Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Subject: Re: [PATCH v2 1/4] tpm: Use a threaded interrupt handler
To:     Jarkko Sakkinen <jarkko@kernel.org>
Cc:     peterhuewe@gmx.de, jgg@ziepe.ca, stefanb@linux.vnet.ibm.com,
        James.Bottomley@hansenpartnership.com, keescook@chromium.org,
        jsnitsel@redhat.com, ml.linux@elloe.vision,
        linux-integrity@vger.kernel.org, linux-kernel@vger.kernel.org
References: <1619394440-30646-1-git-send-email-LinoSanfilippo@gmx.de>
 <1619394440-30646-2-git-send-email-LinoSanfilippo@gmx.de>
 <YIikDCTBcMMxjots@kernel.org>
From:   Lino Sanfilippo <LinoSanfilippo@gmx.de>
Message-ID: <495e816a-afba-4ea0-560c-bc748df26337@gmx.de>
Date:   Thu, 29 Apr 2021 00:37:40 +0200
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101
 Thunderbird/68.10.0
MIME-Version: 1.0
In-Reply-To: <YIikDCTBcMMxjots@kernel.org>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: quoted-printable
Precedence: bulk


Hi,

On 28.04.21 at 01:53, Jarkko Sakkinen wrote:
> On Mon, Apr 26, 2021 at 01:47:17AM +0200, Lino Sanfilippo wrote:
>> Interrupt handling at least includes reading and writing the interrupt
>> status register from the interrupt routine. However over SPI those acce=
sses
>> require a sleepable context, since a mutex is used in the concerning
>> functions.
>> For this reason request a threaded interrupt handler which is running i=
n
>> (sleepable) process context.
>>
>> Signed-off-by: Lino Sanfilippo <LinoSanfilippo@gmx.de>
>> ---
>>  drivers/char/tpm/tpm_tis_core.c | 6 ++++--
>>  1 file changed, 4 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/char/tpm/tpm_tis_core.c b/drivers/char/tpm/tpm_tis=
_core.c
>> index e7d1eab..0959559 100644
>> --- a/drivers/char/tpm/tpm_tis_core.c
>> +++ b/drivers/char/tpm/tpm_tis_core.c
>> @@ -781,8 +781,10 @@ static int tpm_tis_probe_irq_single(struct tpm_chi=
p *chip, u32 intmask,
>>  	int rc;
>>  	u32 int_status;
>>
>> -	if (devm_request_irq(chip->dev.parent, irq, tis_int_handler, flags,
>> -			     dev_name(&chip->dev), chip) !=3D 0) {
>> +
>> +	if (devm_request_threaded_irq(chip->dev.parent, irq, NULL,
>> +				      tis_int_handler, IRQF_ONESHOT | flags,
>> +				      dev_name(&chip->dev), chip) !=3D 0) {
>>  		dev_info(&chip->dev, "Unable to request irq: %d for probe\n",
>>  			 irq);
>>  		return -1;
>> --
>> 2.7.4
>>
>
> Why?
>
> https://elixir.bootlin.com/linux/v5.12/source/drivers/char/tpm/tpm_tis_c=
ore.c#L670
>
> I don't see anything that sleeps there.
>
> /Jarkko1
>

The problem are the register read/write functions which we use to access t=
he status register in
the interrupt handler. In case of SPI they result in taking the spi_bus_lo=
ck which is a mutex.

E.g tpm_tis_spi_read32: tpm_tis_spi_read_bytes->tpm_tis_spi_transfer->spi_=
bus_lock->mutex_lock

Using a threaded interrupt handler seemed to me the easiest way to avoid t=
his issue.

Regards,
Lino