ShellPkg: Remove unnecessary Readme.txt

The Readme.txt contains instructions about how to integrate Shell
into Nt32. Actually Nt32 already contains a macro USE_OLD_SHELL to
choose OLD or NEW Shell.

So remove this txt file.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

MdeModulePkg/PciHostBridgeDxe: Make bitwise operands of the same size

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

OvmfPkg/AcpiPlatformDxe: alloc blobs from 64-bit space unless restricted

... by narrower than 8-byte ADD_POINTER references.

Introduce the CollectAllocationsRestrictedTo32Bit() function, which
iterates over the linker/loader script, and collects the names of the
fw_cfg blobs that are referenced by QEMU_LOADER_ADD_POINTER.PointeeFile
fields, such that QEMU_LOADER_ADD_POINTER.PointerSize is less than 8. This
means that the pointee blob's address will have to be patched into a
narrower-than-8 byte pointer field, hence the pointee blob must not be
allocated from 64-bit address space.

In ProcessCmdAllocate(), consult these restrictions when setting the
maximum address for gBS->AllocatePages(). The default is now MAX_UINT64,
unless restricted like described above to the pre-patch MAX_UINT32 limit.

In combination with Ard's QEMU commit cb51ac2ffe36 ("hw/arm/virt: generate
64-bit addressable ACPI objects", 2017-04-10), this patch enables
OvmfPkg/AcpiPlatformDxe to work entirely above the 4GB mark.

(An upcoming / planned aarch64 QEMU machine type will have no RAM under
4GB at all. Plus, moving the allocations higher is beneficial to the
current "virt" machine type as well; in Ard's words: "having all firmware
allocations inside the same 1 GB (or 512 MB for 64k pages) frame reduces
the TLB footprint".)

Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Igor Mammedov <imammedo@redhat.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Suggested-by: Igor Mammedov <imammedo@redhat.com>
Suggested-by: Gerd Hoffmann <kraxel@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Tested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

MdePkg: update Wi-Fi/Supplicant header files to meet UEFI 2.7.

This patch is used to update supplicant.h and wifi2.h
to meet UEFI 2.7 definition. Add EfiSupplicant80211PMK
field in EFI_SUPPLICANT_DATA_TYPE and change **NetworkDesc
to NetworkDesc[1] in EFI_80211_GET_NETWORKS_RESULT.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Wang Fan <fan.wang@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>
Reviewed-by: Ye Ting <ting.ye@intel.com>

BaseTools: Fix the bug use same FMP_PAYLOAD in different capsule file

Fix the bug that use same FMP_PAYLOAD in different capsule file. Because
in previous FMP generation, the FMP already be generated, so we don't
need to regenerate again.

Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

BaseTools: Fix incremental build failure that override file be removed

Fix a Incremental build failure. The case is: Both A and B package will
include a same .h file, and in the driver's packages section, A
package is listed before B package, so we will use the .h file in the A
package and build success, then we directly delete the .h file in package
A, it cause increment build failure since in the AutoGenTimeStamp file
the .h file in A can't be found.

Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

ShellBinPkg: Ia32/X64 Shell binary update.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>

ShellPkg/parse: Handle Unicode stream from pipe correctly

The original code expects the Unicode stream from pipe doesn't
contains the Unicode BOM.
But that's not true.
Commit [9ed21946c76e430097e9c4e59b419af928e0cb8c] changes
CreateFileInterfaceMem() to add the BOM for Unicode stream.

When parse pipe support was firstly added, a private implementation
ParseReturnStdInLine() was created to specially handle
the Unicode stream without BOM. Since now the Unicode steam contains
BOM, the private implementation can be removed and
ShellFileHandleReturnLine() can be used directly.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Tapan Shah <tapandshah@hpe.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

ShellPkg/alias: Return status for alias deletion

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Tapan Shah <tapandshah@hpe.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

MdePkg: Add BluetoothAttribute.h and BluetoothLeConfig.h

UEFI Spec 2.7 introduces BluetoothAttribute and BluetoothLeConfig
protocols. The patch adds the definitions for them.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

MdePkg/BluetoothIo: Formalize function header comments.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

MdePkg/BluetoothHc: Add detailed function header comments

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

MdePkg/BluetoothConfig: Add new EFI_BLUETOOTH_CONFIG_DATA_TYPE types

UEFI spec 2.7 adds new EFI_BLUETOOTH_CONFIG_DATA_TYPE types.
The patch adds them to the header file.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

MdePkg/DevicePath: Add BluetoothLe device path node support

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Hao A Wu <hao.a.wu@intel.com>

MdePkg SmmIoLib: Use NULL pointer check instead of useless Status check

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=587

The Status check in "if (!EFI_ERROR (Status))" condition is useless,
it should be NULL pointer check. And this patch also fixes a typo
"continous" to "continuous".

Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

MdePkg SmmMemLib: Remove ASSERT in SmmIsBufferOutsideSmmValid

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=577

Currently the SmmIsBufferOutsideSmmValid() function in SmmMemLib.c will
ASSERT in certain conditions. Since this function is a "test" function,
it should not be making decisions on how to handle a failure.
Handling a failure should be left to the caller.

This patch is to remove ASSERT(FALSE) at line 178 of SmmMemLib.c.

Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Bret Barkelew <bret.barkelew@microsoft.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

MdeModulePkg/UefiPxeBcDxe: Refine the PXE boot displayed information

This path is to refine the PXE boot displayed information so as to
in line with NetworkPkg/UefiPxeBcDxe driver.

