we've started getting the warnings for newly created file llvm/lib/Transforms/Utils/MemoryOpRemark.cpp on the cross toolchain builders:

Thank you @aeubanks, got it.

these changes breaks the new-pm-lto-defaults.ll test on the cross win-x-arm linux cross builders:

• https://lab.llvm.org/buildbot/#/builders/119/builds/3775/steps/8/logs/FAIL__LLVM__new-pm-lto-defaults_ll
• https://lab.llvm.org/buildbot/#/builders/60/builds/3287/steps/8/logs/FAIL__LLVM__new-pm-lto-defaults_ll

got it, perfect. Thank you.

@ldionne, @curdeius thank you for review.
Updated diff accordingly

• reworded comment
• %t.grep => %t.escaped.grep
• do not remove temporary file in the test.

these changes should work on Windows, Linux and Darwin platforms, but I'm not sure about the others. In case of any problems we can disable this test for Windows with https://reviews.llvm.org/D102045.

Hi @Quuxplusone, understood, thank you! I agree with you that temporary disabling of the test for some configurations would be optimal for now.

the same assert happens on the ARM cross toolchain builders:

• aarch64 https://lab.llvm.org/buildbot/#/builders/119/builds/2911/steps/14/logs/FAIL__libc____convert_copy_pass_cpp
• armv7 https://lab.llvm.org/buildbot/#/builders/60/builds/2497/steps/14/logs/FAIL__libc____convert_copy_pass_cpp

Hi @phosek,
I have tested the updated diff on Windows-x-ARM cross builder and it works fine. Thank you.

we got the same thing on the win-x-arm cross toolchain builders as described by @uabelho

CHECK: "-internal-isystem" "/usr/local/include" didn't break the test on the cross builders before.

sorry, but these changes break the Clang::gcc-toolchain.cpp test on the ARMv7/AArch64 cross builders:

Updated diff:

• updated doc parts

Updated diff:

Thank you @aganea

Fixed comment.

Updated diff in according of the Lint/clang-format recommendations.

@sepavloff would you add more persons in additional to the reviewers?

@ldionne would you take a look at the updated diff?

Updated main() -> main(int, char**) for the tests.

You are right, we don't need it. I did a copy/paste from another test.
Updated diff without cxxabi.h.

ref https://reviews.llvm.org/D95200

got it, thank you.

@MaskRay , ping

@ikudrin thank you for your help.

these tests get failed on Windows to ARMv7 Linux cross builder with the following error:

C:/buildbot/as-builder-1/x-armv7l/llvm-project/libcxxabi/test/forced_unwind1.pass.cpp:37:42: error: typedef '_Unwind_Exception' cannot be referenced with a struct specifier

http://lab.llvm.org:8011/#/builders/60/builds/1855

thank you for the fix.

here is a problem with extern-templates.sh.cpp test on the windows-to-linux ARM cross builders

LGTM

OK, it is not a problem.
I'll prepare another patch with the string formatting fixes.

LGTM

LGTM

also tested with the aarch64/armv7 compiler-rt tests on Windows cross builders.

LGTM

UNSUPPORTED: * is not working actually. Please check a build log for linux > Builtins-x86_64-linux.Builtins-x86_64-linux::gcc_personality_test.c above. It says

ValueError: Error in UNSUPPORTED list:
couldn't parse text: '*'
in expression: '*'

jfi for updated diff: the cross toolchain build on Windows has passed. check-llvm/check-clang also looks ok.

Hi @AndrewLitteken,
sorry, but this commit fails the cross toolchain builders on Windows platform with the following errors:

