Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 2 Dec 2019 12:33:19 +0000
From:   Mark Rutland <mark.rutland@arm.com>
To:     Heyi Guo <guoheyi@huawei.com>
Cc:     linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org,
        wanghaibin.wang@huawei.com,
        Catalin Marinas <catalin.marinas@arm.com>,
        Will Deacon <will@kernel.org>
Subject: Re: [PATCH] arm64/kernel/entry: refine comment of stack overflow
 check
Message-ID: <20191202123319.GA25809@lakrids.cambridge.arm.com>
References: <20191202113702.34158-1-guoheyi@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20191202113702.34158-1-guoheyi@huawei.com>
User-Agent: Mutt/1.11.1+11 (2f07cb52) (2018-12-01)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, Dec 02, 2019 at 07:37:02PM +0800, Heyi Guo wrote:
> Stack overflow checking can be done by testing
> sp & (1 << THREAD_SHIFT)
> only for the stacks are aligned to (2 << THREAD_SHIFT) with size of
> (1 << THREAD_SIZE), and this is the case when CONFIG_VMAP_STACK is
> set.

Good point, I was sloppy with this comment.

> 
> Fix the code comment to avoid confusion.
> 
> Signed-off-by: Heyi Guo <guoheyi@huawei.com>
> Cc: Catalin Marinas <catalin.marinas@arm.com>
> Cc: Will Deacon <will@kernel.org>
> ---
>  arch/arm64/kernel/entry.S | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S
> index cf3bd2976e57..9e8ba507090f 100644
> --- a/arch/arm64/kernel/entry.S
> +++ b/arch/arm64/kernel/entry.S
> @@ -76,7 +76,8 @@ alternative_else_nop_endif
>  #ifdef CONFIG_VMAP_STACK
>  	/*
>  	 * Test whether the SP has overflowed, without corrupting a GPR.
> -	 * Task and IRQ stacks are aligned to (1 << THREAD_SHIFT).
> +	 * Task and IRQ stacks are aligned to (2 << THREAD_SHIFT) with size of
> +	 * (1 << THREAD_SHIFT).
>  	 */

Can we make that:

	Task and IRQ stacks are aligned so that SP & (1 << THREAD_SHIFT)
	should always be zero.

... which I think is a bit clearer.

With that wording:

Acked-by: Mark Rutland <mark.rutland@arm.com>

Mark.

>  	add	sp, sp, x0			// sp' = sp + x0
>  	sub	x0, sp, x0			// x0' = sp' - x0 = (sp + x0) - x0 = sp
> -- 
> 2.19.1
>