2019-04-01 19:07:03

by Kees Cook

[permalink] [raw]
Subject: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

A "get random" may fail with a TPM error, but those codes were returned
as-is to the caller, which assumed the result was the number of bytes
that had been written to the target buffer, which could lead to a kernel
heap memory exposure and over-read.

This fixes tpm1_get_random() to mask positive TPM errors into -EIO, as
before.

[ 18.092103] tpm tpm0: A TPM error (379) occurred attempting get random
[ 18.092106] usercopy: Kernel memory exposure attempt detected from SLUB object 'kmalloc-64' (offset 0, size 379)!

Link: https://bugzilla.redhat.com/show_bug.cgi?id=1650989
Reported-by: Phil Baker <[email protected]>
Reported-by: Craig Robson <[email protected]>
Fixes: 7aee9c52d7ac ("tpm: tpm1: rewrite tpm1_get_random() using tpm_buf structure")
Cc: Laura Abbott <[email protected]>
Cc: Tomas Winkler <[email protected]>
Cc: Jarkko Sakkinen <[email protected]>
Cc: [email protected]
Signed-off-by: Kees Cook <[email protected]>
---
v3: fix never-succeed, limit checks to tpm cmd return (James, Jason)
v2: also fix tpm2 implementation (Jason Gunthorpe)
---
drivers/char/tpm/tpm1-cmd.c | 7 +++++--
drivers/char/tpm/tpm2-cmd.c | 7 +++++--
2 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/char/tpm/tpm1-cmd.c b/drivers/char/tpm/tpm1-cmd.c
index 85dcf2654d11..faacbe1ffa1a 100644
--- a/drivers/char/tpm/tpm1-cmd.c
+++ b/drivers/char/tpm/tpm1-cmd.c
@@ -510,7 +510,7 @@ struct tpm1_get_random_out {
*
* Return:
* * number of bytes read
- * * -errno or a TPM return code otherwise
+ * * -errno (positive TPM return codes are masked to -EIO)
*/
int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
{
@@ -531,8 +531,11 @@ int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max)

rc = tpm_transmit_cmd(chip, &buf, sizeof(out->rng_data_len),
"attempting get random");
- if (rc)
+ if (rc) {
+ if (rc > 0)
+ rc = -EIO;
goto out;
+ }

out = (struct tpm1_get_random_out *)&buf.data[TPM_HEADER_SIZE];

diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c
index e74c5b7b64bf..8ffa6af61580 100644
--- a/drivers/char/tpm/tpm2-cmd.c
+++ b/drivers/char/tpm/tpm2-cmd.c
@@ -301,7 +301,7 @@ struct tpm2_get_random_out {
*
* Return:
* size of the buffer on success,
- * -errno otherwise
+ * -errno otherwise ((positive TPM return codes are masked to -EIO)
*/
int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
{
@@ -328,8 +328,11 @@ int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
offsetof(struct tpm2_get_random_out,
buffer),
"attempting get random");
- if (err)
+ if (err) {
+ if (err > 0)
+ err = -EIO;
goto out;
+ }

out = (struct tpm2_get_random_out *)
&buf.data[TPM_HEADER_SIZE];
--
2.17.1


--
Kees Cook


2019-04-01 19:12:19

by Jason Gunthorpe

[permalink] [raw]
Subject: Re: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

