For function SplitRecord() in file PropertiesTable.c, there is a
potential subtract underflow case for line:
return TotalNewRecordCount - 1;
However, such case will not happen since the logic in function
SplitTable() ensure that when calling SplitRecord(), the variable
'TotalNewRecordCount' will not be zero when performing the subtraction.
It will be handled in the previous if statement:
if (MaxSplitRecordCount == 0) {
CopyMem (NewRecord, OldRecord, DescriptorSize);
return 0;
}
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>
This commit makes sure that in function CoreStartImage(), module
variable 'mCurrentImage' is restored to the current start image context
on all code paths.
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>
Currently, the PE/COFF image memory protection code uses the same code
paths for protecting and unprotecting an image. This is strange, since
unprotecting an image involves a single call into the CPU arch protocol
to clear the permission attributes of the entire range, and there is no
need to parse the PE/COFF headers again.
So let's store the ImageRecord entries in a linked list, so we can find
it again at unprotect time, and simply clear the permissions.
Note that this fixes a DEBUG hang on an ASSERT() that occurs when the
PE/COFF image fails to load, which causes UnprotectUefiImage() to be
invoked before the image is fully loaded.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
UEFI executables that consist of a single read+write+exec PE/COFF section
trigger this message, but such a binary layout isn't actually an error.
The image can be launched alright, only image protection cannot be applied
to it fully.
One example that elicits the message is (some) Linux kernels (with the EFI
stub of course).
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
When attempting to perform page allocations using AllocateAddress, we
fail to check whether the entire region is free before splitting the
region. This may lead to memory being leaked further into the routine,
when it turns out that one of the memory map entries intersected by the
region is already occupied. In this case, prior conversions are not rolled
back.
For instance, starting from this situation
0x000040000000-0x00004007ffff [ConventionalMemory ]
0x000040080000-0x00004009ffff [Boot Data ]
0x0000400a0000-0x000047ffffff [ConventionalMemory ]
a failed EfiLoaderData allocation @ 0x40000000 that covers the BootData
region will fail, but leave the first part of the allocation converted,
so we end up with
0x000040000000-0x00004007ffff [Loader Data ]
0x000040080000-0x00004009ffff [Boot Data ]
0x0000400a0000-0x000047ffffff [ConventionalMemory ]
even though the AllocatePages() call returned an error.
So let's check beforehand that AllocateAddress allocations are covered
by a single memory map entry, so that it either succeeds or fails
completely, rather than leaking allocations.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Star Zeng <star.zeng@intel.com>
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=370
Use GLOBAL_REMOVE_IF_UNREFERENCED for some memory profile global variables,
then their symbols could be removed when memory profile is disabled.
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Feng Tian <feng.tian@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>
Remove the local definitions for the default and runtime page allocation
granularity macros, and switch to the new MdePkg versions.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
There are cases that the operands of an expression are all with rank less
than UINT64/INT64 and the result of the expression is explicitly cast to
UINT64/INT64 to fit the target size.
An example will be:
UINT32 a,b;
// a and b can be any unsigned int type with rank less than UINT64, like
// UINT8, UINT16, etc.
UINT64 c;
c = (UINT64) (a + b);
Some static code checkers may warn that the expression result might
overflow within the rank of "int" (integer promotions) and the result is
then cast to a bigger size.
The commit refines codes by the following rules:
1). When the expression is possible to overflow the range of unsigned int/
int:
c = (UINT64)a + b;
2). When the expression will not overflow within the rank of "int", remove
the explicit type casts:
c = a + b;
3). When the expression will be cast to pointer of possible greater size:
UINT32 a,b;
VOID *c;
c = (VOID *)(UINTN)(a + b); --> c = (VOID *)((UINTN)a + b);
4). When one side of a comparison expression contains only operands with
rank less than UINT32:
UINT8 a;
UINT16 b;
UINTN c;
if ((UINTN)(a + b) > c) {...} --> if (((UINT32)a + b) > c) {...}
For rule 4), if we remove the 'UINTN' type cast like:
if (a + b > c) {...}
The VS compiler will complain with warning C4018 (signed/unsigned
mismatch, level 3 warning) due to promoting 'a + b' to type 'int'.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
This implements a DXE memory protection policy that ensures that regions
that don't require executable permissions are mapped with the non-exec
attribute set.
