hi, experts:
i found 2 protocols definitions in EmbeddedPkg\Include\Protocol directory:
_EMBEDDED_EXTERNAL_DEVICE and EMBEDDED_DEVICE_PROTOCOL.
so, my question is:
Which component on an ARM board could produce _EMBEDDED_EXTERNAL_DEVICE
protocol?
Which component on an ARM board could produce EMB
Hi Nicolas,
Looking at which modules use those protocols, they look as two different
protocols.
EMBEDDED_EXTERNAL_DEVICE: the only module today that implements this protocol
is the OMAP3 PMIC.
A reason to use this driver could be for instance you write a driver that
depends on a
hi, experts:
i found IPV4 related drivers located at MdeModulePkg\Universal\Network
directory.
But IPV6 related drivers located at NetworkPkg.
Why not put IPV4 related drivers into NetworkPkg?
best wishes--
Live Security
hi, experts:
i found a EDK_GLOBAL marco definition in some python files.
but not find any code to use it.
so, it has been deprecated?!
if not deprecated, so how to use it?
best wishes,--
Live Security Virtual Conference
E
hi, experts:
SMM_ENTRY struct definition in IntelFrameworkPkg/Include/Protocol/LegacyBios.h.
SmmPort is defined to UINT8 type.(it seems CSM spec defined it to UINT16 length)
but some IHV's smm port is UINT16 length.
best wishes,-
hi, experts:
there are some pcd declarations in PcAtChipsetPkg.dec file:
gPcAtChipsetPkgTokenSpaceGuid.PcdIsaAcpiCom1Enable|TRUE|BOOLEAN|0x0003
gPcAtChipsetPkgTokenSpaceGuid.PcdIsaAcpiCom2Enable|TRUE|BOOLEAN|0x0004
i overrided these pcd's value in MyProject.dsc :
gPcAtChipsetPkgTokenSpa
own
SMM_TABLE in LegacyBiosPlatform module according to your platform
SmmAttributes, then build your own SmmTable and finally assign its pointer to
EFI_TO_COMPATIBILITY16_BOOT_TABLE->SmmTable.
Thanks
Elvin
From: nicolas [mailto:nicolas1...@163.com]
Sent: Monday, August 13, 2012 4:32 PM
hi, experts:
in Tiano code, 0x FFE0 saved the address of entry point of the PEI core.
But in ResetVectorVtf0.asm, this position's content is "jmp
EarlyApInitReal16".
the comments described:
This allows the Local APIC Startup IPI to be used to wake up the application
processors.
So, my que
hi, experts:
I am studying UEFI Spec.
It described PCI Option ROMs in chapter 13.4.2 .
it said a PCI OpROM could contain an IA-32 UEFI Driver.
Based on my understanding, UEFI Spec does not support IA32 platform.
Only EFI Spec has ever supported IA32 platform.
So, a PCI OpROM could only contain a
hi, Andrew:
Thanks!
Your explaination is very clear to me!
best wishes
发件人: Andrew Fish [mailto:af...@apple.com]
发送时间:2012年8月14日 14:44
收件人: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] FFE0 code in ResetVectorVtf0.asm
On Aug 13, 2012, at 11:00 PM, nicolas wrote:
hi
@lists.sourceforge.net
主题: Re: [edk2] IA-32 UEFI Driver question
The UEFI spec support IA-32, X64, ARM, and IPF processor bindings.
Andrew Fish
On Aug 14, 2012, at 8:02 PM, nicolas wrote:
hi, experts:
I am studying UEFI Spec.
It described PCI Option ROMs in chapter 13.4.2
hi, experts:
Does EDKII source code support console redirection?
Take my platform as an example, its GOP driver is not ready now.
This platform's com port is ready, and i could debug EDKII post debug message
to a terminal through this COM port.
I want to boot to shell through a terminal, and could
hi, experts:
I found a section in Chapter 4.6 in Build Spec 1.22 version:
..
The EDK II Compatibility Package provides libraries and header files to permit
building
some* EDK Libraries and EDK Components refere nced in an EDK II platform (DSC)
file.
Note: * indicates any EDK libraries or c
hi, experts:
I have some c-language functions, they were programmed with c standard lib
functions.
such as : strcpy, strcmp, printf etc.
Is there a possiblity to compile these c functions without modified in EDKII
environment?
I learned some compiler options letting edkii not depend system c li
hi, experts:
I found using "#pragma pack()" pre-compile syntax for some data structs, but
not for some other data struct in MdePkg\Include\Protocol\Dhcp6.h .
So, is there some considerations for using this pre-compile syntx?