On Mon, Apr 01, 2019 at 12:06:07PM -0700, Kees Cook wrote:
> A "get random" may fail with a TPM error, but those codes were returned
> as-is to the caller, which assumed the result was the number of bytes
> that had been written to the target buffer, which could lead to a kernel
> heap memory exposure and over-read.
>
> This fixes tpm1_get_random() to mask positive TPM errors into -EIO, as
> before.
>
> [ 18.092103] tpm tpm0: A TPM error (379) occurred attempting get random
> [ 18.092106] usercopy: Kernel memory exposure attempt detected from SLUB object 'kmalloc-64' (offset 0, size 379)!
>
> Link: https://bugzilla.redhat.com/show_bug.cgi?id=1650989
> Reported-by: Phil Baker <[email protected]>
> Reported-by: Craig Robson <[email protected]>
> Fixes: 7aee9c52d7ac ("tpm: tpm1: rewrite tpm1_get_random() using tpm_buf structure")
> Cc: Laura Abbott <[email protected]>
> Cc: Tomas Winkler <[email protected]>
> Cc: Jarkko Sakkinen <[email protected]>
> Cc: [email protected]
> Signed-off-by: Kees Cook <[email protected]>
> v3: fix never-succeed, limit checks to tpm cmd return (James, Jason)
> v2: also fix tpm2 implementation (Jason Gunthorpe)
> drivers/char/tpm/tpm1-cmd.c | 7 +++++--
> drivers/char/tpm/tpm2-cmd.c | 7 +++++--
> 2 files changed, 10 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/char/tpm/tpm1-cmd.c b/drivers/char/tpm/tpm1-cmd.c
> index 85dcf2654d11..faacbe1ffa1a 100644
> +++ b/drivers/char/tpm/tpm1-cmd.c
> @@ -510,7 +510,7 @@ struct tpm1_get_random_out {
> *
> * Return:
> * * number of bytes read
> - * * -errno or a TPM return code otherwise
> + * * -errno (positive TPM return codes are masked to -EIO)
> */
> int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
> {
> @@ -531,8 +531,11 @@ int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
>
> rc = tpm_transmit_cmd(chip, &buf, sizeof(out->rng_data_len),
> "attempting get random");
> - if (rc)
> + if (rc) {
> + if (rc > 0)
> + rc = -EIO;
> goto out;
> + }
>
> out = (struct tpm1_get_random_out *)&buf.data[TPM_HEADER_SIZE];
>
> diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c
> index e74c5b7b64bf..8ffa6af61580 100644
> +++ b/drivers/char/tpm/tpm2-cmd.c
> @@ -301,7 +301,7 @@ struct tpm2_get_random_out {
> *
> * Return:
> * size of the buffer on success,
> - * -errno otherwise
> + * -errno otherwise ((positive TPM return codes are masked to -EIO)

Extra bracket, but otherwise looks fine to me

Jason

2019-04-01 19:15:41

by Kees Cook

[permalink] [raw]
Subject: Re: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

On Mon, Apr 1, 2019 at 12:09 PM Jason Gunthorpe <[email protected]> wrote:
> > + * -errno otherwise ((positive TPM return codes are masked to -EIO)
>
> Extra bracket, but otherwise looks fine to me

Thanks! (And I guess it's time for lunch so I can take a break and
learn how to type again...)

--
Kees Cook

2019-04-01 19:42:38

by Winkler, Tomas

[permalink] [raw]
Subject: RE: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

>
> A "get random" may fail with a TPM error, but those codes were returned as-is
> to the caller, which assumed the result was the number of bytes that had been
> written to the target buffer, which could lead to a kernel heap memory
> exposure and over-read.
>
> This fixes tpm1_get_random() to mask positive TPM errors into -EIO, as before.
>
> [ 18.092103] tpm tpm0: A TPM error (379) occurred attempting get random
> [ 18.092106] usercopy: Kernel memory exposure attempt detected from SLUB
> object 'kmalloc-64' (offset 0, size 379)!
>
> Link: https://bugzilla.redhat.com/show_bug.cgi?id=1650989
> Reported-by: Phil Baker <[email protected]>
> Reported-by: Craig Robson <[email protected]>
> Fixes: 7aee9c52d7ac ("tpm: tpm1: rewrite tpm1_get_random() using tpm_buf
> structure")
> Cc: Laura Abbott <[email protected]>
> Cc: Tomas Winkler <[email protected]>
> Cc: Jarkko Sakkinen <[email protected]>
> Cc: [email protected]
> Signed-off-by: Kees Cook <[email protected]>
> ---
> v3: fix never-succeed, limit checks to tpm cmd return (James, Jason)
> v2: also fix tpm2 implementation (Jason Gunthorpe)
Looks good to me.
Thanks
Tomas

