v2 changes include:
1. Use MSR data structure instead of redefin local macro.
Cc: Jeff Fan
Cc: Ruiyu Ni
Cc: David Wei
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Eric Dong
---
Cc: Jeff Fan
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Eric Dong
---
Maintainers.txt | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/Maintainers.txt b/Maintainers.txt
index 39b5b67..9d1b9fb 100644
---
Ray,
I check the spec, found the Quark definition has some small difference with
normal MSR. So I prefer not do this change for Quark.
Thanks,
Eric
-Original Message-
From: Ni, Ruiyu
Sent: Thursday, August 3, 2017 6:22 PM
To: Dong, Eric ; edk2-devel@lists.01.org
It is generally hard to debug across x86 processor mode transitions, and to
debug 16-bit real mode code with modern tools. There are a few places in the
x86 that still require 16-bit real mode for handoffs (like the reset vector) so
you tend to hit this issue more in debugging firmware.
In a form, some new menus may be dynamically inserted between highlight
menu and previous top of screen menu when some question are refreshed.
So the highlight menu and previous top of screen menu perhaps can't be
shown in one page. Existing codes miss to handle this case then will
cause incorrect
Reviewed-by: Jeff Fan
-Original Message-
From: edk2-devel [mailto:edk2-devel-boun...@lists.01.org] On Behalf Of Eric Dong
Sent: Friday, August 04, 2017 11:37 AM
To: edk2-devel@lists.01.org
Cc: Ni, Ruiyu; Fan, Jeff
Subject: [edk2] [Patch v4 2/3] UefiCpuPkg: Add
Reviewed-by: Jeff Fan
-Original Message-
From: Dong, Eric
Sent: Friday, August 04, 2017 11:37 AM
To: edk2-devel@lists.01.org
Cc: Fan, Jeff; Ni, Ruiyu
Subject: [Patch v4 1/3] UefiCpuPkg: Add Pcds used by processor trace feature.
Cc: Jeff Fan
Cc:
Hi Andrew,
THe problem is solved, after the SEC code switch to protected mode, gdb can
work well now
At 2017-08-04 11:26:16, "wang xiaofeng" wrote:
HI Andrew,
How can I adjust the debugger to correct mode? Or I have to enable the
debug after swtich to
Cc: Jeff Fan
Cc: Ruiyu Ni
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong
---
.../CpuCommonFeaturesLib/CpuCommonFeatures.h | 66 +++
.../CpuCommonFeaturesLib/CpuCommonFeaturesLib.c| 11 +
Enable processor trace feature.
V2 update:
1. Add device capability check logic in Support function.
2. Correct the function comments.
V3 update:
1. Get device capability in Suport function instead of
GetConfigData function.
V4 update:
1. Base on the global point to allocate buffer instead
Cc: Jeff Fan
Cc: Ruiyu Ni
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong
---
UefiCpuPkg/UefiCpuPkg.dec | 8
1 file changed, 8 insertions(+)
diff --git a/UefiCpuPkg/UefiCpuPkg.dec
Cc: Jeff Fan
Cc: Ruiyu Ni
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong
---
UefiCpuPkg/Include/Library/RegisterCpuFeaturesLib.h | 1 +
1 file changed, 1 insertion(+)
diff --git
HI Andrew,
How can I adjust the debugger to correct mode? Or I have to enable the
debug after swtich to protected mode?
At 2017-08-04 10:57:16, "Andrew Fish" wrote:
>The reset vector is 16-bit real mode, so you have the debugger in the wrong
>mode. The code should
The reset vector is 16-bit real mode, so you have the debugger in the wrong
mode. The code should transition to 32 bit protected early in the flow.
Sent from my iPhone
> On Aug 3, 2017, at 7:47 PM, wang xiaofeng wrote:
>
> Hello,
> I am tring to add my own SEC code
Hello,
I am tring to add my own SEC code base on OVMF and run on QEMU. Since
the code cannot run I need to step to step trace the assembly code .
