In some cases, (e.g., when running QEMU with TrustZone emulation), the
DT may contain DT nodes whose status is set to 'secure'. Similarly, the
status may be set to 'disabled' if the consumer of the DT image is
expected to treat it as if it weren't there.
So check whether a 'status' property is present, and if so, ignore the
node if the status is not 'okay'.
In some cases, (e.g., when running QEMU with TrustZone emulation), the
DT may contain DT nodes whose status is set to 'secure'. Similarly, the
status may be set to 'disabled' if the consumer of the DT image is
expected to treat it as if it weren't there.
So check whether a 'status' property is present, and if so, ignore the
node if the status is not 'okay'.
In some cases, (e.g., when running QEMU with TrustZone emulation), the
DT may contain memory nodes whose status is set to 'secure'. Similarly,
the status may be set to 'disabled' if the consumer of the DT image is
expected to treat it as if it weren't there.
So check whether a 'status' property is present, and if so, ignore the
node if the status is not 'okay'.
Jeff Fan [Tue, 28 Mar 2017 06:01:24 +0000 (14:01 +0800)]
UefiCpuPkg/PiSmmCpuDxeSmm: Update saved SMM ranges check in SmmProfile
SmmProfile feature required to protect all SMM ranges by structure
mProtectionMemRangeTemplate. This update is to add additonal save SMM ranges
into mProtectionMemRangeTemplate besides the range specified by
mCpuHotPlugData.SmrrBase/mCpuHotPlugData.SmrrSiz.
Cc: Jiewen Yao <jiewen.yao@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Cc: Feng Tian <feng.tian@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>
Jeff Fan [Tue, 28 Mar 2017 05:51:52 +0000 (13:51 +0800)]
UefiCpuPkg/PiSmmCpuDxeSmm: Add IsInSmmRanges() to check SMM range
Internal function IsInSmmRanges() is added t check SMM range by saved SMM ranges
beside by mCpuHotPlugData.SmrrBase/mCpuHotPlugData.SmrrSiz.
Cc: Jiewen Yao <jiewen.yao@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Cc: Feng Tian <feng.tian@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>
Jeff Fan [Tue, 28 Mar 2017 00:48:17 +0000 (08:48 +0800)]
UefiCpuPkg/PiSmmCpuDxeSmm: Save SMM ranges info into global variables
v2:
Add #define SMRR_MAX_ADDRESS to clarify SMRR requirement.
Cc: Jiewen Yao <jiewen.yao@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com> Cc: Feng Tian <feng.tian@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>
Chen A Chen [Wed, 29 Mar 2017 02:23:09 +0000 (10:23 +0800)]
ShellPkg/setvar: Support data format in Shell 2.2 spec
Shell 2.2 spec defines =0x/=0X, =H/=h, =S, =L and =P for
hex number, hex array, ascii string, unicode string and
device path data.
The patch adds such support.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Chen A Chen <chen.a.chen@intel.com> Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com> Cc: Michael D Kinney <michael.d.kinney@intel.com> Cc: Jaben Carsey <jaben.carsey@intel.com> Cc: Jeff Fan <jeff.fan@intel.com>
Ard Biesheuvel [Wed, 29 Mar 2017 11:05:01 +0000 (12:05 +0100)]
EmbeddedPkg/DtPlatformDxe: load platform DTB via new library
To give platforms some room to decide which DTB is suitable and where
to load it from, load the DTB image indirectly via the new
DtPlatformDtbLoaderLib library class.
Ard Biesheuvel [Fri, 31 Mar 2017 10:35:13 +0000 (11:35 +0100)]
EmbeddedPkg: add base DtPlatformDtbLoaderLib implementation
Introduce an implementation of the new DtPlatformDtbLoaderLib library
class that simply retrieves the first raw section of an FV file named
'gDtPlatformDefaultDtbFileGuid'.
Ard Biesheuvel [Fri, 31 Mar 2017 10:32:36 +0000 (11:32 +0100)]
EmbeddedPkg: add DtPlatformDtbLoaderLib library class
To abstract the way a platform reasons about which DTB is appropriate,
and the way it ultimately supplies the DTB image, introduce a new library
class to encapsulate this functionality.
In general, we should not present two separate (and inevitably different)
hardware descriptions to the OS, in the form of ACPI tables and a device
tree blob. For this reason, we recently added the logic to ArmVirtQemu to
only expose the ACPI 2.0 entry point if no DT binary is being passed, and
vice versa.