> ---
> drivers/char/tpm/tpm1-cmd.c | 7 +++++-- drivers/char/tpm/tpm2-cmd.c | 7
> +++++--
> 2 files changed, 10 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/char/tpm/tpm1-cmd.c b/drivers/char/tpm/tpm1-cmd.c
> index 85dcf2654d11..faacbe1ffa1a 100644
> --- a/drivers/char/tpm/tpm1-cmd.c
> +++ b/drivers/char/tpm/tpm1-cmd.c
> @@ -510,7 +510,7 @@ struct tpm1_get_random_out {
> *
> * Return:
> * * number of bytes read
> - * * -errno or a TPM return code otherwise
> + * * -errno (positive TPM return codes are masked to -EIO)
> */
> int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max) { @@ -
> 531,8 +531,11 @@ int tpm1_get_random(struct tpm_chip *chip, u8 *dest,
> size_t max)
>
> rc = tpm_transmit_cmd(chip, &buf, sizeof(out->rng_data_len),
> "attempting get random");
> - if (rc)
> + if (rc) {
> + if (rc > 0)
> + rc = -EIO;
> goto out;
> + }
>
> out = (struct tpm1_get_random_out
> *)&buf.data[TPM_HEADER_SIZE];
>
> diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c
> index e74c5b7b64bf..8ffa6af61580 100644
> --- a/drivers/char/tpm/tpm2-cmd.c
> +++ b/drivers/char/tpm/tpm2-cmd.c
> @@ -301,7 +301,7 @@ struct tpm2_get_random_out {
> *
> * Return:
> * size of the buffer on success,
> - * -errno otherwise
> + * -errno otherwise ((positive TPM return codes are masked to -EIO)
> */
> int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max) { @@ -
> 328,8 +328,11 @@ int tpm2_get_random(struct tpm_chip *chip, u8 *dest,
> size_t max)
> offsetof(struct tpm2_get_random_out,
> buffer),
> "attempting get random");
> - if (err)
> + if (err) {
> + if (err > 0)
> + err = -EIO;
> goto out;
> + }
>
> out = (struct tpm2_get_random_out *)
> &buf.data[TPM_HEADER_SIZE];
> --
> 2.17.1
>
>
> --
> Kees Cook

2019-04-01 23:48:27

by Jarkko Sakkinen

[permalink] [raw]
Subject: Re: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

On Mon, Apr 01, 2019 at 12:06:07PM -0700, Kees Cook wrote:
> A "get random" may fail with a TPM error, but those codes were returned
> as-is to the caller, which assumed the result was the number of bytes
> that had been written to the target buffer, which could lead to a kernel
> heap memory exposure and over-read.
>
> This fixes tpm1_get_random() to mask positive TPM errors into -EIO, as
> before.
>
> [ 18.092103] tpm tpm0: A TPM error (379) occurred attempting get random
> [ 18.092106] usercopy: Kernel memory exposure attempt detected from SLUB object 'kmalloc-64' (offset 0, size 379)!
>
> Link: https://bugzilla.redhat.com/show_bug.cgi?id=1650989
> Reported-by: Phil Baker <[email protected]>
> Reported-by: Craig Robson <[email protected]>
> Fixes: 7aee9c52d7ac ("tpm: tpm1: rewrite tpm1_get_random() using tpm_buf structure")
> Cc: Laura Abbott <[email protected]>
> Cc: Tomas Winkler <[email protected]>
> Cc: Jarkko Sakkinen <[email protected]>
> Cc: [email protected]
> Signed-off-by: Kees Cook <[email protected]>
> ---
> v3: fix never-succeed, limit checks to tpm cmd return (James, Jason)
> v2: also fix tpm2 implementation (Jason Gunthorpe)
> ---
> drivers/char/tpm/tpm1-cmd.c | 7 +++++--
> drivers/char/tpm/tpm2-cmd.c | 7 +++++--
> 2 files changed, 10 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/char/tpm/tpm1-cmd.c b/drivers/char/tpm/tpm1-cmd.c
> index 85dcf2654d11..faacbe1ffa1a 100644
> --- a/drivers/char/tpm/tpm1-cmd.c
> +++ b/drivers/char/tpm/tpm1-cmd.c
> @@ -510,7 +510,7 @@ struct tpm1_get_random_out {
> *
> * Return:
> * * number of bytes read
> - * * -errno or a TPM return code otherwise
> + * * -errno (positive TPM return codes are masked to -EIO)
> */
> int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
> {
> @@ -531,8 +531,11 @@ int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
>
> rc = tpm_transmit_cmd(chip, &buf, sizeof(out->rng_data_len),
> "attempting get random");
> - if (rc)
> + if (rc) {
> + if (rc > 0)
> + rc = -EIO;
> goto out;
> + }
>
> out = (struct tpm1_get_random_out *)&buf.data[TPM_HEADER_SIZE];
>
> diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c
> index e74c5b7b64bf..8ffa6af61580 100644
> --- a/drivers/char/tpm/tpm2-cmd.c
> +++ b/drivers/char/tpm/tpm2-cmd.c
> @@ -301,7 +301,7 @@ struct tpm2_get_random_out {
> *
> * Return:
> * size of the buffer on success,
> - * -errno otherwise
> + * -errno otherwise ((positive TPM return codes are masked to -EIO)
> */
> int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
> {
> @@ -328,8 +328,11 @@ int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
> offsetof(struct tpm2_get_random_out,
> buffer),
> "attempting get random");
> - if (err)
> + if (err) {
> + if (err > 0)
> + err = -EIO;
> goto out;
> + }
>
> out = (struct tpm2_get_random_out *)
> &buf.data[TPM_HEADER_SIZE];
> --
> 2.17.1
>
>
> --
> Kees Cook