Cc: Ye Ting <ting.ye@intel.com>
Cc: Fu Siyuan <siyuan.fu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Ye Ting <ting.ye@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>

MdeModulePkg/UefiPxeBcDxe: Fix the PXE BootMenu selection issue

Currently implementation doesn't accept the input during the user
is trying to select the PXE BootMenu from option 43. This path is
to fix that problem.

Cc: Ye Ting <ting.ye@intel.com>
Cc: Fu Siyuan <siyuan.fu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Ye Ting <ting.ye@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>

ShellPkg/alias: Fix bug to support upper-case alias

alias in UEFI Shell is case insensitive.
Old code saves the alias to variable storage without
converting the alias to lower-case, which results
upper case alias setting doesn't work.
The patch converts the alias to lower case before saving
to variable storage.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Cc: Tapan Shah <tapandshah@hpe.com>

EmulatorPkg/build.sh: Merge GCC toolchain detection from OVMF

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

MdeModulePkg/Xhci: Correct the indention of comments

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>

ShellPkg/UefiShellLib: Check correct variable for NULL

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Jaben Carsey <jaben.carsey@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Westfahl <jeff.westfahl@ni.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

MdeModulePkg/Xhci: Remove TRB when canceling Async Int Transfer

Some USB devices don't report data periodically through Int
Transfer. They report data only when be asked. If the TRB
is not removed from the XHCI HW, when next time HOST asks
data again, the data is reported but consumed by the previous
TRB, which results the HOST thinks data never comes.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao A Wu <hao.a.wu@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>

MdeModulePkg/MnpDxe: Fix EBC build hang issue.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>

MdeModulePkg/UsbBus: Correct debug message

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>

MdeModulePkg/UsbBus: Fix system hang when failed to uninstall UsbIo

When "reconnect -r" is typed in shell, UsbFreeInterface() is called
to uninstall the UsbIo and DevicePath. But When a UsbIo is opened
by a driver and that driver rejects to close the UsbIo in Stop(),
the uninstall doesn't succeed.
But UsbFreeInterface () frees the DevicePath memory without check
whether the uninstall succeeds.
It leads to the DXE core database contain a DevicePath instance but
that instance's memory is freed.
Assertion happens when someone calls InstallProtocol(DevicePath)
because the InstallProtocol() checks all DevicePath instance to
find whether the same one exits in database.

We haven't seen any USB device driver which rejects to close UsbIo
in Stop(), but it's very likely.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Cc: Hao A Wu <hao.a.wu@intel.com>

BaseTools/GCC ARM/AARCH64: Force disable PIE

After Debian's toolchain switched to PIE by default, our edk2 builds began
to fail to build (GCC49 w/ gcc 6.3). This patch fixes the build by forcing
off PIE for both ARM and AARCH64 builds.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: dann frazier <dannf@debian.org>
Add -fno-pic as well for ARM.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Liming Gao <liming.gao@intel.com>

Maintainers.txt: Update maintainers for DuetPkg & Nt32Pkg

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>

MdeModulePkg/Xhci: Fill the 'interval' field for ISO endpoint context

https://bugzilla.tianocore.org/show_bug.cgi?id=510

The commit fills the 'Interval' field of the Endpoint Context data for
isochronous endpoints. It will resolve the error when a Configure
Endpoint Command is sent to an isochronous endpoint.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>

MdeModulePkg/PrintLib: Avoid reading content beyond the format string

https://bugzilla.tianocore.org/show_bug.cgi?id=567

In functions DxePrintLibPrint2ProtocolVaListToBaseList() and
InternalPrintLibSPrintMarker(), when processing ASCII format strings, if
the format string walker pointer 'Format' is pointing at the end of the
format string (i.e. '\0'), the following expression:
*(Format + 1)
will read an undefined value.

Though this value won't affect the functionality, since it will be masked
by variable 'FormatMask':
(*(Format + 1) << 8)) & FormatMask
(FormatMask is 0xff for ASCII format string)

This commit adds additional logic to avoid reading undefined content.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

MdePkg/BasePrintLib: Avoid reading content beyond the format string

https://bugzilla.tianocore.org/show_bug.cgi?id=567

In function BasePrintLibSPrintMarker(), when processing ASCII format
strings, if the format string walker pointer 'Format' is pointing at the
end of the format string (i.e. '\0'), the following expression:
*(Format + 1)
will read an undefined value.

Though this value won't affect the functionality, since it will be masked
by variable 'FormatMask':
(*(Format + 1) << 8)) & FormatMask
(FormatMask is 0xff for ASCII format string)

This commit adds additional logic to avoid reading undefined content.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

MdePkg/DevicePathLib: Reverse the byte order of BD_ADDR for Bluetooth

For the following two functions:
DevPathFromTextBluetooth()
DevPathToTextBluetooth()

The Bluetooth device address "UINT8  Address[6]" is displayed with the
order from Address[5] to Address[0]. This commit reverses the order.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>

QuarkPlatformPkg/SpiFvbServices: correct NumOfLba vararg type in EraseBlocks()

According to the PI spec, Volume 3,
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.EraseBlocks():

> The variable argument list is a list of tuples. Each tuple describes a
> range of LBAs to erase and consists of the following:
> * An EFI_LBA that indicates the starting LBA
> * A UINTN that indicates the number of blocks to erase

(NB, in edk2, EFI_FIRMWARE_VOLUME_BLOCK_PROTOCOL is a typedef to
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.)