However, this is arguably a regression for those who relied on DT
descriptions being available, even if the former behavior can be
restored by passing the -no-acpi switch to QEMU.
So allow a secret handshake with the UEFI Shell, to set a variable that
will result in ACPI to be disabled on subsequent boots even if -no-acpi
was not passed on the QEMU command line.
Recent changes to the PlatformBootManagerLib implementation in ArmPkg
require its users to define a resolution for BootLogoLib. So add this
missing resolution to Beagle.
Ard Biesheuvel [Fri, 31 Mar 2017 08:22:56 +0000 (09:22 +0100)]
BaseTools/GCC AARCH64: force disable PIC code generation
As a security measure, some distro toolchains now default to PIC code
generation, allowing executables (as opposed to shared libraries) using
the objects to be built as PIE binaries, which can be loaded at a random
virtual offset.
However, our ELF to PE/COFF generation code does not deal with the
resulting relocation types (i.e., GOT based), and so the use of PIC code
leads to GenFw errors.
Given that
a) our non-PIC PE/COFF executables are already relocatable,
b) PIC code leads to all symbol references to be indirected via GOT
entries containing absolute addresses, each requiring an entry in the
relocation table,
c) the AArch64 ISA makes it perfectly feasible to built PIE executables
from non-PIC code,
there is absolutely no upside to using PIC code for building EDK2 modules,
and so we're better off simply disabling it unconditionally.
Note that when running under the OS, the GOT has an additional advantage,
i.e., that all .text/.rodata pages remain clean and so can be shared between
processes. This does not apply to the UEFI environment, however.
Ruiyu Ni [Fri, 2 Sep 2016 04:09:54 +0000 (12:09 +0800)]
UefiCpuPkg/MtrrLib: Use a better algorithm to calculate MTRR
The new algorithm finds out the more optimal MTRR solution for
current memory type settings.
Compare against the original algorithm, the new one guarantees
to find the correct MTRR solution, but doesn't guarantee to
find the most optimal MTRR solution.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Jiewen Yao [Mon, 27 Mar 2017 15:01:02 +0000 (23:01 +0800)]
MdeModulePkg/SmmCore: Fix memory leak on Profile unregistered.
Issue reported at bugzillar 445.
Cc: Jeff Fan <jeff.fan@intel.com> Cc: Feng Tian <feng.tian@intel.com> Cc: Star Zeng <star.zeng@intel.com> Cc: Bret Barkelew <Bret.Barkelew@microsoft.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Jiewen Yao <jiewen.yao@intel.com> Reviewed-by: Jeff Fan <jeff.fan@intel.com>
OvmfPkg: Allow multiple add-pointer linker commands to same ACPI table
ACPI tables may contain multiple fields which point to the same
destination table. For example, in some revisions, the FADT contains
both DSDT and X_DSDT fields, and they may both point to the DSDT.
Previously, if Qemu created QEMU_LOADER_ADD_POINTER linker commands for
such instances, the linking process would attempt to install the same
pointed-to table repeatedly. For tables of which there must only be one
instance, the call to AcpiProtocol->InstallAcpiTable() would fail during
the second linker command pointing to the same table, thus entirely
aborting the ACPI table linking process. In the case of tables of which
there may be multiple instances, the table would end up duplicated.
This change adds a memoisation data structure which tracks the table
pointers that have already been processed; even if the same pointer is
encountered multiple times, it is only processed once.
Qin Long [Thu, 30 Mar 2017 07:53:21 +0000 (15:53 +0800)]
CryptoPkg/BaseCryptLib: Fix Build Warning issue in PEI Module
The memory free operation is empty function in PEI. The compiler
optimization will bring the build warning in openssl/crypto/mem.c:
warning C4718: 'CRYPTO_free': recursive call has no side
effects, deleting
This patch uses '/wd4718' to silence the build warning for PEI
module building.