Reviewed-by: Jarkko Sakkinen <[email protected]>

/Jarkko

2019-04-02 19:15:02

by Jarkko Sakkinen

[permalink] [raw]
Subject: Re: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

On Tue, Apr 02, 2019 at 02:46:25AM +0300, Jarkko Sakkinen wrote:
> On Mon, Apr 01, 2019 at 12:06:07PM -0700, Kees Cook wrote:
> > A "get random" may fail with a TPM error, but those codes were returned
> > as-is to the caller, which assumed the result was the number of bytes
> > that had been written to the target buffer, which could lead to a kernel
> > heap memory exposure and over-read.
> >
> > This fixes tpm1_get_random() to mask positive TPM errors into -EIO, as
> > before.
> >
> > [ 18.092103] tpm tpm0: A TPM error (379) occurred attempting get random
> > [ 18.092106] usercopy: Kernel memory exposure attempt detected from SLUB object 'kmalloc-64' (offset 0, size 379)!
> >
> > Link: https://bugzilla.redhat.com/show_bug.cgi?id=1650989
> > Reported-by: Phil Baker <[email protected]>
> > Reported-by: Craig Robson <[email protected]>
> > Fixes: 7aee9c52d7ac ("tpm: tpm1: rewrite tpm1_get_random() using tpm_buf structure")
> > Cc: Laura Abbott <[email protected]>
> > Cc: Tomas Winkler <[email protected]>
> > Cc: Jarkko Sakkinen <[email protected]>
> > Cc: [email protected]
> > Signed-off-by: Kees Cook <[email protected]>
> > ---
> > v3: fix never-succeed, limit checks to tpm cmd return (James, Jason)
> > v2: also fix tpm2 implementation (Jason Gunthorpe)
> > ---
> > drivers/char/tpm/tpm1-cmd.c | 7 +++++--
> > drivers/char/tpm/tpm2-cmd.c | 7 +++++--
> > 2 files changed, 10 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/char/tpm/tpm1-cmd.c b/drivers/char/tpm/tpm1-cmd.c
> > index 85dcf2654d11..faacbe1ffa1a 100644
> > --- a/drivers/char/tpm/tpm1-cmd.c
> > +++ b/drivers/char/tpm/tpm1-cmd.c
> > @@ -510,7 +510,7 @@ struct tpm1_get_random_out {
> > *
> > * Return:
> > * * number of bytes read
> > - * * -errno or a TPM return code otherwise
> > + * * -errno (positive TPM return codes are masked to -EIO)
> > */
> > int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
> > {
> > @@ -531,8 +531,11 @@ int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
> >
> > rc = tpm_transmit_cmd(chip, &buf, sizeof(out->rng_data_len),
> > "attempting get random");
> > - if (rc)
> > + if (rc) {
> > + if (rc > 0)
> > + rc = -EIO;
> > goto out;
> > + }
> >
> > out = (struct tpm1_get_random_out *)&buf.data[TPM_HEADER_SIZE];
> >
> > diff --git a/drivers/char/tpm/tpm2-cmd.c b/drivers/char/tpm/tpm2-cmd.c
> > index e74c5b7b64bf..8ffa6af61580 100644
> > --- a/drivers/char/tpm/tpm2-cmd.c
> > +++ b/drivers/char/tpm/tpm2-cmd.c
> > @@ -301,7 +301,7 @@ struct tpm2_get_random_out {
> > *
> > * Return:
> > * size of the buffer on success,
> > - * -errno otherwise
> > + * -errno otherwise ((positive TPM return codes are masked to -EIO)
> > */
> > int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
> > {
> > @@ -328,8 +328,11 @@ int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max)
> > offsetof(struct tpm2_get_random_out,
> > buffer),
> > "attempting get random");
> > - if (err)
> > + if (err) {
> > + if (err > 0)
> > + err = -EIO;
> > goto out;
> > + }
> >
> > out = (struct tpm2_get_random_out *)
> > &buf.data[TPM_HEADER_SIZE];
> > --
> > 2.17.1
> >
> >
> > --
> > Kees Cook
>
> Reviewed-by: Jarkko Sakkinen <[email protected]>

