tpm_tis: verify locality released before returning from release_locality
Subsystems (2)
PATCH 1/1 tpm_tis: verify locality released before returning from release_locality
For certain tpm chips releasing locality can take long enough that a subsequent call to request_locality will see the locality as being active when the access register is read in check_locality. So check that the locality has been released before returning from release_locality.
--- drivers/char/tpm/tpm_tis_core.c | 47 ++++++++++++++++++++++++++++++++++++++++- 1 file changed, 46 insertions(+), 1 deletion(-)
drivers/char/tpm/tpm_tis_core.c +46 -1
@@ -143,13 +143,58 @@ static bool check_locality(struct tpm_chip *chip, int l)return false; } +static bool locality_inactive (struct tpm_chip *chip,int l) +{ +struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +int rc; +u8 access; + + rc =tpm_tis_read8 (priv,TPM_ACCESS (l), &access); +if (rc <0 ) +return false; + +if ((access & (TPM_ACCESS_VALID |TPM_ACCESS_ACTIVE_LOCALITY )) + ==TPM_ACCESS_VALID ) +return true; + +return false; +} +static int release_locality (struct tpm_chip *chip,int l) {struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +unsigned long stop, timeout; +long rc;tpm_tis_write8 (priv,TPM_ACCESS (l),TPM_ACCESS_ACTIVE_LOCALITY ); -return 0 ; + stop = jiffies + chip->timeout_a; + +if (chip->flags &TPM_CHIP_FLAG_IRQ ) { +again: + timeout = stop - jiffies; +if ((long )timeout <=0 ) +return -1 ; + + rc =wait_event_interruptible_timeout (priv->int_queue, + (locality_inactive (chip, l)), + timeout); + +if (rc >0 ) +return 0 ; + +if (rc == -ERESTARTSYS &&freezing (current)) { +clear_thread_flag (TIF_SIGPENDING ); + goto again; + } + }else { +do { +if (locality_inactive (chip, l)) +return 0 ; +tpm_msleep (TPM_TIMEOUT ); + }while (time_before (jiffies, stop)); + } +return -1 ; }static int request_locality (struct tpm_chip *chip,int l)
Signature · 1 line
2.15.0
Quoted 82 lines · Jerry Snitselaar
For certain tpm chips releasing locality can take long enough that a subsequent call to
request_localitywill see the locality as being active when the access register is read incheck_locality. So check that the locality has been released before returning fromrelease_locality.Cc Jarkko Sakkinen <[email protected]>Cc Peter Huewe <[email protected]>Cc Jason Gunthorpe <[email protected]>Reported-by Laurent Bigonville <[email protected]>Signed-off-by Jerry Snitselaar <[email protected]>---
drivers/char/tpm/tpm_tis_core.c| 47 ++++++++++++++++++++++++++++++++++++++++-1 file changed, 46 insertions(+), 1 deletion(-)drivers/char/tpm/tpm_tis_core.c +46
-1@@ -143,13 +143,58 @@ static bool check_locality(struct tpm_chip *chip, int l)return false; } +static bool locality_inactive (struct tpm_chip *chip,int l) +{ +struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +int rc; +u8 access; + + rc =tpm_tis_read8 (priv,TPM_ACCESS (l), &access); +if (rc <0 ) +return false; + +if ((access & (TPM_ACCESS_VALID |TPM_ACCESS_ACTIVE_LOCALITY )) + ==TPM_ACCESS_VALID ) +return true; + +return false; +} +static int release_locality (struct tpm_chip *chip,int l) {struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +unsigned long stop, timeout; +long rc;tpm_tis_write8 (priv,TPM_ACCESS (l),TPM_ACCESS_ACTIVE_LOCALITY ); -return 0 ; + stop = jiffies + chip->timeout_a; + +if (chip->flags &TPM_CHIP_FLAG_IRQ ) { +again: + timeout = stop - jiffies; +if ((long )timeout <=0 ) +return -1 ; + + rc =wait_event_interruptible_timeout (priv->int_queue, + (locality_inactive (chip, l)), + timeout); + +if (rc >0 ) +return 0 ; + +if (rc == -ERESTARTSYS &&freezing (current)) { +clear_thread_flag (TIF_SIGPENDING ); + goto again; + } + }else { +do { +if (locality_inactive (chip, l)) +return 0 ; +tpm_msleep (TPM_TIMEOUT ); + }while (time_before (jiffies, stop)); + } +return -1 ; }static int request_locality (struct tpm_chip *chip,int l)Signature · 1 line
2.15.0
Laurent,
Can you try this patch with your system since it is the one that has exhibited the problem so far. I've tested on a tpm2.0 and tpm1.2 system here.