Cc: Ting Ye <ting.ye@intel.com> Cc: Eric Dong <eric.dong@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: Eric Dong <eric.dong@intel.com> Reviewed-by: Ye Ting <ting.ye@intel.com>
Jiaxin Wu [Fri, 24 Mar 2017 01:53:50 +0000 (09:53 +0800)]
MdePkg/UefiDevicePathLib: Refine the DevPathFromTextiSCSI protocol parsing
For current iSCSI protocol parsing, UINT16 truncation may be happened. Since
the Spec already have declaimed that 0 is TCP Protocol and 1+ is reserved, the
parsing can be refined as below:
Jiaxin Wu [Thu, 23 Mar 2017 03:35:14 +0000 (11:35 +0800)]
NetworkPkg/DnsDxe: Fix zero StationIp configuration failure of DNSv6
According UEFI Spec, set to zero StationIp means to let the underlying
IPv6 driver choose a source address. But currently, DNSv6 always return
EFI_NO_MAPPING. The issue is caused by below bugs in DnsDxe:
* Incorrect TPL(TPL_CALLBACK) usage during UDP configuration.
* Failed to create the timer used to get IPv6 mapping
* Doesn't check the Ip6Mode.IsStarted flag.
Ard Biesheuvel [Mon, 27 Mar 2017 12:22:16 +0000 (13:22 +0100)]
MdeModulePkg/PeiCore: avoid EFI_IMAGE_MACHINE_TYPE_SUPPORTED to check arch
The EFI_IMAGE_MACHINE_TYPE_SUPPORTED() macro is abused in the PeiCore
code to decide whether the system we are compiling for can deal with
executable code being copied elsewhere and executed from there.
As stated in the comment, this is fundamentally a property of the compiler
target, and so this should be made dependent on MDE_CPU_xxx preprocessor
defines, and not on whether or not the runtime target can deal with
PE/COFF images of a certain machine type.
On X86/IA32, this mostly boils down to the same thing, but not on other
architectures, so let's clean this up.
Qin Long [Tue, 21 Mar 2017 14:58:07 +0000 (22:58 +0800)]
CryptoPkg: Update PK Cipher Wrappers work with opaque objects.
OpenSSL-1.1.xx makes most data structures opaque.
This patch updates Public Key Cipher Wrapper implementations in
BaseCryptLib to use the accessor APIs for opaque object access.
The impacted interfaces includes RSA, DH, X509, PKCS7, etc.
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com> Tested-by: Laszlo Ersek <lersek@redhat.com> Tested-by: Gary Lin <glin@suse.com>
Qin Long [Thu, 23 Mar 2017 12:47:34 +0000 (20:47 +0800)]
CryptoPkg: Update HMAC Wrapper with opaque HMAC_CTX object.
OpenSSL-1.1.xx makes most data structures opaque.
This patch updated HMAC Wrapper implementation with opaque
HMAC_CTX object.
The HmacXXGetContextSize() is marked as deprecated, and updated
to use the fixed HMAC_CTX size, which is just kept for compatibility.
New APIs (HmacXXNew(), HmacXXFree()) were added as the recommended
HMAC_CTX usage interfaces for HMAC-XXXX operations.
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com> Tested-by: Laszlo Ersek <lersek@redhat.com>
Qin Long [Thu, 23 Mar 2017 12:26:45 +0000 (20:26 +0800)]
CryptoPkg: Add extra build option to disable VS build warning
openssl/include/openssl/lhash.h will bring C4090 build warning
issue, which is one known issue for OpenSSL under Visual Studio
toolchain.
Refer to https://github.com/openssl/openssl/issues/2214 for more
discussions against this.
Use /wd4090 to silence this build warning until OpenSSL fix this.
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com>
Qin Long [Thu, 23 Mar 2017 12:19:39 +0000 (20:19 +0800)]
CryptoPkg: Clean-up CRT Library Wrapper.
Cleaning-up CRT Library Wrapper for the third-party cryptography
library building. The changes includes
1. Rename OpenSslSupport.h to CrtLibSupport.h for future alternative
crypto provider support.
2. Remove all un-referenced CRT APIs and headers.
(NOTE: More cleans-up could be possible after OpenSSL integrate the
extra PR request: https://github.com/openssl/openssl/pull/2961)
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com> Tested-by: Laszlo Ersek <lersek@redhat.com> Tested-by: Gary Lin <glin@suse.com>
Qin Long [Thu, 23 Mar 2017 12:09:17 +0000 (20:09 +0800)]
CryptoPkg: Fix handling of &strcmp function pointers
In a couple of places, OpenSSL code uses the address of the
strcmp() function, and assigns it to another comparator function
pointer.
