On 02/23/17 02:14, Gao, Liming wrote:
> Laszlo:
>   Yonghong has sent the another patch to its regression issue. Could you 
> verify it?

Yes, thanks, I'll check that out soon.

(Also, I'm sorry about reporting this after Ard's report; I was very
busy last night and sort of looked at Thunderbird in write-only mode...)

Thanks
Laszlo



> -----Original Message-----
> From: Laszlo Ersek [mailto:ler...@redhat.com] 
> Sent: Thursday, February 23, 2017 9:12 AM
> To: Zhu, Yonghong <yonghong....@intel.com>; edk2-de...@ml01.01.org
> Cc: Gao, Liming <liming....@intel.com>
> Subject: Re: [edk2] [Patch] BaseTools: add error check for Macro usage in the 
> INF file
> 
> On 02/23/17 02:02, Laszlo Ersek wrote:
>> Hi,
>>
>> On 02/21/17 02:18, Yonghong Zhu wrote:
>>> Use of MACRO statements in the EDK II INF files is limited to local
>>> usage only; global or external macros are not permitted. This patch
>>> add the check for not defined macros.
>>>
>>> Cc: Liming Gao <liming....@intel.com>
>>> Contributed-under: TianoCore Contribution Agreement 1.0
>>> Signed-off-by: Yonghong Zhu <yonghong....@intel.com>
>>> ---
>>>  BaseTools/Source/Python/Workspace/MetaFileParser.py    | 9 ++++++++-
>>>  BaseTools/Source/Python/Workspace/WorkspaceDatabase.py | 4 +++-
>>>  2 files changed, 11 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/BaseTools/Source/Python/Workspace/MetaFileParser.py 
>>> b/BaseTools/Source/Python/Workspace/MetaFileParser.py
>>> index 1a5fdf5..37a7f5d 100644
>>> --- a/BaseTools/Source/Python/Workspace/MetaFileParser.py
>>> +++ b/BaseTools/Source/Python/Workspace/MetaFileParser.py
>>> @@ -1,9 +1,9 @@
>>>  ## @file
>>>  # This file is used to parse meta files
>>>  #
>>> -# Copyright (c) 2008 - 2016, Intel Corporation. All rights reserved.<BR>
>>> +# Copyright (c) 2008 - 2017, Intel Corporation. All rights reserved.<BR>
>>>  # (C) Copyright 2015-2016 Hewlett Packard Enterprise Development LP<BR>
>>>  # This program and the accompanying materials
>>>  # are licensed and made available under the terms and conditions of the 
>>> BSD License
>>>  # which accompanies this distribution.  The full text of the license may 
>>> be found at
>>>  # http://opensource.org/licenses/bsd-license.php
>>> @@ -349,10 +349,17 @@ class MetaFileParser(object):
>>>              EdkLogger.error('Parser', FORMAT_INVALID, "No value specified",
>>>                              ExtraData=self._CurrentLine, 
>>> File=self.MetaFile, Line=self._LineIndex + 1)
>>>  
>>>          self._ValueList = [ReplaceMacro(Value, self._Macros) for Value in 
>>> self._ValueList]
>>>          Name, Value = self._ValueList[1], self._ValueList[2]
>>> +        MacroUsed = GlobalData.gMacroRefPattern.findall(Value)
>>> +        if len(MacroUsed) != 0:
>>> +            for Macro in MacroUsed:
>>> +                if Macro in GlobalData.gGlobalDefines:
>>> +                    EdkLogger.error("Parser", FORMAT_INVALID, "Global 
>>> macro %s is not permitted." % (Macro), ExtraData=self._CurrentLine, 
>>> File=self.MetaFile, Line=self._LineIndex + 1)
>>> +            else:
>>> +                EdkLogger.error("Parser", FORMAT_INVALID, "%s not defined" 
>>> % (Macro), ExtraData=self._CurrentLine, File=self.MetaFile, 
>>> Line=self._LineIndex + 1)
>>>          # Sometimes, we need to make differences between EDK and EDK2 
>>> modules 
>>>          if Name == 'INF_VERSION':
>>>              if re.match(r'0[xX][\da-f-A-F]{5,8}', Value):
>>>                  self._Version = int(Value, 0)   
>>>              elif re.match(r'\d+\.\d+', Value):
>>> diff --git a/BaseTools/Source/Python/Workspace/WorkspaceDatabase.py 
>>> b/BaseTools/Source/Python/Workspace/WorkspaceDatabase.py
>>> index e7bc87d..0686721 100644
>>> --- a/BaseTools/Source/Python/Workspace/WorkspaceDatabase.py
>>> +++ b/BaseTools/Source/Python/Workspace/WorkspaceDatabase.py
>>> @@ -1,9 +1,9 @@
>>>  ## @file
>>>  # This file is used to create a database used by build tool
>>>  #
>>> -# Copyright (c) 2008 - 2016, Intel Corporation. All rights reserved.<BR>
>>> +# Copyright (c) 2008 - 2017, Intel Corporation. All rights reserved.<BR>