In this driver, the NumOfLba local variable is defined with type UINTN,
but the TYPE argument passed to VA_ARG() is UINT32. Fix the mismatch.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Kelly Steele <kelly.steele@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Reported-by: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Michael D Kinney <michael.d.kinney@intel.com>

Nt32Pkg/FvbServicesRuntimeDxe: correct NumOfLba vararg type in EraseBlocks()

According to the PI spec, Volume 3,
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.EraseBlocks():

> The variable argument list is a list of tuples. Each tuple describes a
> range of LBAs to erase and consists of the following:
> * An EFI_LBA that indicates the starting LBA
> * A UINTN that indicates the number of blocks to erase

(NB, in edk2, EFI_FIRMWARE_VOLUME_BLOCK_PROTOCOL is a typedef to
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.)

In this driver, the NumOfLba local variable is defined with type UINTN,
but the TYPE argument passed to VA_ARG() is UINT32. Fix the mismatch.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Reported-by: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>

DuetPkg/FvbRuntimeService: correct NumOfLba vararg type in EraseBlocks()

According to the PI spec, Volume 3,
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.EraseBlocks():

> The variable argument list is a list of tuples. Each tuple describes a
> range of LBAs to erase and consists of the following:
> * An EFI_LBA that indicates the starting LBA
> * A UINTN that indicates the number of blocks to erase

(NB, in edk2, EFI_FIRMWARE_VOLUME_BLOCK_PROTOCOL is a typedef to
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.)

In this driver, the NumOfLba local variable is defined with type UINTN,
but the TYPE argument passed to VA_ARG() is UINT32. Fix the mismatch.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Reported-by: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>

BaseTools/Scripts: fix GccBase.lds line endings

Replace a <LF> line ending that was introduced inadvertently by a
recent commit with the correct <CR><LF>

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>

UefiCpuPkg/MpInitLib: Force to enable X2APIC if CPU number > 255

Cc: Michael D Kinney <michael.d.kinney@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>

UefiCpuPkg/MpInitLib: Check APIC mode change around AP function

If APIC ID values are changed during AP functions execution, we need to update
new APIC ID values in local data structure accordingly.

But if APIC mode change happened during AP function execution, we do not support
APIC ID value changed.

Cc: Michael D Kinney <michael.d.kinney@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>

UefiCpuPkg/CpuCommonFeaturesLib: Support X2APIC enable

Current X2APIC is enabled in MpInitLib (used by CpuMpPei and CpuDxe) to follow
SDM suggestion. That means we only enable X2APIC if we found there are any
initial CPU ID value >= 255.

This patch is to provide one chance for platform to enable X2APIC even there is
no any initial CPU ID value >= 255.

Cc: Michael D Kinney <michael.d.kinney@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>

MdeModulePkg: Fix potential memory leaks in DxeHttpLib.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Paulo Alcantara <pcacjr@gmail.com>
Reviewed-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>

Vlv2TbltDevicePkg/FvbRuntimeDxe: correct NumOfLba vararg type in EraseBlocks()

According to the PI spec, Volume 3,
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.EraseBlocks():

> The variable argument list is a list of tuples. Each tuple describes a
> range of LBAs to erase and consists of the following:
> * An EFI_LBA that indicates the starting LBA
> * A UINTN that indicates the number of blocks to erase

(NB, in edk2, EFI_FIRMWARE_VOLUME_BLOCK_PROTOCOL is a typedef to
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.)

In this driver, the NumOfLba local variable is defined with type UINTN,
but the TYPE argument passed to VA_ARG() is UINT32. Fix the mismatch.

Cc: David Wei <david.wei@intel.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Mang Guo <mang.guo@intel.com>
Reported-by: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>

BaseTools: Correct if condition expression for DatumType == 'VOID*'

Correct the if condition expression for DatumType == 'VOID*'. Current
this condition is not work since the DatumType is changed before we do
the value judgement.

Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

MdeModulePkg/DebugSupportDxe: Fix XCODE5 build failure

https://bugzilla.tianocore.org/show_bug.cgi?id=572

Update X64 AsmFuncs.S to resolve a build failure using the
XCODE5 tool chain.  This change updates AsmFuncs.S to match
AsmFuncs.asm and AsmFuncs.nasm.

The error generated in XCODE5 build is:

error: invalid instruction mnemonic 'movzxw'
    movzxw 8(%rax), %rax
    ^~~~~~

The correct instruction is

    movzwq 8(%rax), %rax

Cc: Star Zeng <star.zeng@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Andrew Fish <afish@apple.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Andrew Fish <afish@apple.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>

MdeModulePkg/RegularExpressionDxe: Fix XCODE5 build failure

https://bugzilla.tianocore.org/show_bug.cgi?id=572

The ErrorMessage local variable in OnigurumaMatch() should
be type OnigUChar instead of type CHAR8.  This resolves
a build failure with the XCODE5 tool chain.

Cc: Andrew Fish <afish@apple.com>
Cc: Star Zeng <star.zeng@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Andrew Fish <afish@apple.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>

BaseTools/Scripts: discard .gnu.hash section in GCC builds

Some builds of GCC/binutils will default to using the GNU flavor of
the symbol hash table, and will emit it into a section called .gnu.hash
rather than .hash. We have no use for its contents, and GenFw ignores
it anyway, so it shouldn't really matter what we do with it.

However, due to a workaround for AARCH64 we have in GenFw to deal with
older GCCs that corrupt section-based relocations when merging sections
during the final link, we need the ELF and PE/COFF views of the binary
to be identical. Since we don't place the .gnu.hash section explicitly,
it may end up at the beginning of the ELF binary, causing other sections
to be shifted in the ELF view but not in the PE/COFF view.

