linux-integrity.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Amir Mizinski <amirmizi6@gmail.com>
To: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Cc: Eyal.Cohen@nuvoton.com, oshrialkoby85@gmail.com,
	alexander.steffen@infineon.com, robh+dt@kernel.org,
	mark.rutland@arm.com, peterhuewe@gmx.de, jgg@ziepe.ca,
	arnd@arndb.de, gregkh@linuxfoundation.org,
	devicetree@vger.kernel.org, linux-kernel@vger.kernel.org,
	linux-integrity@vger.kernel.org, oshri.alkoby@nuvoton.com,
	tmaimon77@gmail.com, gcwilson@us.ibm.com, kgoldman@us.ibm.com,
	Dan.Morav@nuvoton.com, oren.tanami@nuvoton.com,
	shmulik.hager@nuvoton.com, amir.mizinski@nuvoton.com,
	Christophe Ricard <christophe-h.ricard@st.com>
Subject: Re: [PATCH v6 2/7] tpm: tpm_tis: Add check_data handle to tpm_tis_phy_ops
Date: Tue, 21 Apr 2020 13:34:44 +0000	[thread overview]
Message-ID: <0cfa0486-8ccb-d7d1-acf2-ca103f723b3a@gmail.com> (raw)
In-Reply-To: <20200408183324.GB33486@linux.intel.com>


Hello jarkko,
I reconfigure my email client by the instructions you've sent, and
re-responsing as you requested.
please tell me if there are still any issues. thank you.

On 2020-04-08 18:33, Jarkko Sakkinen wrote:
> On Tue, Apr 07, 2020 at 07:20:39PM +0300, amirmizi6@gmail.com wrote:
>> From: Amir Mizinski <amirmizi6@gmail.com>
>>
>> In order to validate data integrity we need to compute the crc over the data
>> sent in lower layer (I2C for instance).
>
> s/crc/CRC/
>
>> To do that tpm_tis_check_data() calls a "check_data" operation (if available).
>
> "check_data" does not exist.
>

it is added in this commit to "tpm_tis_phy_ops" struct in
"tpm_tis_core.h", which is inherited in "tpm_tis_i2c.c" on later patch
(7/7).

>> If data integrity check fails, a retry to save the sent/received
>> data is implemented in tpm_tis_send_main()/tpm_tis_recv() functions.
>>
>> Considering this commit, the following steps are done when sending a command:
>>    1. Host writes to TPM_STS.commandReady.
>>    2. Host writes command.
>>    3. Host checks that TPM received data is valid.
>>    4. If data is currupted go to step 1.
>>
>> When receiving data:
>>    1. Host checks that TPM_STS.dataAvail is set.
>>    2. Host saves received data.
>>    3. Host checks that received data is correct.
>>    4. If data is currupted Host writes to TPM_STS.responseRetry and go to
>>       step 1.
>
> These sequences in the commit message look somewhat uselss. Maybe
> just remove them.
>

Their main porpose is to describe how the retry attempt is implemented
in case of currupted data.
should i just describe that with a few words or that's unnecessary?

>>
>> Co-developed-by: Christophe Ricard <christophe-h.ricard@st.com>
>> Signed-off-by: Christophe Ricard <christophe-h.ricard@st.com>
>> Signed-off-by: Amir Mizinski <amirmizi6@gmail.com>
>> ---
>>  drivers/char/tpm/tpm_tis_core.c | 102 +++++++++++++++++++++++++---------------
>>  drivers/char/tpm/tpm_tis_core.h |   3 ++
>>  2 files changed, 67 insertions(+), 38 deletions(-)
>>
>> diff --git a/drivers/char/tpm/tpm_tis_core.c b/drivers/char/tpm/tpm_tis_core.c
>> index 27c6ca0..6c4f232 100644
>> --- a/drivers/char/tpm/tpm_tis_core.c
>> +++ b/drivers/char/tpm/tpm_tis_core.c
>> @@ -242,6 +242,15 @@ static u8 tpm_tis_status(struct tpm_chip *chip)
>>      return status;
>>  }
>>  
>> +static bool tpm_tis_check_data(struct tpm_chip *chip, const u8 *buf, size_t len)
>
> Not sure if this is the best possible function name, "check" can
> mean almost anything.
>

Ok, i'm changing it to "verify_data_integrity". is that ok?

