vvereschaka wrote:
@clayborg ,
the test is still flaky with these changes on the windows build/target host.
Here are two failed test outputs. They are a bit different.
failure 1:
```
# executed command:
'd:\projects\llvm-project\lldb\build-lldb-win\bin\filecheck.exe'
'D:\projects\llvm-projec
vvereschaka wrote:
Hi @Awfa,
we start getting failures for some LLDB API tests after these changes. The
tests are running on Jetson AGX/Cortex a78 (aarch64) board with Ubuntu Linux
22.04 on it. The failed tests get failed with the following error message:
```
AssertionError: No value is not tr
https://github.com/vvereschaka created
https://github.com/llvm/llvm-project/pull/93366
These environment variables avoid some side effects during execution of the
remote API tests on Windows platform.
One of the side effect is a creating of weird folders, such as
`\%SystemDrive%\...`, within
https://github.com/vvereschaka created
https://github.com/llvm/llvm-project/pull/93367
Because the 'make' tool is required to run the lldb testsuite a detection of
this program was added to the CMake script. If 'make' tool is missed on the
host CMake raises a fatal error now.
It is possible t
https://github.com/vvereschaka created
https://github.com/llvm/llvm-project/pull/93639
These changes do the following:
* avoids usage of $(findstring) within $(CC) to detect a type of specified C
compiler. Also they change a way to get a counterpart C++ compiler instead of
modifying the origi
https://github.com/vvereschaka created
https://github.com/llvm/llvm-project/pull/93643
Some of the API tests use Makefiles with the custom rules to manipulate the
source code files. This patch normalizes their file pathes to Unix-style on the
Windows build host and avoids compiler/linker error
https://github.com/vvereschaka created
https://github.com/llvm/llvm-project/pull/93645
Allow passing any custom 'strip' tool, such as 'llvm-strip', to use with the
LLDB API tests.
The tool can be provided via LLDB_TEST_COMMON_ARGS_VAR CMake variable:
`..;--env STRIP=${CMAKE_STRIP};...`.
Depe
https://github.com/vvereschaka created
https://github.com/llvm/llvm-project/pull/93648
Specify ENABLE_THREADS := YES within test's Makefile instead passing -lpthread
throuhg the compiler's CFLAGS options.
Updated the following tests:
*
lldb/test/API/commands/register/register/aarch64_sme_z_r
https://github.com/vvereschaka created
https://github.com/llvm/llvm-project/pull/93649
The test expects 'libstdc++' SO module in the module list. Force using StdC++
by specifying USE_LIBSTDCPP := 1 in the test's Makefile.
>From 741f69159b88c208ec0848b9c6d97be78f64383f Mon Sep 17 00:00:00 2001
https://github.com/vvereschaka edited
https://github.com/llvm/llvm-project/pull/93649
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
vvereschaka wrote:
Hi @Endilll ,
these changes break MSVC build of the projects including LLDB. The `cl`
compiler gets unsupported gcc/clang options, such as
`-Wno-deprecated-declarations`, `-Wno-unknown-pragmas` and
`-Wno-strict-aliasing`, and gets failed because of it.
Here is the command
vvereschaka wrote:
Hi @Jlalond ,
here is still a problem with the window builds:
```
FAILED:
tools/lldb/source/Plugins/ObjectFile/Minidump/CMakeFiles/lldbPluginObjectFileMinidump.dir/ObjectFileMinidump.cpp.obj
ccache C:\PROGRA~1\LLVM\bin\clang-cl.exe /nologo -TP -DGTEST_HAS_RTTI=0
-DUNICODE
vvereschaka wrote:
@Jlalond , perfect, thank you. It is ok for me if you'll fix the problem during
today.
https://github.com/llvm/llvm-project/pull/95312
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/
vvereschaka wrote:
Thank you @Jlalond , I started checking the local builds on Windows/Linux hosts
with these changes.
https://github.com/llvm/llvm-project/pull/96724
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi
https://github.com/vvereschaka approved this pull request.
LGTM.
https://github.com/llvm/llvm-project/pull/96724
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
vvereschaka wrote:
This build got the same failed test on the windows host -
https://lab.llvm.org/buildbot/#/builders/141/builds/318
```
Failed Tests (1):
lldb-shell :: SymbolFile/DWARF/x86/dwp-foreign-type-units.cpp
```
```
# .---command stderr
# |
C:\Users\tcwg\llvm-worker\lld
vvereschaka wrote:
@HaohaiWen ,
looks like these changes break some builders with the failed test
`UNRESOLVED: LLVM::windows-seh-EHa-PreserveCFG.s`
https://lab.llvm.org/buildbot/#/builders/234
https://lab.llvm.org/buildbot/#/builders/58
https://lab.llvm.org/buildbot/#/builders/104
https://lab.ll
vvereschaka wrote:
@HaohaiWen
>We may need to manually delete LLVM::windows-seh-EHa-PreserveCFG.s on those
>machines.
ok, got it. I'll do it for those builders.
>Do you know who can help us to delete it?
if you see some other failed builders because of this problem you can find
their owners o
vvereschaka wrote:
also, you can try to start a `Force Build` with cleaning of the source/build
directories for these builders. It should clean up the current source folder
and reload it.
https://github.com/llvm/llvm-project/pull/77608
___
lldb-commi
vvereschaka wrote:
@dzhidzhoev is sick. I'll merge this MR for him.
https://github.com/llvm/llvm-project/pull/111977
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
https://github.com/vvereschaka closed
https://github.com/llvm/llvm-project/pull/111977
___
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
vvereschaka wrote:
@chelcassanova `lldb/API/SBLanguages.h` is generated file. I suppose its
generation must be enabled for the `lldb-server` target in that case; otherwise
we get the following build errors:
```
In file included from
/home/buildbot/worker/as-builder-9/lldb-remote-linux-ubuntu/l
vvereschaka wrote:
@ashgti ,
these changes break the windows build with the following errors:
```
C:\buildbot\as-builder-10\lldb-x-aarch64\llvm-project\lldb\tools\lldb-dap\IOStream.h(41):
error C2061: syntax error: identifier 'SOCKET'
C:\buildbot\as-builder-10\lldb-x-aarch64\llvm-project\lldb\t
vvereschaka wrote:
Hi @splhack ,
looks like these changes break `lldb-api::TestLoadUnload.py` test on the LLDB
remote-linux builders:
* https://lab.llvm.org/buildbot/#/builders/195/builds/1435
* https://lab.llvm.org/staging/#/builders/197/builds/663
https://github.com/llvm/llvm-project/pull/11
24 matches
Mail list logo
