Re: [PATCH v2 2/2] arm64: Don't use KPTI where we have E0PD

From: Suzuki K Poulose <suzuki.poulose@arm•com>
To: Mark Brown <broonie@kernel•org>
Cc: Catalin Marinas <catalin.marinas@arm•com>,
	Will Deacon <will@kernel•org>,
	linux-arm-kernel@lists•infradead.org
Subject: Re: [PATCH v2 2/2] arm64: Don't use KPTI where we have E0PD
Date: Thu, 10 Oct 2019 11:24:55 +0100	[thread overview]
Message-ID: <2d53ed87-34f3-33e1-5501-77895662c47a@arm.com> (raw)
In-Reply-To: <20191009175205.GN2036@sirena.org.uk>

Hi Mark

On 09/10/2019 18:52, Mark Brown wrote:
> On Tue, Sep 24, 2019 at 10:13:18AM +0100, Suzuki K Poulose wrote:
>> On 16/08/2019 13:10, Mark Brown wrote:
> 
>>> We definitely need some variable I think, and I think you're right that
>>> making the decision on the boot CPU would simplify things a lot.  The
> 
>> relocating the kernel. So, we may be able to perform "raw cpuid check" on
>> the boot CPU with MMU turned on, before we re-write the pagetables for KASLR
>> displacement and nG if that is needed (by maybe updating SWWAPPER_MMU_FLAGS) for
>> the boot CPU and store this information somewhere. Thus we may be able to
>> avoid another re-write of the pagetables after we have booted the secondaries.
> 
> The boot CPU is straightforward, there is only an issue on the
> secondaries where IIRC the rewrite code needs some updates as we
> get left with non-global mappings lying around.
> 
>> Discussing this with Catalin, he suggests to use a variable for the status
>> of "nG" flag for PTE/PMD_MAYBE_NG, to avoid calling the helper function
>> all the time. By using the per-CPU check we can make sure the flag is uptodate.
> 
> That was the discussion about the variable above.  We need one
> for non-optimization reasons anyway since we can't rely on
> checking the state on the current CPU.
> 
>> Also, we can continue to fail the hotplugged CPUs if we detect that the
>> pagetables are Global and the new CPU requires nG (for heterogeneous
>> systems).
> 
> There's no continuing to reject those CPUs unfortunately, we
> don't reject anything currently.  Any such systems would

In fact we do reject the hotplugged CPUs, after we have finalised
the capabilities for KPTI. So, I don't see how the behavior is different.

Cheers
Suzuki

> experience a regression when moving to a kernel where E0PD is
> enabled which doesn't seem ideal.

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists•infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel