linux-integrity.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jarkko Sakkinen <jarkko@kernel.org>
To: Borys Movchan <borysmn@axis.com>, Peter Huewe <peterhuewe@gmx.de>,
	Jason Gunthorpe <jgg@ziepe.ca>
Cc: kernel@axis.com, linux-integrity@vger.kernel.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v6] tpm: Add Upgrade/Reduced mode support for TPM2 modules
Date: Thu, 30 Sep 2021 19:28:34 +0300	[thread overview]
Message-ID: <59f8d5a835ff0f02c9efe0ecff0abbe1b4f08111.camel@kernel.org> (raw)
In-Reply-To: <20210930160241.9691-1-borysmn@axis.com>

On Thu, 2021-09-30 at 18:02 +0200, Borys Movchan wrote:
> If something went wrong during the TPM firmware upgrade, like power
> failure or the firmware image file get corrupted, the TPM might end
> up in Upgrade or Failure mode upon the next start. The state is
> persistent between the TPM power cycle/restart.
> 
> According to TPM specification:
>  * If the TPM is in Upgrade mode, it will answer with TPM2_RC_UPGRADE
>    to all commands except Field Upgrade related ones.

How many such command are there and what they are? This patch is missing
the part that filters in only these commands.

>  * If the TPM is in Failure mode, it will allow performing TPM
>    initialization but will not provide any crypto operations.
>    Will happily respond to Field Upgrade calls.
> 
> Change the behavior of the tpm2_auto_startup(), so it detects the active
> running mode of the TPM by adding the following checks.  If
> tpm2_do_selftest() call returns TPM2_RC_UPGRADE, the TPM is in Upgrade
> mode.
> If the TPM is in Failure mode, it will successfully respond to both
> tpm2_do_selftest() and tpm2_startup() calls. Although, will fail to
> answer to tpm2_get_cc_attrs_tbl(). Use this fact to conclude that TPM is
> in Failure mode.
> 
> If detected that the TPM is in the Upgrade or Failure mode, the function
> sets TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE flag.

I think we should shorten this to TPM_CHIP_FLAG_FIRMWARE_MODE. It's clear
enough and just easier for eyes, really.

> The limited mode flag is used later during driver

You probably have forgot to update commit message over here? Please try to
align it with the code changes.

> initialization/deinitialization to disable functionality which makes no
> sense or will fail in the current TPM state. Following functionality is
> affected:
>  * do not register TPM as a hwrng

Nit: "do" -> "Do"

>  * do not register sysfs entries which provide information impossible to
>    obtain in limited mode
>  * do not register resource managed character device

Ditto.

> 
> Signed-off-by: Borys Movchan <borysmn@axis.com>
> ---
> 
> Notes:
>     v2:
>      * Commit message updated.
>     
>     v3:
>      * Commit message reworked.
>     
>     v4:
>      * Description of how tpm2_auto_startup() detects the mode added to
>        commit message.
>     
>     v5:
>      * Introduce global flag: TPM_CHIP_FLAG_LIMITED_MODE.
>      * Add checks for the flag in places that will not work properly when TPM
>        functionality is limited.
>      * Avoid registering sysfs and character device entries that have no useful
>        function in limited mode.
>      * Do not register TPM as a hwrng.
>      * Do not try to obtain any crypto-related properties from TPM as it will fail
>        in limited mode.
>     
>     v6:
>      * Rename the TPM_CHIP_FLAG_LIMITED_MODE to TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE
>        to reduce confusion. Update info messages.
> 
>  drivers/char/tpm/tpm-chip.c  | 23 ++++++++++++++++-------
>  drivers/char/tpm/tpm-sysfs.c |  3 +++
>  drivers/char/tpm/tpm2-cmd.c  | 14 +++++++++++++-
>  include/linux/tpm.h          |  2 ++
>  4 files changed, 34 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/char/tpm/tpm-chip.c b/drivers/char/tpm/tpm-chip.c
> index ddaeceb7e109..3378bfe06006 100644
> --- a/drivers/char/tpm/tpm-chip.c
> +++ b/drivers/char/tpm/tpm-chip.c
> @@ -444,7 +444,8 @@ static int tpm_add_char_device(struct tpm_chip *chip)
>  		return rc;
>  	}
>  
> -	if (chip->flags & TPM_CHIP_FLAG_TPM2) {
> +	if (chip->flags & TPM_CHIP_FLAG_TPM2 &&
> +	    !(chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE)) {

Please do this instead in tpm.h, give that the check is everywhere:

/*
 * Check if TPM device is in the firmware upgrade mode.
 */
static inline bool tpm_is_firmware_upgrade(void)
{
	return chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE;
}


Makes the code less stressing to read (and find the decision points
later on with tools).

>  		rc = cdev_device_add(&chip->cdevs, &chip->devs);
>  		if (rc) {
>  			dev_err(&chip->devs,
> @@ -488,7 +489,8 @@ static void tpm_del_legacy_sysfs(struct tpm_chip *chip)
>  {
>  	struct attribute **i;
>  
> -	if (chip->flags & (TPM_CHIP_FLAG_TPM2 | TPM_CHIP_FLAG_VIRTUAL))
> +	if (chip->flags & (TPM_CHIP_FLAG_TPM2 | TPM_CHIP_FLAG_VIRTUAL) ||
> +	    chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE)
>  		return;
>  
>  	sysfs_remove_link(&chip->dev.parent->kobj, "ppi");
> @@ -506,7 +508,8 @@ static int tpm_add_legacy_sysfs(struct tpm_chip *chip)
>  	struct attribute **i;
>  	int rc;
>  
> -	if (chip->flags & (TPM_CHIP_FLAG_TPM2 | TPM_CHIP_FLAG_VIRTUAL))
> +	if (chip->flags & (TPM_CHIP_FLAG_TPM2 | TPM_CHIP_FLAG_VIRTUAL) ||
> +		chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE)
>  		return 0;
>  
>  	rc = compat_only_sysfs_link_entry_to_kobj(
> @@ -536,7 +539,8 @@ static int tpm_hwrng_read(struct hwrng *rng, void *data, size_t max, bool wait)
>  
>  static int tpm_add_hwrng(struct tpm_chip *chip)
>  {
> -	if (!IS_ENABLED(CONFIG_HW_RANDOM_TPM))
> +	if (!IS_ENABLED(CONFIG_HW_RANDOM_TPM) ||
> +	    chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE)
>  		return 0;
>  
>  	snprintf(chip->hwrng_name, sizeof(chip->hwrng_name),
> @@ -550,6 +554,9 @@ static int tpm_get_pcr_allocation(struct tpm_chip *chip)
>  {
>  	int rc;
>  
> +	if (chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE)
> +		return 0;
> +
>  	rc = (chip->flags & TPM_CHIP_FLAG_TPM2) ?
>  	     tpm2_get_pcr_allocation(chip) :
>  	     tpm1_get_pcr_allocation(chip);
> @@ -612,7 +619,7 @@ int tpm_chip_register(struct tpm_chip *chip)
>  	return 0;
>  
>  out_hwrng:
> -	if (IS_ENABLED(CONFIG_HW_RANDOM_TPM))
> +	if (IS_ENABLED(CONFIG_HW_RANDOM_TPM) && !(chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE))
>  		hwrng_unregister(&chip->hwrng);
>  out_ppi:
>  	tpm_bios_log_teardown(chip);
> @@ -637,10 +644,12 @@ EXPORT_SYMBOL_GPL(tpm_chip_register);
>  void tpm_chip_unregister(struct tpm_chip *chip)
>  {
>  	tpm_del_legacy_sysfs(chip);
> -	if (IS_ENABLED(CONFIG_HW_RANDOM_TPM))
> +	if (IS_ENABLED(CONFIG_HW_RANDOM_TPM) &&
> +	    !(chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE))
>  		hwrng_unregister(&chip->hwrng);
>  	tpm_bios_log_teardown(chip);
> -	if (chip->flags & TPM_CHIP_FLAG_TPM2)
> +	if (chip->flags & TPM_CHIP_FLAG_TPM2 &&
> +	    !(chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE))
>  		cdev_device_del(&chip->cdevs, &chip->devs);
>  	tpm_del_char_device(chip);
>  }
> diff --git a/drivers/char/tpm/tpm-sysfs.c b/drivers/char/tpm/tpm-sysfs.c
> index 63f03cfb8e6a..2842f3c667a5 100644
> --- a/drivers/char/tpm/tpm-sysfs.c
> +++ b/drivers/char/tpm/tpm-sysfs.c
> @@ -480,6 +480,9 @@ void tpm_sysfs_add_device(struct tpm_chip *chip)
>  
>  	WARN_ON(chip->groups_cnt != 0);
>  
> +	if (chip->flags & TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE)
> +		return;
> +
>  	if (chip->flags & TPM_CHIP_FLAG_TPM2)
>  		chip->groups[chip->groups_cnt++] = &tpm2_dev_group;
>  	else
> diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c
> index a25815a6f625..76245c861e59 100644
> --- a/drivers/char/tpm/tpm2-cmd.c
> +++ b/drivers/char/tpm/tpm2-cmd.c
> @@ -729,7 +729,12 @@ int tpm2_auto_startup(struct tpm_chip *chip)
>  		goto out;
>  
>  	rc = tpm2_do_selftest(chip);
> -	if (rc && rc != TPM2_RC_INITIALIZE)
> +	if (rc == TPM2_RC_UPGRADE) {
> +		dev_info(&chip->dev, "TPM requires firmware upgrade\n");
> +		chip->flags |= TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE;
> +		rc = 0;
> +		goto out;
> +	} else if (rc && rc != TPM2_RC_INITIALIZE)
>  		goto out;
>  
>  	if (rc == TPM2_RC_INITIALIZE) {
> @@ -743,6 +748,13 @@ int tpm2_auto_startup(struct tpm_chip *chip)
>  	}
>  
>  	rc = tpm2_get_cc_attrs_tbl(chip);
> +	if (rc) {
> +		dev_info(&chip->dev,
> +			 "TPM requires firmware recovery/upgrade\n");