If i used "#pragma pack()" pre-compile syntax for all my own defined d
hi, Andrew:
Thanks!
best wishes,
发件人: Andrew Fish [mailto:af...@apple.com]
发送时间:2012年8月24日 10:55
收件人: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] #pragma pack(1) question
Nicolas,
#pragma pack() aligns to the C ABI convention. #pragma pack(1) byte packs the
structure
hi, experts:
i want to disable /WX build option for a dedicated module, so i changed this
module's inf file:
..
[BuildOptions]
MSFT:*_*_IA32_CC_FLAGS = /Oi- /WX:NO
MSFT:*_*_X64_CC_FLAGS = /Oi- /WX:NO
..
but it had no effect!
so, how to disable /WX build option for a dedicated mo
hi, experts:
i am studying edkii build spec.
It talked about static link/dynamic link in chapter 9.4 .
It said:
1. for lib modules, it used static link as the last step.
2. for non-lib modules, such as drivers/app, static link step is optional.
so, it made me confused!
based on my understanding,
hi, Andrew:
Thanks for your explaination!
best wishes,
发件人: Andrew Fish [mailto:af...@apple.com]
发送时间:2012年9月1日 1:14
收件人: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] Static Link/Dynamic link question
Nicolas,
I think the terminology in the build spec. is confusing. Using the
hi, experts:
OvmfPkg compiled error:
... ...
Architecture(s) = IA32 X64
Build target = DEBUG
Toolchain= VS2008
Active Platform = e:\edkii\OvmfPkg\OvmfPkgIa32X64.dsc
Flash Image Definition = e:\edkii\OvmfPkg\OvmfPkgIa32X64.fdf
Processing meta-data .
Build...
e:\edkii\OvmfPkg\
hi, Ersek:
Thank you very much!
this patch is good for me!
i have finished compiling OvmfPkg.
best wishes,
-邮件原件-
发件人: Laszlo Ersek [mailto:ler...@redhat.com]
发送时间: 2012年9月5日 16:12
收件人: nicolas
抄送: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] OvmfPkg compiled error
On
hi, experts:
I tried to use Qemu+OVMF to test secure boot feature.
I followed the guide described in "Signing UEFI applications and drivers for
UEFI Secure Boot.pdf".
but failed to run HelloWorld.efi(signed with microsoft Windows SDK tools), this
error msg appeared in the shell:
Error reported :
hi, experts:
In UEFI Spec, User Identification topic is described in Chapter 31.
It's aslo related to Secure Boot.
So, could it be tested with OVMF or Nt32 Emulator?
best wishes,--
Live Security Virtual Conference
Exclusi
hi, experts:
It seams BlockIo driver in EDKII had supported asynchronous block io.
It seams this feature could reduce post time, so the effect will be obviously?
Why this feature could reduce post time?
During block io read/write, cpu could do other things in parallel?
best wishes
vision DB, KEK, & PK.
Lee
From: nicolas [mailto:nicolas1...@163.com]
Sent: Wednesday, September 05, 2012 3:13 AM
To: edk2-devel@lists.sourceforge.net
Subject: [edk2] Ovmf Secure Boot test question
hi, experts:
I tried to use Qemu+OVMF to test secure boot feature.
I followed the guide
hi, experts:
I added User identification related drivers into Nt32Pkg.dsc.
I followed these instructions described in this webpage:
http://sourceforge.net/apps/mediawiki/tianocore/index.php?title=How_to_Enable_Security
I could find an "User Manager" item in Device Manager.
But i could go into thi
hi, experts:
Sorry for sent it again!
--
I added User identification related drivers into Nt32Pkg.dsc.
I followed these instructions described in this webpage:
http://sourceforge.net/apps/mediawiki/tianocore/index.php?title=How_to_Enable_Security
I could find an "User Manager" ite
n codes with Nt32Pkg
Hi Nicolas,
I verify the User Identification base on the tip code, it works. Do you base on
the tip code to do your patch?
If yes, could you supply the patch you used, I can help to check it with your
patch.
Thanks,
Eric
From: nicolas [mailto:nicolas1...@163.
seems need to create a default user profile, so how to create it?
best wishes,
发件人: Dong, Guo [mailto:guo.d...@intel.com]
发送时间:2012年9月20日 13:21
收件人: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] Study User Identification codes with Nt32Pkg
Hi, Nicolas
From the log, Identify() in
hi, experts:
I tried to compile BaseTools in Ubuntu 11.10, but failed:
root@ubuntu1110:/home/lion/Origen0921# make -C edk2/BaseTools
... ...
make[2]: Entering directory
`/home/lion/Origen0921/edk2/BaseTools/Source/C/Common'
gcc -c -MD -fshort-wchar -fno-strict-aliasing -fno-merge-constants -nostd
e an issue occurred in setting ARCH in
BaseTools/Source/C/GNUmakefile.
-Jordan
On Sun, Sep 23, 2012 at 8:59 PM, nicolas wrote:
> hi, experts:
> I tried to compile BaseTools in Ubuntu 11.10, but failed:
> root@ubuntu1110:/home/lion/Origen0921# make -C edk2/BaseTools
> ... ..
