2019-09-26 09:29:02

by Jarkko Sakkinen

[permalink] [raw]
Subject: [PATCH v2 1/2] efi+tpm: Don't access event->count when it isn't mapped.

From: Peter Jones <[email protected]>

Some machines generate a lot of event log entries. When we're
iterating over them, the code removes the old mapping and adds a
new one, so once we cross the page boundary we're unmapping the page
with the count on it. Hilarity ensues.

This patch keeps the info from the header in local variables so we don't
need to access that page again or keep track of if it's mapped.

Fixes: 44038bc514a2 ("tpm: Abstract crypto agile event size calculations")
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Signed-off-by: Peter Jones <[email protected]>
Tested-by: Lyude Paul <[email protected]>
Reviewed-by: Jarkko Sakkinen <[email protected]>
Acked-by: Matthew Garrett <[email protected]>
Acked-by: Ard Biesheuvel <[email protected]>
Signed-off-by: Jarkko Sakkinen <[email protected]>
---
include/linux/tpm_eventlog.h | 14 +++++++++++---
1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/include/linux/tpm_eventlog.h b/include/linux/tpm_eventlog.h
index 63238c84dc0b..12584b69a3f3 100644
--- a/include/linux/tpm_eventlog.h
+++ b/include/linux/tpm_eventlog.h
@@ -170,6 +170,7 @@ static inline int __calc_tpm2_event_size(struct tcg_pcr_event2_head *event,
u16 halg;
int i;
int j;
+ u32 count, event_type;

marker = event;
marker_start = marker;
@@ -190,16 +191,22 @@ static inline int __calc_tpm2_event_size(struct tcg_pcr_event2_head *event,
}

event = (struct tcg_pcr_event2_head *)mapping;
+ /*
+ * the loop below will unmap these fields if the log is larger than
+ * one page, so save them here for reference.
+ */
+ count = READ_ONCE(event->count);
+ event_type = READ_ONCE(event->event_type);

efispecid = (struct tcg_efi_specid_event_head *)event_header->event;

/* Check if event is malformed. */
- if (event->count > efispecid->num_algs) {
+ if (count > efispecid->num_algs) {
size = 0;
goto out;
}

- for (i = 0; i < event->count; i++) {
+ for (i = 0; i < count; i++) {
halg_size = sizeof(event->digests[i].alg_id);

/* Map the digest's algorithm identifier */
@@ -256,8 +263,9 @@ static inline int __calc_tpm2_event_size(struct tcg_pcr_event2_head *event,
+ event_field->event_size;
size = marker - marker_start;

- if ((event->event_type == 0) && (event_field->event_size == 0))
+ if (event_type == 0 && event_field->event_size == 0)
size = 0;
+
out:
if (do_mapping)
TPM_MEMUNMAP(mapping, mapping_size);
--
2.20.1


2019-09-26 09:29:24

by Ard Biesheuvel

[permalink] [raw]
Subject: Re: [PATCH v2 1/2] efi+tpm: Don't access event->count when it isn't mapped.

On Wed, 25 Sep 2019 at 12:16, Jarkko Sakkinen
<[email protected]> wrote:
>
> From: Peter Jones <[email protected]>
>
> Some machines generate a lot of event log entries. When we're
> iterating over them, the code removes the old mapping and adds a
> new one, so once we cross the page boundary we're unmapping the page
> with the count on it. Hilarity ensues.
>
> This patch keeps the info from the header in local variables so we don't
> need to access that page again or keep track of if it's mapped.
>
> Fixes: 44038bc514a2 ("tpm: Abstract crypto agile event size calculations")
> Cc: [email protected]
> Cc: [email protected]
> Cc: [email protected]
> Signed-off-by: Peter Jones <[email protected]>
> Tested-by: Lyude Paul <[email protected]>
> Reviewed-by: Jarkko Sakkinen <[email protected]>
> Acked-by: Matthew Garrett <[email protected]>
> Acked-by: Ard Biesheuvel <[email protected]>
> Signed-off-by: Jarkko Sakkinen <[email protected]>

Thanks Jarkko.

Shall I take these through the EFI tree?


