Received-SPF: pass (google.com: domain of linux-kernel+bounces-101467-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Feedback-ID: i36794607:Fastmail
User-Agent: Cyrus-JMAP/3.11.0-alpha0-251-g8332da0bf6-fm-20240305.001-g8332da0b
Precedence: bulk
MIME-Version: 1.0
Message-Id: <b08aaae4-8c69-47eb-9658-5f3f5c8e4056@app.fastmail.com>
In-Reply-To: <1a12c8ed-0cb5-5650-24a2-84b021c444c3@huawei.com>
References: <20240313084707.3292300-1-arnd@kernel.org>
 <4f945f4a-ac29-3ef7-9e15-123962f2a0e9@huawei.com>
 <e0684202-6926-4bd1-86f6-2bb682524712@app.fastmail.com>
 <1a12c8ed-0cb5-5650-24a2-84b021c444c3@huawei.com>
Date: Wed, 13 Mar 2024 13:21:26 +0100
From: "Arnd Bergmann" <arnd@kernel.org>
To: "Zhihao Cheng" <chengzhihao1@huawei.com>, "Arnd Bergmann" <arnd@arndb.de>,
 "Richard Weinberger" <richard@nod.at>,
 "Miquel Raynal" <miquel.raynal@bootlin.com>,
 "Vignesh Raghavendra" <vigneshr@ti.com>,
 "Daniel Golle" <daniel@makrotopia.org>
Cc: linux-mtd@lists.infradead.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] mtd: ubi: avoid expensive do_div() on 32-bit machines
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable

On Wed, Mar 13, 2024, at 13:10, Zhihao Cheng wrote:
> =E5=9C=A8 2024/3/13 19:53, Arnd Bergmann =E5=86=99=E9=81=93:
>> On Wed, Mar 13, 2024, at 12:29, Zhihao Cheng wrote:
>>=20
>> The way it usually goes is that someone adds an open-coded
>> 64-bit division that causes a link failure, which prompts
> I'm a little confused, what kind of link failure? Could you show an ex=
ample?

The open-coded 64-bit division without using do_div() shows up as

x86_64-linux-ld: drivers/mtd/ubi/nvmem.o: in function `ubi_nvmem_reg_rea=
d':
nvmem.c:(.text+0x10a): undefined reference to `__umoddi3'
x86_64-linux-ld: nvmem.c:(.text+0x11f): undefined reference to `__udivdi=
3'
x86_64-linux-ld: drivers/mtd/ubi/nvmem.o: in function `ubi_nvmem_reg_rea=
d.cold':
nvmem.c:(.text.unlikely+0x2d): undefined reference to `__umoddi3'

The idea is that gcc expects __umoddi3 to be provided by libgcc,
but Linux intentionally leaves it out in order to catch accidental
64-bit divisions.

     Arnd