[edk2-devel] [PATCH v1 04/12] CryptoPkg: Fix conditionally uninitialized variable

Michael D Kinney michael.d.kinney at intel.com
Thu Nov 24 01:37:41 UTC 2022


Hi Michael,

Comments below.

Mike

> -----Original Message-----
> From: devel at edk2.groups.io <devel at edk2.groups.io> On Behalf Of Michael Kubacki
> Sent: Wednesday, November 9, 2022 9:33 AM
> To: devel at edk2.groups.io
> Cc: Erich McMillan <emcmillan at microsoft.com>; Jiang, Guomin <guomin.jiang at intel.com>; Wang, Jian J
> <jian.j.wang at intel.com>; Yao, Jiewen <jiewen.yao at intel.com>; Michael Kubacki <mikuback at linux.microsoft.com>; Lu, Xiaoyu1
> <xiaoyu1.lu at intel.com>
> Subject: [edk2-devel] [PATCH v1 04/12] CryptoPkg: Fix conditionally uninitialized variable
> 
> From: Michael Kubacki <michael.kubacki at microsoft.com>
> 
> Fixes CodeQL alerts for CWE-457:
> https://cwe.mitre.org/data/definitions/457.html
> 
> Checks the return value from `ASN1_get_object()` to verify values
> set by the function are valid.
> 
> Note that the function returns literal `0x80`:
>     `return (0x80);`
> 
> That is used to check the return value is as the case in other areas
> of the code.
> 
> Cc: Erich McMillan <emcmillan at microsoft.com>
> Cc: Guomin Jiang <guomin.jiang at intel.com>
> Cc: Jian J Wang <jian.j.wang at intel.com>
> Cc: Jiewen Yao <jiewen.yao at intel.com>
> Cc: Michael Kubacki <mikuback at linux.microsoft.com>
> Cc: Xiaoyu Lu <xiaoyu1.lu at intel.com>
> Co-authored-by: Erich McMillan <emcmillan at microsoft.com>
> Signed-off-by: Michael Kubacki <michael.kubacki at microsoft.com>
> ---
>  CryptoPkg/Library/BaseCryptLib/Pk/CryptX509.c | 21 +++++++++++---------
>  1 file changed, 12 insertions(+), 9 deletions(-)
> 
> diff --git a/CryptoPkg/Library/BaseCryptLib/Pk/CryptX509.c b/CryptoPkg/Library/BaseCryptLib/Pk/CryptX509.c
> index 2333157e0d17..f867656e888c 100644
> --- a/CryptoPkg/Library/BaseCryptLib/Pk/CryptX509.c
> +++ b/CryptoPkg/Library/BaseCryptLib/Pk/CryptX509.c
> @@ -807,6 +807,7 @@ X509GetTBSCert (
>    UINT32       Asn1Tag;
>    UINT32       ObjClass;
>    UINTN        Length;
> +  UINTN        Inf;
> 
>    //
>    // Check input parameters.
> @@ -836,9 +837,9 @@ X509GetTBSCert (
>    //
>    Temp   = Cert;
>    Length = 0;
> -  ASN1_get_object (&Temp, (long *)&Length, (int *)&Asn1Tag, (int *)&ObjClass, (long)CertSize);
> +  Inf    = ASN1_get_object (&Temp, (long *)&Length, (int *)&Asn1Tag, (int *)&ObjClass, (long)CertSize);
> 
> -  if (Asn1Tag != V_ASN1_SEQUENCE) {
> +  if (!(Inf & 0x80) && (Asn1Tag != V_ASN1_SEQUENCE)) {

The result of bitwise operation (Inf & 0x80) is a value, not a BOOLEAN.
I think the more correct way to do this check is ((Inf & 0x80) == 0x00).

>      return FALSE;
>    }
> 
> @@ -848,7 +849,7 @@ X509GetTBSCert (
>    //
>    // Verify the parsed TBSCertificate is one correct SEQUENCE data.
>    //
> -  if (Asn1Tag != V_ASN1_SEQUENCE) {
> +  if (!(Inf & 0x80) && (Asn1Tag != V_ASN1_SEQUENCE)) {

The result of bitwise operation (Inf & 0x80) is a value, not a BOOLEAN.
I think the more correct way to do this check is ((Inf & 0x80) == 0x00).

>      return FALSE;
>    }
> 
> @@ -1888,18 +1889,20 @@ Asn1GetTag (
>    IN     UINT32   Tag
>    )
>  {
> -  UINT8  *PtrOld;
> -  INT32  ObjTag;
> -  INT32  ObjCls;
> -  long   ObjLength;
> +  UINT8   *PtrOld;
> +  INT32   ObjTag;
> +  INT32   ObjCls;
> +  long    ObjLength;
> +  UINT32  Inf;
> 
>    //
>    // Save Ptr position
>    //
>    PtrOld = *Ptr;
> 
> -  ASN1_get_object ((CONST UINT8 **)Ptr, &ObjLength, &ObjTag, &ObjCls, (INT32)(End - (*Ptr)));
> -  if ((ObjTag == (INT32)(Tag & CRYPTO_ASN1_TAG_VALUE_MASK)) &&
> +  Inf = ASN1_get_object ((CONST UINT8 **)Ptr, &ObjLength, &ObjTag, &ObjCls, (INT32)(End - (*Ptr)));
> +  if (!(Inf & 0x80) &&

The result of bitwise operation (Inf & 0x80) is a value, not a BOOLEAN.
I think the more correct way to do this check is ((Inf & 0x80) == 0x00).

> +      (ObjTag == (INT32)(Tag & CRYPTO_ASN1_TAG_VALUE_MASK)) &&
>        (ObjCls == (INT32)(Tag & CRYPTO_ASN1_TAG_CLASS_MASK)))
>    {
>      *Length = (UINTN)ObjLength;
> --
> 2.28.0.windows.1
> 
> 
> 
> -=-=-=-=-=-=
> Groups.io Links: You receive all messages sent to this group.
> View/Reply Online (#96150): https://edk2.groups.io/g/devel/message/96150
> Mute This Topic: https://groups.io/mt/94918089/1643496
> Group Owner: devel+owner at edk2.groups.io
> Unsubscribe: https://edk2.groups.io/g/devel/unsub [michael.d.kinney at intel.com]
> -=-=-=-=-=-=
> 



-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#96595): https://edk2.groups.io/g/devel/message/96595
Mute This Topic: https://groups.io/mt/94918089/1813853
Group Owner: devel+owner at edk2.groups.io
Unsubscribe: https://edk2.groups.io/g/devel/unsub [edk2-devel-archive at redhat.com]
-=-=-=-=-=-=-=-=-=-=-=-




More information about the edk2-devel-archive mailing list