First of all, it iterates over all entries in the UEFI memory map, and
removes executable permissions according to the configured DXE memory
protection policy, as recorded in PcdDxeNxMemoryProtectionPolicy.
Secondly, it sets or clears the non-executable attribute when allocating
or freeing pages, both for page based or pool based allocations.
Note that this complements the image protection facility, which applies
strict permissions to BootServicesCode/RuntimeServicesCode regions when
the section alignment allows it. The memory protection configured by this
patch operates on non-code regions only.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
In preparation of adding memory permission attribute management to
the pool allocator, split off the locking of the pool metadata into
a separate lock. This is an improvement in itself, given that pool
allocations can only interfere with the page allocation bookkeeping
if pool pages are allocated or released. But it is also required to
ensure that the permission attribute management does not deadlock,
given that it may trigger page table splits leading to additional
page tables being allocated.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
Instead of assuming that a PE/COFF section of type EFI_IMAGE_SCN_CNT_CODE
can always be mapped read-only, classify a section as a code section only
if it has the executable attribute set and the writable attribute cleared.
This adheres more closely to the PE/COFF spec, and avoids issues with
Linux OS loaders that may consist of a single read/write/execute section.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
If the UEFI image is page aligned, the image code section is set to read
only and the image data section is set to non-executable.
1) This policy is applied for all UEFI image including boot service driver,
runtime driver or application.
2) This policy is applied only if the UEFI image meets the page alignment
requirement.
3) This policy is applied only if the Source UEFI image matches the
PcdImageProtectionPolicy definition.
4) This policy is not applied to the non-PE image region.
The DxeCore calls CpuArchProtocol->SetMemoryAttributes() to protect
the image. If the CpuArch protocol is not installed yet, the DxeCore
enqueues the protection request. Once the CpuArch is installed, the
DxeCore dequeues the protection request and applies policy.
Once the image is unloaded, the protection is removed automatically.
The UEFI runtime image protection is teared down at ExitBootServices(),
the runtime image code relocation need write code segment at
SetVirtualAddressMap(). We cannot assume OS/Loader has taken over
page table at that time.
NOTE: It is per-requisite that code section and data section
should not be not merged. That is same criteria for SMM/runtime driver.
We are not able to detect during BIOS boot, because
we can only get LINK warning below:
"LINK : warning LNK4254: section '.data' (C0000040) merged into
'.text' (60000020) with different attributes"
But final attribute in PE code section is same.
Cc: Star Zeng <star.zeng@intel.com>
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: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Tested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
When gDS->SetMemorySpaceCapabilities() is called,
current DXE core will sync all GCD attributes to memory map
attributes, including RUNTIME attributes.
It is wrong, because RUNTIME attributes should be set for
runtime memory only.
This fix clears the RUNTIME attributes before convert to UEFI
memory map. So that the UEFI memory map is good after
gDS->SetMemorySpaceCapabilities() is called.
Cc: Star Zeng <star.zeng@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
If a BaseAddress of NULL is passed into DXE Core services
CoreAllocateIoSpace() or CoreAllocateMemorySpace(), and
DEBUG() messages are enabled, then a NULL pointer reference
is made. The parameter check for BaseAddress is performed
in the function CoreAllocateSpace() after the DEBUG()
messages. A check is added in the DEBUG() messages to
prevent the NULL pointer reference.
This issue was found with PI SCTs with DEBUG messages
enabled in the DXE Core.
Cc: Feng Tian <feng.tian@intel.com>
Cc: Star Zeng <star.zeng@intel.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>
Reviewed-by: Feng Tian <feng.tian@Intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
Instead of comparing a GUID with gZeroGuid via the CompareGuid API, the
commit uses the IsZeroGuid API to check if the given GUID is a zero GUID.
Cc: Feng Tian <feng.tian@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@Intel.com>
Code logic ensures that the pointer 'DriverInfoData' will not be NULL when
it is used.
Add ASSERT as warning for case that will not happen.
