On Tue, Oct 16, 2018 at 2:39 PM Sebastian Andrzej Siewior <bige...@linutronix.de> wrote: > > On 2018-10-16 14:25:07 [-0700], Andy Lutomirski wrote: > > > diff --git a/arch/x86/mm/pageattr.c b/arch/x86/mm/pageattr.c > > > index 51a5a69ecac9f..fe6b21f0a6631 100644 > > > --- a/arch/x86/mm/pageattr.c > > > +++ b/arch/x86/mm/pageattr.c > > > @@ -2088,7 +2088,9 @@ void __kernel_map_pages(struct page *page, int > > > numpages, int enable) > > > * We should perform an IPI and flush all tlbs, > > > * but that can deadlock->flush only current cpu: > > > */ > > > + preempt_disable(); > > > __flush_tlb_all(); > > > + preempt_enable(); > > > > > > > Depending on your CPU, __flush_tlb_all() is either > > __native_flush_tlb_global() or __native_flush_tlb(). Only > > __native_flush_tlb() could have any problem with preemption, but it > > has a WARN_ON_ONCE(preemptible()); in it. Can you try to figure out > > why that's not firing for you? > > It is firing, it is the warning that was introduced in commit > decab0888e6e (as mention in the commit message; I just noticed it way > later because it popped early in the boot log). > > > I suspect that a better fix would be to put preempt_disable() into > > __native_flulsh_tlb(), but I'd still like to understand why the > > warning isn't working. > > __native_flulsh_tlb() just had its preempt_disable() removed in > decab0888e6e and __kernel_map_pages() is only called from the debug > code. The other caller of __native_flulsh_tlb() seem to hold a lock or > run with disabled interrupts. > > Sebastian
Fair enough. But can you copy the warning to __flush_tlb_all() so it's checked on all systems (but make it VM_WARN_ON_ONCE)? --Andy
