Re: [edk2] [PATCH] MdeModulePkg DebugSupportDxe: Fix build error with GNU assembler

> On Oct 21, 2014, at 11:27 PM, Jordan Justen wrote: > > On Sun, Oct 19, 2014 at 7:09 PM, Fan, Jeff > wrote: >> SourceLevelDebugPkg/DebugAgentLib x64 also used movw %es, %rax. >> We need to fix it also. >> >> And UefiCpuPkg/Library/CpuExceptionLib x64 usedmovl

Re: [edk2] [PATCH] MdeModulePkg DebugSupportDxe: Fix build error with GNU assembler

On Sun, Oct 19, 2014 at 7:09 PM, Fan, Jeff wrote: > SourceLevelDebugPkg/DebugAgentLib x64 also used movw %es, %rax. > We need to fix it also. > > And UefiCpuPkg/Library/CpuExceptionLib x64 usedmovl%es, %eax. > We could clean it to make a consistence in code base. More than that, it appear

Re: [edk2] EDK2 Developer Tools for Windows

On Tue, Oct 21, 2014 at 1:22 AM, Gao, Liming wrote: > Scott: > I will download them and try again. > > For the patch, I provide another compatible version in previous mail. Have > you any comment for it? > > Linux user doesn't need to set GCC49_BIN, GCC49_DLL env, because /usr/bin/ is > the

Re: [edk2] [EDK2] Multi-processor Init in UEFI code

Hi, 1. Bay Trail does not belong to Xeon processors. So, It is not 2) and not 3). 2. Hardware may place APs in wait for SIPI state from IA32 Manual when power-on, and only BSP runs the code. >From BIOS review, BSP may send SIPI to wake up APs and place APs in hlt state >in one

Re: [edk2] [EDK2] Multi-processor Init in UEFI code

Hi, Johnson: Thanks for your reply! I reviewed IA32 Manual, same as you said: BSP is hardware selection when powering on a Multi-processor cores, software need do nothing! Best wishes, -邮件原件- 发件人: Brian J. Johnson [mailto:bjohn...@sgi.com] 发送时间: 2014年10月22日 0:36 收件人: edk2-devel@lists.so

Re: [edk2] EDK2 Developer Tools for Windows

Hello Liming, Ah, yes. Why didn't I think of that! I retested on Windows and Linux using your patch below and it all passes. The GenMake.py patch and the corresponding update of build.exe are still needed. Thanks, Scott -Original Message- From: Gao, Liming [mailto:liming@intel.com]

Re: [edk2] PPIs passed into the PEI Core entry point get zero'ed out when permanent memory is installed?

> On Oct 21, 2014, at 1:45 AM, Gao, Liming wrote: > > Andrew: > After memory is ready, you locates this PPI again. Then, you find this PPI > can't be found or this PPI is found, but its pointer is to Zero? If this PPI > could be found, could you check whether the pointer is migrated? > The

Re: [edk2] Firmware Volume Block Protocol

> On Oct 21, 2014, at 8:08 PM, Narinder Dhillon wrote: > > Hi Andrew, > > It is a controller on PCIe bus, behind a PCIe root. > If the device only supports a single eMMC/SD card, then you can use the PCI IO handle device path. If it supports multiple eMMC/SD cards then you can use the PCI I

Re: [edk2] EDK2 Developer Tools for Windows

Scott: Could we configure GCC49_DLL like below to resolve this issue? set GCC49_DLL=D:\edk2build\tools\gcc491-x86\dll; D:\edk2build\tools\gcc491-x86\bin Thanks Liming -Original Message- From: Scott Duplichan [mailto:sc...@notabs.org] Sent: Wednesday, October 22, 2014 3:28 AM To: edk2-

Re: [edk2] Firmware Volume Block Protocol

Hi Andrew, It is a controller on PCIe bus, behind a PCIe root. Thanx, On Tue, Oct 21, 2014 at 5:52 PM, Tian, Feng wrote: > As far as I know, UEFI doesn’t define such device path for eMMC or SD > device. So it’s hard to work out a generic/cross-platform solution. > > > > But you could develop