Unfortunately, this falls foul of the inconsistent function ABI
that we use in EDKII. We '#define strcmp AsciiStrCmp' but AsciiStrCmp
is an EFIAPI function with the Microsoft ABI. And we're assigning its
address to a non-EFIAPI function, which may well have a different ABI.
Fix this by providing an actual strcmp() function in the default ABI.
We already *had* a prototype for it in OpenSslSupport.h, which was
then superseded by the #define strcmp AsciiStrCmp.
Now, OpenSSL code *can* use &strcmp without problems.
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Tested-by: Laszlo Ersek <lersek@redhat.com> Tested-by: Gary Lin <glin@suse.com>
Qin Long [Thu, 23 Mar 2017 12:06:00 +0000 (20:06 +0800)]
CryptoPkg/OpensslLib: Add new OpenSSL-HOWTO document.
Add one new OpenSSL-HOWTO.txt to introduce how to clone / download
the latest OpenSSL release source for build.
ALso update buildinf.h to reflect the latest update time.
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Tested-by: Laszlo Ersek <lersek@redhat.com> Acked-by: Gary Lin <glin@suse.com> Tested-by: Gary Lin <glin@suse.com>
Qin Long [Thu, 23 Mar 2017 11:37:00 +0000 (19:37 +0800)]
CryptoPkg/OpensslLib: Add new Perl script for file list generation.
OpenSSL-1.1.0xx configure mechanism was updated with new configdata.
This patch update process_file.sh script to new Perl-based script for
auto generation of file list and openssl config file (opensslconf.h).
This only needs to be done once by a developer when updating to a new
version of OpenSSL (or changing options, etc.). Normal users do not
need to do this, since the results are already stored in the EDK2 git
repository.
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com> Acked-by: Laszlo Ersek <lersek@redhat.com>
Qin Long [Thu, 23 Mar 2017 08:21:31 +0000 (16:21 +0800)]
CryptoPkg: Update .gitignore for OpenSSL source masking
Updates .gitignore that masks the OpenSSL source:
1. Remove "Include/openssl" from .gitignore since we needn't duplicate
openssl headers now
2. Update "openssl-*" to "openssl*", since we use "openssl" instead of
"openssl-x.x.xx" as main source directory.
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Tested-by: Laszlo Ersek <lersek@redhat.com>
Qin Long [Thu, 23 Mar 2017 08:14:16 +0000 (16:14 +0800)]
CryptoPkg/OpensslLib: Update INF files to support OpenSSL-1.1.0x build
Update OpensslLib INF files to support OpenSSL-1.1.0x source build.
The file list was generated from the latest OpenSSL-1.1.0e release.
Main changes to support OpensslLib build in this patch include:
1. Use "openssl" instead of "openssl-x.x.xx" as main source directory,
Also update include path in CryptoPkg.dec
2. Enable warnings in GCC builds;
3. Update Visual Studio build options to silence current possible build
warnings.
4. Move the default opensslconf.h to Include/openssl, and add one dummy
dso_conf.h for native UEFI build.
The OpensslLib module build was validated as
build -t VSXXXX -a XX -p CryptoPkg/CryptoPkg.dsc
-m CryptoPkg/Library/OpensslLib/OpensslLib.inf
(NOTE: The extra build options for ARM/RVCT/XCODE were kept, which expect
further optimizations from community)
Cc: Ting Ye <ting.ye@intel.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Gary Lin <glin@suse.com> Cc: Ronald Cron <ronald.cron@arm.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: David Woodhouse <dwmw2@infradead.org> Signed-off-by: Qin Long <qin.long@intel.com> Reviewed-by: Ting Ye <ting.ye@intel.com> Acked-by: Laszlo Ersek <lersek@redhat.com> Tested-by: Laszlo Ersek <lersek@redhat.com> Tested-by: Gary Lin <glin@suse.com>
Ard Biesheuvel [Mon, 27 Mar 2017 09:57:31 +0000 (10:57 +0100)]
EmbeddedPkg: add DT platform driver to select between DT and ACPI
As a follow up to the changes proposed by Laszlo to make ACPI and DT
mutually exclusive on ArmVirtQemu, this patch proposes a DT platform
DXE driver that either installs the NULL protocol PlatformHasAcpiGuid,
or installs the FV embedded DTB binary as a configuration table under
the appropriate GUID, depending on a preference setting recorded as
a UEFI variable, and configurable via a HII screen.
