Jeff Brasen authored Aug 23, 2021 and G Edhaya Chandran committed Aug 23, 2021

SctAPrint uses boot services functions but is called after ExitBootServices. Replace with call to Printf which is safe for use in runtime.
Signed-off-by: Jeff Brasen <jbrasen@nvidia.com>

Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

Heinrich Schuchardt authored Jun 30, 2021 and G Edhaya Chandran committed Aug 09, 2021

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3469



The Self Certification Test (SCT) II Case Specification, 2017 requires
in 5.2.1.4.5. that QueryVariableInfo() shall fail for

    attributes = EFI_VARIABLE_NON_VOLATILE.

Add EFI_VARIABLE_NON_VOLATILE to tested values in function
QueryVariableInfoConfTestSub5().
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>

Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

devel@edk2.groups.io authored Aug 09, 2021 and G Edhaya Chandran committed Aug 09, 2021

This is to fix the SCT BS.AllocatePages failures (not found) with the case that the Start address is not aligned to 64k.
For example,
  The following is available memory region for testing:
    0000000082012000-00000000EB6D9FFF 00000000000696C8
  With the current page alignment calculation, we will get:
    Start address is 0x82020000
    PageNum is 0x696B8
  In BS.AllocatePages, it will make the end address align with 64k,
  so PageNum will be changed from 0x696B8 to 0x696C0. Therefore, the
  end address will become 0xEB6E0000 which is larger than 0xEB6D9FFF,
  so we get not found error in the end.

Therefore, the calculation for getting the PageNum should be updated to PageNum - (2 * EFI_SIZE_TO_PAGES(0x10000)) so that we won't get a wrong PageNum to allocate a memory with a size larger than available space's size.

With this solution, the example above will get 0x696A8 as calculated PageNum. Then, in BS.AllocatePages, the PageNum will be changed from
0x696A8 to 0x696B0. Therefore, the end address will become 0xEB6D0000 that is smaller than 0xEB6D9FFF, so we get not found error in the end.

I also tested this solution on two ARM platforms (NXP1046A and RPi4).

Cc: Samer El-Haj-Mahmoud <samer.el-haj-mahmoud@arm.com>
Cc: G Edhaya Chandran <edhaya.chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Signed-off-by: Sunny Wang <sunny.wang@arm.com>

Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>
Reviewed-by: Tuan Phan <tuanphan@os.amperecomputing.com>

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3246



MdeLibs.dsc.inc was added for some basic/default library
instances provided by MdePkg, RegisterFilterLibNull which will be
consumed by IoLib and BaseLib was added in MdeLibs.dsc.inc.

To build UefiSct with edk2-stable202105 and later, this file
must be included in dsc file.

Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Cc: Arvin Chen <arvinx.chen@intel.com>
Signed-off-by: Barton Gao <gaojie@byosoft.com.cn>
Reviewed-by: Sunny Wang <sunny.wang@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

The commits a9d1fb58 and ae7e5477 caused SCT BS.CreateEvent failures.

Section 7.1 of the UEFI Spec states that TPL_HIGH_LEVEL is designed for
exclusive use by the firmware. The creation of events by UEFI
applications, UEFI drivers, and UEFI OS Loaders should not use this TPL
level.

Therefore, revert TPL_HIGH_LEVEL change in commits a9d1fb58 and
ae7e5477 to not create event with TPL_HIGH_LEVEL to be compliant
with UEFI Spec and fix the failures.

For more information, https://edk2.groups.io/g/devel/message/76338



Cc: Samer El-Haj-Mahmoud <samer.el-haj-mahmoud@arm.com>
Cc: G Edhaya Chandran <edhaya.chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Signed-off-by: Sunny Wang <sunny.wang@arm.com>
Reviewed-By: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

Heinrich Schuchardt authored Mar 10, 2021 and G Edhaya Chandran committed Jun 08, 2021

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3257



SctPrint() requires that %x and %d refers to an UINTN parameter.

SimpleNetworkBBTestFunction.c tries to print INT32 using %x, %d without
converting to UINTN resulting in corrupted output like:

SimpleNetworkBBTestFunction.c:891:
Status - Unsupported, Filter - AAAA00000000

Mode->ReceiveFilterSetting has only 32 bit. The true value is 0.

Convert the parameters to UINTN.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Sunny Wang <sunny.wang@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

Heinrich Schuchardt authored Mar 10, 2021 and G Edhaya Chandran committed May 12, 2021

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3253



SctPrint() requires that %d refers to an UINTN parameter.

