on mismatch?
If you pull a clean copy out of svn does it work?
Thanks,
Andrew Fish
> Updating was working fine up to the prior update (Feb
> 2014). Then this latest update has the EDK2 main items (source)
> updating with no problem. It's when it went to deal with the
> extern
On Oct 12, 2014, at 12:42 AM, David F. wrote:
> That didn't work - creating the fake folders
> (C:\edk2\BaseTools\Bin\Win32\BaseTools\Bin\Win32\Arm\RELEASE_RVCT31,
> etc..) gets rid of the error:
>
> C:\edk2\BaseTools\Bin\Win32
> The node
> 'C:\edk2\BaseTools\Bin\Win32\
> BaseTools\Bin\Win32\Ar
in svn by default.
Thanks,
Andrew Fish
On Oct 11, 2014, at 10:40 PM, David F. wrote:
> As a follow up to my prior message below - if I put the old c:\edk2
> folder back and attempt to build, it works fine. So definitely
> something in one of the updates since 2/25/2014 (that is the
s the PCD could be useful for debugging a specific issue as well.
As long as the reasoning behind the default value for the “canary” is
documented in the PCD default I don’t see an issue using a FixedAtBuildPcd.
Thanks,
Andrew Fish
> Olivier
>
> From: Andrew Fish [mailto:af...@apple
compatible way to get a
random number?
The per boot random number makes it harder for attacking code to figure out
where the “canary” is so it can be defeated. It is like ASLR, and a constant
PCD value is not going to help with that.
Thanks,
Andrew Fish
> Olivier
>
> From:
I’d like to see more flexibility too, but how do we tell the difference between
this caae and a bug where the DSC and FDF are out of sync?
Thanks,
Andrew Fish
> Thanks,
>
> Mike
>
> From: [email protected] [mailto:[email protected]]
> Sent: Thursday, July 17, 2014 9:
l/UefiShell.inf
[Binaries.Ia32]
PE32|Ia32/Shell.efi|*
[Binaries.X64]
PE32|X64/Shell.efi|*
[Binaries.ARM]
PE32|Arm/Shell.efi|*
[Binaries.AArch64]
PE32|AArch64/Shell.efi|*
Thanks,
Andrew Fish
> It seems that GenFds quits when it can't find the first source file listed
> in the fi
it and it
builds the tools, if needed, and the platform.
> Thanks,
>
> Mike
>
> From: Andrew Fish [mailto:[email protected]]
> Sent: Monday, July 14, 2014 2:42 PM
> To: Kinney, Michael D
> Cc: [email protected];
> [email protected]
rnal. If you don’t pull the external I
guess it is un-versioned.
Thanks,
Andrew Fish
> Thanks,
>
> Mike
>
> From: Andrew Fish [mailto:[email protected]]
> Sent: Monday, July 14, 2014 12:21 PM
> To: Kinney, Michael D
> Cc: [email protected];
> edk2-bu
to disconnecting from the network. It would be part of the download
instructions.
Thanks,
Andrew Fish
--
Want fast and easy access to all the code in your enterprise? Index and
search up to 200,000 lines of code wit
ll be generated in a
> directory that is not under source control, so no file differences will be
> shown.
>
But what happens with the previous plan. It seems like if you build binaries
you would get svn status notifications for all the binaries.
> Best regards,
>
>
rsion of svn to pull for the tools.
I like this as:
1) Checking in binaries is evil, and not allowed in some production
environments, so folks end up solving this problem anyway.
2) It make the instructions for using the edk2setup.sh path easier.
Thanks,
Andrew Fish
PS What happens to th
sume that git works out
> of the box on OS X.))
Yes it does.
Thanks,
Andrew Fish--
___
edk2-buildtools-devel mailing list
[email protected]
https:/
lopers that do not want Win32 binaries can opt-out by ignoring
> externs.
How does this work with a git mirror? Is it possible to setup an officially
supported git mirror for the svn?
Thanks,
Andrew Fish
> c. Date: TBD. Goal is immediately after build server is stable.
> 4) M
I ran into some 3rd party INF files that use mixed case for the FILE_GUID and
this was confusing my debugger Python scripts. I think it is better for the
Guid.xref file to output the GUID strings in upper case.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Andrew Fish
ev.org/GCC_Stack_Smashing_Protector. So it helps contains read
based overruns.