The DTB binary can be embedded in the firmware image by adding the
following to the platform .fdf file:
FILE FREEFORM = 25462CDA-221F-47DF-AC1D-259CFAA4E326 {
SECTION RAW = SomePkg/path/to/foo.dtb
}
Laszlo Ersek [Fri, 17 Mar 2017 16:50:24 +0000 (17:50 +0100)]
ArmVirtPkg/PlatformHasAcpiDtDxe: don't expose DT if QEMU provides ACPI
This will let QEMU's "-no-acpi" option exclusively expose DT vs. ACPI to
the guest. Showing both is never needed (it is actually detrimental to the
adoption of standards, such as SBSA / SBBR).
* Without "-no-acpi", the firmware logs (from PlatformHasAcpiDtDxe)
> Found FwCfg @ 0x9020008/0x9020000
> Found FwCfg DMA @ 0x9020010
> InstallProtocolInterface: [EdkiiPlatformHasAcpi] 0
plus the usual messages. Later the guest kernel logs
In addition, in the guest, the "/sys/firmware/devicetree/*" shell pattern
matches no files, while the "/sys/firmware/acpi/tables/*" pattern matches
the ACPI tables.
* With "-no-acpi", the firmware logs:
> PlatformHasAcpiDtDxe | Found FwCfg @ 0x9020008/0x9020000
> PlatformHasAcpiDtDxe | Found FwCfg DMA @ 0x9020010
> PlatformHasAcpiDtDxe | InstallProtocolInterface:
> PlatformHasAcpiDtDxe | [EdkiiPlatformHasDeviceTree] 0
> FdtClientDxe | OnPlatformHasDeviceTree: exposing DTB @
> FdtClientDxe | 0x13FFBF000 to OS
> ...
> DXE_CORE | Driver [AcpiTableDxe] was discovered but not
> DXE_CORE | loaded!!
> DXE_CORE | Driver [QemuFwCfgAcpiPlatform] was discovered but
> DXE_CORE | not loaded!!
> ...
> RamDiskDxe | RamDiskAcpiCheck: Cannot locate the EFI ACPI
> RamDiskDxe | Table Protocol, unable to publish RAM disks to
> RamDiskDxe | NFIT.
(BootGraphicsResourceTableDxe's ReadyToBoot callback --
InstallBootGraphicsResourceTable() -- handles the lack of
EFI_ACPI_TABLE_PROTOCOL silently.) Later the guest kernel logs
In addition, in the guest, the "/sys/firmware/devicetree/*" shell pattern
matches the directory "/sys/firmware/devicetree/base", which contains a
large number of DT nodes, while the "/sys/firmware/acpi/tables/*" pattern
matches no files.
Laszlo Ersek [Fri, 17 Mar 2017 15:56:04 +0000 (16:56 +0100)]
ArmVirtPkg/FdtClientDxe: install DT as sysconfig table in protocol notify
Replace the dependency on PcdPureAcpiBoot with a Platform Has Device Tree
notification callback. Move the sysconfig table installation from the
entry point function to the callback.
Laszlo Ersek [Fri, 17 Mar 2017 14:46:40 +0000 (15:46 +0100)]
ArmVirtPkg: enable AcpiTableDxe and EFI_ACPI_TABLE_PROTOCOL dynamically
In this patch, the ACPI protocol / driver chain is enabled dynamically,
when appropriate. This is being done in one larger patch, because
ArmVirt.dsc.inc, where AcpiTableDxe is built, is used by all the platform
DSCs.
No change in behavior should be observable after this patch on any
ArmVirtPkg platform.
Laszlo Ersek [Fri, 17 Mar 2017 14:12:28 +0000 (15:12 +0100)]
ArmVirtPkg: add XenPlatformHasAcpiDtDxe
This driver produces the EDKII Platform Has ACPI and Platform Has Device
Tree protocols, exactly matching the current ACPI / DT exposure on Xen,
according to ARM vs. AARCH64. At this point it differs from the QEMU
driver PlatformHasAcpiDtDxe in that this one always installs the DT.
Laszlo Ersek [Fri, 17 Mar 2017 14:12:28 +0000 (15:12 +0100)]
ArmVirtPkg: add PlatformHasAcpiDtDxe
This driver produces the EDKII Platform Has ACPI and Platform Has Device
Tree protocols, exactly matching the current ACPI / DT exposure on QEMU,
according to ARM vs. AARCH64, and (in the latter case) to PcdPureAcpiBoot.
