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;
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: quoted-printable
From:   Qian Cai <cai@lca.pw>
Mime-Version: 1.0 (1.0)
Subject: Re: [PATCH -rcu] kcsan: Make KCSAN compatible with lockdep
Date:   Fri, 17 Jan 2020 11:59:39 -0500
Message-Id: <3760F60F-4133-4FE1-9A4C-F335A8230285@lca.pw>
References: <20200117164017.GA21582@paulmck-ThinkPad-P72>
Cc:     Marco Elver <elver@google.com>,
        Alexander Potapenko <glider@google.com>,
        Andrey Konovalov <andreyknvl@google.com>,
        Dmitriy Vyukov <dvyukov@google.com>,
        kasan-dev <kasan-dev@googlegroups.com>,
        LKML <linux-kernel@vger.kernel.org>,
        Peter Zijlstra <peterz@infradead.org>,
        Ingo Molnar <mingo@redhat.com>, Will Deacon <will@kernel.org>
In-Reply-To: <20200117164017.GA21582@paulmck-ThinkPad-P72>
To:     paulmck@kernel.org
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


> On Jan 17, 2020, at 11:40 AM, Paul E. McKenney <paulmck@kernel.org> wrote:=

>=20
> True enough, but even if we reach the nirvana state where there is general=

> agreement on what constitutes a data race in need of fixing and KCSAN
> faithfully checks based on that data-race definition, we need to handle
> the case where someone introduces a bug that results in a destructive
> off-CPU access to a per-CPU variable, which is exactly the sort of thing
> that KCSAN is supposed to detect.  But suppose that this variable is
> frequently referenced from functions that are inlined all over the place.
>=20
> Then that one bug might result in huge numbers of data-race reports in
> a very short period of time, especially on a large system.

It sounds like the case with debug_pagealloc where it prints a spam of those=
, and then the system is just dead.

[   28.992752][  T394] Reported by Kernel Concurrency Sanitizer on:=20
[   28.992752][  T394] CPU: 0 PID: 394 Comm: pgdatinit0 Not tainted 5.5.0-rc=
6-next-20200115+ #3=20
[   28.992752][  T394] Hardware name: HP ProLiant XL230a Gen9/ProLiant XL230=
a Gen9, BIOS U13 01/22/2018=20
[   28.992752][  T394] =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=20
[   28.992752][  T394] =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=20
[   28.992752][  T394] BUG: KCSAN: data-race in __change_page_attr / __chang=
e_page_attr=20
[   28.992752][  T394] =20
[   28.992752][  T394] read to 0xffffffffa01a6de0 of 8 bytes by task 395 on c=
pu 16:=20
[   28.992752][  T394]  __change_page_attr+0xe81/0x1620=20
[   28.992752][  T394]  __change_page_attr_set_clr+0xde/0x4c0=20
[   28.992752][  T394]  __set_pages_np+0xcc/0x100=20
[   28.992752][  T394]  __kernel_map_pages+0xd6/0xdb=20
[   28.992752][  T394]  __free_pages_ok+0x1a8/0x730=20
[   28.992752][  T394]  __free_pages+0x51/0x90=20
[   28.992752][  T394]  __free_pages_core+0x1c7/0x2c0=20
[   28.992752][  T394]  deferred_free_range+0x59/0x8f=20
[   28.992752][  T394]  deferred_init_max21d=20
[   28.992752][  T394]  deferred_init_memmap+0x14a/0x1c1=20
[   28.992752][  T394]  kthread+0x1e0/0x200=20
[   28.992752][  T394]  ret_from_fork+0x3a/0x50=20
[   28.992752][  T394] =20
[   28.992752][  T394] write to 0xffffffffa01a6de0 of 8 bytes by task 394 on=
 cpu 0:=20
[   28.992752][  T394]  __change_page_attr+0xe9c/0x1620=20
[   28.992752][  T394]  __change_page_attr_set_clr+0xde/0x4c0=20
[   28.992752][  T394]  __set_pages_np+0xcc/0x100=20
[   28.992752][  T394]  __kernel_map_pages+0xd6/0xdb=20
[   28.992752][  T394]  __free_pages_ok+0x1a8/0x730=20
[   28.992752][  T394]  __free_pages+0x51/0x90=20
[   28.992752][  T394]  __free_pages_core+0x1c7/0x2c0=20
[   28.992752][  T394]  deferred_free_range+0x59/0x8f=20
[   28.992752][  T394]  deferred_init_maxorder+0x1d6/0x21d=20
[   28.992752][  T394]  deferred_init_memmap+0x14a/0x1c1=20
[   28.992752][  T394]  kthread+0x1e0/0x200=20
[   28.992752][  T394]  ret_from_fork+0x3a/0x50=20

It point out to this,

		pgprot_val(new_prot) &=3D ~pgprot_val(cpa->mask_clr);
		pgprot_val(new_prot) |=3D pgprot_val(cpa->mask_set);

		cpa_inc_4k_install();
		/* Hand in lpsize =3D 0 to enforce the protection mechanism=
 */
		new_prot =3D static_protections(new_prot, address, pfn, 1, 0=
,
					      CPA_PROTECT);

In static_protections(),

	/*
	 * There is no point in checking RW/NX conflicts when the requested=

	 * mapping is setting the page !PRESENT.
	 */
	if (!(pgprot_val(prot) & _PAGE_PRESENT))
		return prot;

Is there a data race there?