> ---
> include/linux/tpm_eventlog.h | 14 +++++++++++---
> 1 file changed, 11 insertions(+), 3 deletions(-)
>
> diff --git a/include/linux/tpm_eventlog.h b/include/linux/tpm_eventlog.h
> index 63238c84dc0b..12584b69a3f3 100644
> --- a/include/linux/tpm_eventlog.h
> +++ b/include/linux/tpm_eventlog.h
> @@ -170,6 +170,7 @@ static inline int __calc_tpm2_event_size(struct tcg_pcr_event2_head *event,
> u16 halg;
> int i;
> int j;
> + u32 count, event_type;
>
> marker = event;
> marker_start = marker;
> @@ -190,16 +191,22 @@ static inline int __calc_tpm2_event_size(struct tcg_pcr_event2_head *event,
> }
>
> event = (struct tcg_pcr_event2_head *)mapping;
> + /*
> + * the loop below will unmap these fields if the log is larger than
> + * one page, so save them here for reference.
> + */
> + count = READ_ONCE(event->count);
> + event_type = READ_ONCE(event->event_type);
>
> efispecid = (struct tcg_efi_specid_event_head *)event_header->event;
>
> /* Check if event is malformed. */
> - if (event->count > efispecid->num_algs) {
> + if (count > efispecid->num_algs) {
> size = 0;
> goto out;
> }
>
> - for (i = 0; i < event->count; i++) {
> + for (i = 0; i < count; i++) {
> halg_size = sizeof(event->digests[i].alg_id);
>
> /* Map the digest's algorithm identifier */
> @@ -256,8 +263,9 @@ static inline int __calc_tpm2_event_size(struct tcg_pcr_event2_head *event,
> + event_field->event_size;
> size = marker - marker_start;
>
> - if ((event->event_type == 0) && (event_field->event_size == 0))
> + if (event_type == 0 && event_field->event_size == 0)
> size = 0;
> +
> out:
> if (do_mapping)
> TPM_MEMUNMAP(mapping, mapping_size);
> --
> 2.20.1
>

2019-09-26 09:38:28

by Jerry Snitselaar

[permalink] [raw]
Subject: Re: [PATCH v2 1/2] efi+tpm: Don't access event->count when it isn't mapped.

On Wed Sep 25 19, Jarkko Sakkinen wrote:
>On Wed, Sep 25, 2019 at 12:25:05PM +0200, Ard Biesheuvel wrote:
>> On Wed, 25 Sep 2019 at 12:16, Jarkko Sakkinen
>> <[email protected]> wrote:
>> >
>> > From: Peter Jones <[email protected]>
>> >
>> > Some machines generate a lot of event log entries. When we're
>> > iterating over them, the code removes the old mapping and adds a
>> > new one, so once we cross the page boundary we're unmapping the page
>> > with the count on it. Hilarity ensues.
>> >
>> > This patch keeps the info from the header in local variables so we don't
>> > need to access that page again or keep track of if it's mapped.
>> >
>> > Fixes: 44038bc514a2 ("tpm: Abstract crypto agile event size calculations")
>> > Cc: [email protected]
>> > Cc: [email protected]
>> > Cc: [email protected]
>> > Signed-off-by: Peter Jones <[email protected]>
>> > Tested-by: Lyude Paul <[email protected]>
>> > Reviewed-by: Jarkko Sakkinen <[email protected]>
>> > Acked-by: Matthew Garrett <[email protected]>
>> > Acked-by: Ard Biesheuvel <[email protected]>
>> > Signed-off-by: Jarkko Sakkinen <[email protected]>
>>
>> Thanks Jarkko.
>>
>> Shall I take these through the EFI tree?
>
>Would be great, if you could because I already sent one PR with fixes for
>v5.4-rc1 yesterday.
>
>/Jarkko

My patch collides with this, so I will submit a v3 that applies on top of
these once I've run a test with all 3 applied on this t480s.

2019-09-26 09:45:18

by Jerry Snitselaar

[permalink] [raw]
Subject: Re: [PATCH v2 1/2] efi+tpm: Don't access event->count when it isn't mapped.

