Signed-off-by: Graeme Gregory <graeme@xora.org.uk>
Reviewed-by: Michael D Kinney <michael.d.kinney@intel.com>
Reviewed-by: Leif Lindholm <quic_llindhol@quicinc.com>

There have been reports[1] of failures to boot due to unicode collation
protocols not being available at Ext4Dxe load time. Therefore, attempt
to initialize unicode collation at Start() time, like done previously in
FatPkg/EnhancedFatDxe. By doing so, we move collation initialization
to BDS, where the module responsible for protocol installation should
have already been loaded and ran.

[1]: https://edk2.groups.io/g/devel/message/100312



Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Marvin Häuser <mhaeuser@posteo.de>
Fixes: d9ceedca ("Ext4Pkg: Add Ext4Dxe driver.")
Signed-off-by: Pedro Falcato <pedro.falcato@gmail.com>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

The current SPCR table revision is 2. Instead of using a hard-coded
value, use the symbolic constant
EFI_ACPI_SERIAL_PORT_CONSOLE_REDIRECTION_TABLE_REVISION.

Signed-off-by: Rebecca Cran <quic_rcran@quicinc.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>

ACPI 6.0 introduced support for ASL 2.0 includes standard operators.
Convert the DSDT from legacy to ASL 2.0 syntax.

Signed-off-by: Rebecca Cran <quic_rcran@quicinc.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>

If PcdArmArchTimerFreqInHz is zero, the value of the timer frequency
will be read from CNTFRQ_EL0. Avoid hard-coding the value in ArmJuno.dsc
and instead let the ArmArchTimerLib driver read it from the SoC.

Signed-off-by: Rebecca Cran <rebecca@quicinc.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>

The MetronomeDxe driver uses the PCD PcdMetronomeTickPeriod to calculate
how many ticks to wait in MicroSecondDelay. Given that the timer clock
on Juno runs at 50 MHz, it ticks every 20 ns; therefore, a setting of
1000 is wrong: for example it causes a call to gBS->Stall (1) to take
10 us.

The driver in MdeModulePkg/Universal/Metronome assumes the clock ticks
at least every 100 ns, which is the minimum allowed by the Metronome
protocol. Since that's the case on Juno, switch from
EmbeddedPkg/MetronomeDxe to MdeModulePkg/Universal/Metronome.

Signed-off-by: Rebecca Cran <rebecca@quicinc.com>
Reviewed-by: Ard Biesheuvel <ardb@kernel.org>

The SymlinkTmp was deallocated unconditionally, so we shouldn't free it
again on EFI_ERROR

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: e81432fb ("Ext4Pkg: Add symbolic links support")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>

The ARM memory region types with the NONSECURE qualification should only
be used in places where this makes a difference, i.e., in the secure
world. The Marvell firmware generally runs at EL2, which is non-secure
only (in ARM v8.0) so the NONSECURE marking is unnecessary here, and
should be dropped.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Leif Lindholm <quic_llindhol@quicinc.com>

BeagleBoard is the only user of the so-called 'PE/COFF protocol', which
just exposes the PE/COFF loader library API via a protocol, presumably
to avoid duplicating this code in the PrePi SEC component as well as the
DXE core.

This is a rather questionable practice, and it would be better to drop
this code so we can remove it from the EDK2 main repo as well.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: Leif Lindholm <quic_llindhol@quicinc.com>

Updates the Standalone MM module to have the necessary INF changes
to build with the following two recent commits made to rewrite the
the variable store header in the MM SPI FVB service.

  - e95c7988
  - 88d44c56

Cc: Ashraf Ali S <ashraf.ali.s@intel.com>
Cc: Isaac Oram <isaac.w.oram@intel.com>
Cc: Rangasai V Chaganty <rangasai.v.chaganty@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Chasel Chiu <chasel.chiu@intel.com>
Signed-off-by: Michael Kubacki <michael.kubacki@microsoft.com>
Reviewed-by: Isaac Oram <isaac.w.oram@intel.com>
Reviewed-by: Chasel Chiu <chasel.chiu@intel.com>

Fix the compilation description error in the Readme file.