SimpleTextOutBBTestFunction_uefi.c has a lot of
StandardLib->RecordAssertion() calls where an INT32 is passed
as argument for a '%d' print code.

This leads to incorrect output like:

MaxMode=-549755813885,

-549755813885 is 0x0xFFFFFF8000000003. So MaxMode actually
is an INT32 with value 3 in this example.

Convert the parameters to UINTN.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-by: Sunny Wang <sunny.wang@arm.com>

Heinrich Schuchardt authored May 12, 2021 and G Edhaya Chandran committed May 12, 2021

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3257



SctPrint() requires that %x and %d refers to an UINTN parameter.

SimpleNetworkBBTestFunction.c tries to print INT32 using %x, %d without converting to UINTN resulting in corrupted output like:

SimpleNetworkBBTestFunction.c:891:
Status - Unsupported, Filter - AAAA00000000

Mode->ReceiveFilterSetting has only 32 bit. The true value is 0.

Convert the parameters to UINTN.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>

Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

Eric Jin authored May 08, 2021 and G Edhaya Chandran committed May 08, 2021

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=3272



The goal of the Support Routine: CheckForSupportGetImage is to detect the
current FMP instance's capability for supporting GetImage().

In current code, "ImageIndex = 0" is the input parameter, but it is
inconsistent with spec -"A unique number identifying the firmware image(s)
within the device. The number is between 1 and DescriptorCount."

It is a bug and need the fix.
The similar errors exist in CheckForSupportSetImage/CheckForSupportCheckImage.

Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Cc: Arvin Chen <arvinx.chen@intel.com>
Signed-off-by: Eric Jin <eric.jin@intel.com>
Reviewed-by: Barton Gao <gaojie@byosoft.com.cn>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

Grant Likely authored May 08, 2021 and G Edhaya Chandran committed May 08, 2021

CapsuleApp.efi is necessary for testing capsule updates of the firmware.
Add it into the default build.

Cc: G Edhaya Chandran <edhaya.chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Signed-off-by: Grant Likely <grant.likely@arm.com>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=3042



Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Cc: Arvin Chen <arvinx.chen@intel.com>
Cc: Leif Lindholm <leif@nuviainc.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Abner Chang <abner.chang@hpe.com>
Signed-off-by: Daniel Schaefer <daniel.schaefer@hpe.com>
Reviewed-by: Abner Chang <abner.chang@hpe.com>
Reviewed-by: Sunny Wang <sunny.wang@arm.com>

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=3042



Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Cc: Arvin Chen <arvinx.chen@intel.com>
Cc: Leif Lindholm <leif@nuviainc.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Abner Chang <abner.chang@hpe.com>
Signed-off-by: Daniel Schaefer <daniel.schaefer@hpe.com>
Reviewed-by: Abner Chang <abner.chang@hpe.com>
Reviewed-by: Sunny Wang <sunny.wang@arm.com>

The sources were copied from Aarch64. Follow-up commit will adapt them
for Riscv64.

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=3042



Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Cc: Arvin Chen <arvinx.chen@intel.com>
Cc: Leif Lindholm <leif@nuviainc.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Abner Chang <abner.chang@hpe.com>
Signed-off-by: Daniel Schaefer <daniel.schaefer@hpe.com>
Reviewed-by: Abner Chang <abner.chang@hpe.com>
Reviewed-by: Sunny Wang <sunny.wang@arm.com>

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3253



SctPrint() requires that %d refers to an UINTN parameter.

SimpleTextOutBBTestFunction_uefi.c has a lot of
StandardLib->RecordAssertion() calls where an INT32 is passed
as argument for a '%d' print code.

This leads to incorrect output like:

MaxMode=-549755813885,

-549755813885 is 0x0xFFFFFF8000000003. So MaxMode actually
is an INT32 with value 3 in this example.