Regards,
Jerry
Le 05/05/18 a 22:03, Jerry Snitselaar a ecrit :
Quoted 5 lines · Jerry Snitselaar
Quoted 2 lines
[...]
Laurent,
Hello Jerry,
Quoted 3 lines
Can you try this patch with your system since it is the one that has exhibited the problem so far. I've tested on a tpm2.0 and tpm1.2 system here.
I just tested the patch and the driver is loading fine again and the device in /dev is present again, so it seems to work.
But for some reason the tpm is again locked (for no visible reason) due to "dictionary attack" so I cannot test further ATM :/
Regards,
Laurent
Quoted 3 lines
Regards,
Jerry
Quoted 16 lines · Laurent Bigonville
Le 05/05/18 a 22:03, Jerry Snitselaar a ecrit :
Quoted 5 lines · Jerry Snitselaar
Quoted 2 lines
[...]
Laurent,
Hello Jerry,
Quoted 3 lines
Can you try this patch with your system since it is the one that has exhibited the problem so far. I've tested on a tpm2.0 and tpm1.2 system here.
I just tested the patch and the driver is loading fine again and the device in /dev is present again, so it seems to work.
Quoted 11 lines
But for some reason the tpm is again locked (for no visible reason) due to "dictionary attack" so I cannot test further ATM :/
Regards,
Laurent
Quoted 3 lines
Regards,
Jerry
Can you response with the tested-by tag as soon as you can? Before we have tested-by, I cannot land the fix.
/Jarkko
Laurent Bigonville
Tested-byLe 05/05/18 a 22:03, Jerry Snitselaar a ecrit :
Quoted 7 lines · Jerry Snitselaar
Quoted 6 lines
For certain tpm chips releasing locality can take long enough that a subsequent call to
request_localitywill see the locality as being active when the access register is read incheck_locality. So check that the locality has been released before returning fromrelease_locality.
Quoted 5 lines
Quoted 5 lines
Cc Jarkko Sakkinen <[email protected]>Cc Peter Huewe <[email protected]>Cc Jason Gunthorpe <[email protected]>Reported-by Laurent Bigonville <[email protected]>Signed-off-by Jerry Snitselaar <[email protected]>
Quoted 82 lines
Quoted 69 lines
---
drivers/char/tpm/tpm_tis_core.c| 47
++++++++++++++++++++++++++++++++++++++++-1 file changed, 46 insertions(+), 1 deletion(-)drivers/char/tpm/tpm_tis_core.c +46
-1@@ -143,13 +143,58 @@ static bool check_locality(struct tpm_chip *chip, int l)return false; } +static bool locality_inactive (struct tpm_chip *chip,int l) +{ +struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +int rc; +u8 access; + + rc =tpm_tis_read8 (priv,TPM_ACCESS (l), &access); +if (rc <0 ) +return false; + +if ((access & (TPM_ACCESS_VALID |TPM_ACCESS_ACTIVE_LOCALITY )) + ==TPM_ACCESS_VALID ) +return true; + +return false; +} +static int release_locality (struct tpm_chip *chip,int l) {struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +unsigned long stop, timeout; +long rc;tpm_tis_write8 (priv,TPM_ACCESS (l),TPM_ACCESS_ACTIVE_LOCALITY ); -return 0 ; + stop = jiffies + chip->timeout_a; + +if (chip->flags &TPM_CHIP_FLAG_IRQ ) { +again: + timeout = stop - jiffies; +if ((long )timeout <=0 ) +return -1 ; + + rc =wait_event_interruptible_timeout (priv->int_queue, + (locality_inactive (chip, l)), + timeout); + +if (rc >0 ) +return 0 ; + +if (rc == -ERESTARTSYS &&freezing (current)) { +clear_thread_flag (TIF_SIGPENDING ); + goto again; + } + }else { +do { +if (locality_inactive (chip, l)) +return 0 ; +tpm_msleep (TPM_TIMEOUT ); + }while (time_before (jiffies, stop)); + } +return -1 ; }Quoted 4 lines
static int
request_locality(struct tpm_chip *chip, int l)Signature · 1 line
2.15.0Laurent,
Can you try this patch with your system since it is the one that has exhibited the problem so far. I've tested on a tpm2.0 and tpm1.2 system here.
