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;
From:   Song Liu <songliubraving@fb.com>
To:     Peter Zijlstra <peterz@infradead.org>,
        Steven Rostedt <rostedt@goodmis.org>,
        "sbsiddha@gmail.com" <sbsiddha@gmail.com>
CC:     Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Linux MM <linux-mm@kvack.org>,
        Kernel Team <Kernel-team@fb.com>,
        "stable@vger.kernel.org" <stable@vger.kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Dave Hansen <dave.hansen@intel.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Andy Lutomirski <luto@amacapital.net>
Subject: Re: [PATCH] x86/mm: Do not split_large_page() for
 set_kernel_text_rw()
Thread-Topic: [PATCH] x86/mm: Do not split_large_page() for
 set_kernel_text_rw()
Thread-Index: AQHVWXL45hcGbMxBFEmxITCR8fhmlKcIeZ6AgARkOQA=
Date:   Mon, 26 Aug 2019 04:40:23 +0000
Message-ID: <164D1F08-80F7-4E13-94FC-78F33B3E299F@fb.com>
References: <20190823052335.572133-1-songliubraving@fb.com>
 <20190823093637.GH2369@hirez.programming.kicks-ass.net>
In-Reply-To: <20190823093637.GH2369@hirez.programming.kicks-ass.net>
Accept-Language: en-US
Content-Language: en-US
received-spf: None (protection.outlook.com: fb.com does not designate
 permitted sender hosts)
Content-Type: text/plain; charset="us-ascii"
Content-ID: <AD68C27445FA53479B507173F2FC3246@namprd15.prod.outlook.com>
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-Network-Message-Id: b3bf2984-2de2-46db-3f11-08d729df8234
X-MS-Exchange-CrossTenant-originalarrivaltime: 26 Aug 2019 04:40:23.3573
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: 8ae927fe-1255-47a7-a2af-5f3a069daaa2
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: bNO7kCSk2A40h5DwTvV5/XrffHYCdw5w7SVwMf7LIIXCYC2lm9jSJO9wyPcg3SJ4K6yQOc/dnPVfT95LRHNQJA==
X-MS-Exchange-Transport-CrossTenantHeadersStamped: MWHPR15MB1952
X-OriginatorOrg: fb.com
X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:5.22.84,1.0.8
 definitions=2019-08-26_02:2019-08-23,2019-08-26 signatures=0
X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 clxscore=1011
 malwarescore=0 impostorscore=0 suspectscore=0 phishscore=0
 lowpriorityscore=0 mlxscore=0 bulkscore=0 mlxlogscore=999 adultscore=0
 spamscore=0 priorityscore=1501 classifier=spam adjust=0 reason=mlx
 scancount=1 engine=8.12.0-1906280000 definitions=main-1908260049
X-FB-Internal: deliver
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

Cc: Steven Rostedt and Suresh Siddha

Hi Peter,=20

> On Aug 23, 2019, at 2:36 AM, Peter Zijlstra <peterz@infradead.org> wrote:
>=20
> On Thu, Aug 22, 2019 at 10:23:35PM -0700, Song Liu wrote:
>> As 4k pages check was removed from cpa [1], set_kernel_text_rw() leads t=
o
>> split_large_page() for all kernel text pages. This means a single kprobe
>> will put all kernel text in 4k pages:
>>=20
>>  root@ ~# grep ffff81000000- /sys/kernel/debug/page_tables/kernel
>>  0xffffffff81000000-0xffffffff82400000     20M  ro    PSE      x  pmd
>>=20
>>  root@ ~# echo ONE_KPROBE >> /sys/kernel/debug/tracing/kprobe_events
>>  root@ ~# echo 1 > /sys/kernel/debug/tracing/events/kprobes/enable
>>=20
>>  root@ ~# grep ffff81000000- /sys/kernel/debug/page_tables/kernel
>>  0xffffffff81000000-0xffffffff82400000     20M  ro             x  pte
>>=20
>> To fix this issue, introduce CPA_FLIP_TEXT_RW to bypass "Text RO" check
>> in static_protections().
>>=20
>> Two helper functions set_text_rw() and set_text_ro() are added to flip
>> _PAGE_RW bit for kernel text.
>>=20
>> [1] commit 585948f4f695 ("x86/mm/cpa: Avoid the 4k pages check completel=
y")
>=20
> ARGH; so this is because ftrace flips the whole kernel range to RW and
> back for giggles? I'm thinking _that_ is a bug, it's a clear W^X
> violation.

Thanks for your comments. Yes, it is related to ftrace, as we have
CONFIG_KPROBES_ON_FTRACE. However, after digging around, I am not sure
what is the expected behavior.=20

Kernel text region has two mappings to it. For x86_64 and four-level=20
page table, there are:=20

	1. kernel identity mapping, from 0xffff888000100000;=20
	2. kernel text mapping, from 0xffffffff81000000,=20

Per comments in arch/x86/mm/init_64.c:set_kernel_text_rw():

        /*
         * Make the kernel identity mapping for text RW. Kernel text
         * mapping will always be RO. Refer to the comment in
         * static_protections() in pageattr.c
         */
	set_memory_rw(start, (end - start) >> PAGE_SHIFT);

kprobe (with CONFIG_KPROBES_ON_FTRACE) should work on kernel identity
mapping.=20

However, my experiment shows that kprobe actually operates on the=20
kernel text mapping (0xffffffff81000000-). It is the same w/ and w/o=20
CONFIG_KPROBES_ON_FTRACE. Therefore, I am not sure whether the comment
is out-dated (10-year old), or the kprobe is doing something wrong.=20


More information about the issue we are looking at.=20

We found with 5.2 kernel (no CONFIG_PAGE_TABLE_ISOLATION, w/=20
CONFIG_KPROBES_ON_FTRACE), a single kprobe will split _all_ PMDs in=20
kernel text mapping into pte-mapped pages. This increases iTLB=20
miss rate from about 300 per million instructions to about 700 per
million instructions (for the application I test with).=20

Per bisect, we found this behavior happens after commit 585948f4f695=20
("x86/mm/cpa: Avoid the 4k pages check completely"). That's why I=20
proposed this PATCH to fix/workaround this issue. However, per
Peter's comment and my study of the code, this doesn't seem the=20
real problem or the only here.=20

I also tested that the PMD split issue doesn't happen w/o=20
CONFIG_KPROBES_ON_FTRACE.=20


In summary, I have the following questions:

1. Which mapping should kprobe work on? Kernel identity mapping or=20
   kernel text mapping?
2. FTRACE causes split of PMD mapped kernel text. How should we fix
   this?=20

Thanks,
Song