__stack_chk_guard is the “canary” placed on the stack by the compiler.
__stack_check_fail() is called if the “canary” has been over written. These are
both compiler intrinsics.
Thanks,
Andrew Fish
~/work/Compil
On Jun 27, 2014, at 3:08 PM, Jordan Justen wrote:
> On Fri, Jun 27, 2014 at 1:54 PM, Andrew Fish wrote:
>> I though I would share some useful git aliases with folks.
>>
>> Have you ever wondered how many instances (different implementations) of a
>> library clas
moryLib.inf:25: LIBRARY_CLASS
= BaseMemoryLib|DXE_DRIVER DXE_RUNTIME_DRIVER DXE_SAL_DRIVER DXE_SMM_DRIVER
UEFI_APPLICATION UEFI_DRIVER
Thanks,
Andrew Fish
PS The “cd ${GIT_PREFIX:-.}” is need for the commands run from the current
working directory. If you always want to search the entir
It is a little unclear from the FDF spec if FvBaseAddress and FvForceRebase
work in a stand alone [FV] section, or if they require an [FV] to be part of an
FD to get relocated. What modes are supported?
Thanks,
Andrew Fish
an
example for current systems.
Thanks,
Andrew Fish
--
"Accelerate Dev Cycles with Automated Cross-Browser Testing - For FREE
Instantly run your Selenium tests across 300+ browser/OS combos.
Get unparalleled scala
evision 2660.
Thanks,
Andrew Fish
>Reviewed-by: Gao, Liming
>
> Thanks
> Liming
> From: Andrew Fish [mailto:[email protected]]
> Sent: Monday, April 07, 2014 7:16 AM
> To: [email protected]
> Cc: [email protected]
> Sub
fast. I went for readable (to a C
programmer) code. You could probably also use
GuidDict.update(Module.Protocols) like syntax too?
Thanks,
Andrew Fish
>svn diff --diff-cmd diff Source/Python/GenFds/GenFds.py
Index: Source/Python/GenFds/GenFds
he cache that the build system uses.
Try removing the Conf/.cache directory and rebuilding.
Thanks,
Andrew Fish
>
> ``` plain
> Generating UNSIGNEDFV FV
>
>
> GenFds.EXE...
> : error C0DE: Tools code failure
> Please send email to edk2-buildtools-dev
rties set:
~/work/edk2/BaseTools/Bin/Win32>svn proplist build.exe
Properties on 'build.exe':
svn:executable
svn:mime-type
~/work/edk2/BaseTools/Bin/Win32>svn proplist LzmaF86Compress.bat
~/work/edk2/BaseTools/Bin/Win32>
Thanks,
Andrew Fish
> Thanks,
> Hot
>
> F
scripts with CRLF crash on Mach OS X. So they need to be LF to function.
In general it seems there are a lot of editors that get confused with a mixed
scheme, so cleaning up the miss matches is a good idea. But to me it seems the
requiremen
be used from the command line, or build
script, to set the tools version. The actual tools now live under the Xcode
Applications directory structure.
Thanks,
Andrew Fish
On Jan 15, 2014, at 4:43 AM, Olivier Martin wrote
ython it is really more about the OS your are building from.
Cross compilers do exist, so it is best not to make assumptions.
Thanks,
Andrew Fish
> GARRETT KIRKENDALL
> SMTS Firmware Engineer | AMD Technology & Engineering
> 7171 Southwest Parkway, Austin, TX 78735 USA
>
On Jan 8, 2014, at 10:00 PM, Liu, Yingke D wrote:
> Hi Andrew,
>
> Thanks for your explanation. The patch is attached according to your comments.
>
Dennis,
The patch looks good.
Thanks,
Andrew Fish
> T
This statement is not portable in Python and it crashes on a Mac since it tries
to read 20 bytes.
struct.unpack("LHH", DepexFile.read(16))
The problem with unpack() is this form uses native size and alignment and
unsigned long is a different size on Windows and Mac OS (64-bit Python).