The hang point is very early before I can use either UDK or debug serial
output. I tried to use gdb to connect to QEMU.I start gdb in
Reviewed-by: Hao Wu
Best Regards,
Hao Wu
> -Original Message-
> From: Zhu, Yonghong
> Sent: Thursday, August 03, 2017 5:00 PM
> To: edk2-devel@lists.01.org
> Cc: Ni, Ruiyu; Wu, Hao A; Andrew Fish
> Subject: [Patch V2 2/4] DuetPkg: Fix Xcode 9 Beta treating 32-bit
The patch applies on top of Laszlo's IoMmuDxe cleanup series [1].
Commit is also available at https://github.com/codomania/edk2/tree/qemufwcfg-sev
[1] https://lists.01.org/pipermail/edk2-devel/2017-August/012808.html
Changes since v2:
* Changes to address v2 feedbacks.
Changes since v1:
* Drop
Commit 09719a01b11b (OvmfPkg/QemuFwCfgLib: Implement SEV internal function
for Dxe phase) uses IOMMU protocol to allocate and free FW_CFG_DMA_ACCESS
buffer when SEV is active. During initial commits we made assumption that
IOMMU.AllocateBuffer() will provide PlainTextAddress (i.e C-bit cleared).
On 08/04/17 00:07, Brijesh Singh wrote:
> Hi Laszlo,
>
> Thanks for the detail review, I will soon send v3 with all your feedback
> addressed. I must admit that I have constant struggle with formating issues
> in EDKII contributions. While browsing the code, several packages have
> code and
Hi Laszlo,
Thanks for the detail review, I will soon send v3 with all your feedback
addressed. I must admit that I have constant struggle with formating issues
in EDKII contributions. While browsing the code, several packages have
code and comment exceeding 79 char. But looking at your previous
Leif,
Thanks for all the work on this!
I have updated the default branch on edk2-platforms to
be the master branch.
Mike
> -Original Message-
> From: edk2-devel [mailto:edk2-devel-boun...@lists.01.org] On
> Behalf Of Leif Lindholm
> Sent: Thursday, August 3, 2017 7:09 AM
> To:
On 08/03/17 18:10, Brijesh Singh wrote:
> Commit 09719a01b11b (OvmfPkg/QemuFwCfgLib: Implement SEV internal function
> for Dxe phase)
> uses IOMMU protocol to allocate and free FW_CFG_DMA_ACCESS buffer
> when SEV is active. During initial commits we made assumption that
> IOMMU.AllocateBuffer()
Tahnia,
My guess is that you are not using NASM.
It appears that InternalMemIsZeroBuffer is only implemented for NASM and not
ASM.
Which is a bug since building with ASM results in a missing function.
Your tools_def.txt file is missing this statement. Please make sure this
statement
is
This patch adds IsNodeInList() to BaseLib, which verifies the given
Node is part of the doubly-linked List provided.
V2:
- Rename "List" to "FirstEntry" and "Node" to "SecondEntry" to clarify that
"FirstEntry" does not need to be the doubly-linked list's head node.
V3:
- Remove ASSERTs
1) Replace InternalBaseLibIsNodeInList() with
InternalBaseLibIsListValid().
- The verification whether Node is within the doubly-linked List
is now done by IsNodeInList().
- Whether the list is valid is returned.
2) The comments within InsertHeadList() and InsertTailList() stated
Ruiyu,
Sorry for replying to old patch, I was cleaning out some old sandboxes
when I stumbled onto this issue.
Why is EFIAPI removed from InternalCharToUpper in
UefiShellLevel2CommandsLib.c? It is present in both BaseLibInternals.h and
String.c. Without EFIAPI, GCC builds
Add the python script RunMakefile.py that can be used
in a PREBUILD/POSTBUIILD action to invoke a makefile
passing in context as makefile defines. The command
line arguments passed into RunMakefile.py are converted
to the following set of defines.
* ACTIVE_PLATFORM
* TARGET_ARCH
* TOOL_CHAIN_TAG
* Extend PREBUILD/POSTBUILD define values to support more than
one argument.
* Delay normalization of PREBUILD/POSTBUILD define values
until all arguments in the define values can be processed.
* Convert PREBUILD/POSTBUILD build define value arguments
that are WORKSPACE or PACKAGES_PATH
Add sample makefile that can be used to test RunMakefile.py
script and can also be used as a template to start a new
PREBUILD/POSTBUILD makefile.
