[PATCH for-4.22 v2] acpi: Set TPM2 LAML to actual log area size

[PATCH for-4.22 v2] acpi: Set TPM2 LAML to actual log area size

[Tu Dinh]

The LAML field should follow the TCG PFP specification:

The TCG ACPI specification uses the field name "Log Area Minimum
Length", but the field value is the actual log area length reserved by
Platform Firmware, not a lower bound.

Remove the now-unused TPM_LOG_AREA_MINIMUM_SIZE.

No functional change.

Signed-off-by: Tu Dinh <ngoc-tu.dinh@vates.tech>
---
v2: Remove the now-unused TPM_LOG_AREA_MINIMUM_SIZE.
---
 tools/libacpi/acpi2_0.h | 1 -
 tools/libacpi/build.c   | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)

diff --git a/tools/libacpi/acpi2_0.h b/tools/libacpi/acpi2_0.h
index 6dfa939a8c..51623e2a8a 100644
--- a/tools/libacpi/acpi2_0.h
+++ b/tools/libacpi/acpi2_0.h
@@ -140,7 +140,6 @@ struct acpi_20_tpm2 {

 #define TPM_LOG_AREA_ADDRESS        0xFED50000

-#define TPM_LOG_AREA_MINIMUM_SIZE   (64 << 10)
 #define TPM_LOG_SIZE                (64 << 10)

 /*
diff --git a/tools/libacpi/build.c b/tools/libacpi/build.c
index 2f29863db1..95188e217e 100644
--- a/tools/libacpi/build.c
+++ b/tools/libacpi/build.c
@@ -473,7 +473,7 @@ static int construct_secondary_tables(struct acpi_ctxt *ctxt,
             tpm2->platform_class = TPM2_ACPI_CLASS_CLIENT;
             tpm2->control_area_address = TPM_CRB_CTRL_REQ;
             tpm2->start_method = TPM2_START_METHOD_CRB;
-            tpm2->log_area_minimum_length = TPM_LOG_AREA_MINIMUM_SIZE;
+            tpm2->log_area_minimum_length = TPM_LOG_SIZE;
             tpm2->log_area_start_address = TPM_LOG_AREA_ADDRESS;

             set_checksum(tpm2,
--
2.43.0


--
Ngoc Tu Dinh | Vates XCP-ng Developer

XCP-ng & Xen Orchestra - Vates solutions

web: https://vates.tech

Re: [PATCH for-4.22 v2] acpi: Set TPM2 LAML to actual log area size

[Jan Beulich]

On 05.11.2025 01:03, Tu Dinh wrote:
> The LAML field should follow the TCG PFP specification:
> 
> The TCG ACPI specification uses the field name "Log Area Minimum
> Length", but the field value is the actual log area length reserved by
> Platform Firmware, not a lower bound.

And this is said where? Even the field description (ACPI 6.5) reads "Identifies
the minimum length (in bytes) of the system’s preboot CC event log area." (I
agree it makes little sense that way, but the change still needs proper
explanation.)

Jan

Re: [PATCH for-4.22 v2] acpi: Set TPM2 LAML to actual log area size

[Tu Dinh]

On 06/11/2025 13:10, Jan Beulich wrote:
> On 05.11.2025 01:03, Tu Dinh wrote:
>> The LAML field should follow the TCG PFP specification:
>>
>> The TCG ACPI specification uses the field name "Log Area Minimum
>> Length", but the field value is the actual log area length reserved by
>> Platform Firmware, not a lower bound.
>
> And this is said where? Even the field description (ACPI 6.5) reads "Identifies
> the minimum length (in bytes) of the system’s preboot CC event log area." (I
> agree it makes little sense that way, but the change still needs proper
> explanation.)
>
> Jan

The above is an exact quote from the "TCG PC Client Platform Firmware
Profile Specification Version 1.06 Revision 52".


--
Ngoc Tu Dinh | Vates XCP-ng Developer

XCP-ng & Xen Orchestra - Vates solutions

web: https://vates.tech

Re: [PATCH for-4.22 v2] acpi: Set TPM2 LAML to actual log area size

[Jan Beulich]

On 06.11.2025 13:13, Tu Dinh wrote:
> On 06/11/2025 13:10, Jan Beulich wrote:
>> On 05.11.2025 01:03, Tu Dinh wrote:
>>> The LAML field should follow the TCG PFP specification:
>>>
>>> The TCG ACPI specification uses the field name "Log Area Minimum
>>> Length", but the field value is the actual log area length reserved by
>>> Platform Firmware, not a lower bound.
>>
>> And this is said where? Even the field description (ACPI 6.5) reads "Identifies
>> the minimum length (in bytes) of the system’s preboot CC event log area." (I
>> agree it makes little sense that way, but the change still needs proper
>> explanation.)
> 
> The above is an exact quote from the "TCG PC Client Platform Firmware 
> Profile Specification Version 1.06 Revision 52".

Interesting. It didn't become clear to me that this actually is a quotation.
Perhaps:

The LAML field should follow the TCG PC Client Platform Firmware Profile
Specification Version. Version 1.06 Revision 52 says:

'The TCG ACPI specification uses the field name "Log Area Minimum
 Length", but the field value is the actual log area length reserved by
 Platform Firmware, not a lower bound.'

Remove the now-unused TPM_LOG_AREA_MINIMUM_SIZE.

? Then
Acked-by: Jan Beulich <jbeulich@suse.com>
and I could make the adjustment while committing (after 4.21 was branched off,
of course).

Jan

Re: [PATCH for-4.22 v2] acpi: Set TPM2 LAML to actual log area size

[Tu Dinh]

On 06/11/2025 13:19, Jan Beulich wrote:
> On 06.11.2025 13:13, Tu Dinh wrote:
>> On 06/11/2025 13:10, Jan Beulich wrote:
>>> On 05.11.2025 01:03, Tu Dinh wrote:
>>>> The LAML field should follow the TCG PFP specification:
>>>>
>>>> The TCG ACPI specification uses the field name "Log Area Minimum
>>>> Length", but the field value is the actual log area length reserved by
>>>> Platform Firmware, not a lower bound.
>>>
>>> And this is said where? Even the field description (ACPI 6.5) reads "Identifies
>>> the minimum length (in bytes) of the system’s preboot CC event log area." (I
>>> agree it makes little sense that way, but the change still needs proper
>>> explanation.)
>>
>> The above is an exact quote from the "TCG PC Client Platform Firmware
>> Profile Specification Version 1.06 Revision 52".
>
> Interesting. It didn't become clear to me that this actually is a quotation.
> Perhaps:
>
> The LAML field should follow the TCG PC Client Platform Firmware Profile
> Specification Version. Version 1.06 Revision 52 says:
>
> 'The TCG ACPI specification uses the field name "Log Area Minimum
>   Length", but the field value is the actual log area length reserved by
>   Platform Firmware, not a lower bound.'
>
> Remove the now-unused TPM_LOG_AREA_MINIMUM_SIZE.
>
> ? Then
> Acked-by: Jan Beulich <jbeulich@suse.com>
> and I could make the adjustment while committing (after 4.21 was branched off,
> of course).
>
> Jan

That's clearer, thanks.


--
Ngoc Tu Dinh | Vates XCP-ng Developer

XCP-ng & Xen Orchestra - Vates solutions

web: https://vates.tech