Re: Tianocore and TPM2 pcr values

[Jorge Fernandez Monteagudo <jorgefm@cirsa.com>]

Yes, from log I see:


Loading driver at 0x0008F3F2000 EntryPoint=0x0008F3F2240 Tcg2Dxe.efi
InstallProtocolInterface: BC62157E-3E33-4FEC-9920-2D3B36D750DF 8F410C18
ProtectUefiImageCommon - 0x8F4107C0
  - 0x000000008F3F2000 - 0x000000000000D800
PROGRESS CODE: V03040002 I0
InterfaceId - 0xFFFFFFFF
  InterfaceType    - 0x0F
InterfaceCapability - 0x300000FF
  InterfaceVersion - 0x3
StatusEx - 0xFF
  TpmFamily - 0x3
PtpInterface - 0
VID - 0x15D1
DID - 0x001A
RID - 0x10
Tcg2.ProtocolVersion  - 01.01
Tcg2.StructureVersion - 01.01
Tpm2GetCapabilityManufactureID - 00584649
Tpm2GetCapabilityFirmwareVersion - 00050000 00044102
Tpm2GetCapabilityMaxCommandResponseSize - 00000500, 00000500
GetSupportedAndActivePcrs - Count = 00000002
Tcg2.SupportedEventLogs - 0x00000003
Tcg2.HashAlgorithmBitmap - 0x00000003
Tcg2.NumberOfPCRBanks      - 0x00000002
Tcg2.ActivePcrBanks        - 0x00000003
...


________________________________
De: Yao, Jiewen <jiewen.yao@intel.com>
Enviado: miércoles, 26 de septiembre de 2018 8:44:54
Para: Jorge Fernandez Monteagudo; Zhang, Chao B; edk2-devel@lists.01.org
Asunto: RE: Tianocore and TPM2 pcr values


ProtectUefiImageCommon is not related.



Below code is the Tcg2Dxe entrypoint, I expect you can see some message there:



====================================

DriverEntry()

  if (CompareGuid (PcdGetPtr(PcdTpmInstanceGuid), &gEfiTpmDeviceInstanceNoneGuid) ||

      CompareGuid (PcdGetPtr(PcdTpmInstanceGuid), &gEfiTpmDeviceInstanceTpm12Guid)){

    DEBUG ((DEBUG_INFO, "No TPM2 instance required!\n"));

    return EFI_UNSUPPORTED;

  }



  if (GetFirstGuidHob (&gTpmErrorHobGuid) != NULL) {

    DEBUG ((EFI_D_ERROR, "TPM2 error!\n"));

    return EFI_DEVICE_ERROR;

  }



  Status = Tpm2RequestUseTpm ();

  if (EFI_ERROR (Status)) {

    DEBUG ((EFI_D_ERROR, "TPM2 not detected!\n"));

    return Status;

  }



  //

  // Fill information

  //

  ASSERT (TCG_EVENT_LOG_AREA_COUNT_MAX == sizeof(mTcg2EventInfo)/sizeof(mTcg2EventInfo[0]));



  mTcgDxeData.BsCap.Size = sizeof(EFI_TCG2_BOOT_SERVICE_CAPABILITY);

  mTcgDxeData.BsCap.ProtocolVersion.Major = 1;

  mTcgDxeData.BsCap.ProtocolVersion.Minor = 1;

  mTcgDxeData.BsCap.StructureVersion.Major = 1;

  mTcgDxeData.BsCap.StructureVersion.Minor = 1;



  DEBUG ((EFI_D_INFO, "Tcg2.ProtocolVersion  - %02x.%02x\n", mTcgDxeData.BsCap.ProtocolVersion.Major, mTcgDxeData.BsCap.ProtocolVersion.Minor));

  DEBUG ((EFI_D_INFO, "Tcg2.StructureVersion - %02x.%02x\n", mTcgDxeData.BsCap.StructureVersion.Major, mTcgDxeData.BsCap.StructureVersion.Minor));



  Status = Tpm2GetCapabilityManufactureID (&mTcgDxeData.BsCap.ManufacturerID);

  if (EFI_ERROR (Status)) {

    DEBUG ((EFI_D_ERROR, "Tpm2GetCapabilityManufactureID fail!\n"));