On Wed Sep 25 19, Jerry Snitselaar wrote:
>On Wed Sep 25 19, Jarkko Sakkinen wrote:
>>On Wed, Sep 25, 2019 at 12:25:05PM +0200, Ard Biesheuvel wrote:
>>>On Wed, 25 Sep 2019 at 12:16, Jarkko Sakkinen
>>><[email protected]> wrote:
>>>>
>>>> From: Peter Jones <[email protected]>
>>>>
>>>> Some machines generate a lot of event log entries. When we're
>>>> iterating over them, the code removes the old mapping and adds a
>>>> new one, so once we cross the page boundary we're unmapping the page
>>>> with the count on it. Hilarity ensues.
>>>>
>>>> This patch keeps the info from the header in local variables so we don't
>>>> need to access that page again or keep track of if it's mapped.
>>>>
>>>> Fixes: 44038bc514a2 ("tpm: Abstract crypto agile event size calculations")
>>>> Cc: [email protected]
>>>> Cc: [email protected]
>>>> Cc: [email protected]
>>>> Signed-off-by: Peter Jones <[email protected]>
>>>> Tested-by: Lyude Paul <[email protected]>
>>>> Reviewed-by: Jarkko Sakkinen <[email protected]>
>>>> Acked-by: Matthew Garrett <[email protected]>
>>>> Acked-by: Ard Biesheuvel <[email protected]>
>>>> Signed-off-by: Jarkko Sakkinen <[email protected]>
>>>
>>>Thanks Jarkko.
>>>
>>>Shall I take these through the EFI tree?
>>
>>Would be great, if you could because I already sent one PR with fixes for
>>v5.4-rc1 yesterday.
>>
>>/Jarkko
>
>My patch collides with this, so I will submit a v3 that applies on top of
>these once I've run a test with all 3 applied on this t480s.

Tested with Peter's patches, and that was the root cause on this 480s.

I think there should still be a check for tbl_size to make sure we
aren't sticking -1 into efi_tpm_final_log_size though, which will be
the case right now if it fails to parse an event.

2019-09-26 10:56:28

by Jarkko Sakkinen

[permalink] [raw]
Subject: Re: [PATCH v2 1/2] efi+tpm: Don't access event->count when it isn't mapped.

On Wed, Sep 25, 2019 at 12:25:05PM +0200, Ard Biesheuvel wrote:
> On Wed, 25 Sep 2019 at 12:16, Jarkko Sakkinen
> <[email protected]> wrote:
> >
> > From: Peter Jones <[email protected]>
> >
> > Some machines generate a lot of event log entries. When we're
> > iterating over them, the code removes the old mapping and adds a
> > new one, so once we cross the page boundary we're unmapping the page
> > with the count on it. Hilarity ensues.
> >
> > This patch keeps the info from the header in local variables so we don't
> > need to access that page again or keep track of if it's mapped.
> >
> > Fixes: 44038bc514a2 ("tpm: Abstract crypto agile event size calculations")
> > Cc: [email protected]
> > Cc: [email protected]
> > Cc: [email protected]
> > Signed-off-by: Peter Jones <[email protected]>
> > Tested-by: Lyude Paul <[email protected]>
> > Reviewed-by: Jarkko Sakkinen <[email protected]>
> > Acked-by: Matthew Garrett <[email protected]>
> > Acked-by: Ard Biesheuvel <[email protected]>
> > Signed-off-by: Jarkko Sakkinen <[email protected]>
>
> Thanks Jarkko.
>
> Shall I take these through the EFI tree?

Would be great, if you could because I already sent one PR with fixes for
v5.4-rc1 yesterday.

/Jarkko

2019-09-27 15:31:16

by Jarkko Sakkinen

[permalink] [raw]
Subject: Re: [PATCH v2 1/2] efi+tpm: Don't access event->count when it isn't mapped.

