Re: [PATCH] SecurityPkg: fix sha256 signature check

["Long, Qin" <qin.long@intel.com>]

> -----Original Message-----
> From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org] On Behalf Of
> Laszlo Ersek
> Sent: Thursday, May 10, 2018 8:36 PM
> To: James Bottomley <James.Bottomley@HansenPartnership.com>; edk2-
> devel@lists.01.org
> Cc: Zhang Lubo <lubo.zhang@intel.com>
> Subject: Re: [edk2] [PATCH] SecurityPkg: fix sha256 signature check
> 
> On 05/10/18 00:09, James Bottomley wrote:
> > commit c035e37335ae43229d7e68de74a65f2c01ebc0af
> > Author: Zhang Lubo <lubo.zhang@intel.com>
> > Date:   Thu Jan 5 14:58:05 2017 +0800
> >
> >     SecurityPkg: enhance secure boot Config Dxe & Time Based AuthVariable.
> >
> > Added a check for sha256 being the ownly allowed signature hash.
> > Unfortuantely this commit assumed the form of the signature data was a
> > raw SignedData sequence.  Most tools actually generate a ContentInfo
> > sequence instead which contains a header identifying the content as
> > pkcs7-SignedData.  Fix this check to allow either format to work.
> >
> > This fix is needed at least for efitools because we generate signed
> > variable updates with the ContentInfo header.
> >
> > Signed-off-by: James Bottomley
> <James.Bottomley@HansenPartnership.com>
> > ---
> >  CryptoPkg/Library/OpensslLib/openssl              |  2 +-
> >  SecurityPkg/Library/AuthVariableLib/AuthService.c | 11 ++++++++++-
> >  2 files changed, 11 insertions(+), 2 deletions(-)
> >
> > diff --git a/CryptoPkg/Library/OpensslLib/openssl
> > b/CryptoPkg/Library/OpensslLib/openssl
> > index b2758a2292..d4e4bd2a81 160000
> > --- a/CryptoPkg/Library/OpensslLib/openssl
> > +++ b/CryptoPkg/Library/OpensslLib/openssl
> > @@ -1 +1 @@
> > -Subproject commit b2758a2292aceda93e9f44c219b94fe21bb9a650
> > +Subproject commit d4e4bd2a8163f355fa8a3884077eaec7adc75ff7
> 
> This hunk should not be necessary; please see edk2 commit b85b20fba42e
> ("CryptoPkg/OpensslLib: Update OpenSSL version to 1.1.0h", 2018-04-15).
> 
> (I'll let the SecurityPkg maintainers review the rest.)
> 
> Thanks,
> Laszlo

[Long, Qin] I think so. 
OpenSSL submodule was already upgraded to 1.1.0h (d4e4bd2a8...)

> 
> > diff --git a/SecurityPkg/Library/AuthVariableLib/AuthService.c
> > b/SecurityPkg/Library/AuthVariableLib/AuthService.c
> > index 213a524f27..855ea3350a 100644
> > --- a/SecurityPkg/Library/AuthVariableLib/AuthService.c
> > +++ b/SecurityPkg/Library/AuthVariableLib/AuthService.c
> > @@ -1908,10 +1908,19 @@ VerifyTimeBasedPayload (
> >    //    in VARIABLE_AUTHENTICATION_2 descriptor.
> >    //    This field has the fixed offset (+13) and be calculated based on two
> bytes of length encoding.
> >    //
> > +  // However the data may also begin
> > +  //   ContentInfo ::= SEQUENCE {
> > +  //	   contentType ContentType,
> > +  //	   content
> > +  //	     [0] EXPLICIT ANY DEFINED BY contentType OPTIONAL }
> > +  //
> > +  // In which case the fixed offset is +32  //
> >    if ((Attributes &
> EFI_VARIABLE_TIME_BASED_AUTHENTICATED_WRITE_ACCESS) != 0) {
> >      if (SigDataSize >= (13 + sizeof (mSha256OidValue))) {
> >        if (((*(SigData + 1) & TWO_BYTE_ENCODE) != TWO_BYTE_ENCODE) ||
> > -           (CompareMem (SigData + 13, &mSha256OidValue, sizeof
> (mSha256OidValue)) != 0)) {
> > +	  (CompareMem (SigData + 13, &mSha256OidValue, sizeof
> (mSha256OidValue)) != 0 &&
> > +	   CompareMem (SigData + 32, &mSha256OidValue, sizeof
> > +(mSha256OidValue)) != 0)) {
> >            return EFI_SECURITY_VIOLATION;
> >          }
> >      }
> >
[Long, Qin]  This part looks good to me. 
	I prefer to add this to make both formats (with or without contentType) to work.

> 
> _______________________________________________
> edk2-devel mailing list
> edk2-devel@lists.01.org
> https://lists.01.org/mailman/listinfo/edk2-devel