According to the Debug Port Table 2 (DBG2) specification,
February 17, 2021, the NamespaceString is a NULL terminated
ASCII string that consists of a fully qualified reference
to the object that represents the serial port device in the
ACPI namespace.
The DBG2 table generator did not populate the full device
path for the serial port device, and this results in a FWTS
test failure.
Therefore, populate the full namespace device path for the
serial port in DBG2 table.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com>
Reviewed-by: Pierre Gondois <pierre.gondois@arm.com>
Tested-by: Jagadeesh Ujja <Jagadeesh.Ujja@arm.com>
Tested-by: Sunny Wang <sunny.wang@arm.com>
When build option ABOVE_4G_MEMORY is set to true, nothing will change
and EDKII will use all available memory.
Setting it to false will create memory type information HOB in
payload entry, so that EDKII will reserve enough memory below 4G
for EDKII modules. This option is useful for bootloaders that are not
fully 64-bit aware such as Qubes R4.0.4 bootloader, Zorin and Proxmox.
Cc: Guo Dong <guo.dong@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Maurice Ma <maurice.ma@intel.com>
Cc: Benjamin You <benjamin.you@intel.com>
Signed-off-by: Sean Rhodes <sean@starlabs.systems>
Reviewed-by: Guo Dong <guo.dong@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Add some documentation to the CloudHv target in order to clarify how to
use it and what to expect from it.
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Instead of using the CMOS, the CloudHv platform relies on the list of
memmap entries provided through the PVH boot protocol to determine the
last RAM address below 4G.
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Instead of hardcoding the address of the RSDP in the firmware, let's
rely on the PVH structure hvm_start_info to retrieve this information.
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Following the model from the Xen target, CloudHv is generated as a PVH
ELF binary to take advantage of the PVH specification, which requires
less emulation from the VMM.
The fdf include file CloudHvElfHeader.fdf.inc has been generated from
the following commands:
$ gcc -D PVH64 -o elf_gen OvmfPkg/OvmfXenElfHeaderGenerator.c
$ ./elf_gen 4194304 OvmfPkg/CloudHv/CloudHvElfHeader.fdf.inc
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
CloudHv doesn't need any VARS store, and it doesn't need the CODE
section to be generated separately either. The only thing needed is to
generate a firmware binary that can be used by Cloud Hypervisor.
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Updating the fdf include file based on the run of the ELF header
generator. The diff from this patch is the result of:
$ gcc -o elf_gen OvmfPkg/OvmfXenElfHeaderGenerator.c
$ ./elf_gen 2097152 OvmfPkg/XenElfHeader.fdf.inc
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Instead of having the PVH ELF header part of the fdf file directly, we
move it to a dedicated include file. This is the first step in
automating the generation of the header.
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Adding some flexibility to the program through optional parameters and
global define, so that other targets can use the generator.
* A global define is added so that we can choose at build time if we
want to use 32-bit or 64-bit base structures.
* A first optional parameter is added so the user can provide the
expected blob size of the generated binary.
* A second optional parameter is added so the user can specify an output
file to which the generated output will be printed.
The default behavior isn't modified.
Acked-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Sebastien Boeuf <sebastien.boeuf@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
No need to check the interface protocol then conditionally setting,
just set it to BOOT_PROTOCOL and check for error.
This is what Linux does for HID devices as some don't follow the USB spec.
One example is the Aspeed BMC HID keyboard device, which adds a massive
boot delay without this patch as it doesn't respond to
'GetProtocolRequest'.
Cc: Hao A Wu <hao.a.wu@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Signed-off-by: Matt DeVillier <matt.devillier@gmail.com>
Signed-off-by: Patrick Rudolph <patrick.rudolph@9elements.com>
Signed-off-by: Sean Rhodes <sean@starlabs.systems>
Reviewed-by: Hao A Wu <hao.a.wu@intel.com>
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=3818
It will have some potential issue when memory larger than 2G because
the high memory address will be fill with 0xFFFFFFFF when do the
operation of INTN + INT64 but it is 32 bit normal data in fact.
Should use same data type INT64 + INT64.
V3:
1. Use INT64 as input parameter because all date type is 64 bit
V2:
1. Force the data type to UINTN to avoid high dword be filled with
0xFFFFFFFF
2. Keep INTN because the offset may postive or negative.
Reviewed-by: Guo Dong <guo.dong@intel.com>
Reviewed-by: Ray Ni <ray.ni@intel.com>
Signed-off-by: Guomin Jiang <guomin.jiang@intel.com>
Move the logic that stores starting PCI attributes and sets the
EFI_PCI_IO_ATTRIBUTE_DUAL_ADDRESS_CYCLE attribute to
DriverBindingStart() before the memory that backs the
DMA engine is allocated.
This ensures that the DMA-backing memory is not forcibly allocated
below 4G in system address map. Otherwise the allocation fails on
platforms that do not have any memory below the 4G mark and the drive
initialisation fails.
Leave the PCI device enabling attribute logic in NvmeControllerInit()
to ensure that the device is re-enabled on reset in case it was
disabled via PCI attributes.