Laszlo Ersek [Fri, 17 Mar 2017 12:47:35 +0000 (13:47 +0100)]
EmbeddedPkg: introduce EDKII Platform Has Device Tree GUID
The presence of this GUID in the PPI database, and/or in the DXE protocol
database (as dictated by the platform's needs in these firmware phases)
implies that the platform provides the operating system with a Device
Tree-based hardware description. This is not necessarily exclusive with
other types of hardware description (for example, an ACPI-based one).
A platform PEIM and/or DXE driver is supposed to produce a single instance
of the PPI and/or protocol (with NULL contents), if appropriate. The
decision to produce the PPI and/or protocol is platform specific; for
example, in the DXE phase, it could depend on an HII checkbox / underlying
non-volatile UEFI variable.
In the DXE phase, the protocol is meant to be consumed by the platform
driver that
- owns the Device Tree description of the hardware, and
- is responsible for installing it as a system configuration table.
Said FDT-owner driver can wait for the protocol via DEPEX or protocol
notify.
Because this GUID is not standard, it is prefixed with EDKII / Edkii, as
seen elsewhere (for example in MdeModulePkg and SecurityPkg).
Laszlo Ersek [Fri, 17 Mar 2017 12:47:35 +0000 (13:47 +0100)]
EmbeddedPkg: introduce PlatformHasAcpiLib
Add the shorter-term library instance outlined in the previous patch to
EmbeddedPkg, so that we can imbue AcpiTableDxe with a protocol dependency
on EDKII_PLATFORM_HAS_ACPI_GUID.
Laszlo Ersek [Fri, 17 Mar 2017 12:47:35 +0000 (13:47 +0100)]
EmbeddedPkg: introduce EDKII Platform Has ACPI GUID
The presence of this GUID in the PPI database, and/or in the DXE protocol
database (as dictated by the platform's needs in these firmware phases)
implies that the platform provides the operating system with an ACPI-based
hardware description. This is not necessarily exclusive with other types
of hardware description (for example, a Device Tree-based one).
A platform PEIM and/or DXE driver is supposed to produce a single instance
of the PPI and/or protocol (with NULL contents), if appropriate. The
decision to produce the PPI and/or protocol is platform specific; for
example, in the DXE phase, it could depend on an HII checkbox / underlying
non-volatile UEFI variable.
In the DXE phase, the protocol is meant to be depended-upon by
"MdeModulePkg/Universal/Acpi/AcpiTableDxe", indirectly:
* In the long term, interested platforms will establish this dependency by
hooking an (upcoming) NULL-class DepexLib instance into AcpiTableDxe in
their DSC files, pointing DepexLib's DEPEX through a FixedAtBuild PCD to
the GUID introduced here. (For the prerequisite BaseTools feature, refer
to <https://bugzilla.tianocore.org/show_bug.cgi?id=443>).
* In the short term, an interested platform may hook a private NULL-class
library instance (called e.g. "PlatformHasAcpiLib") into AcpiTableDxe.
Such a library instance would be a specialization of the above described
generic DepexLib, with the DEPEX open-coded on the GUID introduced here.
Either way, the platform makes EFI_ACPI_TABLE_PROTOCOL and (if enabled)
EFI_ACPI_SDT_PROTOCOL dependent on the platform's dynamic decision to
produce or not to produce a NULL protocol instance with this GUID.
In turn, other (platform and universal) DXE drivers that produce ACPI
tables will wait for EFI_ACPI_TABLE_PROTOCOL / EFI_ACPI_SDT_PROTOCOL, via
DEPEX, protocol notify, or a simple gBS->LocateProtocol() in a "late
enough" callback (such as Ready To Boot).
Because this GUID is not standard, it is prefixed with EDKII / Edkii, as
seen elsewhere in MdeModulePkg and SecurityPkg. In addition, an effort is
made to avoid the phrase "AcpiPlatform", as that belongs to drivers /
libraries that produce platform specific ACPI content (as opposed to
deciding whether the entire firmware will have access to
EFI_ACPI_TABLE_PROTOCOL, or any similar facilities in the PEI phase).