Regards,
Jerry
Hello,
Top posting, sorry.
I don't know if I did it well to include the "Tested-by" tag because I don't see that the patch has landed in linus branch already.
And as far as I understand, this will not be in the upcoming 4.17 release as we are already late in the cycle?
Kind regards,
Laurent Bigonville
Le 11/05/18 a 21:02, Laurent Bigonville a ecrit :
Quoted 97 lines
Le 05/05/18 a 22:03, Jerry Snitselaar a ecrit :
Quoted 7 lines · Jerry Snitselaar
Quoted 6 lines
For certain tpm chips releasing locality can take long enough that a subsequent call to
request_localitywill see the locality as being active when the access register is read incheck_locality. So check that the locality has been released before returning fromrelease_locality.Quoted 5 lines
Quoted 5 lines
Cc Jarkko Sakkinen <[email protected]>Cc Peter Huewe <[email protected]>Cc Jason Gunthorpe <[email protected]>Reported-by Laurent Bigonville <[email protected]>Signed-off-by Jerry Snitselaar <[email protected]>Tested-by Laurent Bigonville <[email protected]>Quoted 82 lines
Quoted 69 lines
---
drivers/char/tpm/tpm_tis_core.c| 47
++++++++++++++++++++++++++++++++++++++++-1 file changed, 46 insertions(+), 1 deletion(-)drivers/char/tpm/tpm_tis_core.c +46
-1@@ -143,13 +143,58 @@ static bool check_locality(struct tpm_chip *chip, int l)return false; } +static bool locality_inactive (struct tpm_chip *chip,int l) +{ +struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +int rc; +u8 access; + + rc =tpm_tis_read8 (priv,TPM_ACCESS (l), &access); +if (rc <0 ) +return false; + +if ((access & (TPM_ACCESS_VALID |TPM_ACCESS_ACTIVE_LOCALITY )) + ==TPM_ACCESS_VALID ) +return true; + +return false; +} +static int release_locality (struct tpm_chip *chip,int l) {struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +unsigned long stop, timeout; +long rc;tpm_tis_write8 (priv,TPM_ACCESS (l),TPM_ACCESS_ACTIVE_LOCALITY ); -return 0 ; + stop = jiffies + chip->timeout_a; + +if (chip->flags &TPM_CHIP_FLAG_IRQ ) { +again: + timeout = stop - jiffies; +if ((long )timeout <=0 ) +return -1 ; + + rc =wait_event_interruptible_timeout (priv->int_queue, + (locality_inactive (chip, l)), + timeout); + +if (rc >0 ) +return 0 ; + +if (rc == -ERESTARTSYS &&freezing (current)) { +clear_thread_flag (TIF_SIGPENDING ); + goto again; + } + }else { +do { +if (locality_inactive (chip, l)) +return 0 ; +tpm_msleep (TPM_TIMEOUT ); + }while (time_before (jiffies, stop)); + } +return -1 ; }Quoted 4 lines
static int
request_locality(struct tpm_chip *chip, int l)Signature · 1 line
2.15.0Laurent,
Can you try this patch with your system since it is the one that has exhibited the problem so far. I've tested on a tpm2.0 and tpm1.2 system here.