On Wed, Sep 25, 2019 at 08:16:16AM -0700, Jerry Snitselaar wrote:
> On Wed Sep 25 19, Jarkko Sakkinen wrote:
> > On Wed, Sep 25, 2019 at 12:25:05PM +0200, Ard Biesheuvel wrote:
> > > On Wed, 25 Sep 2019 at 12:16, Jarkko Sakkinen
> > > <[email protected]> wrote:
> > > >
> > > > From: Peter Jones <[email protected]>
> > > >
> > > > Some machines generate a lot of event log entries. When we're
> > > > iterating over them, the code removes the old mapping and adds a
> > > > new one, so once we cross the page boundary we're unmapping the page
> > > > with the count on it. Hilarity ensues.
> > > >
> > > > This patch keeps the info from the header in local variables so we don't
> > > > need to access that page again or keep track of if it's mapped.
> > > >
> > > > Fixes: 44038bc514a2 ("tpm: Abstract crypto agile event size calculations")
> > > > Cc: [email protected]
> > > > Cc: [email protected]
> > > > Cc: [email protected]
> > > > Signed-off-by: Peter Jones <[email protected]>
> > > > Tested-by: Lyude Paul <[email protected]>
> > > > Reviewed-by: Jarkko Sakkinen <[email protected]>
> > > > Acked-by: Matthew Garrett <[email protected]>
> > > > Acked-by: Ard Biesheuvel <[email protected]>
> > > > Signed-off-by: Jarkko Sakkinen <[email protected]>
> > >
> > > Thanks Jarkko.
> > >
> > > Shall I take these through the EFI tree?
> >
> > Would be great, if you could because I already sent one PR with fixes for
> > v5.4-rc1 yesterday.
> >
> > /Jarkko
>
> My patch collides with this, so I will submit a v3 that applies on top of
> these once I've run a test with all 3 applied on this t480s.

Great, thanks.

/Jarkko

2019-09-27 15:32:31

by Jarkko Sakkinen

[permalink] [raw]
Subject: Re: [PATCH v2 1/2] efi+tpm: Don't access event->count when it isn't mapped.

On Wed, Sep 25, 2019 at 09:41:33AM -0700, Jerry Snitselaar wrote:
> On Wed Sep 25 19, Jerry Snitselaar wrote:
> > On Wed Sep 25 19, Jarkko Sakkinen wrote:
> > > On Wed, Sep 25, 2019 at 12:25:05PM +0200, Ard Biesheuvel wrote:
> > > > On Wed, 25 Sep 2019 at 12:16, Jarkko Sakkinen
> > > > <[email protected]> wrote:
> > > > >
> > > > > From: Peter Jones <[email protected]>
> > > > >
> > > > > Some machines generate a lot of event log entries. When we're
> > > > > iterating over them, the code removes the old mapping and adds a
> > > > > new one, so once we cross the page boundary we're unmapping the page
> > > > > with the count on it. Hilarity ensues.
> > > > >
> > > > > This patch keeps the info from the header in local variables so we don't
> > > > > need to access that page again or keep track of if it's mapped.
> > > > >
> > > > > Fixes: 44038bc514a2 ("tpm: Abstract crypto agile event size calculations")
> > > > > Cc: [email protected]
> > > > > Cc: [email protected]
> > > > > Cc: [email protected]
> > > > > Signed-off-by: Peter Jones <[email protected]>
> > > > > Tested-by: Lyude Paul <[email protected]>
> > > > > Reviewed-by: Jarkko Sakkinen <[email protected]>
> > > > > Acked-by: Matthew Garrett <[email protected]>
> > > > > Acked-by: Ard Biesheuvel <[email protected]>
> > > > > Signed-off-by: Jarkko Sakkinen <[email protected]>
> > > >
> > > > Thanks Jarkko.
> > > >
> > > > Shall I take these through the EFI tree?
> > >
> > > Would be great, if you could because I already sent one PR with fixes for
> > > v5.4-rc1 yesterday.
> > >
> > > /Jarkko
> >
> > My patch collides with this, so I will submit a v3 that applies on top of
> > these once I've run a test with all 3 applied on this t480s.
>
> Tested with Peter's patches, and that was the root cause on this 480s.
>
> I think there should still be a check for tbl_size to make sure we
> aren't sticking -1 into efi_tpm_final_log_size though, which will be
> the case right now if it fails to parse an event.

You could sent a follow up patch for that I think. The current
ones are kind of already "went through the process" and do right
things but I do agree that a sanity check would make sense just
in case.

/Jarkko