CAP = TRUE
+WRITE_LOCK_STATUS = TRUE
+READ_DISABLED_CAP = TRUE
+READ_ENABLED_CAP = TRUE
+READ_STATUS= TRUE
+READ_LOCK_CAP = TRUE
+READ_LOCK_STATUS = TRUE
#
Thanks,
Andrew Fish
PS I was work
lude-dir=/usr/include/c++/4.2.1
Apple LLVM version 5.0 (clang-500.2.79) (based on LLVM 3.3svn)
Target: x86_64-apple-darwin13.0.2
Thread model: posix
On Dec 5, 2013, at 10:33 AM, Andrew Fish wrote:
>
> On Dec 5, 2013, at 9:51 AM, Michael Berman
> wrote:
>
>> Hello,
>>
o"
>
> Before I jump into antlr.c and err.c, I was hoping someone could perhaps
> point me at the recipe for building on current OSX.
>
I’ll take a look when I get a chance.
I had to add -Wno-self-assign -Wno-deprecated-declarations to turn off some
warnings. And I also seem t
ight thing.
Thanks,
Andrew Fish
On Dec 4, 2013, at 10:05 AM, Andrew Fish wrote:
> When I build a log file I noticed that it has line ending issues. Most of the
> file seems to have 0x0d 0x0a, but some cases have 0x0a 0x0d 0x0a which my
> editor thinks is a upside-down question mark
3 | 0x85, 0x4E, 0x3|
3430 37 20 7d 0d 0a 20 20 20 20 50 63 64 49 73 61 42 |7 }..PcdIsaB|
Thanks,
Andrew Fish
PS
(master)>python
Python 2.7.5 (default, Sep 12 2013, 21:33:34)
[GCC 4.2.1 Compatible Apple LLVM 5.0 (clang-500.0.68)] on darwin
Type "help", "copyright&q
On Nov 6, 2013, at 11:41 PM, Andrew Fish wrote:
>
> On Nov 6, 2013, at 11:34 PM, Gao, Liming wrote:
>
>> Andrew:
>> EDKII build system bases on file postfix name to handle them. For new
>> postfix name, you may extend build_rule.txt file to add new rule for it.
do this with the rules that exist. It seems
libraries are special cased in the Python code.
Thanks,
Andrew Fish
> Thanks
> Liming
> -Original Message-
> From: Andrew Fish [mailto:[email protected]]
> Sent: Thursday, November 07, 2013 11:54 AM
> To: edk2-buildtools-deve
to do this with the edk2 build system.
Thanks,
Andrew Fish
--
November Webinars for C, C++, Fortran Developers
Accelerate application performance with scalable programming models. Explore
techniques for threading, error
Robert,
Try deleting your Conf/.cache directory and building again.
Thanks,
Andrew Fish
On Oct 11, 2013, at 9:07 AM, "Fendt, Robert (GE Germany)"
wrote:
>
--
October Webinars: Code for Performan
t looked at, I'm guessing that the
OUTPUT_DIRECTORY = value in the [Defines] section has a space in it? Or there
are \ path separators in the file and not /.
Thanks,
Andrew Fish
--
LIMITED TIME SALE - Full Year of
Sent from my iPhone
On Sep 11, 2013, at 6:41 AM, Anthony Ross wrote:
> Hello one and all
> I have been working on building the Coreboot Package from EDK 2 source code
> in Ubuntu 12.04 but it has been failing repeatedly. Well earlier after
> setting up the EDK 2 environment and entering it (
Tools/Conf/tools_def.template.
Thanks,
Andrew Fish
> Thanks
> Liming
> -Original Message-
> From: Andrew Fish [mailto:[email protected]]
> Sent: Thursday, August 22, 2013 12:04 AM
> To: Gao, Liming
> Cc: Justen, Jordan L; [email protected];
> ed
y does it take a year to add a new complier to the edk2, or even a new
version of a compiler?