Regards,
Jerry
Quoted 4 lines · Laurent Bigonville
Hello,
Top posting, sorry.
Quoted 8 lines
I don't know if I did it well to include the "Tested-by" tag because I don't see that the patch has landed in linus branch already.
And as far as I understand, this will not be in the upcoming 4.17 release as we are already late in the cycle?
Kind regards,
Laurent Bigonville
It should go into his branch during the merge window for 4.18.
Quoted 100 lines
Le 11/05/18 a 21:02, Laurent Bigonville a ecrit :
Quoted 97 lines
Le 05/05/18 a 22:03, Jerry Snitselaar a ecrit :
Quoted 7 lines · Jerry Snitselaar
Quoted 6 lines
For certain tpm chips releasing locality can take long enough that a subsequent call to
request_localitywill see the locality as being active when the access register is read incheck_locality. So check that the locality has been released before returning fromrelease_locality.Quoted 5 lines
Quoted 5 lines
Cc Jarkko Sakkinen <[email protected]>Cc Peter Huewe <[email protected]>Cc Jason Gunthorpe <[email protected]>Reported-by Laurent Bigonville <[email protected]>Signed-off-by Jerry Snitselaar <[email protected]>Tested-by Laurent Bigonville <[email protected]>Quoted 82 lines
Quoted 69 lines
---
drivers/char/tpm/tpm_tis_core.c| 47
++++++++++++++++++++++++++++++++++++++++-1 file changed, 46 insertions(+), 1 deletion(-)drivers/char/tpm/tpm_tis_core.c +46
-1@@ -143,13 +143,58 @@ static bool check_locality(struct tpm_chip *chip, int l)return false; } +static bool locality_inactive (struct tpm_chip *chip,int l) +{ +struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +int rc; +u8 access; + + rc =tpm_tis_read8 (priv,TPM_ACCESS (l), &access); +if (rc <0 ) +return false; + +if ((access & (TPM_ACCESS_VALID |TPM_ACCESS_ACTIVE_LOCALITY )) + ==TPM_ACCESS_VALID ) +return true; + +return false; +} +static int release_locality (struct tpm_chip *chip,int l) {struct tpm_tis_data *priv =dev_get_drvdata (&chip->dev); +unsigned long stop, timeout; +long rc;tpm_tis_write8 (priv,TPM_ACCESS (l),TPM_ACCESS_ACTIVE_LOCALITY ); -return 0 ; + stop = jiffies + chip->timeout_a; + +if (chip->flags &TPM_CHIP_FLAG_IRQ ) { +again: + timeout = stop - jiffies; +if ((long )timeout <=0 ) +return -1 ; + + rc =wait_event_interruptible_timeout (priv->int_queue, + (locality_inactive (chip, l)), + timeout); + +if (rc >0 ) +return 0 ; + +if (rc == -ERESTARTSYS &&freezing (current)) { +clear_thread_flag (TIF_SIGPENDING ); + goto again; + } + }else { +do { +if (locality_inactive (chip, l)) +return 0 ; +tpm_msleep (TPM_TIMEOUT ); + }while (time_before (jiffies, stop)); + } +return -1 ; }Quoted 4 lines
static int
request_locality(struct tpm_chip *chip, int l)Signature · 1 line
2.15.0Laurent,
Can you try this patch with your system since it is the one that has exhibited the problem so far. I've tested on a tpm2.0 and tpm1.2 system here.
Regards,
Jerry
Quoted 11 lines · Jerry Snitselaar
For certain tpm chips releasing locality can take long enough that a subsequent call to
request_localitywill see the locality as being active when the access register is read incheck_locality. So check that the locality has been released before returning fromrelease_locality.Cc Jarkko Sakkinen <[email protected]>Cc Peter Huewe <[email protected]>Cc Jason Gunthorpe <[email protected]>Reported-by Laurent Bigonville <[email protected]>Signed-off-by Jerry Snitselaar <[email protected]>
/Jarkko