So let's add .gnu.hash to the GCC linker script. We don't care about its
contents so add it to the /DISCARD/ section.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Tested-by: Leif Lindholm <leif.lindholm@linaro.org>
Reviewed-by: Liming Gao <liming.gao@intel.com>

MdeModulePkg MemoryProfileInfo: Use PdbStringOffset to get PDB info

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=570
Use PdbStringOffset to get PDB info, it will have no functional impact,
and will just make the code more generic.

Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

MdeModulePkg SmiHandlerProfile: Use fixed data type in data structure

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=568

Use fixed data type in data structure and make the structure
be natural aligned.
Without this update, the code must assume DXE and SMM are using
same data type (same size of UINTN), but it may be not true at
some case, for example, after standalone SMM feature is enabled.
With this update, the data structure will be phase independent
and convenient for consumer to parse the data.

Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

MdeModulePkg SmiHandlerProfile: Fix no PDB case handling incorrectly

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=569

The PdbStringOffset should be set to 0 for no PDB case,
then SmiHandlerProfileInfo can use it to know whether
there is PCD info or not.

Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

ShellPkg/UefiShellLib: Avoid reading undefined content before string

https://bugzilla.tianocore.org/show_bug.cgi?id=566

In function InternalShellPrintWorker(), if the string in variable
'mPostReplaceFormat2' starts with character L'%', the following
expression:

*(ResumeLocation-1) == L'^' at line 2831

will read an undefined value before the starting of string
'mPostReplaceFormat2'.

This commit adds additional logic to avoid reading undefined content.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

BaseTools: Fix the bug that different DSC file use same build output

We meet a corner case that build different DSC file, but the DSC file use
same build output directory, and the different DSC file use a same PCD
with different Pcd Type, it cause build failure.

Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

BaseTools: Fix incremental build bug on DynamicPcd Token Generation

During incremental build, we meet the bug that the different drivers use
the different token for the same DynamicPcd.

Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

OvmfPkg: make the 4MB flash size the default (again)

Xen gained support for the 4MB flash image in Xen commit 0d6968635ce5
("hvmloader: avoid tests when they would clobber used memory",
2017-05-19), which is part of Xen 4.9.0-rc6.

The previously default 2MB can be explicitly selected with

  -D FD_SIZE_2MB

or

  -D FD_SIZE_IN_KB=2048

Cc: Jordan Justen <jordan.l.justen@intel.com>
Suggested-by: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
(cherry picked from commit bba8dfbec3bbc4fba7fa6398ba3cf76593e0725e)
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
[lersek@redhat.com: reference Xen commit in commit message]
Signed-off-by: Laszlo Ersek <lersek@redhat.com>

UefiCpuPkg/MpInitLib: Fix X64 XCODE5/NASM compatibility issues

https://bugzilla.tianocore.org/show_bug.cgi?id=565

Fix NASM compatibility issues with XCODE5 tool chain.
The XCODE5 tool chain for X64 builds using PIE (Position
Independent Executable).  For most assembly sources using
PIE mode does not cause any issues.

However, if assembly code is copied to a different address
(such as AP startup code in the MpInitLib), then the
X64 assembly source must be implemented to be compatible
with PIE mode that uses RIP relative addressing.

The specific changes in this patch are:

* Use LEA instruction instead of MOV instruction to lookup
  the addresses of functions.

* The assembly function RendezvousFunnelProc() is copied
  below 1MB so it can be executed as part of the MpInitLib
  AP startup sequence.  RendezvousFunnelProc() calls the
  external function InitializeFloatingPointUnits().  The
  absolute address of InitializeFloatingPointUnits() is
  added to the MP_CPU_EXCHANGE_INFO structure that is passed
  to RendezvousFunnelProc().

Cc: Andrew Fish <afish@apple.com>
Cc: Jeff Fan <jeff.fan@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Andrew Fish <afish@apple.com>

UefiCpuPkg/DxeMpInitLib.inf: Add missing SynchronizationLib

Contributed-under: TianoCore Contribution Agreement 1.0
Cc: Eric Dong <eric.dong@intel.com>
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>

MdeModulePkg/LogoDxe: Return error if HII Package not present

https://bugzilla.tianocore.org/show_bug.cgi?id=554

Update LogoDxe module to print a DEBUG() message and exit
with an error instead of ASSERT_EFI_ERROR() if the HII
Image Package with the logo image is not present.

If a tool chain does not support generation of PE/COFF
resource sections, then this module can not produce the logo
from an HII Image Package.  XCODE5 is an example of a tool
chain that does not currently support generation of PE/COFF
resource sections.

Cc: Star Zeng <star.zeng@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Andrew Fish <afish@apple.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael Kinney <michael.d.kinney@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>

ArmPlatformPkg: Timer access for non-secure EL1/0

According to Section 2.3.6 of the "UEFI Specification 2.6 Errata A";
the primary CPU must be configured such that 'Timer access must be
provided to non-secure EL1 and EL0 by setting bits EL1PCTEN and
EL1PCEN in register CNTHCTL_EL2.'

This commit adds this missing set-up to the PrePi and PrePeiCore
modules.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com>
Signed-off-by: Evan Lloyd <evan.lloyd@arm.com>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>

ArmPkg: Add CNTHCTL_EL2 support functions

Added helper functions for reading and writing the
CNTHCTL_EL2 register.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com>
Signed-off-by: Evan Lloyd <evan.lloyd@arm.com>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>

MdeModulePkg/BDS: Fix a buffer overflow bug

KeyOption points to a buffer holding the content of Key####.
So its size is smaller than EFI_BOOT_MANAGER_KEY_OPTION.
Old code to assign value to KeyOption->OptionNumber modifies
the memory outside of the KeyOption buffer.