Convert the parameters to UINTN.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3244

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3239



A superfluous carriage return leads to an error:

uefi-sct/SctPkg/build.sh: line 61:
syntax error near unexpected token `fi'
uefi-sct/SctPkg/build.sh: line 61:
`    fi'
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>
Reviewed-by: Barton Gao <gaojie@byosoft.com.cn>

gcc compilers versions 5.* also need to be converted from string to integer before testing

Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Cc: Irene Park <ipark@nvidia.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Samer El-Haj-Mahmoud <samer.el-haj-mahmoud@arm.com>
Cc: Stuart Yoder <Stuart.Yoder@arm.com>
Signed-off-by: Joseph Hemann <Joseph.heman@arm.com>
Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-By: G Edhaya Chandran <edhaya.chandran@arm.com>

Grant Likely authored Feb 22, 2021 and G Edhaya Chandran committed Feb 22, 2021

The buildzip.sh script builds the SCT and the EFI Shell, and then zips
them up in a file structure that can be booted when unzipped to a USB
drive or other block storage.
Signed-off-by: Grant Likely <grant.likely@arm.com>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

Grant Likely authored Feb 22, 2021 and G Edhaya Chandran committed Feb 22, 2021

The script was using a big list of numerical argument to pass on extra
parameters. Use the bash 'shift' command to carve of arguments for the
script so that $@ can be used for the remainder.
Signed-off-by: Grant Likely <grant.likely@arm.com>

Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

Grant Likely authored Feb 22, 2021 and G Edhaya Chandran committed Feb 22, 2021

The build.sh script is very useful for setting up the build environment
before calling the package build. Sometimes additional packages are
needed when building the SCT. (e.g., it is useful to build ShellPkg).
Refactor the build code to allow additional DSCs to be added to the
build.

This patch is useful when building a full SCT image that includes the
EDK2 shell.
Signed-off-by: Grant Likely <grant.likely@arm.com>
Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Cc: Samer El-Haj-Mahmoud <samer.el-haj-mahmoud@arm.com>

Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3057



According to table Table 27. TPL Restrictions in UEFI spec 2.8 B:

Event Notification Levels must be "> TPL_APPLICATION" and "<= TPL_HIGH_LEVEL".

In the conformance test for CreateEvent() check that TPL_APPLICATION
results in EFI_INVALID_PARAMETER if the event can be notified.

In the functional tests check that events can be created with
TPL_HIGH_LEVEL.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3057



According to table Table 27. TPL Restrictions in UEFI spec 2.8 B:

Event Notification Levels must be "> TPL_APPLICATION" and "<= TPL_HIGH_LEVEL".

In the conformance test for CreateEvent() check that TPL_APPLICATION
results in EFI_INVALID_PARAMETER if the event can be notified.

Run the other conformance tests for TPL_HIGH_LEVEL, too.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3146



EFI_UNICODE_COLLATION_PROTOCOL.FatToStr() expects as parameter Fat an ASCII
string not a wide string.

In function BBTestFatToStrFunctionAutoTest() this string is supplied as
TestData[Index].

When printing this string we cannot use "%s" but we must use "%a".
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-By: G Edhaya Chandran <edhaya.chandran@arm.com>

Request to move Arvin Chen to reviewer of uefi-sct

Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Signed-off-by: ArvinX Chen <arvinx.chen@intel.com>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3142



The parameter name ExtendedVerification has repeatedly been misspelled.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-By: G Edhaya Chandran <edhaya.chandran@arm.com>

Change Eric Jin from Maintainer to Reviewer.

Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Cc: Arvin Chen <arvinx.chen@intel.com>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Signed-off-by: Eric Jin <eric.jin@intel.com>
Reviewed-By: Barton Gao <gaojie@byosoft.com.cn>

Header.Length is of type uint8[2], this cause incompatibility with compare with sizeof()

Request to add Arvin Chen from Intel as maintainer of uefi-sct

Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Cc: Barton Gao <gaojie@byosoft.com.cn>
Signed-off-by: ArvinX Chen <arvinx.chen@intel.com>
Reviewed-by: Eric Jin <eric.jin@intel.com>

Heinrich Schuchardt authored Dec 07, 2020 and G Edhaya Chandran committed Dec 07, 2020

The functional tests for OpenEx() use RecordAssertion() statements that lack a print code for the Tpl argument. This leads to a segmentation violation.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Grant Likely <grant.likely@arm.com>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

ArvinX Chen authored Dec 07, 2020 and G Edhaya Chandran committed Dec 07, 2020

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3022



The storage device erase behavior may have two possibilities:
 1.Write all data into "0"
 2.Write all data into "1"
but now tool behavior can only check case 1 (Write all data into "0"),
so we need add the other case into SCT tool to correct the test behavior.

Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Signed-off-by: ArvinX Chen <arvinx.chen@intel.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

devel@edk2.groups.io authored Dec 07, 2020 and G Edhaya Chandran committed Dec 07, 2020

Now, in our SCT test case "CheckImage" behavior have some problems.Once
tool need to check the "EFI_SECURITY_VIOLATION" have correct return from
"EFI_FIRMWARE_MANAGEMENT_PROTOCOL->CheckImage" function, the function wil
because tool behavior probability return EFI_BUFFER_TOO_SMALL,so we should
gave it a correct header info to make the test item can process correctly.

Cc: ArvinX Chen <arvinx.chen@intel.com>
Cc: G Edhaya Chandran <Edhaya.Chandran@arm.com>
Cc: Eric Jin <eric.jin@intel.com>
Cc: Wei6 Xu <wei6.xu@intel.com>
Signed-off-by: ArvinX Chen <arvinx.chen@intel.com>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

build.sh contains a mixture of tabs and spaces, with differing
numbers of spaces used for indents. Update it so any single block only
uses one style of indentation.
Signed-off-by: Rebecca Cran <rebecca@nuviainc.com>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

The mailing list has changed from lists.01.org to groups.io.
Update Maintainers.txt and Readme.md to match.
Signed-off-by: Rebecca Cran <rebecca@nuviainc.com>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

build.sh usage indicates that "GCC5" can be specified, but it's currently
rejected due to a case statement only matching "GCC" or "gcc".
Fix this by adding a wildcard match.
Signed-off-by: Rebecca Cran <rebecca@nuviainc.com>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>

Heinrich Schuchardt authored Nov 30, 2020 and G Edhaya Chandran committed Nov 30, 2020

Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=3029

After commit 75c92b85 ("uefi-sct/SctPkg: NULL deref in DevicePathToText
test") a build failure was observed.

Length is a field in MEDIA_OFFSET_DEVICE_PATH.Header and not in MEDIA_OFFSET_DEVICE_PATH itself.

Fixes: 75c92b85

 ("uefi-sct/SctPkg: NULL deref in DevicePathToText test")
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

devel@edk2.groups.io authored Nov 25, 2020 and G Edhaya Chandran committed Nov 25, 2020

On systems with memory protection enabled the modification of local function initialization data results in permission issue. Make a copy of data prior to modification.
Signed-off-by: Jeff Brasen <jbrasen@nvidia.com>
Reviewed-By: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

Heinrich Schuchardt authored Nov 24, 2020 and G Edhaya Chandran committed Nov 24, 2020

Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=3029



Function DevicePathToTextConvertDeviceNodeToTextCoverageTest() tests if
DeviceNodeToText() correctly converts a Relative Offset Range node. After calling SctConvertTextToDeviceNode() it tries to set the field Reserved of the returned device node to 0.

If the tested firmware does not return the expected text
SctConvertTextToDeviceNode() may return NULL or a device node that is shorter than expected. In both cases it is not possible to access the field Reserved.

So we must check both that the returned node is not NULL and that it has the exepected size.

Due to the missing check a NULL dereference was observed when running the SCT on U-Boot.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>
Reviewed-by: Grant Likely <grant.likely@arm.com>

devel@edk2.groups.io authored Nov 24, 2020 and G Edhaya Chandran committed Nov 24, 2020

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3065



If LocateProtocol() has failed, variable Ebc is not valid and we may not dereference it.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-by: Grant Likely <grant.likely@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>

Grant Likely authored Nov 24, 2020 and G Edhaya Chandran committed Nov 24, 2020

EDK2 has removed some deprecated function wrappers. The BlackBoxTest
still uses StrnCpy() and UnicodeValueToString(). This patch fixes the
build by moving to UnicodeValueToStringS(). However, the code using
StrnCpy() is simply #ifdef'd out because on reading the code it is
apparent that the testcase is non-functional and the fix is not
straightforward. As I understand it, there are no functional
implementations of the interface in the wild anyway. If/when it does
become important someone can do the work of making a proper test case.

Fixes: Bug 2068
Signed-off-by: Grant Likely <grant.likely@...>
Cc: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@...>
Cc: G Edhaya Chandran <@edhay>
Cc: Eric Jin <eric.jin@...>
Cc: Barton Gao <gaojie@...>
Reviewed-by: Samer El-Haj-Mahmoud <Samer.El-Haj-Mahmoud@arm.com>
Reviewed-by: G Edhaya Chandran<edhaya.chandran@arm.com>