Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Will Deacon <will@kernel.org>
To:     catalin.marinas@arm.com, Nikhil V <quic_nprakash@quicinc.com>
Cc:     kernel-team@android.com, Will Deacon <will@kernel.org>,
        quic_pkondeti@quicinc.com, quic_guptap@quicinc.com,
        quic_charante@quicinc.com, linux-arm-kernel@lists.infradead.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] arm64: mm: Make hibernation aware of KFENCE
Date:   Thu, 20 Jul 2023 13:13:46 +0100
Message-Id: <168984989072.750295.13436985718136646972.b4-ty@kernel.org>
In-Reply-To: <20230713070757.4093-1-quic_nprakash@quicinc.com>
References: <20230713070757.4093-1-quic_nprakash@quicinc.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: 8bit
Precedence: bulk

On Thu, 13 Jul 2023 12:37:57 +0530, Nikhil V wrote:
> In the restore path, swsusp_arch_suspend_exit uses copy_page() to
> over-write memory. However, with features like KFENCE enabled, there could
> be situations where it may have marked some pages as not valid, due to
> which it could be reported as invalid accesses.
> 
> Consider a situation where page 'P' was part of the hibernation image.
> Now, when the resume kernel tries to restore the pages, the same page 'P'
> is already in use in the resume kernel and is kfence protected, due to
> which its mapping is removed from linear map. Since restoring pages happens
> with the resume kernel page tables, we would end up accessing 'P' during
> copy and results in kernel pagefault.
> 
> [...]

Applied to arm64 (for-next/fixes), thanks!

[1/1] arm64: mm: Make hibernation aware of KFENCE
      https://git.kernel.org/arm64/c/a8bd38dbc57c

Cheers,
-- 
Will

https://fixes.arm64.dev
https://next.arm64.dev
https://will.arm64.dev