Status of the hotspot repo in the build-infra forest

Fri Oct 26 05:34:12 PDT 2012

This looks like the cl.exe hasn't been found. Would be interesting to 
look at spec.gmk and possibly config.log from this run.

/Erik

On 2012-10-25 22:23, Kelly O'Hair wrote:
> If the build-infra/jdk8/hotspot changes are NOT integrated, then the build-infra building will break all windows builds.
> The old builds still work fine, but the build-infra windows builds would be broken.
> So until the hotspot changes are either integrated through hotspot-rt, then integrated into jdk8/jdk8, then pulled down into jdk8/build
> we cannot integrate with a build regression on all windows builds.
>
> Someone needs to figure out why this is happening and adjust our common makefile logic to
> break this hotspot change dependency, then we can integrate.
>
> This is why I no longer want ANY hotspot changes being integrated into build-infra/jdk8/hotspot, none.
>
> This is the error I am seeing on all windows builds:
>
> 	rm -f x86_64.ad
> 	cat  C:\jprt\t\p1\181446~1.koh\s\hotspot/src/cpu/x86/vm/x86_64.ad  C:\jprt\t\p1\181446~1.koh\s\hotspot/src/cpu/x86/vm/x86.ad  C:\jprt\t\p1\181446~1.koh\s\hotspot/src/os_cpu/windows_x86/vm/windows_x86_64.ad>x86_64.ad
> 	C:\jprt\t\p1\181446~1.koh\s\build\window~1\fixpath.exe /nologo /W3 /WX /Zi /D "_LP64" /D "AMD64" /D "WIN32" /D "_WINDOWS" /D "VM_LITTLE_ENDIAN" /D TARGET_OS_FAMILY_windows /D TARGET_ARCH_x86 /D TARGET_ARCH_MODEL_x86_64 /D TARGET_OS_ARCH_windows_x86 /D TARGET_OS_ARCH_MODEL_windows_x86_64 /D TARGET_COMPILER_visCPP /MD /D _STATIC_CPPLIB /D _DISABLE_DEPRECATE_STATIC_CPPLIB /D _CRT_SECURE_NO_WARNINGS /D _CRT_SECURE_NO_DEPRECATE /EHsc /MD /D _STATIC_CPPLIB /D _DISABLE_DEPRECATE_STATIC_CPPLIB  /I "..\generated"  /I "C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm"  /I "C:\jprt\t\p1\181446~1.koh\s\hotspot\src\os\windows\vm"  /I "C:\jprt\t\p1\181446~1.koh\s\hotspot\src\cpu\x86\vm" /c C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\main.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\adlparse.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\archDesc.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\arena.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\dfa.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\dict2.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\filebuff.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\forms.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\formsopt.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\formssel.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\output_c.cpp C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\adlc\output_h.cpp
> Usage: fixpath -c|m<path at path@...>  /cygdrive/c/WINDOWS/notepad.exe /cygdrive/c/x/test.txt	C:\jprt\t\p1\181446~1.koh\s\build\window~1\fixpath.exe /nologo /W3 /WX /Zi /D "_LP64" /D "AMD64" /D "WIN32" /D "_WINDOWS" /D "VM_LITTLE_ENDIAN" /D TARGET_OS_FAMILY_windows /D TARGET_ARCH_x86 /D TARGET_ARCH_MODEL_x86_64 /D TARGET_OS_ARCH_windows_x86 /D TARGET_OS_ARCH_MODEL_windows_x86_64 /D TARGET_COMPILER_visCPP /MD /D _STATIC_CPPLIB /D _DISABLE_DEPRECATE_STATIC_CPPLIB /D _CRT_SECURE_NO_WARNINGS /D _CRT_SECURE_NO_DEPRECATE /EHsc /MD /D _STATIC_CPPLIB /D _DISABLE_DEPRECATE_STATIC_CPPLIB  /I "..\generated"  /I "C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm"  /I "C:\jprt\t\p1\181446~1.koh\s\hotspot\src\os\windows\vm"  /I "C:\jprt\t\p1\181446~1.koh\s\hotspot\src\cpu\x86\vm" /c C:\jprt\t\p1\181446~1.koh\s\hotspot\src\share\vm\opto\opcodes.cpp