Could SamsungPlatformPkg boot linux kernel now?
Does linaro provide new kernel version?
best wishes,
发件人: Olivier Martin [mailto:olivier.mar...@arm.com]
发送时间:2012年2月21日 19:32
收件人: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] UEFI ARM Studies
Hi Nicolas,
I hav
hi, Andrew:
I am running BeagleBoard's edk2 code with qemu.
The first 4 bytes of the FV header is a branch to the entry point of the SEC
PE/COFF Image.
So, which tool pads these 4 bytes into the FV header during producing the final
FD image?
I could not find them in the BaseTools directory.
b
hi, Andrew:
Got it!
Thank you very much!
best wishes,
发件人: Andrew Fish [mailto:af...@apple.com]
发送时间:2012年10月9日 16:27
收件人: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] The first 4 bytes in BeagleBoard's FV header question
On Oct 9, 2012, at 1:17 AM, nicolas wrote:
hi, Oliver:
Why not implement a Reset Ppi for Versatile Express board?
best wishes,--
Don't let slow site performance ruin your business. Deploy New Relic APM
Deploy New Relic app performance management and know exactly
w
hi, Oliver:
I have a problem about stack setup for an ARM platform.
Based on an OrigenBoard(with a DualCore , supports TrustZone), and use
SamsungPlatformPkg code.
During the boot procedure, there are 3 places to setup stacks:
1. _ModuleEntryPoint
Used these pcds to setup stack:
gArmPlat
hi, Oliver:
I build ArmVExpress-CTA9x4.dsc:
build -a ARM -p ArmPlatformPkg/ArmVExpressPkg/ArmVExpress-CTA9x4.dsc -t ARMGCC
-b DEBUG.
and tried to run it with qemu:
qemu-system-arm -M vexpress-a9 -m 1024 -serial stdio -mtdblock
SEC_ARMVEXPRESS_EFI.fd -kernel vmlinuz -initrd initrd.img.
It could
express boot
Hi Nicolas,
I have not tried to use qEmu to boot UEFI. But I would suggest you to remove
the parameters ‘-kernel vmlinuz -initrd initrd.img’ from your command line.
It is possible qEmu directly starts Linux if these parameters are passed.
Olivier
From: nicolas [mailto:nicolas1
ns you cannot run UEFI on qEmu VExpress L
From: nicolas [mailto:nicolas1...@163.com]
Sent: 15 October 2012 10:52
To: edk2-devel@lists.sourceforge.net
Subject: Re: [edk2] qemu emulates ARM versatile express boot
hi, Oliver:
My qemu version is: 1.2.0
If not pass -kernel param, it would tip
hi, Oliver:
I have a problem about stack setup for an ARM platform.
Based on an ArmRealViewEb-RTSM-A9x2.dsc.
During the boot procedure, there are 3 places to setup stacks:
1. _ModuleEntryPoint
Used these pcds to setup stack:
gArmPlatformTokenSpaceGuid.PcdCPUCoresSecStackBase|0x4B00
hi, experts:
In DxeIpl PEIM, before jumping DxeMain(), it will call UpdateStackHob()
function.
So, my question is:
1. When to create a Hob whose name is gEfiHobMemoryAllocStackGuid?
SEC phase or PEI earlier phase?
2. gEfiHobMemoryAllocStackGuid Hob's function?
Just saving this region to
hi, ARM experts:
I am studying ARM code.
I found ArmPlatformSec code had call ArmDisableInterrupts().
So, when to enable interrupts?
It seemed PEI phase did not need timer interrupt.
But DXE phase need timer interrupt to schedule tasks.
best wishes,--
hi, ARM experts:
I have a question about gArmGlobalVariableGuid hob.
It seemed PlatformPeim would create a gArmGlobalVariableGuid hob to describe
global variable area.
Based on dsc/dec file, it seemed only pre-reserved 0x04 bytes for global
variable on the top of primary core stack.
So, who wil
hi, Oliver:
I have question about InitializeExceptions function in
ArmPkg\Drivers\CpuDxe\Exception.c .
It will call ArmWriteVBar function.
But if an ARM CPU did not implement security extension, and CP15's C12 not
exist.
So how to initialize vector table at 0x (or 0x )?
It seems
hi, Oliver:
I have a question about Nor flash boot method.
Based on ArmPlatformLib-Full-Boot.png, it seems ARM Versatile PB supports total
EFI boot sequence.
So, my question is:
1. Cold reset will be treated as an exception.
So if user presses power button, it will search exception vector tabl
hi, Rony:
The problem still existed! (git clone still failed)
My os is : Ubuntu 11.10 / git version is 1.7.5.4
Girish has ever created a git website before:
git clone git://github.com/girishks/Origen-EDK-II-Package.git
So, what is the difference between linaro's and Girish's package?
dler".
best wishes,
At 2012-11-07 16:22:40,nicolas wrote:
hi, Oliver:
I have a question about Nor flash boot method.