Applied to my master branch. Jason, Tomas, do you want me to add
reviewed-by's?

/Jarkko

2019-04-02 19:49:26

by Winkler, Tomas

[permalink] [raw]
Subject: RE: [PATCH v3] tpm: Actually fail on TPM errors during "get random"



> On Tue, Apr 02, 2019 at 02:46:25AM +0300, Jarkko Sakkinen wrote:
> > On Mon, Apr 01, 2019 at 12:06:07PM -0700, Kees Cook wrote:
> > > A "get random" may fail with a TPM error, but those codes were
> > > returned as-is to the caller, which assumed the result was the
> > > number of bytes that had been written to the target buffer, which
> > > could lead to a kernel heap memory exposure and over-read.
> > >
> > > This fixes tpm1_get_random() to mask positive TPM errors into -EIO,
> > > as before.
> > >
> > > [ 18.092103] tpm tpm0: A TPM error (379) occurred attempting get
> random
> > > [ 18.092106] usercopy: Kernel memory exposure attempt detected from
> SLUB object 'kmalloc-64' (offset 0, size 379)!
> > >
> > > Link: https://bugzilla.redhat.com/show_bug.cgi?id=1650989
> > > Reported-by: Phil Baker <[email protected]>
> > > Reported-by: Craig Robson <[email protected]>
> > > Fixes: 7aee9c52d7ac ("tpm: tpm1: rewrite tpm1_get_random() using
> > > tpm_buf structure")
> > > Cc: Laura Abbott <[email protected]>
> > > Cc: Tomas Winkler <[email protected]>
> > > Cc: Jarkko Sakkinen <[email protected]>
> > > Cc: [email protected]
> > > Signed-off-by: Kees Cook <[email protected]>
> > > ---
> > > v3: fix never-succeed, limit checks to tpm cmd return (James, Jason)
> > > v2: also fix tpm2 implementation (Jason Gunthorpe)
> > > ---
> > > drivers/char/tpm/tpm1-cmd.c | 7 +++++--
> > > drivers/char/tpm/tpm2-cmd.c | 7 +++++--
> > > 2 files changed, 10 insertions(+), 4 deletions(-)
> > >
> > > diff --git a/drivers/char/tpm/tpm1-cmd.c
> > > b/drivers/char/tpm/tpm1-cmd.c index 85dcf2654d11..faacbe1ffa1a
> > > 100644
> > > --- a/drivers/char/tpm/tpm1-cmd.c
> > > +++ b/drivers/char/tpm/tpm1-cmd.c
> > > @@ -510,7 +510,7 @@ struct tpm1_get_random_out {
> > > *
> > > * Return:
> > > * * number of bytes read
> > > - * * -errno or a TPM return code otherwise
> > > + * * -errno (positive TPM return codes are masked to -EIO)
> > > */
> > > int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max) {
> > > @@ -531,8 +531,11 @@ int tpm1_get_random(struct tpm_chip *chip, u8
> > > *dest, size_t max)
> > >
> > > rc = tpm_transmit_cmd(chip, &buf, sizeof(out->rng_data_len),
> > > "attempting get random");
> > > - if (rc)
> > > + if (rc) {
> > > + if (rc > 0)
> > > + rc = -EIO;
> > > goto out;
> > > + }
> > >
> > > out = (struct tpm1_get_random_out
> *)&buf.data[TPM_HEADER_SIZE];
> > >
> > > diff --git a/drivers/char/tpm/tpm2-cmd.c
> > > b/drivers/char/tpm/tpm2-cmd.c index e74c5b7b64bf..8ffa6af61580
> > > 100644
> > > --- a/drivers/char/tpm/tpm2-cmd.c
> > > +++ b/drivers/char/tpm/tpm2-cmd.c
> > > @@ -301,7 +301,7 @@ struct tpm2_get_random_out {
> > > *
> > > * Return:
> > > * size of the buffer on success,
> > > - * -errno otherwise
> > > + * -errno otherwise ((positive TPM return codes are masked to -EIO)
> > > */
> > > int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max) {
> > > @@ -328,8 +328,11 @@ int tpm2_get_random(struct tpm_chip *chip, u8
> *dest, size_t max)
> > > offsetof(struct tpm2_get_random_out,
> > > buffer),
> > > "attempting get random");
> > > - if (err)
> > > + if (err) {
> > > + if (err > 0)
> > > + err = -EIO;
> > > goto out;
> > > + }
> > >
> > > out = (struct tpm2_get_random_out *)
> > > &buf.data[TPM_HEADER_SIZE];
> > > --
> > > 2.17.1
> > >
> > >
> > > --
> > > Kees Cook
> >
> > Reviewed-by: Jarkko Sakkinen <[email protected]>
>
> Applied to my master branch. Jason, Tomas, do you want me to add reviewed-
> by's?
Sure, it fixes my patch.

