Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From:   Andy Lutomirski <luto@amacapital.net>
Mime-Version: 1.0 (1.0)
Subject: Re: [PATCH v3] x86/fault: Send a SIGBUS to user process always for hwpoison page access.
Date:   Mon, 8 Mar 2021 10:14:44 -0800
Message-Id: <8F3F763F-59CC-4E25-B4DE-89CD0632F754@amacapital.net>
References: <20210308174912.4ac9029a@alex-virtual-machine>
Cc:     "Luck, Tony" <tony.luck@intel.com>,
        Andy Lutomirski <luto@kernel.org>,
        HORIGUCHI NAOYA <naoya.horiguchi@nec.com>,
        Dave Hansen <dave.hansen@linux.intel.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        "H. Peter Anvin" <hpa@zytor.com>, X86 ML <x86@kernel.org>,
        yangfeng1@kingsoft.com, Linux-MM <linux-mm@kvack.org>,
        LKML <linux-kernel@vger.kernel.org>
In-Reply-To: <20210308174912.4ac9029a@alex-virtual-machine>
To:     Aili Yao <yaoaili@kingsoft.com>
Precedence: bulk


> On Mar 8, 2021, at 1:49 AM, Aili Yao <yaoaili@kingsoft.com> wrote:
>=20
> =EF=BB=BFOn Sun, 7 Mar 2021 11:16:24 -0800
> Andy Lutomirski <luto@amacapital.net> wrote:
>=20
>>>>>>> Some programs may use read(2), write(2), etc as ways to check if
>>>>>>> memory is valid without getting a signal.  They might not want
>>>>>>> signals, which means that this feature might need to be configurable=
. =20
>>>>>>=20
>>>>>> That sounds like an appalling hack. If users need such a mechanism
>>>>>> we should create some better way to do that.
>>>>>>=20
>>>>>=20
>>>>> Appalling hack or not, it works. So, if we=E2=80=99re going to send a s=
ignal to user code that looks like it originated from a bina fide architectu=
ral recoverable fault, it needs to be recoverable.  A load from a failed NVD=
IMM page is such a fault. A *kernel* load is not. So we need to distinguish i=
t somehow. =20
>>>>=20
>>>> Sorry for my previous mis-understanding, and i have some questions:
>>>> if programs use read,write to check if if memory is valid, does it real=
ly want to cover the poison case? =20
>>=20
>> I don't know.
>>=20
>>>> When for such a case, an error is returned,  can the program realize it=
's hwposion issue not other software error and process correctly? =20
>>=20
>> Again, I don't know.  But changing the API like this seems potentialy
>> dangerous and needs to be done with care.
>>=20
>>>>=20
>>>> if this is the proper action, the original posion flow in current code f=
rom read and write need to change too.
>>>>=20
>>>=20
>>> Sorry, another question:
>>>  When programs use read(2), write(2) as ways to check if memory is valid=
, does it really want to check if the user page the program provided is vali=
d, not the destination or disk space valid? =20
>>=20
>> They may well be trying to see if their memory is valid.
>=20
> Thanks for your reply, and I don't know what to do.
> For current code, if user program write to a block device(maybe a test try=
) and if its user copy page corrupt when in kernel copy, the process is kill=
ed with a SIGBUS.
> And for the page fault case in this thread, the process is error returned.=


Can you point me at that SIGBUS code in a current kernel?

>=20
> --=20
> Thanks!
> Aili Yao