The patch fixes this bug.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Steven Shi <steven.shi@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>

CryptoPkg/BaseCryptLib: Add NULL pointer checks in DH and P7Verify

Add more NULL pointer checks before using them in DhGenerateKey and
Pkcs7GetCertificatesList functions to eliminate possible dereferenced
pointer issue.

Cc: Ting Ye <ting.ye@intel.com>
Cc: Hao Wu <hao.a.wu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qin Long <qin.long@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Ting Ye <ting.ye@intel.com>

edk2: Add .DS_Store to .gitignore for macOS

https://bugzilla.tianocore.org/show_bug.cgi?id=558

macOS may generate .DS_Store files in directories.
The .gitignore file is updated to ignore these
.DS_Store files.

Cc: Andrew Fish <afish@apple.com>
Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Andrew Fish <afish@apple.com>

BaseTools: Clean up tools_def.template for XCODE5

Reorganize the statements for XCODE5 to match other tool
chains and remove dependency on XCLANG and XCODE32

Cc: Andrew Fish <afish@apple.com>
Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Andrew Fish <afish@apple.com>

BaseTools: Add -D NO_MSABI_VARGS to X64 XCODE5 CC_FLAGS

https://bugzilla.tianocore.org/show_bug.cgi?id=561

Update BaseTools/Conf/tools_def.template to add the define

-D NO_MSABI_VAARGS

To CC_FLAGS for X64 XCODE5 builds.

The llvm/clang compiler used in XCODE5 builds supports the
_ms_ versions of the vararg builtins, but the compiler
generates build errors.

The recommendation from the XCODE5 experts is to never use
the _ms_ version of the vararg builtins.  The define
NO_MSABI_VARARGS is already supported in MdePkg/Include/Base.h
and forces the use the standard vararg builtins.

Cc: Andrew Fish <afish@apple.com>
Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Andrew Fish <afish@apple.com>

OvmfPkg: Add XCODE5 statements to fix build break

https://bugzilla.tianocore.org/show_bug.cgi?id=559

The XCODE5 tool chain has a FAMILY of GCC.  The
GCC statements in the [BuildOptions] section add
flags that are not compatible with XCODE5.  Add
empty XCODE5 statements in [BuildOptions] sections
to prevent the use of the GCC flags in XCODE5
builds.

Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Andrew Fish <afish@apple.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

UefiCpuPkg: Use FINIT instead of hex values

https://bugzilla.tianocore.org/show_bug.cgi?id=560

Update X64 NASM file to match IA32 NASM file
and use FINIT instruction instead of hand
assembled hex values for the FINIT instruction.

Cc: Jeff Fan <jeff.fan@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>

UefiCpuPkg/BaseUefiCpuLib: Use NASM read-only data section name

https://bugzilla.tianocore.org/show_bug.cgi?id=556

NASM requires read-only data sections to use the section
name .rodata.  This fix changes .rdata to .rodata.

The build failure from use of .rdata is seen when using
the XCODE5 tool chain.

Section "7.8.1 macho extensions to the SECTION Directive"
of the NASM documentation at http://www.nasm.us/doc/
describes the section name requirements.

Cc: Jeff Fan <jeff.fan@intel.com>
Cc: Andrew Fish <afish@apple.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael Kinney <michael.d.kinney@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>

UefiCpuPkg/PiSmmCpuDxeSmm: Add missing JMP instruction

https://bugzilla.tianocore.org/show_bug.cgi?id=555

Add JMP instruction in SmiEntry.S file that is missing.  This
updates SmiEntry.S to match the logic in SmiEntry.asm and
SmiEntry.nasm.

The default BUILDRULEORDER has .nasm higher priority than
.asm or .S, so this issue was not seen with MSFT or GCC
tool chain families.  The XCODE5 tool chain overrides the
BUILDRULEORDER with .S higher than .nasm, so this issue
was only seen when using XCODE5 tool chain when IA32 SMM
is enabled.

Cc: Jeff Fan <jeff.fan@intel.com>
Cc: Andrew Fish <afish@apple.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael Kinney <michael.d.kinney@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

PcAtChipsetPkg/SerialIoLib: Remove negative value shift

https://bugzilla.tianocore.org/show_bug.cgi?id=553

Remove left shift of negative values that always evaluate
to 0 to address build errors from the llvm/clang compiler
used in the XCODE5 tool chain.

Clang rightfully complains about left-shifting ~DLAB. DLAB is #defined
as 0x01 (an "int"), hence ~DLAB has value (-2) on all edk2 platforms.
Left-shifting a negative int is undefined behavior.

Rather than replacing ~DLAB with ~(UINT32)DLAB, realize that the nonzero
bits of (~(UINT32)DLAB << 7) would all be truncated away in the final
conversion to UINT8 anyway. So just remove (~DLAB << 7).

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Andrew Fish <afish@apple.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>

MdeModulePkg PCD: Fix TmpTokenSpaceBufferCount not assigned correctly

When DynamicEx PCD is only used in PEI code, but not DXE code,
current implementation of DxePcdGetNextTokenSpace does not assign
TmpTokenSpaceBufferCount correctly, but leaves it as initial value,
then DxePcdGetNextTokenSpace may return incorrect token space guid
and status.

This patch is to fix this issue.

Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

OvmfPkg/README: document 4MB flash layout

Cc: Jordan Justen <jordan.l.justen@intel.com>
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=527
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg/PlatformPei: align EmuVariableNvStore at any page boundary

EmuVariableFvbRuntimeDxe now uses a 4KB (EFI_PAGE_SIZE) block size.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Regression-tested-by: Gary Lin <glin@suse.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg/EmuVariableFvbRuntimeDxe: change block size to 4KB

EmuVariableFvbRuntimeDxe currently produces a Firmware Volume Block
protocol that is based on a block map of two blocks, each block having
PcdFlashNvStorageFtwSpareSize for size.