> /Jarkko

2019-04-03 17:53:13

by Jarkko Sakkinen

[permalink] [raw]
Subject: Re: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

On Tue, Apr 02, 2019 at 07:13:52PM +0000, Winkler, Tomas wrote:
>
>
> > On Tue, Apr 02, 2019 at 02:46:25AM +0300, Jarkko Sakkinen wrote:
> > > On Mon, Apr 01, 2019 at 12:06:07PM -0700, Kees Cook wrote:
> > > > A "get random" may fail with a TPM error, but those codes were
> > > > returned as-is to the caller, which assumed the result was the
> > > > number of bytes that had been written to the target buffer, which
> > > > could lead to a kernel heap memory exposure and over-read.
> > > >
> > > > This fixes tpm1_get_random() to mask positive TPM errors into -EIO,
> > > > as before.
> > > >
> > > > [ 18.092103] tpm tpm0: A TPM error (379) occurred attempting get
> > random
> > > > [ 18.092106] usercopy: Kernel memory exposure attempt detected from
> > SLUB object 'kmalloc-64' (offset 0, size 379)!
> > > >
> > > > Link: https://bugzilla.redhat.com/show_bug.cgi?id=1650989
> > > > Reported-by: Phil Baker <[email protected]>
> > > > Reported-by: Craig Robson <[email protected]>
> > > > Fixes: 7aee9c52d7ac ("tpm: tpm1: rewrite tpm1_get_random() using
> > > > tpm_buf structure")
> > > > Cc: Laura Abbott <[email protected]>
> > > > Cc: Tomas Winkler <[email protected]>
> > > > Cc: Jarkko Sakkinen <[email protected]>
> > > > Cc: [email protected]
> > > > Signed-off-by: Kees Cook <[email protected]>
> > > > ---
> > > > v3: fix never-succeed, limit checks to tpm cmd return (James, Jason)
> > > > v2: also fix tpm2 implementation (Jason Gunthorpe)
> > > > ---
> > > > drivers/char/tpm/tpm1-cmd.c | 7 +++++--
> > > > drivers/char/tpm/tpm2-cmd.c | 7 +++++--
> > > > 2 files changed, 10 insertions(+), 4 deletions(-)
> > > >
> > > > diff --git a/drivers/char/tpm/tpm1-cmd.c
> > > > b/drivers/char/tpm/tpm1-cmd.c index 85dcf2654d11..faacbe1ffa1a
> > > > 100644
> > > > --- a/drivers/char/tpm/tpm1-cmd.c
> > > > +++ b/drivers/char/tpm/tpm1-cmd.c
> > > > @@ -510,7 +510,7 @@ struct tpm1_get_random_out {
> > > > *
> > > > * Return:
> > > > * * number of bytes read
> > > > - * * -errno or a TPM return code otherwise
> > > > + * * -errno (positive TPM return codes are masked to -EIO)
> > > > */
> > > > int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max) {
> > > > @@ -531,8 +531,11 @@ int tpm1_get_random(struct tpm_chip *chip, u8
> > > > *dest, size_t max)
> > > >
> > > > rc = tpm_transmit_cmd(chip, &buf, sizeof(out->rng_data_len),
> > > > "attempting get random");
> > > > - if (rc)
> > > > + if (rc) {
> > > > + if (rc > 0)
> > > > + rc = -EIO;
> > > > goto out;
> > > > + }
> > > >
> > > > out = (struct tpm1_get_random_out
> > *)&buf.data[TPM_HEADER_SIZE];
> > > >
> > > > diff --git a/drivers/char/tpm/tpm2-cmd.c
> > > > b/drivers/char/tpm/tpm2-cmd.c index e74c5b7b64bf..8ffa6af61580
> > > > 100644
> > > > --- a/drivers/char/tpm/tpm2-cmd.c
> > > > +++ b/drivers/char/tpm/tpm2-cmd.c
> > > > @@ -301,7 +301,7 @@ struct tpm2_get_random_out {
> > > > *
> > > > * Return:
> > > > * size of the buffer on success,
> > > > - * -errno otherwise
> > > > + * -errno otherwise ((positive TPM return codes are masked to -EIO)
> > > > */
> > > > int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max) {
> > > > @@ -328,8 +328,11 @@ int tpm2_get_random(struct tpm_chip *chip, u8
> > *dest, size_t max)
> > > > offsetof(struct tpm2_get_random_out,
> > > > buffer),
> > > > "attempting get random");
> > > > - if (err)
> > > > + if (err) {
> > > > + if (err > 0)
> > > > + err = -EIO;
> > > > goto out;
> > > > + }
> > > >
> > > > out = (struct tpm2_get_random_out *)
> > > > &buf.data[TPM_HEADER_SIZE];
> > > > --
> > > > 2.17.1
> > > >
> > > >
> > > > --
> > > > Kees Cook
> > >
> > > Reviewed-by: Jarkko Sakkinen <[email protected]>
> >
> > Applied to my master branch. Jason, Tomas, do you want me to add reviewed-
> > by's?
> Sure, it fixes my patch.