>>>  # (C) Copyright 2016 Hewlett Packard Enterprise Development LP<BR>
>>>  # This program and the accompanying materials
>>>  # are licensed and made available under the terms and conditions of the 
>>> BSD License
>>>  # which accompanies this distribution.  The full text of the license may 
>>> be found at
>>>  # http://opensource.org/licenses/bsd-license.php
>>> @@ -1828,10 +1828,12 @@ class InfBuildData(ModuleBuildClassObject):
>>>              self.__Macros = {}
>>>              # EDK_GLOBAL defined macros can be applied to EDK module
>>>              if self.AutoGenVersion < 0x00010005:
>>>                  self.__Macros.update(GlobalData.gEdkGlobal)
>>>                  self.__Macros.update(GlobalData.gGlobalDefines)
>>> +            else:
>>> +                self.__Macros.update(self.Defines)
>>>          return self.__Macros
>>>  
>>>      ## Get architecture
>>>      def _GetArch(self):
>>>          return self._Arch
>>>
>>
>> I don't understand how, but this patch (commit dc4c770763d0) breaks OVMF for 
>> me.
>>
>> I bisected it, I didn't want to believe it, then I built the tree right 
>> before it, and that worked.
>>
>> I also built the tree at current master (526f160f311c, "ArmPkg/ArmMmuLib: 
>> AARCH64: enable stack alignment checking", 2017-02-22), with this commit 
>> reverted on top, and that also works.
>>
>> This is the error I see in the OVMF log:
>>
>>> Loading driver E660EA85-058E-4B55-A54B-F02F83A24707
>>> InstallProtocolInterface: 5B1B31A1-9562-11D2-8E3F-00A0C969723B 7E8752C0
>>> Loading driver at 0x0007E4B7000 EntryPoint=0x0007E4B727B DisplayEngine.efi
>>> InstallProtocolInterface: BC62157E-3E33-4FEC-9920-2D3B36D750DF 7E875518
>>> ProtectUefiImageCommon - 0x7E8752C0
>>>   - 0x000000007E4B7000 - 0x000000000001BAE0
>>> ASSERT 
>>> .../MdeModulePkg/Library/CustomizedDisplayLib/CustomizedDisplayLib.c(928): 
>>> mCDLStringPackHandle != ((void *) 0)
>>
>> I don't have the slightest clue what's going on. Apparently, this change 
>> causes BaseTools to mis-build OVMF.
> 
> If it's any help, I think that the second hunk breaks library constructors 
> somehow.
> 
> I compared the build report files between "good" (--> without this patch) and 
> "bad" (--> with this patch), and the differences I see follow this pattern, 
> under every module that links against library instance with a constructor:
> 
>  
> >----------------------------------------------------------------------------------------------------------------------<
>  Library
>  
> ------------------------------------------------------------------------------------------------------------------------
>  MdePkg/Library/PeiMemoryAllocationLib/PeiMemoryAllocationLib.inf
>  {MemoryAllocationLib}
>  
> MdePkg/Library/PeiServicesTablePointerLibIdt/PeiServicesTablePointerLibIdt.inf
>  {PeiServicesTablePointerLib}
>  MdePkg/Library/BasePcdLibNull/BasePcdLibNull.inf
>  {PcdLib}
>  MdePkg/Library/BaseLib/BaseLib.inf
>  {BaseLib}
>  MdePkg/Library/BaseMemoryLibRepStr/BaseMemoryLibRepStr.inf
>  {BaseMemoryLib}
>  MdePkg/Library/PeiHobLib/PeiHobLib.inf
>  {HobLib}
>  MdePkg/Library/PeiServicesLib/PeiServicesLib.inf
>  {PeiServicesLib}
>  MdePkg/Library/BaseDebugPrintErrorLevelLib/BaseDebugPrintErrorLevelLib.inf
>  {DebugPrintErrorLevelLib}
>  MdePkg/Library/BasePrintLib/BasePrintLib.inf
>  {PrintLib}
>  MdePkg/Library/BaseIoLibIntrinsic/BaseIoLibIntrinsic.inf
>  {IoLib}
>  OvmfPkg/Library/PlatformDebugLibIoPort/PlatformDebugLibIoPort.inf
> -{DebugLib:  C = PlatformDebugLibIoPortConstructor}
> +{DebugLib:  C = PlatformDebugLibIoPortConstructor 
> PlatformDebugLibIoPortConstructor}
>  MdePkg/Library/PeimEntryPoint/PeimEntryPoint.inf
>  {PeimEntryPoint}
>  
> <---------------------------------------------------------------------------------------------------------------------->
> 
> Note "DebugLib". In the "bad" case, the constructor name seems to be 
> duplicated.
> 
> This applies to all other constructor functions as well.
> 
> Thanks
> Laszlo
> 

_______________________________________________
edk2-devel mailing list
edk2-devel@lists.01.org
https://lists.01.org/mailman/listinfo/edk2-devel

Reply via email to