>> +{
>> +    struct tpm_tis_data *priv = dev_get_drvdata(&chip->dev);
>> +
>> +    if (priv->phy_ops->check_data)
>> +        return priv->phy_ops->check_data(priv, buf, len);
>
> New line here before the return statement.
>
>> +    return true;
>> +}
>> +
>>  static void tpm_tis_ready(struct tpm_chip *chip)
>>  {
>>      struct tpm_tis_data *priv = dev_get_drvdata(&chip->dev);
>> @@ -308,47 +317,59 @@ static int tpm_tis_recv(struct tpm_chip *chip, u8 *buf, size_t count)
>>  {
>>      struct tpm_tis_data *priv = dev_get_drvdata(&chip->dev);
>>      int size = 0;
>> -    int status;
>> +    int status, i;
>>      u32 expected;
>> +    bool check_data = false;
>>  
>> -    if (count < TPM_HEADER_SIZE) {
>> -        size = -EIO;
>> -        goto out;
>> -    }
>> +    for (i = 0; i < TPM_RETRY; i++) {
>> +        if (count < TPM_HEADER_SIZE) {
>> +            size = -EIO;
>> +            goto out;
>> +        }
>>  
>> -    size = recv_data(chip, buf, TPM_HEADER_SIZE);
>> -    /* read first 10 bytes, including tag, paramsize, and result */
>> -    if (size < TPM_HEADER_SIZE) {
>> -        dev_err(&chip->dev, "Unable to read header\n");
>> -        goto out;
>> -    }
>> +        size = recv_data(chip, buf, TPM_HEADER_SIZE);
>> +        /* read first 10 bytes, including tag, paramsize, and result */
>> +        if (size < TPM_HEADER_SIZE) {
>> +            dev_err(&chip->dev, "Unable to read header\n");
>> +            goto out;
>> +        }
>>  
>> -    expected = be32_to_cpu(*(__be32 *) (buf + 2));
>> -    if (expected > count || expected < TPM_HEADER_SIZE) {
>> -        size = -EIO;
>> -        goto out;
>> -    }
>> +        expected = be32_to_cpu(*(__be32 *) (buf + 2));
>> +        if (expected > count || expected < TPM_HEADER_SIZE) {
>> +            size = -EIO;
>> +            goto out;
>> +        }
>>  
>> -    size += recv_data(chip, &buf[TPM_HEADER_SIZE],
>> -              expected - TPM_HEADER_SIZE);
>> -    if (size < expected) {
>> -        dev_err(&chip->dev, "Unable to read remainder of result\n");
>> -        size = -ETIME;
>> -        goto out;
>> -    }
>> +        size += recv_data(chip, &buf[TPM_HEADER_SIZE],
>> +                  expected - TPM_HEADER_SIZE);
>> +        if (size < expected) {
>> +            dev_err(&chip->dev, "Unable to read remainder of result\n");
>> +            size = -ETIME;
>> +            goto out;
>> +        }
>>  
>> -    if (wait_for_tpm_stat(chip, TPM_STS_VALID, chip->timeout_c,
>> -                &priv->int_queue, false) < 0) {
>> -        size = -ETIME;
>> -        goto out;
>> +        if (wait_for_tpm_stat(chip, TPM_STS_VALID, chip->timeout_c,
>> +                      &priv->int_queue, false) < 0) {
>> +            size = -ETIME;
>> +            goto out;
>> +        }
>> +
>> +        status = tpm_tis_status(chip);
>> +        if (status & TPM_STS_DATA_AVAIL) {    /* retry? */
>> +            dev_err(&chip->dev, "Error left over data\n");
>> +            size = -EIO;
>> +            goto out;
>> +        }
>> +
>> +        check_data = tpm_tis_check_data(chip, buf, size);
>> +        if (!check_data)
>> +            tpm_tis_write8(priv, TPM_STS(priv->locality),
>> +                       TPM_STS_RESPONSE_RETRY);
>> +        else
>> +            break;
>>      }
>> -    status = tpm_tis_status(chip);
>> -    if (status & TPM_STS_DATA_AVAIL) {    /* retry? */
>> -        dev_err(&chip->dev, "Error left over data\n");
>> +    if (!check_data)
>>          size = -EIO;
>> -        goto out;
>> -    }
>> -
>>  out:
>>      tpm_tis_ready(chip);
>>      return size;
>> @@ -453,14 +474,19 @@ static void disable_interrupts(struct tpm_chip *chip)
>>  static int tpm_tis_send_main(struct tpm_chip *chip, const u8 *buf, size_t len)
>>  {
>>      struct tpm_tis_data *priv = dev_get_drvdata(&chip->dev);
>> -    int rc;
>> +    int rc, i;
>>      u32 ordinal;
>>      unsigned long dur;
>> +    bool data_valid = false;
>>  
>> -    rc = tpm_tis_send_data(chip, buf, len);
>> -    if (rc < 0)
>> -        return rc;
>> -
>> +    for (i = 0; i < TPM_RETRY && !data_valid; i++) {
>> +        rc = tpm_tis_send_data(chip, buf, len);
>> +        if (rc < 0)
>> +            return rc;
>> +        data_valid = tpm_tis_check_data(chip, buf, len);
>> +    }
>> +    if (!data_valid)
>> +        return -EIO;
>>      /* go and do it */
>>      rc = tpm_tis_write8(priv, TPM_STS(priv->locality), TPM_STS_GO);
>>      if (rc < 0)
>> diff --git a/drivers/char/tpm/tpm_tis_core.h b/drivers/char/tpm/tpm_tis_core.h
>> index d06c65b..486c2e9 100644
>> --- a/drivers/char/tpm/tpm_tis_core.h
>> +++ b/drivers/char/tpm/tpm_tis_core.h
>> @@ -34,6 +34,7 @@ enum tis_status {
>>      TPM_STS_GO = 0x20,
>>      TPM_STS_DATA_AVAIL = 0x10,
>>      TPM_STS_DATA_EXPECT = 0x08,
>> +    TPM_STS_RESPONSE_RETRY = 0x02,
>>  };
>>  
>>  enum tis_int_flags {
>> @@ -106,6 +107,8 @@ struct tpm_tis_phy_ops {
>>      int (*read16)(struct tpm_tis_data *data, u32 addr, u16 *result);
>>      int (*read32)(struct tpm_tis_data *data, u32 addr, u32 *result);
>>      int (*write32)(struct tpm_tis_data *data, u32 addr, u32 src);
>> +    bool (*check_data)(struct tpm_tis_data *data, const u8 *buf,
>> +               size_t len);
>
> Aren't you validating the contents of the buf?
>
> /Jarkko