We realized that DXE drivers that are independent of AcpiPlatformDxe (that
is, independent of QEMU's ACPI generation), such as RamDiskDxe and
BootGraphicsResourceTableDxe, may produce and/or manipulate ACPI tables,
at driver dispatch or even at Ready To Boot.
This makes it unsafe for us to check for ACPI presence in the UEFI system
config table in a Ready To Boot callback, in order to decide about
exposing the DT.
We realized that DXE drivers that are independent of AcpiPlatformDxe (that
is, independent of QEMU's ACPI generation), such as RamDiskDxe and
BootGraphicsResourceTableDxe, may produce and/or manipulate ACPI tables,
at driver dispatch or even at Ready To Boot.
This makes it unsafe for us to check for ACPI presence in the UEFI system
config table in a Ready To Boot callback, in order to decide about
exposing the DT.
Jeff Fan [Mon, 27 Mar 2017 07:00:00 +0000 (15:00 +0800)]
UefiCpuPkg/MpLib.c: Add checking CR0 PG bit
If CR0 PG bit is not set, it means paging is not enabled on BSP. Thus, Execute
Disable feature is not working actually. Thus, we cannot enable it on APs.
v2:
Correct the commit log.
Cc: Feng Tian <feng.tian@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Jeff Fan <jeff.fan@intel.com> Reviewed-by: Feng Tian <feng.tian@intel.com>
Ard Biesheuvel [Wed, 15 Mar 2017 16:26:41 +0000 (16:26 +0000)]
ArmPkg/ArmExceptionLib: use EL0 stack for synchronous exceptions
In order to be able to produce meaningful diagnostic output when taking
synchronous exceptions that have been caused by corruption of the stack
pointer, prepare the EL0 stack pointer and switch to it when handling the
'Sync exception using SPx' exception class.
Jeff Fan [Thu, 23 Mar 2017 05:19:49 +0000 (13:19 +0800)]
UefiCpuPkg/AcpiCpuData.h: Support >4GB MMIO address
The current CPU_REGISTER_TABLE_ENTRY structure only defined UINT32 Index to
indicate MSR/MMIO address. It's ok for MSR because MSR address is UINT32 type
actually. But for MMIO address, UINT32 limits MMIO address exceeds 4GB.
This update on CPU_REGISTER_TABLE_ENTRY is to add additional UINT32 field
HighIndex to indicate the high 32bit MMIO address and original Index still
indicate the low 32bit MMIO address.
This update makes use of original padding space between ValidBitLength and
Value to add HighIndex.
Cc: Feng Tian <feng.tian@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Jeff Fan <jeff.fan@intel.com> Reviewed-by: Feng Tian <feng.tian@intel.com>
Jeff Fan [Thu, 23 Mar 2017 04:55:26 +0000 (12:55 +0800)]
UefiCpuPkg/RegisterCpuFeaturesLib: Define Index to UINT64
The input parameter Index of PreSmmCpuRegisterTableWrite() and
CpuRegisterTableWrite() is defined as UINT32. Index is MSR/MMIO address that
will be saved in CPU register table. UINT32 blocks the MMIO address > 4GB.
This fix is to define Index to UINT64 instead of UINT32.
Cc: Feng Tian <feng.tian@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Jeff Fan <jeff.fan@intel.com> Reviewed-by: Feng Tian <feng.tian@intel.com>
Derek Lin [Fri, 24 Feb 2017 07:26:19 +0000 (15:26 +0800)]
BaseTools: Skip module AutoGen by comparing timestamp.
[Introduction]
The BaseTool Build.py AutoGen parse INF meta-file and generate
AutoGen.c/AutoGen.h/makefile. When we only change .c .h code, the
AutoGen might be not necessary, but Build.py spend a lot of time on it.
There's a -u flag to skip all module's AutoGen. In my environment, it save
35%~50% of time in rebuild a ROM.
However, if user change one .INF meta-file, then -u flag is not available.
[Idea]
AutoGen can compare meta-file's timestamp and decide if the module's
AutoGen can be skipped. With this, when a module's INF is changed, we
only run this module's AutoGen, we don't need to run other module's.
[Implementation]
In the end of a module's AutoGen, we create a AutoGenTimeStamp.
The file save a file list that related to this module's AutoGen.
In other word, the file list in AutoGenTimeStamp is INPUT files of
module AutoGen, AutoGenTimeStamp file is OUTPUT.