Based on ArmPlatformLib-Full-Boot.png, it seems ARM Versatile PB supports total
EFI boot sequence.
So, my question is:
1. Cold reset will be treated as an exception.
, nicolas wrote:
hi, Rony:
The problem still existed! (git clone still failed)
My os is : Ubuntu 11.10 / git version is 1.7.5.4
Girish has ever created a git website before:
git clone git://github.com/girishks/Origen-EDK-II-Package.git
So, what is the difference between linaro's and Gir
hi, Oliver:
Based on ArmPlatformPkg\Documentation\ArmPlatformLib-Full-Boot.png, these
functions listed in SEC, should be declared in ArmPlatform.h .
But It seems current ArmPlatformLib.h has some changes.
Some functions listed in png picture, but not declared in ArmPlatformLib.h .
best wishes,--
hi, Oliver:
I have 2 questions about arm compiler options:
1. no_unaligned_access
I found RVCT set --no_unaligned_access in tools_def.txt
But ARMGCC not set?
2. How to ensure current ARM's code is aligned on 4 bytes boundary?
Have current code been aligned on 4 bytes boundary?
bes
-devel@lists.sourceforge.net
主题: Re: [edk2] ARM Compiler option question
Hi Nicolas,
1. From the UEFI specification: “Unaligned access should be enabled if
supported; Alignment faults are enabled otherwise.”
FYI, On ARMv7 with MMU enabled the CPU handles unaligned accesses (at the
hi, Experts:
I am studying ARM UEFI code.
In sec code, it would call ArmDisableInterrupts to disable IRQ/FIQ in CPSR.
After that, i could not find any code to enable IRQ/FIQ interrupts in Dxe phase
code.
Why?
best wishes,---
hi, Oliver:
Based on ArmPlatformPkg\Documentation\ArmPlatformLib-Full-Boot.png, it's a
complete UEFI boot on versatile board, not depend on any 3rd party binary code.
But based on "Secure Board Boot on ARM Reference Hardware System Softeware on
ARM.pdf" (ARM document,doc number is: ARM DEN 0006A)
hi, Oliver:
ArmEnableIrq() function declaration should be added into ArmLib.h .
Maybe user just wants to disable irq only.
best wishes,--
Monitor your physical, virtual and cloud infrastructure from a single
web console.
hi, experts:
I am using ARMGCC toolchains(linux version) to compile an ARM project.
I found the debug version binaries were too big on ARM platform.
Maybe developers just want to add some debug messages output from serial port.
But if developers compiled debug version, these drivers image were stil
hi, experts:
I found a line in tools_def.txt :
DEBUG_*_*_OBJCOPY_ADDDEBUGFLAG =
--add-gnu-debuglink=$(DEBUG_DIR)\$(MODULE_NAME).debug
Will it merge debug info file into efi binary?
Or it will produce a separate debug info file?
best wishes,--
[mailto:af...@apple.com]
发送时间:2012年11月27日 4:05
收件人: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] objcopy option "--add-gnu-debuglink"
Nicolas,
I don't know the specific answer for your tools, but I can give you a hint on
how things work in general.
The general answer is
ArmCpuDxe.dll .
If i want to use gdb to debug ArmCpuDxe.efi, what is the steps?
best wishes,
发件人: Andrew Fish [mailto:af...@apple.com]
发送时间:2012年11月29日 17:39
收件人: edk2-devel@lists.sourceforge.net
主题: Re: [edk2] objcopy option "--add-gnu-debuglink"
On Nov 29, 2012, at 1:14 A
hi, Stephen:
PE header has an item to indicate it's 32bit / x64 / ARM .
Every efi driver binary is PE format.
best wishes,
-邮件原件-
发件人: Stephen Polkowski [mailto:step...@centtech.com]
发送时间: 2012年12月4日 0:03
收件人: edk2-devel@lists.sourceforge.net
主题: [edk2] How to determine UEFI arch
hi, Oliver:
I found there was a potential bug if MPCore not support TrustZone.
CEntryPoint() function in Sec.c :
1. ArmCallSEV() (line 62)
It will signal BOOT_MEM_INIT event.
then, the secondary core's code in SecEntryPoint.S would begin to run.
And, the secondary core could aslo ru
hi, Rony:
I followed these steps described in this web page to compile an EDKII Origen
Pkg:
https://wiki.linaro.org/Boards/Origen/Setup/UEFI
But it booted failed in DXE Phase, the log message is:
Boot firmware (version built at 19:37:38 on Apr 11 2013)
Trust Zone Configuration is disabled
.Time
61 matches
Mail list logo