Cc: Liming Gao
Cc: Yonghong Zhu
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by:
A platform DSC file can provide PREBUILD and POSTBUILD actions in the
[Defines] section of the DSC file. The EDK II DSC specification only allows
a single argument for these two defines, which means the PREBUILD and
POSTBUILD actions are limited to simple batch files, shell scripts, or
I have no preference. Please go ahead.
> -Original Message-
> From: Marvin H?user [mailto:marvin.haeu...@outlook.com]
> Sent: Thursday, August 3, 2017 11:43 PM
> To: edk2-devel@lists.01.org
> Cc: Gao, Liming ; Kinney, Michael D
>
>
Commit 09719a01b11b (OvmfPkg/QemuFwCfgLib: Implement SEV internal function for
Dxe phase)
uses IOMMU protocol to allocate and free FW_CFG_DMA_ACCESS buffer
when SEV is active. During initial commits we made assumption that
IOMMU.AllocateBuffer() will provide PlainTextAddress (i.e C-bit
cleared).
The patch applies on top of Laszlo's IoMmuDxe cleanup series [1].
Commit is also available at https://github.com/codomania/edk2/tree/qemufwcfg-sev
[1] https://lists.01.org/pipermail/edk2-devel/2017-August/012808.html
Changes since v1:
* Drop Patch 1 through 3, they are covered by Laszlo's
Hey Liming,
I noticed that and assumed that a second call is not going to hurt much (only
used in DEBUG afterall).
To work around that, the macro could be modified to either call IsNodeInList()
or InternalBaseLibIsListValid() based on the PCD value,
though one would have to know/check the PCD's
edksetup.sh implements a test that requires the current working
directory to contain the edksetup.sh script. This test has the side
effect of requiring the WORKSPACE to be set to the same directory as
the edksetup.sh. In a multiple workspace configuration, it is
required to be able to have a
Marvin:
If PcdVerifyNodeInList is set TRUE, InternalBaseLibIsListValid() will be
checked twice in two ASSERT(). In the original logic,
InternalBaseLibIsListValid only runs once. Could we work out the same check
logic?
Thanks
Liming
> -Original Message-
> From: edk2-devel
Reviewed-by: Liming Gao for this serials.
> -Original Message-
> From: Zeng, Star
> Sent: Wednesday, August 2, 2017 6:37 PM
> To: edk2-devel@lists.01.org
> Cc: Zeng, Star ; Gao, Liming
> Subject: [PATCH 1/2] MdePkg: Add
Reviewed-by: Liming Gao
> -Original Message-
> From: edk2-devel [mailto:edk2-devel-boun...@lists.01.org] On Behalf Of Chris
> Ruffin
> Sent: Thursday, August 3, 2017 4:30 AM
> To: edk2-devel@lists.01.org
> Subject: [edk2] [PATCH] BaseTools/Conf: apply nasmb, asm16
Jiewen:
The change is in SecFspSecPlatformLibNull. This library instance provides the
dummy function. It should not be used by any real FSP. We meet with the build
issue with /WHOLEARCHIVE option. Because this library depends on the external
TempRamInitApi(), but FspSecCoreS and FspSecCoreT
On 08/03/17 16:35, Brijesh Singh wrote:
> Laszlo,
>
> One minor issue, I got compilation error with GCC48.
>
> /home/brijesh/codomania/edk2-new/edk2/OvmfPkg/IoMmuDxe/AmdSevIoMmu.c: In
> function ‘IoMmuUnmap’:
> /home/brijesh/codomania/edk2-new/edk2/OvmfPkg/IoMmuDxe/AmdSevIoMmu.c:408:25:
> error:
Laszlo,
One minor issue, I got compilation error with GCC48.
/home/brijesh/codomania/edk2-new/edk2/OvmfPkg/IoMmuDxe/AmdSevIoMmu.c: In
function ‘IoMmuUnmap’:
/home/brijesh/codomania/edk2-new/edk2/OvmfPkg/IoMmuDxe/AmdSevIoMmu.c:408:25:
error: ‘CommonBufferHeader’ may be used uninitialized in
On 08/03/17 16:10, Brijesh Singh wrote:
> Hi Laszlo,
>
> On 08/02/2017 04:24 PM, Laszlo Ersek wrote:
>> This series is proposed as a replacement (or a replacement "basis") for
>> patches #1 through #3 of Brijesh's series
>>
>>[PATCH v1 0/4] OvmfPkg : IoMmuDxe: BusMasterCommonBuffer support
Eric,
Typo: " Protected Processor Inventory Number" should be " Intel Processor Trace
"
+ @param[in] StateIf TRUE, then the Protected Processor Inventory
+ Number feature must be enabled.
