Introduce PcdXenGrantFrames to replace a define in XenBusDxe and allow
the same value to be used in a different module.
The reason for the number of page to be 4 doesn't exist anymore, so
simply remove the comment.
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-33-anthony.perard@citrix.com>
On a Xen PVH guest, none of the existing serial or console interface
works, so we add a new one, based on XenConsoleSerialPortLib, and
implemented via SerialDxe.
That is a simple console implementation that can work on both PVH
guest and HVM guests, even if it is rarely going to be used on HVM.
Have PlatformBootManagerLib look for the new console, when running as a
Xen guest.
Since we use VENDOR_UART_DEVICE_PATH, fix its description and coding
style.
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1689
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-32-anthony.perard@citrix.com>
When running as a Xen PVH guest, there is no CMOS to read the memory
size from. Rework GetSystemMemorySize(Below|Above)4gb() so they can
work without CMOS by reading the e820 table.
Rework XenPublishRamRegions to also care for the reserved and ACPI
entry in the e820 table. The region that was added by InitializeXen()
isn't needed as that same entry is in the e820 table provided by
hvmloader.
MTRR settings aren't modified anymore, on HVM it's already done by
hvmloader, on PVH it is supposed to have sane default. MTRR will need
to be done properly but keeping what's already been done by programs
that have run before OVMF will do for now.
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1689
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-24-anthony.perard@citrix.com>
The ACPI Timer isn't present in a PVH guest, but local APIC works on
both PVH and HVM.
Note that the use of SecPeiDxeTimerLibCpu might be an issue with a
driver of type DXE_RUNTIME_DRIVER. I've attempted to find out which of
the DXE_RUNTIME_DRIVER uses the TimerLib at runtime. I've done that by
replacing the TimerLib evaluation in
[LibraryClasses.common.DXE_RUNTIME_DRIVER] by a different one and
checking every module that uses it (with the --report-file=report
build option).
ResetSystemRuntimeDxe is calling the TimerLib API at runtime to do the
operation "EfiResetCold", so this may never complete if the OS have
disabled the Local APIC Timer.
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1689
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-10-anthony.perard@citrix.com>
This patch allows the ResetVector to be run indenpendently from build
time addresses.
The goal of the patch is to avoid having to create RAM just below 4G
when creating a Xen PVH guest while being compatible with the way
hvmloader currently load OVMF, just below 4G.
Only the new PVH entry point will do the calculation.
The ResetVector will figure out its current running address by creating
a temporary stack, make a call and calculate the difference between the
build time address and the address at run time.
This patch copies and make the necessary modification to some other asm
files:
- copy of UefiCpuPkg/.../Flat32ToFlat64.asm:
Allow Transition32FlatTo64Flat to be run from anywhere in memory
- copy of UefiCpuPkg/../SearchForBfvBase.asm:
Add a extra parameter to indicate where to start the search for the
boot firmware volume.
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1689
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-9-anthony.perard@citrix.com>
This patch changes the flash device image of OvmfXen to make it look
like it's an ELF. For this, we replace the empty embedded variable store
by a binary array, which is a ELF file header.
The ELF header explain to a loader to load the binary at the address
1MB, then jump to the PVH entry point which will be created in a later
patch. The header also includes a Xen ELF note that is part of the
PVH ABI.
That patch include OvmfXenElfHeaderGenerator.c which can be use to
regenerate the ELF header, but this will be a manual step.
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1689
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-6-anthony.perard@citrix.com>
Introduce XenPlatformPei, a copy of OvmfPkg/PlatformPei without some
of QEMU specific initialization, Xen does not support QemuFwCfg.
This new module will be adjusted to accommodate Xen PVH.
fw_cfg dependents that have been removed, which are dynamically skipped
when running PlatformPei on Xen:
- GetFirstNonAddress(): controlling the 64-bit PCI MMIO aperture via the
(experimental) "opt/ovmf/X-PciMmio64Mb" file
- GetFirstNonAddress(): honoring the hotplug DIMM area
("etc/reserved-memory-end") in the placement of the 64-bit PCI MMIO
aperture
- NoexecDxeInitialization() is removed, so PcdPropertiesTableEnable and
PcdSetNxForStack are left constant FALSE (not set dynamically from
fw_cfg "opt/ovmf/PcdXxxx")
- MaxCpuCountInitialization(), PublishPeiMemory(): the max CPU count is
not taken from the QemuFwCfgItemSmpCpuCount fw_cfg key;
PcdCpuMaxLogicalProcessorNumber is used intact and
PcdCpuApInitTimeOutInMicroSeconds is never changed or used.