> Usage: fixpath -c|m<path at path@...>  /cygdrive/c/WINDOWS/notepad.exe /cygdrive/c/x/test.txt	C:\jprt\t\p1\181446~1.koh\s\build\window~1\fixpath.exe /manifest  kernel32.lib user32.lib gdi32.lib winspool.lib  comdlg32.lib advapi32.lib shell32.lib ole32.lib oleaut32.lib  uuid.lib Wsock32.lib winmm.lib /nologo /machine:AMD64 /opt:REF  /opt:ICF,8 /map /debug psapi.lib /subsystem:console /out:adlc.exe main.obj adlparse.obj archDesc.obj arena.obj dfa.obj dict2.obj filebuff.obj forms.obj formsopt.obj formssel.obj opcodes.obj output_c.obj output_h.obj
> Usage: fixpath -c|m<path at path@...>  /cygdrive/c/WINDOWS/notepad.exe /cygdrive/c/x/test.txt	C:\jprt\t\p1\181446~1.koh\s\build\window~1\fixpath.exe /manifest adlc.exe.manifest /outputresource:adlc.exe;#1
> Usage: fixpath -c|m<path at path@...>  /cygdrive/c/WINDOWS/notepad.exe /cygdrive/c/x/test.txt	rm -f  ad_x86_64.cpp  ad_x86_64.hpp  ad_x86_64_clone.cpp  ad_x86_64_expand.cpp  ad_x86_64_format.cpp  ad_x86_64_gen.cpp  ad_x86_64_misc.cpp  ad_x86_64_peephole.cpp  ad_x86_64_pipeline.cpp  adGlobals_x86_64.hpp  dfa_x86_64.cpp
> 	if exist adfiles rmdir /s /q adfiles
> 	mkdir adfiles
> 	adlc -q -T -D_LP64 x86_64.ad
> 'adlc' is not recognized as an internal or external command,
> operable program or batch file.
> NMAKE : fatal error U1077: 'adlc' : return code '0x1'
> Stop.
> NMAKE : fatal error U1077: 'cd' : return code '0x2'
> Stop.
> NMAKE : fatal error U1077: '"C:\Program Files (x86)\Microsoft Visual Studio 10.0\VC\BIN\amd64\nmake.EXE"' : return code '0x2'
> Stop.
> make[4]: *** [generic_build2] Error 2
> make[4]: Leaving directory `/cygdrive/c/jprt/t/p1/181446~1.koh/s/hotspot/make'
> make[3]: *** [product] Error 2
> make[3]: Leaving directory `/cygdrive/c/jprt/t/p1/181446~1.koh/s/hotspot/make'
> make[2]: *** [/cygdrive/c/jprt/t/p1/181446~1.koh/s/build/window~1/hotspot/_hotspot.timestamp] Error 2
> make[2]: Leaving directory `/cygdrive/c/jprt/t/p1/181446~1.koh/s/common/makefiles'
> make[1]: *** [hotspot-only] Error 2
> make[1]: Leaving directory `/cygdrive/c/jprt/T/P1/181446.kohair/s/build/windows-x86_64-normal-server-release'
> make: *** [images] Error 2
>
>
> -kto
>
> On Oct 25, 2012, at 6:59 AM, Magnus Ihse Bursie wrote:
>
>> The hotspot repo has for a long time been problematic for build-infra, since we have not really kept it up-to-date -- changes in build-infra/hotspot has not been pushed properly, and we have diverted from upstream hotspot, which made new fixes require additional manual work to sort out, with the added risk of human errors.
>>
>> I have now done a thourough analysis of the hotspot situation. This is my findings: We have two sets of differences, compared to hotspot in the build forest. The differences compared to hotspot-rt/hotspot are the same, apart from the new stuff in hotspot-rt that has not yet reached build/hotspot.
>>
>> The two differences are:
>> * The incomplete Queens/test-in-build patch, which is complete and correct on build-infra, but not the build forest.
>> * A patch to allow build-infra to run on msys.
>>
>> I have sent Kelly a patch that will compensate for the first.
>>
>> I have sent a webrev for the second. If there is any problem with this, we can skip it for now -- the "only" thing that happens is that the new build won't be able to build on msys in mainline. It's unfortunately, but it's not a regression since we've never had msys support there before.
>>
>> /Magnus