Thanks,
Andrew Fish
> One main point to separate BaseTools project is to make sure EDKII BaseTools
> is a stable version. BaseTools is different from Shell. If BaseTools has the
> issu
On Aug 20, 2013, at 4:30 PM, Jordan Justen wrote:
> On Tue, Aug 20, 2013 at 3:47 PM, Andrew Fish wrote:
>> On Aug 20, 2013, at 3:41 PM, Jordan Justen wrote:
>>> On Tue, 2013-08-20 at 09:45 -0700, Andrew Fish wrote:
>>>> The current BaseTools process does not see
On Aug 20, 2013, at 3:41 PM, Jordan Justen wrote:
> On Tue, 2013-08-20 at 09:45 -0700, Andrew Fish wrote:
>> The current BaseTools process does not seem to work for any one
>> who does not have a @intel.com email address, and that seems like
>> a broken thing for a
o I'm
going to run into the same issue as you soon.
The current BaseTools process does not seem to work for any one who does not
have a @intel.com email address, and that seems like a broken thing for an open
source project.
Thanks,
Andrew Fish
On Aug 20, 2013, at 4:24 AM, Olivier Mar
t items.
Given EFI is little-endian it may be better to use "<" vs. "=" so we never have
an endian issue.
Thanks,
Andrew Fish
> But then the code that builds the "DepexFile" needs to be updated as well.
>
> I guess to be truly portable, we would need t
e Clang 4.0 (tags/Apple/clang-418.0.60)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import struct
>>> struct.calcsize ("L")
8
>>> struct.calcsize ("I")
4
Thanks,
Andre
This is not an issue with the edk2 tools. Theses relocations deal with dynamic
linking so this is an issue with your compiler flags.
In a static link image you don't need a PC relative relocation. The Global
Offset Table is about patching globals. These concepts don't exist in PE/COFF
relocati
On May 28, 2013, at 2:32 PM, Jordan Justen wrote:
> On Tue, May 28, 2013 at 12:26 PM, Andrew Fish wrote:
>> On May 28, 2013, at 12:10 PM, Jordan Justen wrote:
>>>> Anyway, I am not sure GCC for ARM architectures supports PE/COFF (I guess
>>>> that would &
ou talking about?
I vote to leave ELF in the name as it is the debug format being used. Well I
guess we could call it DWARF to be pedantic, but Mach-O uses DWARF too, so ELF
better maps what the debugger knows how to deal with.
Thanks,
Andrew Fish
--
; From: Lee mike [mailto:[email protected]]
> Sent: Monday, May 06, 2013 6:29 AM
> To: Andrew Fish
> Cc: [email protected]
> Subject: Re: [edk2-buildtools] Re-define element in [BuildOptions]
>
> Hi Andrew,
>
> Sorry for my unclear s
Andrew Fish
On May 4, 2013, at 11:18 PM, Lee mike wrote:
> Hi,
>
> I wrote code to .dsc as below:
>
> [BuildOptions]
> MSFT:*_*_*_CC_FLAGS = /D DEBUG1
> MSFT:*_*_*_CC_FLAGS = /D DEBUG2
> MSFT:*_*_*_CC_FLAGS = /D DEBUG3
>
> The output mak
The only other generic thing to try is removing the Conf/.cache/ directory, and
trying again.
Maybe one of the tools developers will be able to give more detail?
Andrew Fish
On Apr 3, 2013, at 9:02 PM, "Mielicki, Lukasz"
wrote:
> Nope. Just one build at time (running on mult
Are you starting multiple builds in the same workspace?
Sent from my iPhone
On Apr 3, 2013, at 8:04 PM, "Mielicki, Lukasz"
wrote:
> Hi,
>
> Is there a solution for the following build issue? I'm hitting this almost
> every build.
> Must be some race condition in the build tool.
>
> build...
Tomer,
This list is to talk about the edk2 tools in the BaseTools. So questions like
this should go to the edk2 dev list.
GCC requires correct EFIAPI usage to generate correct code. Visual Studio and
clang generate the correct ABI without requiring EFIAPI, so this is the most
common error to l
after the build we copy our debugger scripts into this
directory. We launch lldb from the Symbols directory and our symbol loading
scripts know how to find the symbols.
Basically I don't mind overriding the Conf/build_rules.txt file, but I'd rather
not have to mess with the Python.
T
subprocess that executes the
$(CC) command and it is not global.
Thanks,
Andrew Fish
> Thanks
> Liming
> From: Andrew Fish [mailto:[email protected]]
> Sent: Friday, March 08, 2013 2:41 AM
> To: [email protected]
> Subject: [edk2-buildtools] Request to add
lease let me know.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Olivier Martin
Andrew Fish
--
Symantec Endpoint Protection 12 positioned as A LEADER in The Forrester
Wave(TM): Endpoint Security,
What is the correct process to update the *.template files?