Cc: Ard Biesheuvel <ardb+tianocore@kernel.org>
Cc: Bibo Mao <maobibo@loongson.cn>
Cc: Chao Li <lichao@loongson.cn>
Cc: Leif Lindholm <quic_llindhol@quicinc.com>
Cc: Liming Gao <gaoliming@byosoft.com.cn>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Signed-off-by: xianglai li <lixianglai@loongson.cn>
Reviewed-by: Chao Li <lichao@loongson.cn>

Platform may implement an additional NVS region following
Regular variable region and in this case SpiFvbService should include
both region size when calculating the total NVS region size.

The PcdFlashNvStorageAdditionalSize is for compatible with legacy
usages that should be deprecated. The new usage model should define
separate regions without implicit connections to UEFI Variable or
FTW regions.

Example NVS flash map for such legacy usage:
Note: PcdFlashNvStorageAdditionalSize is equal to platform
      PcdFlashFvNvStorageEventLogSize.

  ---------------
  |UEFI Variable|
  ---------------
  |EventLog     | <= this is Additional NVS region
  ---------------
  |FTW Working  |
  ---------------
  |FTW Spare    |
  ---------------

Cc: Ashraf Ali S <ashraf.ali.s@intel.com>
Cc: Isaac Oram <isaac.w.oram@intel.com>
Cc: Rangasai V Chaganty <rangasai.v.chaganty@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Michael Kubacki <michael.kubacki@microsoft.com>
Signed-off-by: Chasel Chiu <chasel.chiu@intel.com>
Reviewed-by: Michael Kubacki <michael.kubacki@microsoft.com>
Reviewed-by: Isaac Oram <isaac.w.oram@intel.com>

When invalid VariableStore FV header detected, current SpiFvbService
will erase both FV and VariableStore headers from flash, however,
it will only rewrite FV header back and cause invalid VariableStore
header.

This patch adding the support for rewriting both FV header and
VariableStore header when VariableStore corruption happened.
The Corrupted variable content should be taken care by
FaultTolerantWrite driver later.

Platform has to set PcdFlashVariableStoreType to inform SpiFvbService
which VariableStoreType should be rewritten.

Cc: Ashraf Ali S <ashraf.ali.s@intel.com>
Cc: Isaac Oram <isaac.w.oram@intel.com>
Cc: Rangasai V Chaganty <rangasai.v.chaganty@intel.com>
Cc: Ray Ni <ray.ni@intel.com>
Cc: Michael Kubacki <michael.kubacki@microsoft.com>
Signed-off-by: Chasel Chiu <chasel.chiu@intel.com>
Reviewed-by: Michael Kubacki <michael.kubacki@microsoft.com>
Reviewed-by: S, Ashraf Ali <ashraf.ali.s@intel.com>
Reviewed-by: Isaac Oram <isaac.w.oram@intel.com>

V1: Make the Windows build example current working directory based
instead of arbitrary.
V2: Fix case mismatch in edk2 repo local name.

Cc: Sai Chaganty <rangasai.v.chaganty@intel.com>
Cc: Nate DeSimone <nathaniel.l.desimone@intel.com>
Cc: Liming Gao <gaoliming@byosoft.com.cn>
Cc: Rebecca Cran <quic_rcran@quicinc.com>
Signed-off-by: Isaac Oram <isaac.w.oram@intel.com>
Reviewed-by: Rebecca Cran <rebecca@quicinc.com>

We need to free SymlinkTmp before exiting if SymlinkSizeTmp != ReadSize
condition is true.

Reported-by: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: e81432fb ("Ext4Pkg: Add symbolic links support")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>

The directory entry name conventions forbid having null-terminator
symbols in its body and can lead to undefined behavior conditions
and crashes

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: 89b2bb0d ("Ext4Pkg: Fix and clarify handling regarding non-utf8 dir entries")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

Accessing array using index of uint64 type makes MSVC compiler to
include `__allmul` function in NOOPT which is not referenced in IA32.
So we null-terminate the string using ReadSize, which should be equal to
SymlinkSizeTmp after correct reading. Also add missing MultU64x32
in Ext4Read.