i do.
when sending, the data is written to the buff in "tpm_tis_send_data(chip,buf, len)".
and validated in "data_valid = tpm_tis_check_data(chip, buf, len)".
data is not sent until TPM_STS_GO is set.

when receiving, the data in the buffer is verified after recv_data, and
writing to TPM_STS_RESPONSE_RETRY in case it fails to recive it again.


Thank you
Amir Mizinski


  parent reply	other threads:[~2020-04-21 13:34 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-04-07 16:20 [PATCH v6 0/7] Add tpm i2c ptp driver amirmizi6
2020-04-07 16:20 ` [PATCH v6 1/7] tpm: tpm_tis: Make implementation of read16 read32 write32 optional amirmizi6
2020-04-08 16:46   ` Jarkko Sakkinen
2020-04-07 16:20 ` [PATCH v6 2/7] tpm: tpm_tis: Add check_data handle to tpm_tis_phy_ops amirmizi6
2020-04-08 18:33   ` Jarkko Sakkinen
     [not found]     ` <CAMHTsUUzQ9rpPfsnEtJruAC3THr+XNA=5Wk7OCteqZXOQ0L=UA@mail.gmail.com>
2020-04-12 13:53       ` Jarkko Sakkinen
2020-04-21 13:34     ` Amir Mizinski [this message]
2020-04-21 20:19       ` Jarkko Sakkinen
2020-04-07 16:20 ` [PATCH v6 3/7] tpm: tpm_tis: rewrite "tpm_tis_req_canceled()" amirmizi6
2020-04-07 16:20 ` [PATCH v6 4/7] tpm: tpm_tis: Fix expected bit handling and send all bytes in one shot without last byte in exception amirmizi6
2020-04-07 16:20 ` [PATCH v6 5/7] tpm: Handle an exception for TPM Firmware Update mode amirmizi6
2020-04-07 16:20 ` [PATCH v6 6/7] dt-bindings: tpm: Add YAML schema for TPM TIS I2C options amirmizi6
2020-04-15 15:20   ` Rob Herring
2020-04-22  8:15     ` Amir Mizinski
2020-04-07 16:20 ` [PATCH v6 7/7] tpm: tpm_tis: add tpm_tis_i2c driver amirmizi6
2020-04-08 18:36   ` Jarkko Sakkinen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=0cfa0486-8ccb-d7d1-acf2-ca103f723b3a@gmail.com \
    --to=amirmizi6@gmail.com \
    --cc=Dan.Morav@nuvoton.com \
    --cc=Eyal.Cohen@nuvoton.com \
    --cc=alexander.steffen@infineon.com \
    --cc=amir.mizinski@nuvoton.com \
    --cc=arnd@arndb.de \
    --cc=christophe-h.ricard@st.com \
    --cc=devicetree@vger.kernel.org \
    --cc=gcwilson@us.ibm.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=jarkko.sakkinen@linux.intel.com \
    --cc=jgg@ziepe.ca \
    --cc=kgoldman@us.ibm.com \
    --cc=linux-integrity@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=oren.tanami@nuvoton.com \
    --cc=oshri.alkoby@nuvoton.com \
    --cc=oshrialkoby85@gmail.com \
    --cc=peterhuewe@gmx.de \
    --cc=robh+dt@kernel.org \
    --cc=shmulik.hager@nuvoton.com \
    --cc=tmaimon77@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).