- InitializeXenPlatform(), S3Verification(): S3 is assumed disabled (not
consulting "etc/system-states" via QemuFwCfgS3Enabled()).
- InstallFeatureControlCallback(): the feature control MSR is not set
from "etc/msr_feature_control"
(also removed FeatureControl.c as there is nothing been executed)
Also removed:
- SMRAM/TSEG-related low mem size adjusting (PcdSmmSmramRequire is
assumed FALSE) in PublishPeiMemory(),
- QemuInitializeRam() entirely,
Xen related changes:
- Have removed the module variable mXen, as it should be always true.
- Have the platform PEI initialization fails if Xen has not been
detected.
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1689
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-5-anthony.perard@citrix.com>
Introduce XenResetVector, a copy of OvmfPkg/ResetVector, with one
changes:
- SEC_DEFAULT_CR0: enable cache (bit 30 or CD set to 0)
Xen copies the OVMF code to RAM, there is no need to disable cache.
This new module will later be modified to add a new entry point, more
detail in a following commit "OvmfPkg/XenResetVector: Add new entry point
for Xen PVH"
Value FILE_GUID of XenResetVector have not changed compare to ResetVector
because it is a special value (gEfiFirmwareVolumeTopFileGuid).
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1689
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-4-anthony.perard@citrix.com>
OvmfXen is a copy of OvmfX64, removing VirtIO and some SMM.
This new platform will be changed to make it works on two types of Xen
guest: HVM and PVH.
Compare to OvmfX64, this patch:
- changed: PLATFORM_GUID, OUTPUT_DIRECTORY, FLASH_DEFINITION
- removed: VirtioLib class resolution
- removed: all UEFI_DRIVER modules for virtio devices
- removed: DXE_SMM_DRIVER and SMM_CORE lib class resolutions
- removed: DXE_SMM_DRIVER and SMM_CORE FDF rules
- removed: Everything related to SMM_REQUIRE==true
- removed: Everything related to SECURE_BOOT_ENABLE==true
- removed: Everything related to TPM2_ENABLE==true
- changed: PcdPciDisableBusEnumeration dynamic default flipped to TRUE
- changed: default FD_SIZE_IN_KB to 2M.
- reverted d272449d9e, "OvmfPkg: raise DXEFV size to 11 MB"
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1689
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190813113119.14804-3-anthony.perard@citrix.com>
Specify the firmware to use via the newer '-drive if=pflash' syntax
which allows specifying the raw format parameter. This
avoids warnings with newer version of QEMU.
Signed-off-by: Rebecca Cran <rebecca@bsdio.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Philippe Mathieu-Daude <philmd@redhat.com>
$ADD_QEMU_HDA was added because QEMU used to refuse to run without a
disk. Since newer versions run without any disks, remove it.
Signed-off-by: Rebecca Cran <rebecca@bsdio.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Philippe Mathieu-Daude <philmd@redhat.com>
Enable multithreaded builds by default when building OvmfPkg
using build.sh.
This can drastically reduce build times. For example, on a
modern ThreadRipper system the time required to build decreases
from 3 minutes to 1 minute.
Signed-off-by: Rebecca Cran <rebecca@bsdio.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
(a) OvmfPkg first had to resolve the TpmMeasurementLib class -- for
SECURE_BOOT_ENABLE only -- when the DxeImageVerificationLib instance
became dependent on TpmMeasurementLib. For details, refer to commit
0d28d286bf ("OvmfPkg: resolve TpmMeasurementLib dependency
introduced in r14687", 2013-09-21).
(b) At the time, only one instance of TpmMeasurementLib existed, namely
DxeTpmMeasurementLib. This lib instance didn't do anything -- like it
was desirable for OVMF --, because OVMF didn't include any Tcg / TrEE
protocol implementations.
(c) In commit 308521b133 ("MdeModulePkg: Move TpmMeasurementLib
LibraryClass from SecurityPkg", 2015-07-01), TpmMeasurementLibNull was
introduced.
(d) In commit 285542ebbb ("OvmfPkg: Link AuthVariableLib for following
merged variable driver deploy", 2015-07-01), a TpmMeasurementLib
resolution became necessary regardless of SECURE_BOOT_ENABLE. And so
TpmMeasurementLib was resolved to TpmMeasurementLibNull in OVMF, but
only in the non-SECURE_BOOT_ENABLE case. This step -- possibly, the
larger series containing commit 285542ebbb -- missed an opportunity
for simplification: given (b), the DxeTpmMeasurementLib instance
should have been simply replaced with the TpmMeasurementLibNull
instance, regardless of SECURE_BOOT_ENABLE.