Cc: Star Zeng <star.zeng@intel.com>
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>
- abstrated to abstracted
- accessibla to accessible
- addres to address
- apropriate to appropriate
- arry to array
- availabe to available
- avaliable to available
- becasue to because
- correponding to corresponding
- etablished to established
- exeuction to execution
- extensiable to extensible
- fileds to fields
- loadding to loading
- ptototypes to prototypes
- prococol protocol
- requried to required
- resoruce to resource
- runing to running
- uild to build
Cc: Star Zeng <star.zeng@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Giri P Mudusuru <giri.p.mudusuru@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Code logic ensures that both pointers 'DriverInfoData' and 'AllocInfoData'
will not be NULL when they are used.
Add ASSERTs as warning for cases that will not happen.
Cc: Star Zeng <star.zeng@intel.com>
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>
1. Implement include GetRecordingState/SetRecordingState/Record for
memory profile protocol.
2. Consume PcdMemoryProfilePropertyMask to support disable recording
at the start.
3. Consume PcdMemoryProfileDriverPath to control which drivers need
memory profile 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>
PeiCore EntryPoint library _ModuleEntryPoint() will call PeiCore(), then call
CpuDeadLoop (). When NORETURN is added for PeiCore(), MSVC compiler will report
warning C4702: unreachable code for CpuDeadLoop (). And, the warning is treated
as error and cause build break. DxeMain() has the similar issue.
edk2 uses EntryPoint library to wrap every module entry point function except
for SEC. The module entry point is still called by _ModuleEntryPoint(). So,
there will be negative impact to add NORETURN for the module entry point.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Liming Gao <liming.gao@intel.com>
This patch adds the NORETURN attribute to functions that transfer to
other phases, along with an UNREACHABLE() call at the end to avoid
false warnings.
DxeIpl has been excluded as its main function returns a status.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Marvin Haeuser <Marvin.Haeuser@outlook.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
This patches rename CoreGetMemoryMapPropertiesTable to
CoreGetMemoryMapWithSeparatedImageSection.
The reason is that CoreGetMemoryMapPropertiesTable is called
by MemoryAttributesTable.c to get separated PE image section
information.
It is confusing to use *PropertiesTable, because it is NOT
related to PropertiesTable.c
We rename it to avoid confusing.
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
Set the ImageContext.Handle and ImageContext.ImageRead() fields so that
PeCoffLoaderRelocateImageExtraAction() can invoke PeCoffLoaderGetImageInfo
or PeCoffLoaderGetPeHeader if desired to obtain additional metadata.
We will use this to create a PeCoffLoaderRelocateImageExtraAction()
implementation that applies boot time strict mapping permissions to
PE/COFF modules.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Use 128 bytes as the start size region to be same to previous one.
64 bytes is small as the first range. On X64 arch, POOL_OVERHEAD
takes 40 bytes, the pool data less than 24 bytes can be fit into
it. But, the real allocation is few that can't reduce its free pool
link list. And, the second range (64~128) has more allocation
that also increases the free pool link list of the first range.
Then, the link list will become longer and longer. When LinkList
check enable in DEBUG tip, the long link list will bring the
additional overhead and bad performance. Here is the performance
data collected in our X64 platform with DEBUG enable.
64 byte: 22 seconds in BDS phase
128 byte: 19.6 seconds in BDS phase
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Liming Gao <liming.gao@intel.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Michael Kinney <michael.d.kinney@intel.com>
It can improve profile performance, especially when
PcdMemoryProfileMemoryType configured without EfiBootServicesData.
CoreUpdateProfile() can return quickly, but not depend on the further
code to find the buffer not recorded and then return.
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Feng Tian <feng.tian@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>
The patch uses CoreAcquireLockOrFail() instead of
CoreAcquireProtocolLock() in CoreLocateProtocol() to avoid
assertion when CoreLocateProtocol() is called with the
protocol database locked.
The issue was found when changing PcdDebugPrintErrorLevel to
enable page/pool allocation debug message.
Nt32 platform hangs immediately after DxeCore is loaded.
Investigation shows the following calling stacks:
DxeCore entry point (Install a certain protocol)
0 DxeCore::CoreInstallProtocolInterface // Protocol DB is locked
1 DxeCore::AllocatePool
2 PeiDxeDebugLibReportStatusCode::DebugPrint
3 DxeReportStatusCodeLib::ReportStatusCodeEx // <-------------------|
4 DxeReportStatusCodeLib::InternalGetReportStatusCode |
5 DxeCore::LocateProtocol(StatusCodeRuntimeProtocol) |
// Assertion when locking Protocol DB 2nd time |
6 DxeCore::CoreAcquireProtocolLock |
7 PeiDxeDebugLibReportStatusCode::DebugAssert |
8 DxeReportStatusCodeLib::ReportSatusCodeEx // loop begins ---------
In frame #6 the assertion is triggered due to the protocol database
is already locked. #8 calls #4 and the loop begins.