Re: [edk2] Firmware Volume Block Protocol

As far as I know, UEFI doesn't define such device path for eMMC or SD device. So it's hard to work out a generic/cross-platform solution. But you could develop a platform-specific solution, such as using Vendor Device Path or Controller Device Path, to represent the location of eMMC device. Fro

Re: [edk2] Firmware Volume Block Protocol

> On Oct 21, 2014, at 5:31 PM, Narinder Dhillon wrote: > > Hi All, > > I am trying to implement a non-volatile variable storage at the top end of an > eMMC flash device. From all the reading, I have implemented the firmware > volume block protocol in the eMMC driver. > What I can't figure is,

[edk2] Firmware Volume Block Protocol

Hi All, I am trying to implement a non-volatile variable storage at the top end of an eMMC flash device. From all the reading, I have implemented the firmware volume block protocol in the eMMC driver. What I can't figure is, what kind of device path do I install along with FVB protocol ? I have se

Re: [edk2] EDK2 Developer Tools for Windows

Hello Liming, For your first patch, I have to change lines 4492,4493 from DEF to ENV. After that, Windows hosted x86 builds work. Windows hosted ARM/AARCH fail because "echo" is not found (no echo.exe in the path). Linux hosted builds look OK. Thanks, Scott -Original Message- From: Gao

Re: [edk2] EDK2 Developer Tools for Windows

Hello Liming, The build_rule.template part of this revised patch looks good. It removes the need for the non-ipf symrename echo work around, and makes the build output easier to look at. The new tools_def.template patch almost works. I get an echo fail for Windows hosted ARM/AARCH64 builds: "e

Re: [edk2] [EDK2] Multi-processor Init in UEFI code

On 10/20/2014 05:50 AM, tiger...@via-alliance.com wrote: > Hi, experts: > > I am studying Intel 64 and IA-32 Architectures Software Developer’s > Manual Volume 3A.pdf . > > I have a few questions about Multi-processor management. > > 1.7.5 Multi-processor initialization In Chapter 7 > > …… > > The

[edk2] BDS: should BootOrder or OsIndications take priority?

Section "7.5.4 Exchanging information between the OS and Firmware" in the UEFI spec seems to imply that if bit 0 is set in both OsIndicationsSupported and OsIndications, then at next boot the firmware should stop at the firmware UI. It appears that this OS-side request is independent of, and prevai

Re: [edk2] [PATCH 2/7] OvmfPkg Sec: Convert Ia32/SecEntry.asm to NASM

Jordan: SecMain.inf can list SecEntry.nasm once without tool chain family. For example: [Sources.IA32] - Ia32/SecEntry.asm | MSFT - Ia32/SecEntry.asm | INTEL - Ia32/SecEntry.S | GCC + Ia32/SecEntry.nasm Thanks Liming -Original Message- From: Jordan Justen [mailto:jordan.l.jus...@

Re: [edk2] EDK2 Developer Tools for Windows

Scott: Update the patch for BaseTools/Conf files. 1. Update build_rule.template to add the specific rule for IPF arch. Then, the specific echo is not required. 2. Update tools_def.template to configure GCC49_BIN and GCC49_DLL. It is the compatible change. Index: build_rule.template

Re: [edk2] PPIs passed into the PEI Core entry point get zero'ed out when permanent memory is installed?

Andrew: After memory is ready, you locates this PPI again. Then, you find this PPI can't be found or this PPI is found, but its pointer is to Zero? If this PPI could be found, could you check whether the pointer is migrated? Thanks Liming -Original Message- From: Andrew Fish [mailto:af

Re: [edk2] EDK2 Developer Tools for Windows

Scott: I will download them and try again. For the patch, I provide another compatible version in previous mail. Have you any comment for it? Linux user doesn't need to set GCC49_BIN, GCC49_DLL env, because /usr/bin/ is the default system PATH. Windows user need to set GCC49_BIN, GCC49_DL