+ If FALSE, then the Protected
Reviewed-by: Jeff Fan
-Original Message-
From: Dong, Eric
Sent: Tuesday, August 01, 2017 3:51 PM
To: edk2-devel@lists.01.org
Cc: Fan, Jeff; Ni, Ruiyu
Subject: [Patch 1/3] UefiCpuPkg: Add definition for LMCE feature.
Cc: Jeff Fan
Cc: Ruiyu Ni
Reviewed-by: Jeff Fan
-Original Message-
From: Dong, Eric
Sent: Tuesday, August 01, 2017 3:51 PM
To: edk2-devel@lists.01.org
Cc: Fan, Jeff; Ni, Ruiyu
Subject: [Patch 3/3] UefiCpuPkg PiSmmCpuDxeSmm: Check LMCE capability when wait
for AP.
Cc: Jeff Fan
Hi Laszlo,
On 08/02/2017 04:24 PM, Laszlo Ersek wrote:
This series is proposed as a replacement (or a replacement "basis") for
patches #1 through #3 of Brijesh's series
[PATCH v1 0/4] OvmfPkg : IoMmuDxe: BusMasterCommonBuffer support when
SEV is active
Eric,
Typo: " Protected Processor Inventory Number" should be " Local machine check
exception "
+ @param[in] StateIf TRUE, then the Protected Processor Inventory
+ Number feature must be enabled.
+ If FALSE, then the
Hi all,
A little over two years ago, I set up OpenPlatformPkg [1] as a way to
prototype a single source tree for multiple platforms, built against
edk2.
Last year, we finalised the rules for how such a tree could be
integrated into the edk2-platforms [2] repository, and I have now
completed the
Eric,
In ProcTraceInitialize(), we cannot assume ProcessorNumber 0 is always BSP.
+ if (ProcessorNumber == 0) {
+IsBsp = TRUE;
The recommended method is to allocate memory in ProcTraceGetConfigData().
Or add one buffer point in ConfigData structure.
If buffer point is NULL, allocate
Reviewed-by: Eric Dong
-Original Message-
From: Wu, Hao A
Sent: Wednesday, August 2, 2017 4:30 PM
To: edk2-devel@lists.01.org
Cc: Wu, Hao A ; Dong, Eric ; Fan, Jeff
Subject: [PATCH] UefiCpuPkg/Msr: Add a
Hi
May I know what test you have done to remove this?
Have you built a real FSP and boot it with FSP Wrapper?
Thank you
Yao Jiewen
> -Original Message-
> From: Song, BinX
> Sent: Thursday, August 3, 2017 1:47 PM
> To: edk2-devel@lists.01.org
> Cc: Yao, Jiewen ;
Eric,
Please refer to the below code sample to avoid defining local macros.
MSR_IA32_MTRR_DEF_TYPE_REGISTER DefType;
//
// Enable Cache MTRR
//
DefType.Uint64 = AsmReadMsr64 (MSR_IA32_MTRR_DEF_TYPE);
DefType.Bits.E = 1;
DefType.Bits.FE = 1;
AsmWriteMsr64 (MSR_IA32_MTRR_DEF_TYPE,
Remove the deprecated macro and change code which use the macro.
Eric Dong (7):
QuarkSocPkg MtrrLib: Enhance get mtrr mask logic.
Vlv2TbltDevicePkg: Enhance get mtrr mask logic.
UefiCpuPkg CpuDxe: Enhance get mtrr mask logic.
QuarkSocPkg MtrrLib: Remove reference deprecated macro.
Cc: Jeff Fan
Cc: Ruiyu Ni
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Eric Dong
---
UefiCpuPkg/Include/Library/MtrrLib.h | 27 ---
1 file changed, 27 deletions(-)
diff --git
In order to not use the deprecated macro, refine
get mtrr mask value logic.
