
> > are allowed or not.
> 
> But livepatch isn't a module, it's part of the kernel.

Ah, I remembered that module_init(klp_init) and created a wrong mental link ;-)

> Even if the init
> function returns an error, that doesn't prevent any of the other
> exported functions from getting called.

Well, it still will be covered by that later klp_patch_is_registered()
check. But I am find with leaving it as is.

Best Regards,
Petr