After changing #6 to CoreAcquireLockOrFail(), the assertion is
avoided and the loop is broken.
With the fix, NT32 can boot to Shell even setting
PcdDebugPrintErrorLevel to 0xFFFFFFFF, with all error levels turned
on.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
Current MemoryAttributesTable will be installed on ReadyToBoot event
at TPL_NOTIFY level, it maybe incorrect when PcdHiiOsRuntimeSupport
= TRUE as HiiDatabaseDxe will have runtime memory allocation for HII
OS runtime support on and after ReadyToBoot. The issue was exposed at
http://article.gmane.org/gmane.comp.bios.edk2.devel/10125.
To make sure the correctness of MemoryAttributesTable, this patch is
to enhance MemoryAttributesTable installation to install
MemoryAttributesTable on ReadyToBoot event at TPL_CALLBACK - 1 level
to make sure it is at the last of ReadyToBoot event, and also hook
runtime memory allocation after ReadyToBoot.
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
The following patch for MemoryAttributesTable will need the memory type.
And CoreUpdateProfile() can also use the memory type for check.
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Feng Tian <feng.tian@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>
Originally, the code block for "Report DXE Core image information to the PE/COFF
Extra Action Library" was after ProcessLibraryConstructorList(). To fix an issue,
ProcessLibraryConstructorList() was moved to be right after CoreInitializeGcdServices()
at c5d5379937, but the code block was left.
As there maybe PeCoffExtraActionLib implementation need Constructor executed first,
the patch is to move the code block for "Report DXE Core image information to the PE/COFF
Extra Action Library" to be after ProcessLibraryConstructorList.
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Feng Tian <feng.tian@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>
Check for Type AllocateAddress,
if NumberOfPages is 0 or
if (NumberOfPages << EFI_PAGE_SHIFT) is above MAX_ADDRESS or
if (Start + NumberOfBytes) rolls over 0 or
if Start is above MAX_ADDRESS or
if End is above MAX_ADDRESS,
return EFI_NOT_FOUND.
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Michael Kinney <michael.d.kinney@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Feng Tian <feng.tian@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>
When the DXE core is loaded, it invokes the PeCoffExtraActionLib library
function 'PeCoffLoaderRelocateImageExtraAction' explicitly, which may be
in addition to the same function having been called by the DxeIpl PE/COFF
loader instance.
The ImageContext that DXE core presents to this function is only partially
initialized, which may result in the following output on AArch64 systems:
add-symbol-file ..MdeModulePkg/Core/Dxe/DxeMain/DEBUG/DxeCore.dll 0x5F226240
<some intermediate output>
add-symbol-file ..MdeModulePkg/Core/Dxe/DxeMain/DEBUG/DxeCore.dll 0x5F226000
This is caused by incorrect data in the ImageContext structure, which means
the start of the .text section is calculated incorrectly. In general, it is
the duty of the caller to present a valid ImageContext structure, so let's
add the missing values before invoking PeCoffLoaderRelocateImageExtraAction().
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Star Zeng <star.zeng@intel.com>
If 'CoreGetMemoryMap' return success the 'MemoryMap' shouldn't be NULL.
Add code to make sure this.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Yao Jiewen <Jiewen.Yao@intel.com>
According to the spec, each entry in the Memory
Attributes table shall have the same type as
the region it was carved out of in the UEFI memory map.
The current attribute uses RTData for PE Data, but
it should be RTCode.
This patch fixed the issue. It is validated with or
without PropertiesTable.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: "Yao, Jiewen" <jiewen.yao@intel.com>
Tested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Liming Gao <liming.gao@intel.com>
Hao Wu
Liming Gao
Ard Biesheuvel
Dandan Bi
Laszlo Ersek
Star Zeng
Fu Siyuan
Jiewen Yao
Michael Kinney
Gary Lin
Giri P Mudusuru
Marvin H?user
Ruiyu Ni
Qiu Shumin