(The total size is 2 * PcdFlashNvStorageFtwSpareSize.)

FaultTolerantWriteDxe in turn expects the block size to be a power of two.

In the 4MB build of OVMF, PcdFlashNvStorageFtwSpareSize is 264KB, which is
not a power of two. In order to equip EmuVariableFvbRuntimeDxe for this
build, shrink the block size to 4KB (EFI_PAGE_SIZE), and grow the block
count from 2 to EFI_SIZE_TO_PAGES(2 * PcdFlashNvStorageFtwSpareSize). The
total size remains

  2 * PcdFlashNvStorageFtwSpareSize
  --------------------------------- * EFI_PAGE_SIZE
            EFI_PAGE_SIZE

Right now EmuVariableFvbRuntimeDxe open-codes the block count of 2 in
various limit checks, so introduce a few new macros:
- EMU_FVB_NUM_TOTAL_BLOCKS, for the LHS of the above product,
- EMU_FVB_NUM_SPARE_BLOCKS for the half of that.

Also rework the FVB protocol members to support an arbitrary count of
blocks.

Keep the invariant intact that the first half of the firmware volume hosts
the variable store and the FTW working block, and that the second half
maps the FTW spare area.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=527
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Regression-tested-by: Gary Lin <glin@suse.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg/EmuVariableFvbRuntimeDxe: strip trailing whitespace

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Regression-tested-by: Gary Lin <glin@suse.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg/QemuFlashFvbServicesRuntimeDxe: correct NumOfLba vararg type in EraseBlocks()

According to the PI spec, Volume 3,
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.EraseBlocks():

> The variable argument list is a list of tuples. Each tuple describes a
> range of LBAs to erase and consists of the following:
> * An EFI_LBA that indicates the starting LBA
> * A UINTN that indicates the number of blocks to erase

(NB, in edk2, EFI_FIRMWARE_VOLUME_BLOCK_PROTOCOL is a typedef to
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.)

In this driver, the NumOfLba local variable is defined with type UINTN,
but the TYPE argument passed to VA_ARG() is UINT32. Fix the mismatch.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Reported-by: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg/EmuVariableFvbRuntimeDxe: correct NumOfLba vararg type in EraseBlocks()

According to the PI spec, Volume 3,
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.EraseBlocks():

> The variable argument list is a list of tuples. Each tuple describes a
> range of LBAs to erase and consists of the following:
> * An EFI_LBA that indicates the starting LBA
> * A UINTN that indicates the number of blocks to erase

(NB, in edk2, EFI_FIRMWARE_VOLUME_BLOCK_PROTOCOL is a typedef to
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.)

In this driver, the NumOfLba local variable is defined with type UINTN,
but the TYPE argument passed to VA_ARG() is UINT32. Fix the mismatch.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

EmulatorPkg/FvbServicesRuntimeDxe: correct NumOfLba vararg type in EraseBlocks()

According to the PI spec, Volume 3,
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.EraseBlocks():

> The variable argument list is a list of tuples. Each tuple describes a
> range of LBAs to erase and consists of the following:
> * An EFI_LBA that indicates the starting LBA
> * A UINTN that indicates the number of blocks to erase

(NB, in edk2, EFI_FIRMWARE_VOLUME_BLOCK_PROTOCOL is a typedef to
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.)

In this driver, the NumOfLba local variable is defined with type UINTN,
but the TYPE argument passed to VA_ARG() is UINT32. Fix the mismatch.

Cc: Andrew Fish <afish@apple.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Reported-by: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

ArmPlatformPkg/NorFlashDxe: correct NumOfLba vararg type in EraseBlocks()

According to the PI spec, Volume 3,
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.EraseBlocks():

> The variable argument list is a list of tuples. Each tuple describes a
> range of LBAs to erase and consists of the following:
> * An EFI_LBA that indicates the starting LBA
> * A UINTN that indicates the number of blocks to erase

(NB, in edk2, EFI_FIRMWARE_VOLUME_BLOCK_PROTOCOL is a typedef to
EFI_FIRMWARE_VOLUME_BLOCK2_PROTOCOL.)

In this driver, the NumOfLba local variable is defined with type UINTN,
but the TYPE argument passed to VA_ARG() is UINT32. Fix the mismatch.

In addition, update the DEBUG macro invocation where NumOfLba is formatted
with the %d conversion specifier: UINTN values should be converted to
UINT64 and printed with %Lu or %Lx for portability between 32-bit and
64-bit.

Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Leif Lindholm <leif.lindholm@linaro.org>
Reported-by: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

MdeModulePkg/PciHostBridgeDxe: Fix EBC build failure

Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

OvmfPkg: resolve PcdLib for PEIMs to PeiPcdLib by default

In the previous patch we had to add two explicit Null resolutions, but
here we can remove five PeiPcdLib ones, after setting the default to it.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg: resolve PcdLib for all PEIMs individually

Currently the default (module type independent) PcdLib resolution is to
BasePcdLibNull.inf, which is inherited by all PEIMs. In the next patch,
we'll flip the PEIM default resolution to PeiPcdLib.inf, but in order to
keep that patch both correct and simple to review, we should spell out the
Null resolution for those two PEIMs (ReportStatusCodeRouterPei and
StatusCodeHandlerPei) that are now the only ones that don't specify an
explicit resolution.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg/PlatformPei: don't allocate reserved mem varstore if SMM_REQUIRE

For the emulated variable store, PlatformPei allocates reserved memory (as
early as possible, so that the address remains the same during reboot),
and PcdEmuVariableNvStoreReserved carries the address to
EmuVariableFvbRuntimeDxe.