(e) In commit 1abfa4ce48 ("Add TPM2 support defined in trusted computing
group.", 2015-08-13), the TrEE dependency was replaced with a Tcg2
dependency in DxeTpmMeasurementLib.
(f) Starting with commit 0c0a50d6b3 ("OvmfPkg: include Tcg2Dxe module",
2018-03-09), OVMF would include a Tcg2 protocol implementation,
thereby satisfying DxeTpmMeasurementLib's dependency. With
TPM2_ENABLE, it would actually make sense to consume
DxeTpmMeasurementLib -- however, DxeTpmMeasurementLib would never be
used without SECURE_BOOT_ENABLE.
Therefore, we have the following four scenarios:
- TPM2_ENABLE + SECURE_BOOT_ENABLE: works as expected.
- Neither enabled: works as expected.
- Only TPM2_ENABLE: this build is currently incorrect, because
Variable/RuntimeDxe consumes TpmMeasurementLib directly, but
TpmMeasureAndLogData() will never reach the TPM because we link
TpmMeasurementLibNull into the variable driver. This is a problem from
the larger series containing (f).
- Only SECURE_BOOT_ENABLE: this build works as expected, but it is
wasteful -- given that the protocol database will never contain Tcg2
without TPM2_ENABLE, we should simply use TpmMeasurementLibNull. This is
a problem from (d).
Resolving TpmMeasurementLib to DxeTpmMeasurementLib as a function of
*only* TPM2_ENABLE, we can fix / optimize the last two cases.
v2:
- Amend the title and description suggested by Laszlo
- Move TpmMeasurementLib to the existed TPM2_ENABLE block
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Marc-André Lureau <marcandre.lureau@redhat.com>
Cc: Stefan Berger <stefanb@linux.ibm.com>
Signed-off-by: Gary Lin <glin@suse.com>
Message-Id: <20190704040731.5303-1-glin@suse.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Calling DisconnectController() on children isn't part of the job of
EFI_DRIVER_BINDING_PROTOCOL.Stop() as it only needs to deallocate
resources allocated in Start(). The disconnection will happen when
both DevicePath and XenBus protocols gets uninstalled.
Reported-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Message-Id: <20190701111403.7007-1-anthony.perard@citrix.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
In XenBusDxe, the XenBusAddDevice() opens the gXenIoProtocolGuid on
behalf of child controllers. It is never closed and prevents us from
uninstalling the protocol.
Close it where we stop all the children in XenBusDxe->Stop().
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190701105012.25758-1-anthony.perard@citrix.com>
Mostly, this is only necessary for devices that the CSM might have
native support for, such as VirtIO and NVMe; PciBusDxe will already
degrade devices to 32-bit if they have an OpROM.
However, there doesn't seem to be a generic way of requesting PciBusDxe
to downgrade specific devices.
There's IncompatiblePciDeviceSupportProtocol but that doesn't provide
the PCI class information or a handle to the device itself, so there's
no simple way to just match on all NVMe devices, for example.
Just leave gUefiOvmfPkgTokenSpaceGuid.PcdPciMmio64Size set to zero for
CSM builds, until/unless that can be fixed.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190626113742.819933-5-dwmw2@infradead.org>
QemuVideoDxe installs its own legacy INT 10h handler for the benefit of
systems like Windows 2008r2 which attempt to use INT 10h even when booted
via EFI.
This interacts extremely badly with a CSM actually attempting to install
a real video BIOS.
The last thing done before invoking a legacy OpROM is to call INT 10h to
set a plain text mode. In the case where it's the video BIOS OpROM being
loaded, INT 10h will normally point to an iret stub in the CSM itself.
Unless QemuVideoDxe has changed INT10h to point to a location in the
0xC0000 segment that it didn't allocate properly, so the real OpROM has
been shadowed over them top of it, and the INT 10h vector now points to
some random place in the middle of the newly-shadowed OpROM.
Don't Do That Then. QemuVideoDxe doesn't do any acceleration and just
sets up a linear framebuffer, so we don't lose much by just
unconditionally using BiosVideoDxe instead when CSM is present.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190626113742.819933-4-dwmw2@infradead.org>
Iterate over the available block devices in much the same way as
BdsLibEnumerateAllBootOption() does, but limiting to those devices
which are PCI-backed, which can be represented in the BbsTable.
One day we might need to extend the BbsTable to allow us to distinguish
between different NVMe namespaces on a device.
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20190626113742.819933-3-dwmw2@infradead.org>