Cc: Jeff Fan
Cc: Ruiyu Ni
Cc: David Wei
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Eric Dong
---
Cc: Jeff Fan
Cc: Ruiyu Ni
Cc: David Wei
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Eric Dong
---
Vlv2TbltDevicePkg/PlatformInitPei/MemoryPeim.c | 10 ++
1 file changed, 6
Cc: Jeff Fan
Cc: Ruiyu Ni
Cc: Michael D Kinney
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Eric Dong
---
.../QuarkNorthCluster/Library/MtrrLib/MtrrLib.c| 26
In order to not use the deprecated macro, refine
get mtrr mask value logic.
Cc: Jeff Fan
Cc: Ruiyu Ni
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Eric Dong
---
UefiCpuPkg/CpuDxe/CpuDxe.c | 13 ++---
Cc: Jeff Fan
Cc: Ruiyu Ni
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Eric Dong
---
UefiCpuPkg/CpuDxe/CpuDxe.c | 22 +++---
UefiCpuPkg/CpuDxe/CpuDxe.h | 1 +
2 files changed, 12
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: zwei4
---
Platform/BroxtonPlatformPkg/BuildBios.bat | 2 +-
.../PlatformDsc/Components.IA32.dsc | 3 +++
.../BroxtonPlatformPkg/PlatformDsc/Components.dsc | 4
Bug: https://bugzilla.tianocore.org/show_bug.cgi?id=635
Cc: Liming Gao
Cc: Michael D Kinney
Cc: Jeff Fan
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Andrew Fish
---
Bug: https://bugzilla.tianocore.org/show_bug.cgi?id=635
Cc: Liming Gao
Cc: Michael D Kinney
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Andrew Fish
---
Bug: https://bugzilla.tianocore.org/show_bug.cgi?id=635
Cc: Ruiyu Ni
Cc: Hao Wu
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Andrew Fish
---
DuetPkg/EfiLdr/TianoDecompress.c | 6 +++---
1 file changed, 3
Bug: https://bugzilla.tianocore.org/show_bug.cgi?id=635
Cc: Liming Gao
Cc: Michael D Kinney
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Andrew Fish
---
BaseTools/Source/C/Common/Decompress.c
V2:
Fix build MdeModulePkg failure by use LShiftU64() rather than << operator.
Bug: https://bugzilla.tianocore.org/show_bug.cgi?id=635
Yonghong Zhu (4):
BaseTools: Fix Xcode 9 Beta treating 32-bit left shift as undefined
DuetPkg: Fix Xcode 9 Beta treating 32-bit left shift as undefined
Reviewed-by: Ye Ting
-Original Message-
From: Wu, Jiaxin
Sent: Wednesday, August 02, 2017 4:01 PM
To: edk2-devel@lists.01.org
Cc: Ye, Ting ; Fu, Siyuan ; Wu, Jiaxin
Subject: [Patch] NetworkPkg/HttpDxe:
Ting,
Thanks, I will update it when commit the patch.
Siyuan
-Original Message-
From: Ye, Ting
Sent: Thursday, August 3, 2017 4:37 PM
To: Fu, Siyuan ; edk2-devel@lists.01.org
Cc: Wu, Jiaxin
Subject: RE: [edk2] [Patch] NetworkPkg: iSCSI should
Hi Siyuan,
I would suggest to add more info to below error message, such as below:
Error! Only last 3 bytes are configurable, please input 6 hex numbers for last
3 bytes only or 12 hex numbers for full SSID!
+ L"Error! Only last 3 bytes are configurable, please input 6 or 12 hex
Reviewed-by: Ye Ting
-Original Message-
From: Fu, Siyuan
Sent: Thursday, July 27, 2017 11:44 AM
To: edk2-devel@lists.01.org
Cc: Ye, Ting ; Wu, Jiaxin
Subject: [Patch] NetworkPkg: Display HTTP redirection info to the screen if
The last 3 bytes of ISID should be able to changed by setting the keyword with
a value with length 6 (only last 3 bytes) or 12 (full ISID) according to the
keyword definition in UEFI configuration namespace website.
Cc: Ye Ting
Cc: Wu Jiaxin
V2: Refine the codes to make the logic clear.
In a form, some new menus may be dynamically inserted between highlight
menu and previous top of screen menu when some question are refreshed.
So the highlight menu and previous top of screen menu perhaps can't be
shown in one page. Existing codes
71 matches
Mail list logo