Re: [REGRESSION] x86/efi: Make efi_enter/leave_mm() use the use_/unuse_temporary_mm() machinery (linux-next)

[Hugh Dickins <hughd@google.com>]

On Tue, 29 Apr 2025, Peter Zijlstra wrote:
> On Tue, Apr 29, 2025 at 12:01:22PM +0300, Jani Nikula wrote:
> > On Thu, 24 Apr 2025, "Borah, Chaitanya Kumar" <chaitanya.kumar.borah@intel.com> wrote:
> > > +Andy, Ingo
> > >
> > > Friendly reminder.
> > > Issue is still seen on latest linux-next runs.
> > >
> > > https://intel-gfx-ci.01.org/tree/linux-next/next-20250424/bat-rpls-4/boot0.txt
> > >
> > > Regards
> > >
> > > Chaitanya
> > 
> > Andy, Ingo -
> > 
> > Commit e7021e2fe0b4 ("x86/efi: Make efi_enter/leave_mm() use the
> > use_/unuse_temporary_mm() machinery") on linux-next regresses as
> > reported by Chaitanya
> > 
> > Please look into it.
> 
> Does your kernel include the below?
> 
> ---
> commit aef1d0209ddf127a8069aca5fa3a062be4136b76
> Author: Peter Zijlstra <peterz@infradead.org>
> Date:   Fri Apr 18 11:50:34 2025 +0200
> 
>     x86/mm: Fix {,un}use_temporary_mm() IRQ state
>     
>     As the function switch_mm_irqs_off() implies, it ought to be called with
>     IRQs *off*. Commit 58f8ffa91766 ("x86/mm: Allow temporary MMs when IRQs
>     are on") caused this to not be the case for EFI.
>     
>     Ensure IRQs are off where it matters.
>     
>     Fixes: 58f8ffa91766 ("x86/mm: Allow temporary MMs when IRQs are on")
>     Reported-by: Borislav Petkov (AMD) <bp@alien8.de>
>     Tested-by: Borislav Petkov (AMD) <bp@alien8.de>
>     Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
>     Signed-off-by: Ingo Molnar <mingo@kernel.org>
>     Cc: H. Peter Anvin <hpa@zytor.com>
>     Cc: Andrew Morton <akpm@linux-foundation.org>
>     Cc: Andy Lutomirski <luto@kernel.org>
>     Cc: Linus Torvalds <torvalds@linux-foundation.org>
>     Cc: Rik van Riel <riel@surriel.com>
>     Link: https://lore.kernel.org/r/20250418095034.GR38216@noisy.programming.kicks-ass.net
> 
> diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
> index 79c124f6f3f2..39761c7765bd 100644
> --- a/arch/x86/mm/tlb.c
> +++ b/arch/x86/mm/tlb.c
> @@ -986,6 +986,7 @@ struct mm_struct *use_temporary_mm(struct mm_struct *temp_mm)
>  	struct mm_struct *prev_mm;
>  
>  	lockdep_assert_preemption_disabled();
> +	guard(irqsave)();
>  
>  	/*
>  	 * Make sure not to be in TLB lazy mode, as otherwise we'll end up
> @@ -1018,6 +1019,7 @@ struct mm_struct *use_temporary_mm(struct mm_struct *temp_mm)
>  void unuse_temporary_mm(struct mm_struct *prev_mm)
>  {
>  	lockdep_assert_preemption_disabled();
> +	guard(irqsave)();
>  
>  	/* Clear the cpumask, to indicate no TLB flushing is needed anywhere */
>  	cpumask_clear_cpu(smp_processor_id(), mm_cpumask(this_cpu_read(cpu_tlbstate.loaded_mm)));

Hi Peter, I haven't checked on most recent -nexts, but earlier found that
patch to be not quite enough, at least if you have CONFIG_DEBUG_VM=y:
because switch_mm_irqs_off() contains a

		VM_WARN_ON_ONCE(prev != &init_mm && !cpumask_test_cpu(cpu,
				mm_cpumask(prev)));

which doesn't like what (un)use_temporary_mm() is now doing. I couldn't
be sure who was right or wrong, and just proceeded by commenting out
the warning - ONCE shouldn't be much trouble, except xfstests uses
some nefarious mechanism to resurrect ONCE repeatedly.

Hugh