Messages in this thread | | | Date | Fri, 30 Jul 2010 18:01:39 +0300 | Subject | Re: [PATCH 01/13] AppArmor: misc. base functions and defines | From | Pekka Enberg <> |
| |
On Fri, Jul 30, 2010 at 5:24 PM, Changli Gao <xiaosuo@gmail.com> wrote: > On Fri, Jul 30, 2010 at 6:53 PM, Pekka Enberg <penberg@cs.helsinki.fi> wrote: >> On Fri, Jul 30, 2010 at 1:01 PM, John Johansen >>> >>> vfree can sleep, and skipping the schedule_work parts won't work for >>> apparmor as many of these allocations are being freed via rcu callbacks >>> as most of our object life cycles are dependent on cred refcounting. >> >> Can someone point me to where vfree() actually sleeps? I'm unable to >> find the exact spot. > > http://lxr.linux.no/linux+v2.6.34.1/mm/vmalloc.c#L1405 . vfree -> > __vunmap, vunmap -> __vunmap, and there is a might_sleep() function in > vunmap.
Yes, but that doesn't answer my question. Where's the actual call-site that can sleep? I can't find it!
> BTW: I'll respin the kvmalloc patch later.
Great!
| |