However, EmuVariableFvbRuntimeDxe is excluded from the SMM_REQUIRE build,
and then noone consumes PcdEmuVariableNvStoreReserved. Don't waste
reserved memory whenever that's the case.

(Even a dynamic default for PcdEmuVariableNvStoreReserved would be
unnecessary; but that way the PcdSet64S() call in the
ReserveEmuVariableNvStore() function doesn't compile.)

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg: sync PcdVariableStoreSize with PcdFlashNvStorageVariableSize

"MdeModulePkg/MdeModulePkg.dec" declares PcdVariableStoreSize like this:

> The size of volatile buffer. This buffer is used to store VOLATILE
> attribute variables.

There is no inherent reason why the size of the volatile variable store
should match the same of the non-volatile variable store. Indeed flash
variables in the 4MB build work fine without this equality.

However, OvmfPkg/EmuVariableFvbRuntimeDxe uses PcdVariableStoreSize to
initialize the non-volatile VARIABLE_STORE_HEADER too. (Presumably based
on the fact that ultimately that storage will not be permanent.) When
using EmuVariableFvbRuntimeDxe in the 4MB build, the mismatch between the
two mentioned PCDs (which is apparent through EmuVariableFvbRuntimeDxe's
VARIABLE_STORE_HEADER) triggers an assertion in the variable driver:

> ASSERT MdeModulePkg/Universal/Variable/RuntimeDxe/Variable.c(3772):
> mNvVariableCache->Size == VariableStoreLength

Bringing PcdVariableStoreSize in sync with PcdFlashNvStorageVariableSize
fixes this. It also happens to ensure a volatile store size in the 4MB
build that equals the non-volatile store size, which likely doesn't hurt
for symmetry.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Fixes: b24fca05751f8222acf264853709012e0ab7bf49
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg/PlatformPei: remove unused PcdVariableStoreSize dependency

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg: remove gUefiOvmfPkgTokenSpaceGuid.PcdSecureBootEnable

This PCD is no longer used.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

OvmfPkg/EmuVariableFvbRuntimeDxe: always format an auth varstore header

In this patch, we extend commit d92eaabefbe0 ("OvmfPkg: simplify
VARIABLE_STORE_HEADER generation", 2016-02-05) to
EmuVariableFvbRuntimeDxe.

This is the difference between FvAndVarTemplate and
FvAndAuthenticatedVarTemplate:

> --- non-auth    2017-05-05 22:32:06.001512283 +0200
> +++ auth        2017-05-05 22:32:18.841364882 +0200
> @@ -1,7 +1,7 @@
>    //
> -  // Templates for standard (non-authenticated) variable FV header
> +  // Templates for authenticated variable FV header
>    //
> -  STATIC FVB_FV_HDR_AND_VARS_TEMPLATE FvAndVarTemplate = {
> +  STATIC FVB_FV_HDR_AND_VARS_TEMPLATE FvAndAuthenticatedVarTemplate = {
>      { // EFI_FIRMWARE_VOLUME_HEADER FvHdr;
>        // UINT8                     ZeroVector[16];
>        { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 },
> @@ -34,7 +34,7 @@
>        EFI_FVH_REVISION,
>
>        // EFI_FV_BLOCK_MAP_ENTRY    BlockMap[1];
> -      {
> +      {
>          {
>            2, // UINT32 NumBlocks;
>            EMU_FVB_BLOCK_SIZE  // UINT32 Length;
> @@ -44,8 +44,8 @@
>      // EFI_FV_BLOCK_MAP_ENTRY     EndBlockMap;
>      { 0, 0 }, // End of block map
>      { // VARIABLE_STORE_HEADER      VarHdr;
> -      // EFI_GUID  Signature;
> -      EFI_VARIABLE_GUID,
> +        // EFI_GUID  Signature;     // need authenticated variables for secure boot
> +        EFI_AUTHENTICATED_VARIABLE_GUID,
>
>        // UINT32  Size;
>        (

After this change, using "-bios", the variable driver logs:

- with the SB feature enabled:
> Variable driver will work with auth variable format!
> Variable driver will work with auth variable support!

- with the SB feature disabled:
> Variable driver will work with auth variable format!
> Variable driver will continue to work without auth variable support!

Cc: Jordan Justen <jordan.l.justen@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>

MdeModulePkg/PciBus: Add IOMMU support.

If IOMMU protocol is installed, PciBus need call IOMMU
to set access attribute for the PCI device in Map/Ummap.

Only after the access attribute is set, the PCI device can
access the DMA memory.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Leo Duran <leo.duran@amd.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Previous patch Tested-by: Brijesh Singh <brijesh.singh@amd.com>
Previous patch Tested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Leo Duran <leo.duran@amd.com>

MdeModulePkg/PciHostBridge: Add IOMMU support.

If IOMMU protocol is installed, PciHostBridge just calls
IOMMU AllocateBuffer/FreeBuffer/Map/Unmap.

PciHostBridge does not set IOMMU access attribute,
because it does not know which device request the DMA.
This work is done by PciBus driver.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Leo Duran <leo.duran@amd.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Previous patch Tested-by: Brijesh Singh <brijesh.singh@amd.com>
Previous patch Tested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Leo Duran <leo.duran@amd.com>

MdeModulePkg/Include: Add IOMMU protocol definition.

This protocol is to abstract DMA access from IOMMU.
1) Intel "DMAR" ACPI table.
2) AMD "IVRS" ACPI table
3) ARM "IORT" ACPI table.