Why these two dev_info() messages differ?

You should put a single info message that is unconditionally printed
when in FW upgrade mode, and not sprikle bloat to fallback paths.

> +		chip->flags |= TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE;
> +		rc = 0;
> +		goto out;
> +	}
>  
>  out:
>  	if (rc > 0)
> diff --git a/include/linux/tpm.h b/include/linux/tpm.h
> index aa11fe323c56..498f487d786f 100644
> --- a/include/linux/tpm.h
> +++ b/include/linux/tpm.h
> @@ -207,6 +207,7 @@ enum tpm2_return_codes {
>  	TPM2_RC_INITIALIZE	= 0x0100, /* RC_VER1 */
>  	TPM2_RC_FAILURE		= 0x0101,
>  	TPM2_RC_DISABLED	= 0x0120,
> +	TPM2_RC_UPGRADE		= 0x012D,
>  	TPM2_RC_COMMAND_CODE    = 0x0143,
>  	TPM2_RC_TESTING		= 0x090A, /* RC_WARN */
>  	TPM2_RC_REFERENCE_H0	= 0x0910,
> @@ -277,6 +278,7 @@ enum tpm_chip_flags {
>  	TPM_CHIP_FLAG_HAVE_TIMEOUTS	= BIT(4),
>  	TPM_CHIP_FLAG_ALWAYS_POWERED	= BIT(5),
>  	TPM_CHIP_FLAG_FIRMWARE_POWER_MANAGED	= BIT(6),
> +	TPM_CHIP_FLAG_FIRMWARE_UPGRADE_MODE	= BIT(7),
>  };
>  
>  #define to_tpm_chip(d) container_of(d, struct tpm_chip, dev)

/Jarkko


  reply	other threads:[~2021-09-30 16:28 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-09-30 16:02 [PATCH v6] tpm: Add Upgrade/Reduced mode support for TPM2 modules Borys Movchan
2021-09-30 16:28 ` Jarkko Sakkinen [this message]
2021-09-30 16:31   ` Jarkko Sakkinen
2021-12-08  9:30   ` Axel Jonsson
2021-12-08  9:53   ` [PATCH v7] " Axel Jonsson
2021-12-11 14:15     ` Jarkko Sakkinen
2021-12-13  7:06       ` Axel Jonsson
2021-12-13  7:09       ` [PATCH v8] " Axel Jonsson
2021-12-21  8:35         ` Jarkko Sakkinen
2021-09-30 16:30 ` [PATCH v6] " Borys Movchan

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=59f8d5a835ff0f02c9efe0ecff0abbe1b4f08111.camel@kernel.org \
    --to=jarkko@kernel.org \
    --cc=borysmn@axis.com \
    --cc=jgg@ziepe.ca \
    --cc=kernel@axis.com \
    --cc=linux-integrity@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=peterhuewe@gmx.de \
    /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).