[CMake] Re-add -fno-rtti to llvm-config --cxxflags #174084

aengelke · 2025-12-31T11:25:51Z

If there's no strong reason, we should get rid of per-target RTTI later.

Fix for LLVM_CONFIG_HAS_RTTI.

nikic

I'm somewhat confused by how this code works, including for the existing LLVM_CONFIG_HAS_RTTI flag. If LLVM_REQUIRES_RTTI is used somewhere, does this depend on whether the last target for which llvm_update_compile_flags is called happens to have that set or not?

Should llvm-config only be using the global LLVM_ENABLE_RTTI instead of this second LLVM_CONFIG_HAS_RTTI variable that also uses LLVM_REQUIRES_RTTI? And if so, can we extract determining the corresponding compiler flag into a separate function that gets explicitly called from the llvm-config cmake instead of relying on the side-effect here?

aengelke · 2026-01-01T21:13:04Z

CI failure was a spurious link error, worked fine on second try.

nikic

LGTM

llvm-ci · 2026-01-02T01:03:01Z

LLVM Buildbot has detected a new failure on builder ppc64le-flang-rhel-clang running on ppc64le-flang-rhel-test while building llvm at step 6 "test-build-unified-tree-check-flang".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/157/builds/43393

Here is the relevant piece of the build log for the reference

Step 6 (test-build-unified-tree-check-flang) failure: 1200 seconds without output running [b'ninja', b'check-flang'], attempting to kill
...
PASS: Flang :: Semantics/modfile73.f90 (4061 of 4071)
PASS: Flang :: Semantics/numeric_storage_size.f90 (4062 of 4071)
PASS: Flang :: Driver/mrecip.f90 (4063 of 4071)
PASS: Flang :: Driver/use-module-error.f90 (4064 of 4071)
PASS: Flang :: Driver/mcmodel.f90 (4065 of 4071)
PASS: Flang :: Driver/fopenmp.f90 (4066 of 4071)
PASS: Flang :: Driver/use-module.f90 (4067 of 4071)
PASS: Flang :: Intrinsics/math-codegen.fir (4068 of 4071)
PASS: Flang :: Driver/linker-options.f90 (4069 of 4071)
PASS: Flang :: Driver/omp-driver-offload.f90 (4070 of 4071)
command timed out: 1200 seconds without output running [b'ninja', b'check-flang'], attempting to kill
process killed by signal 9
program finished with exit code -1
elapsedTime=2791.351677

llvm-ci · 2026-01-02T04:30:45Z

LLVM Buildbot has detected a new failure on builder clang-ppc64le-linux-test-suite running on ppc64le-clang-test-suite while building llvm at step 6 "test-build-unified-tree-check-all".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/95/builds/19609

Here is the relevant piece of the build log for the reference

Step 6 (test-build-unified-tree-check-all) failure: test (failure)
******************** TEST 'LeakSanitizer-Standalone-powerpc64le :: TestCases/create_thread_leak.cpp' FAILED ********************
Exit Code: 2

Command Output (stdout):
--
# RUN: at line 3
/home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/./bin/clang  --driver-mode=g++ -O0  -m64 -fno-function-sections  -gline-tables-only -fsanitize=leak -I/home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/../ -pthread /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp -o /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LELsanConfig/TestCases/Output/create_thread_leak.cpp.tmp
# executed command: /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/./bin/clang --driver-mode=g++ -O0 -m64 -fno-function-sections -gline-tables-only -fsanitize=leak -I/home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/../ -pthread /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp -o /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LELsanConfig/TestCases/Output/create_thread_leak.cpp.tmp
# RUN: at line 4
not  /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LELsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 1 0 0 2>&1 | FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK123
# executed command: not /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LELsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 1 0 0
# executed command: FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK123
# RUN: at line 5
not  /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LELsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 0 1 0 2>&1 | FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# executed command: not /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LELsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 0 1 0
# executed command: FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# RUN: at line 6
not  /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LELsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 0 0 1 2>&1 | FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# executed command: not /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/build/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LELsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 0 0 1
# note: command had no output on stdout or stderr
# error: command failed with exit status: 1
# executed command: FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# .---command stderr------------
# | FileCheck error: '<stdin>' is empty.
# | FileCheck command line:  FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-test-suite/clang-ppc64le-test-suite/llvm-project/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# `-----------------------------
# error: command failed with exit status: 2

--

********************

…)" This reverts commit d1b88ca.

…gs (llvm#174084)

aengelke · 2026-01-02T07:45:11Z

Buildbot failures seem unrelated: first is just a timeout, second is spurious (tests pass in next build). @ronlieb you noted that this breaks your build -- what's happening and is this a pure downstream thing?

llvm-ci · 2026-01-02T09:28:57Z

LLVM Buildbot has detected a new failure on builder clang-ppc64le-linux-multistage running on ppc64le-clang-multistage-test while building llvm at step 5 "ninja check 1".