These updates are required to process ACPI files with Xcode.
Thanks,
Andrew Fish
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: [email protected]
~/work/BaseTools>svn diff --diff-cmd diff
Index: C
On Dec 24, 2012, at 1:05 PM, Jordan Justen wrote:
> On Mon, Dec 24, 2012 at 7:35 AM, Ben Schroeder wrote:
>> What is the correct method for building a driver without debug symbols?
>>
>> I imagined RELEASe builds will not include debug symbols.
>>
>> But it seems that changing Conf/target.txt
,
Andrew Fish
On Dec 5, 2012, at 12:03 PM, Hauch, Larry wrote:
> Hi Tim,
> You are correct in what the INF specification states. Unfortunately, this
> line was added accidentally.
> As a future enhancement, we had planned on adding the depex statements “in
> comments” undern
Sounds like a bug to me.
Thanks,
Andrew
On Dec 3, 2012, at 3:00 PM, [email protected] wrote:
> Hi all,
> I have some binary-only drivers with dependency expressions that are not
> being processed. These drivers are not UEFI model drivers.
>
> I see in WorkspaceDatabase.py:
> # If the
On Oct 24, 2012, at 3:37 PM, Szankin, Maciej wrote:
> Is there any build-in config option that allows me to merge EDK output image
> with OS image?
An EFI image is a PE/COFF executable. What do you mean by OS image? Executable
formats vary by OS, ELFMach-O/PE/COFF, and I don't see the point of
Alok,
On a side note, you usually get better compression if you compress the entire
FV. A file in the FV can contain a compressed FV. This could be another
workaround.
Sent from my iPhone
On Sep 12, 2012, at 9:57 AM, "Pant, Alok" wrote:
> Justin, Thanks. Appreciate your suggestions
>
> A)
.
The error reporting from the Python tools is not the greatest and some times it
is good to look at the line before and the line after the error.
Andrew Fish
On Sep 11, 2012, at 3:56 PM, Pant, Alok wrote:
> Hi all,
> I am using UDK.SPI.UP1 source code to build some SMM drivers. Everyth
orking.
Andrew Fish
On Aug 1, 2012, at 5:55 PM, Melwyn DSouza wrote:
> I downloaded and installed EDK 2 development kit from the website.
>
> I tried to follow the instructions here
>
> http://sourceforge.net/apps/mediawiki/tianocore/index.php?title=Xcode
>
> XCode version
)
${d_path}(+)${s_base}.iii
Andrew Fish
On Jul 31, 2012, at 6:32 PM, Melwyn DSouza wrote:
>
>
> On Tue, Jul 31, 2012 at 5:42 PM, Andrew Fish wrote:
>
> On Jul 31, 2012, at 5:29 PM, Melwyn DSouza wrote:
>
>> On Tue, Jul 31, 2012 at 4:41 PM, Andrew Fish wrote:
>
On Jul 31, 2012, at 5:29 PM, Melwyn DSouza wrote:
> On Tue, Jul 31, 2012 at 4:41 PM, Andrew Fish wrote:
>
> On Jul 31, 2012, at 4:32 PM, Melwyn DSouza wrote:
>
>>
>>
>> On Mon, Jul 30, 2012 at 5:44 PM, Andrew Fish wrote:
>>
>> On Jul 30, 2012,
On Jul 31, 2012, at 4:32 PM, Melwyn DSouza wrote:
>
>
> On Mon, Jul 30, 2012 at 5:44 PM, Andrew Fish wrote:
>
> On Jul 30, 2012, at 4:49 PM, Melwyn DSouza wrote:
>
>> I have an (yasm) asm file which i need to have it as a part of Module
>> Application. It is
On Jul 30, 2012, at 4:49 PM, Melwyn DSouza wrote:
> I have an (yasm) asm file which i need to have it as a part of Module
> Application. It is of yasm file format,
> Yasm file format
> http://yasm.tortall.net/
>
> edk 2 build tools point to x86_amd64 format
>
> as seen in tools_def.txt
>
>
70 matches
Mail list logo