Great, I'll add it. Thank you. Just want to be explicit with these
things as I consider them as if I was asking a signature from someone
:-)

/Jarkko

2019-05-28 20:28:08

by Laura Abbott

[permalink] [raw]
Subject: Re: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

On 4/3/19 1:52 PM, Jarkko Sakkinen wrote:
> On Tue, Apr 02, 2019 at 07:13:52PM +0000, Winkler, Tomas wrote:
>>
>>
>>> On Tue, Apr 02, 2019 at 02:46:25AM +0300, Jarkko Sakkinen wrote:
>>>> On Mon, Apr 01, 2019 at 12:06:07PM -0700, Kees Cook wrote:
>>>>> A "get random" may fail with a TPM error, but those codes were
>>>>> returned as-is to the caller, which assumed the result was the
>>>>> number of bytes that had been written to the target buffer, which
>>>>> could lead to a kernel heap memory exposure and over-read.
>>>>>
>>>>> This fixes tpm1_get_random() to mask positive TPM errors into -EIO,
>>>>> as before.
>>>>>
>>>>> [ 18.092103] tpm tpm0: A TPM error (379) occurred attempting get
>>> random
>>>>> [ 18.092106] usercopy: Kernel memory exposure attempt detected from
>>> SLUB object 'kmalloc-64' (offset 0, size 379)!
>>>>>
>>>>> Link: https://bugzilla.redhat.com/show_bug.cgi?id=1650989
>>>>> Reported-by: Phil Baker <[email protected]>
>>>>> Reported-by: Craig Robson <[email protected]>
>>>>> Fixes: 7aee9c52d7ac ("tpm: tpm1: rewrite tpm1_get_random() using
>>>>> tpm_buf structure")
>>>>> Cc: Laura Abbott <[email protected]>
>>>>> Cc: Tomas Winkler <[email protected]>
>>>>> Cc: Jarkko Sakkinen <[email protected]>
>>>>> Cc: [email protected]
>>>>> Signed-off-by: Kees Cook <[email protected]>
>>>>> ---
>>>>> v3: fix never-succeed, limit checks to tpm cmd return (James, Jason)
>>>>> v2: also fix tpm2 implementation (Jason Gunthorpe)
>>>>> ---
>>>>> drivers/char/tpm/tpm1-cmd.c | 7 +++++--
>>>>> drivers/char/tpm/tpm2-cmd.c | 7 +++++--
>>>>> 2 files changed, 10 insertions(+), 4 deletions(-)
>>>>>
>>>>> diff --git a/drivers/char/tpm/tpm1-cmd.c
>>>>> b/drivers/char/tpm/tpm1-cmd.c index 85dcf2654d11..faacbe1ffa1a
>>>>> 100644
>>>>> --- a/drivers/char/tpm/tpm1-cmd.c
>>>>> +++ b/drivers/char/tpm/tpm1-cmd.c
>>>>> @@ -510,7 +510,7 @@ struct tpm1_get_random_out {
>>>>> *
>>>>> * Return:
>>>>> * * number of bytes read
>>>>> - * * -errno or a TPM return code otherwise
>>>>> + * * -errno (positive TPM return codes are masked to -EIO)
>>>>> */
>>>>> int tpm1_get_random(struct tpm_chip *chip, u8 *dest, size_t max) {
