Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
Message-ID: <2d7e011ae604b3d6d952bed5f7ba08a090930fbe.camel@kernel.org>
Subject: Re: [PATCH v4 1/6] tpm, tpm_tis_spi: Request threaded irq
From:   Jarkko Sakkinen <jarkko@kernel.org>
To:     Lino Sanfilippo <LinoSanfilippo@gmx.de>
Cc:     peterhuewe@gmx.de, jgg@ziepe.ca, stefanb@linux.vnet.ibm.com,
        linux@mniewoehner.de, linux-integrity@vger.kernel.org,
        linux-kernel@vger.kernel.org, lukas@wunner.de,
        p.rosenberger@kunbus.com, Lino Sanfilippo <l.sanfilippo@kunbus.com>
Date:   Tue, 17 May 2022 21:23:18 +0300
In-Reply-To: <1b306d56-c6b3-e51f-1d63-2f6725fa7557@gmx.de>
References: <20220509080559.4381-1-LinoSanfilippo@gmx.de>
         <20220509080559.4381-2-LinoSanfilippo@gmx.de> <YnucgDH3I87RI8PN@kernel.org>
         <486cec01-ec02-3f11-0b81-037e0700c503@gmx.de> <Yn6esMyUl+QhECq+@kernel.org>
         <1b306d56-c6b3-e51f-1d63-2f6725fa7557@gmx.de>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.1 
MIME-Version: 1.0
Precedence: bulk

On Mon, 2022-05-16 at 21:52 +0200, Lino Sanfilippo wrote:
> Hi,
>=20
> On 13.05.22 at 20:08, Jarkko Sakkinen wrote:
> > On Wed, May 11, 2022 at 09:18:39PM +0200, Lino Sanfilippo wrote:
> > > Hi,
> > >=20
> > > On 11.05.22 at 13:22, Jarkko Sakkinen wrote:
> > > > On Mon, May 09, 2022 at 10:05:54AM +0200, Lino Sanfilippo wrote:
> > > > > From: Lino Sanfilippo <l.sanfilippo@kunbus.com>
> > > > >=20
> > > > > Interrupt handling at least includes reading and writing the inte=
rrupt
> > > > > status register within the interrupt routine. Since accesses over=
 the SPI
> > > > > bus are synchronized by a mutex, request a threaded interrupt han=
dler to
> > > > > ensure a sleepable context during interrupt processing.
> > > > >=20
> > > > > Fixes: 1a339b658d9d ("tpm_tis_spi: Pass the SPI IRQ down to the d=
river")
> > > > > Signed-off-by: Lino Sanfilippo <l.sanfilippo@kunbus.com>
> > > >=20
> > > > When you state that it needs a sleepable context, you should bring =
a
> > > > context why it needs it. This not to disregard the code change over=
ally but
> > > > you cannot make even the most obvious claim without backing data.
> > > >=20
> > >=20
> > > so what kind of backing data do you have in mind? Would it help to em=
phasize more
> > > that the irq handler is running in hard irq context in the current co=
de and thus
> > > must not access registers over SPI since SPI uses a mutex (I consider=
 it as basic
> > > knowledge that a mutex must not be taken in hard irq context)?
> >=20
> > There's zero mention about specific lock you are talking about. Providi=
ng
> > the basic knowledge what you are trying to do is the whole point of the
> > commit message in the first place. I'd presume this patch is related to=
 the
> > use bus_lock_mutex but it is fully ingored here.
> >=20
>=20
> Ok, understood. I will amend the commit message to make more clear that
> reading and writing registers from the interrupt handler results in
> a call to tpm_tis_spi_transfer() which uses the bus_lock_mutex of the
> spi device and thus requires a sleepable context.

Yeah, please be always as explicit as possible, so that it is impossible
to get it wrong. Then a reader of your patch saves time from seeking e.g.
the current name of the data structure. Just dumb things down as far as
you can.

Commit messages have a dual function:

1. They *lower* the barrier to look into a code change, which helps
the patches get any attention.
2. Proper commit messages save tons of time from the maintainer, when
you have revisit years old commits, e.g. when bisecting a bug.

Comparing to e.g. Github the key difference is this: in Github you have
commits and issues. In kernel commit is both issue and the commit bundled
together. Therefore, every commit also needs to have a "bug report"
included.

BR, Jarkko