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: 7c46116b ("Ext4Pkg: Add ext2/3 support")
Fixes: e81432fb ("Ext4Pkg: Add symbolic links support")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>

Missing EFI_OUT_OF_RESOURCES exit status on failed Ext4CreateDentry
leads to NULL-pointer dereference in Ext4GetFileInfo (passing NULL
buffer in Ext4ReadDir)

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: 21b18538 ("Ext4Pkg: Add a directory entry tree.")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

Missing check in some cases leads to failed StrCpyS call in
Ext4GetVolumeLabelInfo. Also correct condition that checks Inode pointer
for being NULL in Ext4AllocateInode

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: cfbbae59 ("Ext4Pkg: Add handling of EFI_FILE_SYSTEM_VOLUME_LABEL GetInfo().")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

This check already present in the while loop below, but absent for cases
when input file is nameless, so to handle assertion in Ext4ReadFile we
need to add it at the top of function

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: d9ceedca ("Ext4Pkg: Add Ext4Dxe driver.")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

Correct multiplication overflow check code and add additional check
for emptiness of number of blocks and block number.

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: d9ceedca ("Ext4Pkg: Add Ext4Dxe driver.")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>

Missing check for wrong s_log_block_size exponent leads to shift out of
bounds. Limit block size to 2 MiB

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: d9ceedca ("Ext4Pkg: Add Ext4Dxe driver.")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

We need to validate inode number to prevent reading non-existent and
incorrect inodes so we checks that inode number valid across opened
partition before we read it in Ext4ReadInode.

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: d9ceedca ("Ext4Pkg: Add Ext4Dxe driver.")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Acked-by: Marvin Häuser <mhaeuser@posteo.de>

Superblock s_inodes_per_group field can't be zero, it leads to division
by zero in BlockGroup routine Ext4ReadInode

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: d9ceedca ("Ext4Pkg: Add Ext4Dxe driver.")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

We need to check EXT4_FEATURE_RO_COMPAT_METADATA_CSUM in the
FeatureRoCompat field instead of FeaturesCompat. The proper way to do
this is to use macro EXT4_HAS_METADATA_CSUM. 
Also, replace the EXT4_FEATURE_INCOMPAT_CSUM_SEED check with predefined
macro EXT4_HAS_INCOMPAT

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: d9ceedca ("Ext4Pkg: Add Ext4Dxe driver.")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

We need to free buffer on return if BlockRemainder != 0. Also changed
return logic from function to use use common exit to prevent code
duplication.

Cc: Marvin Häuser <mhaeuser@posteo.de>
Cc: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Vitaly Cheptsov <vit9696@protonmail.com>
Fixes: d9ceedca ("Ext4Pkg: Add Ext4Dxe driver.")
Signed-off-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

Fix the ARM_MPCORE_INFO table and incorporate the DXE driver and test
app to the build so that EFI_MP_SERVICES_PROTOCOL can be used and tested
on Raspberry Pi 4. Note that the test app is not added to the image - it
can be taken from the build directory and executed from the UEFI shell.

Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Rebecca Cran <rebecca@quicinc.com>

To follow new DSU PMU specs, the following changes have been made:
- Add a cluster node for each CPM.
- Add DSU node to each CPM.

Signed-off-by: Nhi Pham <nhi@os.amperecomputing.com>

The hob was built but wasn't consumed yet. This change will let the DXE
Main shift the memory space consumed by DXE modules to the region
specified by the PcdSystemMemoryBase and PcdSystemMemorySize.

Signed-off-by: Nhi Pham <nhi@os.amperecomputing.com>

PrePeiCore's Arm module just added a requirement to run all library
constructors that PrePeiCore uses. One of them is included in the
constructor of the ExtractGuidedSectionLib. This constructor will build
a HOB that is not supported "at this time" and causes an error.
We don't need to use this constructor because PrePeiCore doesn't
really need this library.

This is a workaround to avoid PrePeiCore module calling
ExtractGuidedSectionLib's constructor.