>>>>> @@ -531,8 +531,11 @@ int tpm1_get_random(struct tpm_chip *chip, u8
>>>>> *dest, size_t max)
>>>>>
>>>>> rc = tpm_transmit_cmd(chip, &buf, sizeof(out->rng_data_len),
>>>>> "attempting get random");
>>>>> - if (rc)
>>>>> + if (rc) {
>>>>> + if (rc > 0)
>>>>> + rc = -EIO;
>>>>> goto out;
>>>>> + }
>>>>>
>>>>> out = (struct tpm1_get_random_out
>>> *)&buf.data[TPM_HEADER_SIZE];
>>>>>
>>>>> diff --git a/drivers/char/tpm/tpm2-cmd.c
>>>>> b/drivers/char/tpm/tpm2-cmd.c index e74c5b7b64bf..8ffa6af61580
>>>>> 100644
>>>>> --- a/drivers/char/tpm/tpm2-cmd.c
>>>>> +++ b/drivers/char/tpm/tpm2-cmd.c
>>>>> @@ -301,7 +301,7 @@ struct tpm2_get_random_out {
>>>>> *
>>>>> * Return:
>>>>> * size of the buffer on success,
>>>>> - * -errno otherwise
>>>>> + * -errno otherwise ((positive TPM return codes are masked to -EIO)
>>>>> */
>>>>> int tpm2_get_random(struct tpm_chip *chip, u8 *dest, size_t max) {
>>>>> @@ -328,8 +328,11 @@ int tpm2_get_random(struct tpm_chip *chip, u8
>>> *dest, size_t max)
>>>>> offsetof(struct tpm2_get_random_out,
>>>>> buffer),
>>>>> "attempting get random");
>>>>> - if (err)
>>>>> + if (err) {
>>>>> + if (err > 0)
>>>>> + err = -EIO;
>>>>> goto out;
>>>>> + }
>>>>>
>>>>> out = (struct tpm2_get_random_out *)
>>>>> &buf.data[TPM_HEADER_SIZE];
>>>>> --
>>>>> 2.17.1
>>>>>
>>>>>
>>>>> --
>>>>> Kees Cook
>>>>
>>>> Reviewed-by: Jarkko Sakkinen <[email protected]>
>>>
>>> Applied to my master branch. Jason, Tomas, do you want me to add reviewed-
>>> by's?
>> Sure, it fixes my patch.
>
> Great, I'll add it. Thank you. Just want to be explicit with these
> things as I consider them as if I was asking a signature from someone
> :-)
>
> /Jarkko
>
Was this intended to go in for 5.2? I still don't see it in the tree.

Thanks,
Laura

2019-05-29 14:40:52

by Jarkko Sakkinen

[permalink] [raw]
Subject: Re: [PATCH v3] tpm: Actually fail on TPM errors during "get random"

On Tue, May 28, 2019 at 03:02:49PM -0400, Laura Abbott wrote:
> > Great, I'll add it. Thank you. Just want to be explicit with these
> > things as I consider them as if I was asking a signature from someone
> > :-)
> >
> > /Jarkko
> >
> Was this intended to go in for 5.2? I still don't see it in the tree.

Was intended but I failed to notice that I should start to send PRs
to Linus instead of security tree and was waiting for security tree
to be rebased. I'll include to the next PR.

/Jarkko