During rebuild, we compare time stamp between INPUT and OUTPUT, and
decide if we can skip it.
Below is the Input/Output of a module's AutoGen.
[Input]
1. All the DSC/DEC/FDF used by the platform.
2. Macro and PCD defined by Build Options such as "build -D AAA=TRUE
--pcd BbbPcd=0".
3. INF file of a module.
4. Source files of a module, list in [Sources] section of INF.
5. All the library link by the module.
6. All the .h files included by the module's sources.
This patch save my build time. When I make a change without touching
DSC/DEC/FDF, it is absolutely much faster than original rebuild,
35%~50% time saving in my environment
(compare to original tool rebuild time).
If I change any DSC/DEC/FDF, there's no performance improve, because it
can't skip any module's AutoGen.
Please note that if your environment will generate DSC/FDF during prebuild,
it will not skip any AutoGen because of DSC timestamp is changed. This will
require prebuild script not to update metafile when content is not changed.
Ruiyu Ni [Thu, 23 Mar 2017 05:09:45 +0000 (13:09 +0800)]
MdePkg/DevicePathLib: Fix FromText bug for multi-instance devicepath
UefiDevicePathLibConvertTextToDevicePath correctly detects when it
has hit a ',' splicing together multiple paths. However, the code
that tries to cope with it:
{code}
if (IsInstanceEnd) {
DeviceNode = (EFI_DEVICE_PATH_PROTOCOL *) AllocatePool (
END_DEVICE_PATH_LENGTH);
ASSERT (DeviceNode != NULL);
SetDevicePathEndNode (DeviceNode);
NewDevicePath = AppendDevicePathNode (DevicePath, DeviceNode);
FreePool (DevicePath);
FreePool (DeviceNode);
DevicePath = NewDevicePath;
}
{code}
causes a problem. The END node that's appended it the node for the
entire list. So when the node is appended in AppendDevicePathNode,
it winds up disappearing. This leads to the path
'PciRoot(0x0),PciRoot(0x0)' parsing as if 'PciRoot(0x0)/PciRoot(0x0)'
were specified. These are two very different things.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Warner Losh <imp@bsdimp.com> Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Chen A Chen [Thu, 16 Mar 2017 05:22:40 +0000 (13:22 +0800)]
ShellPkg: Add Shell invocation option '-exit'
According to Shell spec 2.2 '-exit' invocation option is used to specify
that after running the command line when launched, the UEFI Shell must
immediately exit.
Jiaxin Wu [Wed, 22 Mar 2017 01:22:07 +0000 (09:22 +0800)]
NetworkPkg/IScsiDxe: Fix the incorrect error handling in DriverEntryPoint
Currently, error handling in IScsiDriverEntryPoint is incorrect. For
example, if IScsiCreateAttempts() return error due to the limited max
variable size, iSCSI will not unload the configuration entries.
Jeff Fan [Thu, 23 Mar 2017 01:37:47 +0000 (09:37 +0800)]
UefiCpuPkg/RegisterCpuFeaturesLib: Set CpuFeatureEntry initial value
CpuFeatureEntry will be set before using it. But VS2012 build reported the build
warning "potentially uninitialized local variable 'CpuFeatureEntry' used".
This fix is to set CpuFeatureEntry initial value and add ASSERT check later.
Cc: Feng Tian <feng.tian@intel.com> Cc: Michael Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Jeff Fan <jeff.fan@intel.com> Reviewed-by: Feng Tian <feng.tian@intel.com>
Ard Biesheuvel [Thu, 16 Mar 2017 13:23:42 +0000 (13:23 +0000)]
ArmPkg/DefaultExceptionHandlerLib: walk call stack unconditionally
Currently, we only attempt to walk the call stack and print a backtrace
if the program counter refers to a location covered by a PE/COFF image.
However, regardless of the value of PC, the frame pointer may still have
a meaningful value, and so we can still produce the remainder of the
backtrace.
Ard Biesheuvel [Sat, 18 Mar 2017 21:20:46 +0000 (21:20 +0000)]
ArmPkg/PlatformBootManagerLib: move to BootLogoLib for boot splash support
Replace the duplicated and outdated code in QuietBoot.c with a reference
to BootLogoLib, which provides the same functionality. This also allows
us to drop all references to IntelFrameworkModulePkg in this module.