Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Feedback-ID: i56a14606:Fastmail
User-Agent: Cyrus-JMAP/3.7.0-alpha0-1115-g8b801eadce-fm-20221102.001-g8b801ead
Mime-Version: 1.0
Message-Id: <43fda82a-f3df-42c8-9eb2-23cf2dce8628@app.fastmail.com>
In-Reply-To: <0903321f-e0cf-fd7b-bbdd-fc4fdc0f05a0@189.cn>
References: <1670229006-4063-1-git-send-email-chensong_2000@189.cn>
 <2a3d3359-a5fd-453b-81f1-35c7a35fc12d@app.fastmail.com>
 <0903321f-e0cf-fd7b-bbdd-fc4fdc0f05a0@189.cn>
Date:   Tue, 06 Dec 2022 15:22:58 +0100
From:   "Arnd Bergmann" <arnd@arndb.de>
To:     "Song Chen" <chensong_2000@189.cn>,
        "Steven Rostedt" <rostedt@goodmis.org>,
        "Masami Hiramatsu" <mhiramat@kernel.org>,
        "Niklas Schnelle" <schnelle@linux.ibm.com>
Cc:     linux-kernel@vger.kernel.org, linux-trace-kernel@vger.kernel.org,
        Linux-Arch <linux-arch@vger.kernel.org>
Subject: Re: [PATCH v3 3/4] include/asm-generic/io.h: remove performing pointer
 arithmetic on a null pointer
Content-Type: text/plain;charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, Dec 6, 2022, at 07:01, Song Chen wrote:
> =E5=9C=A8 2022/12/5 18:04, Arnd Bergmann =E5=86=99=E9=81=93:
>> On Mon, Dec 5, 2022, at 09:30, Song Chen wrote:
>>=20
>> We have discussed this bit multiple times, and Niklas Schnelle
>> last posted his series to fix this as an RFC in [1].
>>=20
>
> Trace triggers the warning accidentally by including io.h indirectly=20
> because of the absence of PCI_IOBASE in hexagon. So what trace can do =
in=20
> this case is either to suppress warning or just ignore it, the warning=20
> will go away as long as hexagon has put PCI_IOBASE in place or=20
> implemented its own inb() etc, i think they will do it sooner or later.

hexagon/riscv/s390 should not implement inb(), there is no reason
for that because no hardware uses it. Half of the other architectures
that currently implement inb() should not do so either.

> Introducing HAS_IOPORT to trace seems no necessary and too much impact.

I don't think that trace has anything to do with it, the asm-generic
header should just not provde the inb() interface on architectures
that don't use it.

    Arnd