[PATCH] D42281: Compile the LLDB tests out-of-tree
Pavel Labath via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Wed Jan 31 08:14:35 PST 2018
labath added a comment.
In https://reviews.llvm.org/D42281#992359, @aprantl wrote:
> In https://reviews.llvm.org/D42281#992295, @aprantl wrote:
>
> > There are a few test failures, but they look manageable. I will need help resolving them though:
>
>
> Updated list:
>
> http://lab.llvm.org:8011/builders/lldb-windows7-android/builds/7843
> FAIL: TestLoadUnload.LoadUnloadTestCase.test_lldb_process_load_and_unload_commands_dwarf
> test3 HT4A1JT02945,aarch64-linux-android-clang,aarch64,tcp,-lldb-mi 8 unexpected failures failed ( 41 mins, 48 secs )
> stdio
> FAIL: TestBreakpointInGlobalConstructor.TestBreakpointInGlobalConstructors.test
> FAIL: TestMoveNearest.TestMoveNearest.test
> FAIL: TestGlobalVariables.GlobalVariablesTestCase.test_c_global_variables_dwarf
> FAIL: TestGlobalVariables.GlobalVariablesTestCase.test_c_global_variables_dwo
> FAIL: TestGlobalVariables.GlobalVariablesTestCase.test_c_global_variables_gmodules
> FAIL: TestLoadUnload.LoadUnloadTestCase.test_lldb_process_load_and_unload_commands_dwarf
> FAIL: TestLoadUnload.LoadUnloadTestCase.test_lldb_process_load_and_unload_commands_dwo
> FAIL: TestLoadUnload.LoadUnloadTestCase.test_lldb_process_load_and_unload_commands_gmodules
>
> http://lab.llvm.org:8011/builders/lldb-x86_64-ubuntu-14.04-android/builds/10506
> test7 emulator-5554,i686-linux-android-clang,i386,abstract,-lldb-mi failed
>
> http://lab.llvm.org:8011/builders/lldb-x86_64-darwin-13.4/builds/8462
> test2 5A24000830,aarch64-linux-android-gcc,aarch64,tcp,-lldb-mi:-watchpoint 2 unexpected failures failed
> FAIL: TestMoveNearest.TestMoveNearest.test
> FAIL: TestBreakpointInGlobalConstructor.TestBreakpointInGlobalConstructors.test
Most of these turned out to be caused by leftover shared libraries in the source tree (fun fact: "svn status" refuses to list the .so files even though they are not present in the svn:ignore list), when I cleaned those up, the tests started passing again.
There is still the TestLoadUnload failure on the windows bot, which I suspect to be a real issue (we've changed the test a bit, and now it passes full paths to lldb, so I suspect a directory-separator issue), but I have not looked into that yet. I'll handle that on our side in the coming days.
All in all, I would say this was rather a painless experience. I think we are ready for phase two.
Repository:
rL LLVM
https://reviews.llvm.org/D42281
More information about the llvm-commits
mailing list