tech-kern archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: More duplicate code, vnode locking question
On Sat, Apr 25, 2009 at 11:44 PM, Andrew Doran <ad%netbsd.org@localhost> wrote:
> On Thu, Apr 23, 2009 at 02:56:49AM +0300, Elad Efrat wrote:
>
>> + error = genfs_can_mount(devvp, VREAD, l->l_cred);
>
> FYI l->l_cred used outside core kernel code (basically in anything that
> could be loaded as a module) should be kauth_cred_get() to reduce exposure
> to 'struct lwp'.
Do we want to include such changes as part of larger commits, or do
them independently? I think the latter, as it's easier to inspect
and make sure they're correct (ie., maybe not all "l->l_cred" is
equal to the result of kauth_cred_get()? think of functions that
take more than one lwp).
That said, I don't mind going over the changes I just did and use
kauth_cred_get() instead of l->l_cred. Btw, maybe we should collect
all those notes and FYIs to a single document? (unless we already
have one that I'm not aware of :)
>> + *
>> + * Permission to update a mount is checked higher, so
>> here we presume
>> + * updating the mount is okay (for example, as far as
>> securelevel goes)
>> + * which leaves us with the normal check.
>> */
>
> http://nxr.netbsd.org/source/xref/share/misc/style#47
Slipped in, sorry about that. Fixed.
Thanks,
-e.
- References:
- More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
- Re: More duplicate code, vnode locking question
Home |
Main Index |
Thread Index |
Old Index