There might be multiple IOMMU engines on one platform.
For example, one for graphic and one for rest PCI devices
(such as ATA/USB).
All IOMMU engines are reported by one ACPI table.

All IOMMU protocol provider should be based upon ACPI table.
This single IOMMU protocol can handle multiple IOMMU engines on one system.

This IOMMU protocol provider can use UEFI device path to distinguish
if the device is graphic or ATA/USB, and find out corresponding
IOMMU engine.

The IOMMU protocol provides 2 capabilities:
A) Set DMA access attribute - such as write/read control.
B) Remap DMA memory - such as remap above 4GiB system memory address
to below 4GiB device address.
It provides AllocateBuffer/FreeBuffer/Map/Unmap for DMA memory.
The remapping can be static (fixed at build time) or dynamic (allocate
at runtime).

4) AMD "SEV" feature.
We can have an AMD SEV specific IOMMU driver to produce IOMMU protocol,
and manage SEV bit.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Leo Duran <leo.duran@amd.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Previous patch Tested-by: Brijesh Singh <brijesh.singh@amd.com>
Previous patch Tested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Leo Duran <leo.duran@amd.com>

MdeModulePkg/UfsPassThruDxe: Fix typo in UfsPassThruGetTargetLun()

For function UfsPassThruGetTargetLun(), the length of the input device
node specified by 'DevicePath' should be compared with the size of
'UFS_DEVICE_PATH' rather than the size of 'SCSI_DEVICE_PATH'.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>

ShellBinPkg: Ia32/X64 Shell binary update.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>

BaseTools: Fix the bug for CArray PCD override in command line

This patch updated the CArray PCD override format from B"{}" to H"{}"
which align to build spec. Besides, it also do the clean up for the
function BuildOptionPcdValueFormat.

Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

BaseTools: Fix the bug that FixedPcdGetPtr failure for CArray Pcd

This patch for the bug FixedPcdGetPtr report failure for the CArray type
Pcd. 1) correct the Fixed Pcd list; 2) correct the Fixed Pcd in Library
AutoGen file to same with Driver AutoGen file format.

Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>

MdeModulePkg SmiHandlerProfile: Fix memory leak in DumpSmiChildContext

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=530

In DumpSmiChildContext() of SmiHandlerProfile.c and
SmiHandlerProfileInfo.c, the return buffer from
ConvertDevicePathToText() should be freed after used.

Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

ShellPkg/memmap: Dump memory map information for all memory types

The patch dumps memory map information for all memory types.
But to follow the SFO format of "memmap" defined in Shell 2.2 spec,
the patch doesn't dump the memory map information for OEM/OS
memory types. But it does include the OEM/OS memory in the total
size in SFO format.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

ShellPkg/memmap: Refine code

The patch changes Buffer to Descriptors, changes
(UINT8 *Walker) to (EFI_MEMORY_DESCRIPTOR *Walker).
The change makes lots of type conversion unnecessary.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

ShellPkg/HandleParsingLib: Show LoadedImageProtocol file name

This patch adds support for showing the file name associated with a
LoadedImageProtocol file path. This is a behavior that was present in
the old shell but has been lost in the new shell.

For example, using 'dh -v' in the old shell:

    Handle D3 (3A552218)
    Image (3A54C918)   File:MicrocodeUpdate
        ParentHandle..: 3A666398

vs. the new shell:

    D3: 3A552218
    LoadedImage
        Revision......: 0x00001000
        ParentHandle..: 3A666398

Here's what the output of 'dh -v' looks like after this patch:

    D3: 3A552218
    LoadedImage
        Name..........: MicrocodeUpdate
        Revision......: 0x00001000
        ParentHandle..: 3A666398

This seems like useful information for the shell to display.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Jaben Carsey <jaben.carsey@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Westfahl <jeff.westfahl@ni.com>
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

ShellPkg/HandleParsingLib: Open LoadedImageProtocol first

This patch changes the order of operations to make sure we can open the
LoadedImageProtocol before getting the format string. This should not
affect functionality, and makes the next patch easier to review.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Jaben Carsey <jaben.carsey@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Westfahl <jeff.westfahl@ni.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

ShellPkg/HandleParsingLib: Show LoadedImageProtocol file path as text

This patch adds support for displaying a text representation of the file
path associated with a LoadedImageProtocol. This is a behavior that was
present in the old shell but has been lost in the new shell.

For example, using 'dh -v' in the old shell:

    FilePath......: FvFile(F3331DE6-4A55-44E4-B767-7453F7A1A021)
    FilePath......: \EFI\BOOT\BOOTX64.EFI

vs. the new shell:

    FilePath......: 3A539018
    FilePath......: 3A728718

This seems like useful information for the shell to display.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Jaben Carsey <jaben.carsey@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Westfahl <jeff.westfahl@ni.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>

UefiCpuPkg/PiSmmCpuDxeSmm: Fix logic check error

Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>

UefiCpuPkg/PiSmmCpuDxeSmm: Check ProcessorId == INVALID_APIC_ID

If PcdCpuHotPlugSupport is TRUE, gSmst->NumberOfCpus will be the
PcdCpuMaxLogicalProcessorNumber. If gSmst->SmmStartupThisAp() is invoked for
those un-existed processors, ASSERT() happened in ConfigSmmCodeAccessCheck().

This fix is to check if ProcessorId is valid before invoke
gSmst->SmmStartupThisAp() in ConfigSmmCodeAccessCheck() and to check if
ProcessorId is valid in InternalSmmStartupThisAp() to avoid unexpected DEBUG
error message displayed.

Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>

UefiCpuPkg/SmmCpuFeaturesLib: Correct print level

Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>