  } else {

    DEBUG ((EFI_D_INFO, "Tpm2GetCapabilityManufactureID - %08x\n", mTcgDxeData.BsCap.ManufacturerID));

  }











From: Jorge Fernandez Monteagudo [mailto:jorgefm@cirsa.com]
Sent: Wednesday, September 26, 2018 2:40 PM
To: Yao, Jiewen <jiewen.yao@intel.com>; Zhang, Chao B <chao.b.zhang@intel.com>; edk2-devel@lists.01.org
Subject: Re: Tianocore and TPM2 pcr values



Hi Yao



> Yes, it is always good to enable serial port debug. There are lots of debug message in Tcg2Dxe driver. We can know what is wrong.



>From the log I've been able to see that "measure" messages start once Tcg2Dxe.efi. From the beggining I can only see "ProtectUefiImageCommon"

messages but I don't know if they are related.



>In your patch, since we are using UEFI as payload, and there is no PEI, I am not clear which driver you expect will extend something to PCR0. Do you think coreboot is CRTM? Or the UEFI payload is the CRTM? Who should be >responsible to extend coreboot image from flash, and who should extend UEFI payload?



I think nothing is implemented in coreboot because when TPM2 was not activated in edk2 PCR0-10 were all 0. It's only checking what device

is available and sending the tpm2_startup command. I'll try to investigate the coreboot project to see if the tianocore payload could be extended

before loading because coreboot should be the CRTM.



> Also, only *3rd part* image will change PCR2 and PCR4. Do you have such case in your platform?



First notice. No I don't have such case in my platform.



Thanks!

Jorge

________________________________

De: Yao, Jiewen <jiewen.yao@intel.com<mailto:jiewen.yao@intel.com>>
Enviado: miércoles, 26 de septiembre de 2018 8:11:58
Para: Jorge Fernandez Monteagudo; Zhang, Chao B; edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
Asunto: RE: Tianocore and TPM2 pcr values



Hi Jorge
Yes, it is always good to enable serial port debug. There are lots of debug message in Tcg2Dxe driver. We can know what is wrong.

In pure UEFI BIOS, the PEI driver extends to PCR0, and DXE image measurement lib extend to PCR2, PCR4, PCR5. The DXE driver extends variable to PCR1/7, and exposes the TCG2 protocol to let OS use it.

In your patch, since we are using UEFI as payload, and there is no PEI, I am not clear which driver you expect will extend something to PCR0. Do you think coreboot is CRTM? Or the UEFI payload is the CRTM? Who should be responsible to extend coreboot image from flash, and who should extend UEFI payload?

Also, only *3rd part* image will change PCR2 and PCR4. Do you have such case in your platform?

Anyway, there should still be something measured - boot variable (PCR1), secure boot variable (PCR7), GPT (5), action (4,5), separator (1~7), if you include Tcg2Dxe driver.

I am not clear if coreboot already extends something to separator according to TCG PFP spec. If that is the case, we probably need a special handing in DXE driver.


I look forward to your serial debug message and design discussion.

