Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1330464 > unrolled thread

Re: [PATCH] efi: runtime-wrappers: run UEFI Runtime Services with interrupts enabled

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [PATCH] efi: runtime-wrappers: run UEFI Runtime Services with  interrupts enabled Ard Biesheuvel <ard.biesheuvel@linaro.org> - 2016-02-09 18:00 +0100
    Re: [PATCH] efi: runtime-wrappers: run UEFI Runtime Services with  interrupts enabled Matt Fleming <matt@codeblueprint.co.uk> - 2016-02-11 17:10 +0100

#1330464 — Re: [PATCH] efi: runtime-wrappers: run UEFI Runtime Services with interrupts enabled

On 8 February 2016 at 20:37, Andy Lutomirski <luto@amacapital.net> wrote:
> On Feb 4, 2016 5:58 AM, "Ard Biesheuvel" <ard.biesheuvel@linaro.org> wrote:
>>
>> OK, since Sai has confirmed that Windows leaves interrupts enabled when
>> calling the EFI variable store related runtime services, we should be able
>> to do the same for Linux, or at least be slightly more confident that we
>> won't have to back out this change later.
>
> Could this use a mutex instead of a spinlock?
>

When I first started working on this code, I proposed using a mutex,
but at the time, we still had the efi-pstore case to worry about
http://article.gmane.org/gmane.linux.kernel.efi/4112

In the mean time, we have modified the efi-pstore code so it simply
gives up when the EFI varstore is busy, and we also got rid of the NMI
special case where locks are ignored. In summary, it sounds to me that
moving to a mutex should be feasible, but I am only really familiar
with the ARM side of the implementation, which is far less complex
than the x86 side, so Matt should confirm.

@Matt?

> Can someone with a mixed mode setup read a variable in a loop and make
> sure it doesn't crash and burn?  It should work fine, but explicit
> testing would be nice.  (It's interesting mainly because doing a mixed
> mode call with interrupts on can result in a non-IST CPL0 to CPL0
> exception delivery, which won't result in a stack switch.  This could
> easily trigger a stack overflow, logic bug, microcode bug, or
> as-yet-unknown CPU "feature".
>
> Hmm.  We should also audit the mixed mode entry code to make sure that
> the high bits of RSP are explicitly clear before switching into compat
> mode.  If I had to make a guess about how CPUs behave, I'd guess
> pessimistically: Intel CPUs clear the high bits of RSP when switching
> into long mode due to interrupt delivery, and AMD CPUs leave them set
> just to mess with us.
>
> Also, a WARN_ON(in_interrupt()) somewhere might be a good sanity check.
>

[toc] | [next] | [standalone]

#1332191

On Tue, 09 Feb, at 05:52:34PM, Ard Biesheuvel wrote:
> On 8 February 2016 at 20:37, Andy Lutomirski <luto@amacapital.net> wrote:
> > On Feb 4, 2016 5:58 AM, "Ard Biesheuvel" <ard.biesheuvel@linaro.org> wrote:
> >>
> >> OK, since Sai has confirmed that Windows leaves interrupts enabled when
> >> calling the EFI variable store related runtime services, we should be able
> >> to do the same for Linux, or at least be slightly more confident that we
> >> won't have to back out this change later.
> >
> > Could this use a mutex instead of a spinlock?
> >
> 
> When I first started working on this code, I proposed using a mutex,
> but at the time, we still had the efi-pstore case to worry about
> http://article.gmane.org/gmane.linux.kernel.efi/4112
> 
> In the mean time, we have modified the efi-pstore code so it simply
> gives up when the EFI varstore is busy, and we also got rid of the NMI
> special case where locks are ignored. In summary, it sounds to me that
> moving to a mutex should be feasible, but I am only really familiar
> with the ARM side of the implementation, which is far less complex
> than the x86 side, so Matt should confirm.

You cannot touch mutexes from interrupt context, not even to perform a
mutex_trylock(), and that's the problem.

Since the efi-pstore code can run from interrupt context we need to
use a locking primitive that works, even though we have all that
non-blocking logic. 

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web