FAILED: lib/Analysis/CMakeFiles/LLVMAnalysis.dir/IRSimilarityIdentifier.cpp.obj 
C:\PROGRA~2\MICROS~1\2017\COMMUN~1\VC\Tools\MSVC\1416~1.270\bin\Hostx64\x64\cl.exe  /nologo /TP -DGTEST_HAS_RTTI=0 -DUNICODE -D_CRT_NONSTDC_NO_DEPRECATE -D_CRT_NONSTDC_NO_WARNINGS -D_CRT_SECURE_NO_DEPRECATE -D_CRT_SECURE_NO_WARNINGS -D_HAS_EXCEPTIONS=0 -D_SCL_SECURE_NO_DEPRECATE -D_SCL_SECURE_NO_WARNINGS -D_UNICODE -D__STDC_CONSTANT_MACROS -D__STDC_FORMAT_MACROS -D__STDC_LIMIT_MACROS -Ilib\Analysis -IC:\buildbot\as-builder-2\x-aarch64\llvm-project\llvm\lib\Analysis -Iinclude -IC:\buildbot\as-builder-2\x-aarch64\llvm-project\llvm\include -D__OPTIMIZE__ /Zc:inline /Zc:__cplusplus /Zc:strictStrings /Oi /Zc:rvalueCast /bigobj /W4 -wd4141 -wd4146 -wd4244 -wd4267 -wd4291 -wd4351 -wd4456 -wd4457 -wd4458 -wd4459 -wd4503 -wd4624 -wd4722 -wd4100 -wd4127 -wd4512 -wd4505 -wd4610 -wd4510 -wd4702 -wd4245 -wd4706 -wd4310 -wd4701 -wd4703 -wd4389 -wd4611 -wd4805 -wd4204 -wd4577 -wd4091 -wd4592 -wd4319 -wd4709 -wd4324 -w14062 -we4238 /Gw /MD /O2 /Ob2    /EHs-c- /GR- -UNDEBUG -std:c++14 /showIncludes /Folib\Analysis\CMakeFiles\LLVMAnalysis.dir\IRSimilarityIdentifier.cpp.obj /Fdlib\Analysis\CMakeFiles\LLVMAnalysis.dir\LLVMAnalysis.pdb /FS -c C:\buildbot\as-builder-2\x-aarch64\llvm-project\llvm\lib\Analysis\IRSimilarityIdentifier.cpp
C:\buildbot\as-builder-2\x-aarch64\llvm-project\llvm\include\llvm/Analysis/IRSimilarityIdentifier.h(154): error C3861: 'hash_value': identifier not found
C:\buildbot\as-builder-2\x-aarch64\llvm-project\llvm\include\llvm/Analysis/IRSimilarityIdentifier.h(154): error C2672: 'llvm::IRSimilarity::hash_value': no matching overloaded function found
C:\buildbot\as-builder-2\x-aarch64\llvm-project\llvm\include\llvm/Analysis/IRSimilarityIdentifier.h(153): error C2672: 'hash_combine': no matching overloaded function found

see more details here:
http://lab.llvm.org:8011/builders/llvm-clang-win-x-aarch64/builds/3016
http://lab.llvm.org:8011/builders/llvm-clang-win-x-armv7l/builds/1265

Hi Sergej,

the LLVM::minsize-call-cse-2.ll test gets failed on llvm-clang-x86_64-expensive-checks-ubuntu builder:
http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-ubuntu/builds/8634

sorry, but looks like your changes break one of libc++ tests on ARM cross toolchain build on Windows: libc++::function_type_default_deleter.fail.cpp
Here is the first failed build: http://lab.llvm.org:8011/builders/llvm-clang-win-x-armv7l/builds/945

@enh,
may be let's commit the fix? This problem keeps the builder in red for a week already.

It works fine now for me

Testing Time: 61.79s
  Unsupported: 12
  Passed     : 51

I understand, not a problem. I'm testing the updated patch.

The test got failed again. I got the following output:

ubuntu@jetson8:~$ ./t.tmp.exe
Testing 29715 symbols.
1ILeeeEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE -> <-0x1.cecececececececececececececep+11983L
Got instead: <-0x1.cecececececececececececececep+11983L>
t.tmp.exe: C:/buildbot/temp/llvm-project/libcxxabi/test/test_demangle.pass.cpp:30001: void testFPLiterals(): Assertion `false' failed.
Aborted (core dumped)

looks like it does not expect trailing >.

I ran the test directly on our aarch64 board. The output is the following

ubuntu@jetson8:~$ ./t.tmp.exe
Testing 29715 symbols.
1ILeeeEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE -> <-0x1.cecececececececececececececep+11983
Got instead: <-0x1.cecececececececececececececep+11983L>
t.tmp.exe: C:/buildbot/temp/llvm-project/libcxxabi/test/test_demangle.pass.cpp:30001: void testFPLiterals(): Assertion `false' failed.
Aborted (core dumped)

Thank you @smeenai,
not a problem.

sorry, but LLVM::universal-output.test test gets failed on the armv7 cross toolchain buider.
Here is a direct link to the failed test log
http://lab.llvm.org:8011/builders/llvm-clang-win-x-armv7l/builds/802

I have tested the patch with VS2017/cl 19.16 build + clang/llvm tests and clang/llvm unit tests.
It looks fine.

Hi Sergei,
sorry for a late answer.
I have tested these changes on Windows cross build to Aarch64 and I found the following problems:

Hi Dmitri, Eduardo,

Hello Eduardo,

Hi Alex,

Hi Marco,

I see that Harbormaster failed the build on Windows, but there is a lot of error: Connection to server timed out errors. Looks like there is some problem with that build system.

I have tested this patch on the ubuntu builder with enabled expensive checks and no-inputs.test test has passed.

Is there maybe a problem with this buildbot and parameters like -out:<value>?

This is the expensive checks builder and I see this crash on all that kind of builders

• http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-debian/builds/6242
• http://lab.llvm.org:8011/builders/llvm-clang-x86_64-expensive-checks-win/builds/23340

Hi Eric,

Hi @compnerd,
here is one more place, which needs a proper detection for Python3/Python2.
I just copied a part of your changes accordingly.