Signed-off-by: Nhi Pham <nhi@os.amperecomputing.com>

The NOR Flash driver was recently moved from the Tianocore\edk2 repository
to the Tianocore\edk2-platforms repository at the following location:
Platform\ARM\Drivers\NorFlashDxe\NorFlashDxe.inf

As part of this move the FILE_GUID for the NorFlashDxe.inf at the new
location was also updated from: 93E34C7E-B50E-11DF-9223-2443DFD72085
to: DE6AE758-D662-4E17-A97C-4C5964DA4C41

Correspondingly, the device paths for the BootMonFs also requires to be
updated in gArmBootMonFsTokenSpaceGuid.PcdBootMonFsSupportedDevicePaths
so that the file system is correctly mounted.

Therefore, update the BootMonFS device path for the platforms that
utilise it.

Reported-by: Sudeep Holla <Sudeep.Holla@arm.com>
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com>
Tested-by: Sudeep Holla <Sudeep.Holla@arm.com>

We have been hitting many issues with ".." ASSERTing on / in internal
code, so make sure that we direct /.. to /. This is safe.

Signed-off-by: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Marvin Häuser <mhaeuser@posteo.de>
Reported-by: Savva Mitrofanov <savvamtr@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

Fix typos discovered by SpellCheck, which were discovered as part of the
initial CI enablement effort.

Cc: Pedro Falcato <pedro.falcato@gmail.com>
Signed-off-by: Marvin Häuser <mhaeuser@posteo.de>
Reviewed-by: Pedro Falcato <pedro.falcato@gmail.com>

Previously, the handling was mixed and/or non-existent regarding non utf-8
dirent names. Clarify it.

Signed-off-by: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Marvin Häuser <mhaeuser@posteo.de>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

Several questions have popped up regarding the ext4 directory entry
layout and contents off-list. Attempt to clarify these issues by adding
some explanatory comments.

Signed-off-by: Pedro Falcato <pedro.falcato@gmail.com>
Cc: Marvin Häuser <mhaeuser@posteo.de>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

Fix an out-of-bounds read inside CompareMem() when
checking for "." or ".." by explicitly bounding name_len
to [0, 2] beforehand.

Reported-by: Savva Mitrofanov <savvamtr@gmail.com>
Fixes: 45e37d85 ("Ext4Pkg: Hide "." and ".." entries from Read() callers.")
Cc: Marvin Häuser <mhaeuser@posteo.de>
Signed-off-by: Pedro Falcato <pedro.falcato@gmail.com>
Reviewed-by: Marvin Häuser <mhaeuser@posteo.de>

defines two PCDs, PcdShellFile and PcdShellFileDesc,
which holds the GUID and description of the UEFI shell file to be loaded.

A PCDs based solution gives flexibility to the user to load different
images, by just overriding the DSC file.

The user can load a diagnostic image or test image during
PCDBootToShellOnly or later stages.

Cc: Isaac Oram <isaac.w.oram@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Liming Gao <gaoliming@byosoft.com.cn>
Signed-off-by: Abdul Lateef Attar <AbdulLateef.Attar@amd.com>
Reviewed-by: Isaac Oram <isaac.w.oram@intel.com>

GCC compiler puts the DevicePath PCDs to the read-only
section. During boot if try to process the device path
after PtrGetPtr it throws a page fault exception.

Hence making a local copy using DuplicateDevicePath()
to avoid the page fault exception.

Cc: Isaac Oram <isaac.w.oram@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Liming Gao <gaoliming@byosoft.com.cn>

Signed-off-by: Abdul Lateef Attar <abdattar@amd.com>
Reviewed-by: : Isaac Oram <isaac.w.oram@intel.com>

This contains binary interface changes and requires FSP 4.2.0.2A or later

Cc: Nate DeSimone <nathaniel.l.desimone@intel.com>
Cc: Chasel Chiu <chasel.chiu@intel.com>
Signed-off-by: Isaac Oram <isaac.w.oram@intel.com>
Reviewed-by: Nate DeSimone <nathaniel.l.desimone@intel.com>

This contains binary interface changes and requires FSP 4.2.0.2A or later

Cc: Nate DeSimone <nathaniel.l.desimone@intel.com>
Cc: Chasel Chiu <chasel.chiu@intel.com>
Signed-off-by: Isaac Oram <isaac.w.oram@intel.com>
Reviewed-by: Nate DeSimone <nathaniel.l.desimone@intel.com>