Cc: Ray Ni <ray.ni@intel.com>
Cc: Leif Lindholm <quic_llindhol@quicinc.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Tomas Pilar <quic_tpilar@quicinc.com>
Acked-by: Hao A Wu <hao.a.wu@intel.com>
The SNP patch series updated the OvmfPkgX64 build but forgot the AmdSev
variant, resulting in a broken OvmfSevMetadata table.
Fixes: cca9cd3dd6 ("OvmfPkg: reserve CPUID page")
Fixes: 707c71a01b ("OvmfPkg: reserve SNP secrets page")
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Reviewed-by: Brijesh Singh <brijesh.singh@amd.com>
Acked-by: Jiewen Yao <jiewen.yao@intel.com>
Use the correct PCD type for PcdPlatformBootTimeOut so it gets wired up
to the Timeout EFI variable automatically, which is how the boot manager
stores the timeout preference.
Note that this changes the default to 5 seconds, which appears to be
common across platforms.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Pierre Gondois <pierre.gondois@arm.com>
Reviewed-by: Sami Mujawar <sami.mujawar@arm.com>
Alex reports that the cache invalidation performed by
ArmVirtMemoryInitPeiLib takes a non-negligible amount of time at boot.
This cache invalidation used to be necessary to avoid inconsistencies
between the CPU's cached and uncached views of the permanent PEI memory
region, given that the PEI phase is where the MMU gets enabled.
The only allocations done from permanent PEI memory with the MMU off are
pages used for page tables, and since commit 748fea6279
("ArmPkg/ArmMmuLib AARCH64: invalidate page tables before populating
them"), each of those is invalidated in the caches explicitly, for
reasons described in the patch's commit log. All other allocations done
in PEI are either from temporary PEI memory, which includes the stack,
or from permanent PEI memory but after the MMU has been enabled.
This means that the cache invalidation in ArmVirtMemoryInitPeiLib is no
longer necessary, and can simply be dropped.
Reported-by: Alexander Graf <graf@amazon.com>
Reviewed-by: Leif Lindholm <leif@nuviainc.com>
Reviewed-by: Alexander Graf <graf@amazon.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=3835
The commit ade62c18f4 caused a boot failure
when OVMF is build with SECURE_BOOT/SMM enabled.
This happen because the above commit extended the BaseMemEncryptSevLib.inf
to include VmgExitLib. The FvbServicesSmm uses the functions provided
by the MemEncryptSevLib to clear the memory encryption mask from the
page table. It created a dependency, as shown below
OvmfPkg/FvbServicesSmm.inf
---> MemEncryptSevLib class
---> "OvmfPkg/BaseMemEncryptSevLib/DxeMemEncryptSevLib.inf" instance
---> VmgExitLib
---> "OvmfPkg/VmgExitLib" instance
---> LocalApicLib class
---> UefiCpuPkg/BaseXApicX2ApicLib/BaseXApicX2ApicLib.inf instance
---> TimerLib class
---> "OvmfPkg/AcpiTimerLib/DxeAcpiTimerLib.inf" instance
---> PciLib class
---> "OvmfPkg/DxePciLibI440FxQ35/DxePciLibI440FxQ35.inf" instance
The LocalApicLib provides a constructor, execution of the constructor
causes an exception. The SEV-ES and SEV-SNP do not support the SMM, so
skip including the VmgExitLib chain. Use the module override to use the
VmgExitLibNull to avoid the inclusion of unneeded LocalApicLib dependency
chain in FvbServicesSmm. We ran similar issue for AmdSevDxe driver,
see commit 19914edc5a
After the patch, the dependency look like this:
OvmfPkg/FvbServicesSmm.inf
---> MemEncryptSevLib class
---> "OvmfPkg/BaseMemEncryptSevLib/DxeMemEncryptSevLib.inf" instance
---> VmgExitLib
---> "UefiCpuPkg/Library/VmgExitLibNull" instance
Fixes: ade62c18f4
Reported-by: Aaron Young <aaron.young@oracle.com>
Cc: Dann Frazier <dann.frazier@canonical.com>
Cc: Michael Roth <michael.roth@amd.com>
Cc: James Bottomley <jejb@linux.ibm.com>
Cc: Min Xu <min.m.xu@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Ard Biesheuvel <ardb+tianocore@kernel.org>
Cc: Erdem Aktas <erdemaktas@google.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
Tested-by: dann frazier <dann.frazier@canonical.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
XCODE5 reported the following warning:
OvmfPkg/Library/VmgExitLib/VmgExitVcHandler.c:1895:12: note:
uninitialized use occurs here
Compacted
^^^^^^^^^
Initialize the 'Compacted' variable to fix the warning.
Signed-off-by: Rebecca Cran <quic_rcran@quicinc.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>
Building with the CLANG35 and CLANG38 toolset fails because of variables
which are set but not otherwise used in the RELEASE build.
GCC added -Wno-unused-but-set-variable back in 2016, and later added
-Wno-unused-const-variable. Add those to CLANG35_WARNING_OVERRIDES and
CLANG38_WARNING_OVERRIDES.
Signed-off-by: Rebecca Cran <quic_rcran@quicinc.com>
Reviewed-by: Michael D Kinney <michael.d.kinney@intel.com>