Thank you
Yao Jiewen


> -----Original Message-----
> From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org] On Behalf Of
> Jorge Fernandez Monteagudo
> Sent: Wednesday, September 26, 2018 1:46 PM
> To: Zhang, Chao B <chao.b.zhang@intel.com<mailto:chao.b.zhang@intel.com>>; edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
> Subject: Re: [edk2] Tianocore and TPM2 pcr values
>
> Hi Chao!
>
>
> Maybe the traces I get from the debug build and
>
>
>   gEfiMdePkgTokenSpaceGuid.PcdReportStatusCodePropertyMask|0x7
>   gEfiMdePkgTokenSpaceGuid.PcdDebugPrintErrorLevel|0x800A044F
>   gEfiMdePkgTokenSpaceGuid.PcdDebugPropertyMask|0x2F
>
> can help.
>
>
> ________________________________
> De: edk2-devel <edk2-devel-bounces@lists.01.org<mailto:edk2-devel-bounces@lists.01.org>> en nombre de Jorge
> Fernandez Monteagudo <jorgefm@cirsa.com<mailto:jorgefm@cirsa.com>>
> Enviado: martes, 25 de septiembre de 2018 16:09:31
> Para: Zhang, Chao B; edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
> Asunto: Re: [edk2] Tianocore and TPM2 pcr values
>
> Hi Chao!
>
>
> PCR0 has not changed in any of the test I've done! What info do you need?
>
>
> I'm using:
>
> coreboot: ae05d095b36ac835a6b1a221e6858065e5486888, master branch
>
> tianocore: 07ecd98ac18d6792181856faca7d4bed1b587261, coreboot
> branch
>
> Attached are the changes I've done to tianocore to get TPM2 support and no
> console.
> PCR0 is always
> 3d458cfe55cc03ea1f443f1562beec8df51c75e14a9fcf9a7234a13f198e7969
>
> Thanks!
> ________________________________
> De: Zhang, Chao B <chao.b.zhang@intel.com<mailto:chao.b.zhang@intel.com>>
> Enviado: martes, 25 de septiembre de 2018 15:41:45
> Para: Jorge Fernandez Monteagudo; edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
> Cc: You, Benjamin
> Asunto: RE: Tianocore and TPM2 pcr values
>
> Hi Jorge:
>     PCR 0 should change if you  use different core boot payload + UEFI. So
> your case seems to be an issue. Can you provide more detailed info?
>
>
> -----Original Message-----
> From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org] On Behalf Of
> Jorge Fernandez Monteagudo
> Sent: Monday, September 24, 2018 5:57 PM
> To: edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
> Subject: [edk2] Tianocore and TPM2 pcr values
>
> Hi all,
>
>
> This is my first message in this list. I'm using tianocore as a payload for a
> Coreboot in order to
>
> boot a custom board I'm working on it. Finally I've been able to enable the
> TPM2 support in
>
> coreboot and in tianocore but I have some questions regarding the values
> I'm seeing in the PCRs.
>
>
> I'm using Tianocore master branch as is selected by coreboot menuconfig
> and x64 architecture.
>
> Once the system is running I can read the PCRs and, if I'm not wrong, PCRs 0
> to 7 are handled
>
> by the Tianocore/Coreboot. I've flashed a coreboot+tianocore in release
> mode and a coreboot+
>
> tianocore in debug mode and the PCRs are the same. Is it ok? I thought that
> any change in the
>
> coreboot.rom will made the PCR values to change...
>
>
> pcr0:
> 3d458cfe55cc03ea1f443f1562beec8df51c75e14a9fcf9a7234a13f198e7969
> pcr1:
> a3a3552caa68c6d9db64bf1ed4dca08080f99b59f1b26debc9abefa59ee8ca28
> pcr2:
> 3d458cfe55cc03ea1f443f1562beec8df51c75e14a9fcf9a7234a13f198e7969
> pcr3:
> 3d458cfe55cc03ea1f443f1562beec8df51c75e14a9fcf9a7234a13f198e7969
> pcr4:
> 74a35102770e65ab94b35135a4bf54c411134ae8059e03df41060a33f573871
> f
> pcr5:
> dfa65561584cb8604b1675c869f3341d0c99c642ce9d91353380361126235ad
> 8
> pcr6:
> 3d458cfe55cc03ea1f443f1562beec8df51c75e14a9fcf9a7234a13f198e7969
> pcr7:
> b5710bf57d25623e4019027da116821fa99f5c81e9e38b87671cc574f9281439
>
> Another test I've done is using the Tianocore stable branch as selected by
> coreboot
> (STABLE_COMMIT_ID=315d9d08fd77db1024ccc5307823da8aaed85e2f) and
> I get the same values from release and build coreboot.roms except that
> PCR1 has the same value as PCR0, 2, 3 and 6, it seems it's not used in this
> version.
>
> Is this the expected behavior?
>
> Thanks!
> Jorge
> _______________________________________________
> edk2-devel mailing list
> edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
> https://lists.01.org/mailman/listinfo/edk2-devel
> _______________________________________________
> edk2-devel mailing list
> edk2-devel@lists.01.org<mailto:edk2-devel@lists.01.org>
> https://lists.01.org/mailman/listinfo/edk2-devel