Re: [4.18.0 rc8 BUG] possible irq lock inversion dependency detected

[Thomas Gleixner <tglx@linutronix.de>]

On Sat, 11 Aug 2018, Matthew Wilcox wrote:

> On Sat, Aug 11, 2018 at 12:28:24PM +0500, Mikhail Gavrilov wrote:
> > Hi guys.
> > I am catched new bug. It occured when I start virtual machine.
> > Can anyone look?
> 
> I'd suggest that st->lock should be taken with irqsave.  Like this;
> please test.

That should fix it, but that's suboptimal because that's an extra
safe/restore in switch_to(). So we better disable interrupts at the other
call site. Patch below.

Thanks,

	tglx

8<------------------

diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
index 30ca2d1a9231..07ce27082a40 100644
--- a/arch/x86/kernel/process.c
+++ b/arch/x86/kernel/process.c
@@ -416,9 +416,11 @@ static __always_inline void __speculative_store_bypass_update(unsigned long tifn
 
 void speculative_store_bypass_update(unsigned long tif)
 {
-	preempt_disable();
+	unsigned long flags;
+
+	local_irq_save(flags);
 	__speculative_store_bypass_update(tif);
-	preempt_enable();
+	local_irq_restore(flags);
 }
 
 void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,