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;
Subject: Re: [PATCH 3/3] mm:fix gup_pud_range
To:     Qiujun Huang <hqjagain@gmail.com>, <akpm@linux-foundation.org>
CC:     <ira.weiny@intel.com>, <jgg@ziepe.ca>, <dan.j.williams@intel.com>,
        <rppt@linux.ibm.com>, <aneesh.kumar@linux.ibm.com>,
        <keith.busch@intel.com>, <linux-mm@kvack.org>,
        <linux-kernel@vger.kernel.org>
References: <1568994684-1425-1-git-send-email-hqjagain@gmail.com>
From:   John Hubbard <jhubbard@nvidia.com>
Message-ID: <1a162778-41b9-4428-1058-82aaf82314b1@nvidia.com>
Date:   Fri, 20 Sep 2019 12:37:04 -0700
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.8.0
MIME-Version: 1.0
In-Reply-To: <1568994684-1425-1-git-send-email-hqjagain@gmail.com>
Content-Type: text/plain; charset="utf-8"
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 9/20/19 8:51 AM, Qiujun Huang wrote:
> __get_user_pages_fast try to walk the page table but the
> hugepage pte is replace by hwpoison swap entry by mca path.

I expect you mean MCE (machine check exception), rather than mca?

> ...
> [15798.177437] mce: Uncorrected hardware memory error in
> 				user-access at 224f1761c0
> [15798.180171] MCE 0x224f176: Killing pal_main:6784 due to
> 				hardware memory corruption
> [15798.180176] MCE 0x224f176: Killing qemu-system-x86:167336
> 				due to hardware memory corruption
> ...
> [15798.180206] BUG: unable to handle kernel
> [15798.180226] paging request at ffff891200003000
> [15798.180236] IP: [<ffffffff8106edae>] gup_pud_range+
> 				0x13e/0x1e0
> ...
> 
> We need to skip the hwpoison entry in gup_pud_range.

It would be nice if this spelled out a little more clearly what's
wrong. I think you and Aneesh are saying that the entry is really
a swap entry, created by the MCE response to a bad page?

> 
> Signed-off-by: Qiujun Huang <hqjagain@gmail.com>
> ---
>  mm/gup.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/mm/gup.c b/mm/gup.c
> index 98f13ab..6157ed9 100644
> --- a/mm/gup.c
> +++ b/mm/gup.c
> @@ -2230,6 +2230,8 @@ static int gup_pud_range(p4d_t p4d, unsigned long addr, unsigned long end,
>  		next = pud_addr_end(addr, end);
>  		if (pud_none(pud))
>  			return 0;
> +		if (unlikely(!pud_present(pud)))
> +			return 0;

If the MCE hwpoison behavior puts in swap entries, then it seems like all
page table walkers would need to check for p*d_present(), and maybe at all
levels too, right?  

thanks,
-- 
John Hubbard
NVIDIA


>  		if (unlikely(pud_huge(pud))) {
>  			if (!gup_huge_pud(pud, pudp, addr, next, flags,
>  					  pages, nr))
>