Full details are available at: https://lab.llvm.org/buildbot/#/builders/76/builds/13697

Here is the relevant piece of the build log for the reference

Step 5 (ninja check 1) failure: stage 1 checked (failure)
******************** TEST 'LeakSanitizer-AddressSanitizer-powerpc64le :: TestCases/create_thread_leak.cpp' FAILED ********************
Exit Code: 2

Command Output (stdout):
--
# RUN: at line 3
/home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/./bin/clang  --driver-mode=g++ -O0  -m64 -fno-function-sections  -gline-tables-only -fsanitize=address -I/home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/../ -pthread /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp -o /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LEAsanConfig/TestCases/Output/create_thread_leak.cpp.tmp
# executed command: /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/./bin/clang --driver-mode=g++ -O0 -m64 -fno-function-sections -gline-tables-only -fsanitize=address -I/home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/../ -pthread /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp -o /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LEAsanConfig/TestCases/Output/create_thread_leak.cpp.tmp
# RUN: at line 4
not  /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LEAsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 1 0 0 2>&1 | FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK123
# executed command: not /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LEAsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 1 0 0
# executed command: FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK123
# RUN: at line 5
not  /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LEAsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 0 1 0 2>&1 | FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# executed command: not /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LEAsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 0 1 0
# executed command: FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# RUN: at line 6
not  /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LEAsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 0 0 1 2>&1 | FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# executed command: not /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/stage1/runtimes/runtimes-bins/compiler-rt/test/lsan/POWERPC64LEAsanConfig/TestCases/Output/create_thread_leak.cpp.tmp 10 0 0 1
# note: command had no output on stdout or stderr
# error: command failed with exit status: 1
# executed command: FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# .---command stderr------------
# | FileCheck error: '<stdin>' is empty.
# | FileCheck command line:  FileCheck /home/buildbots/llvm-external-buildbots/workers/ppc64le-clang-multistage-test/clang-ppc64le-multistage/llvm/compiler-rt/test/lsan/TestCases/create_thread_leak.cpp --check-prefixes=LEAK,LEAK234
# `-----------------------------
# error: command failed with exit status: 2

--

********************

RKSimon · 2026-01-02T10:54:28Z

@aengelke This is spamming MSVC builds with exception override warnings https://lab.llvm.org/buildbot/#/builders/46/builds/28456

[3188/18/878/4083]Building CXX object lib\Analysi...eFiles\LLVMAnalysis.dir\MemoryProfileInfo.cpp.obj
cl : Command line warning D9025 : overriding '/EHs' with '/EHs-'
cl : Command line warning D9025 : overriding '/EHc' with '/EHc-'
[3187/18/879/4083]Building CXX object lib\Analysis\CMakeFiles\LLVMAnalysis.dir\MemoryBuiltins.cpp.obj
cl : Command line warning D9025 : overriding '/EHs' with '/EHs-'
cl : Command line warning D9025 : overriding '/EHc' with '/EHc-'
[3186/18/880/4083]Building CXX object lib\Analysis\CMakeFiles\LLVMAnalysis.dir\LoopInfo.cpp.obj
cl : Command line warning D9025 : overriding '/EHs' with '/EHs-'
cl : Command line warning D9025 : overriding '/EHc' with '/EHc-'
[3185/18/881/4083]Building CXX object lib\Analysis\CMakeFiles\LLVMAnalysis.dir\ObjCARCInstKind.cpp.ob
cl : Command line warning D9025 : overriding '/EHs' with '/EHs-'
cl : Command line warning D9025 : overriding '/EHc' with '/EHc-'
[3184/18/882/4083]Building CXX object lib\Analysi...LLVMAnalysis.dir\MemoryDependenceAnalysis.cpp.obj
cl : Command line warning D9025 : overriding '/EHs' with '/EHs-'
cl : Command line warning D9025 : overriding '/EHc' with '/EHc-'
[3183/18/883/4083]Building CXX object lib\Analysi...les\LLVMAnalysis.dir\ObjCARCAliasAnalysis.cpp.obj
cl : Command line warning D9025 : overriding '/EHs' with '/EHs-'
cl : Command line warning D9025 : overriding '/EHc' with '/EHc-'
[3182/18/884/4083]Building CXX object lib\Analysi...keFiles\LLVMAnalysis.dir\MemorySSAUpdater.cpp.obj
cl : Command line warning D9025 : overriding '/EHs' with '/EHs-'
cl : Command line warning D9025 : overriding '/EHc' with '/EHc-'
[3181/18/885/4083]Building CXX object lib\Analysi...Files\LLVMAnalysis.dir\LoopAccessAnalysis.cpp.obj
cl : Command line warning D9025 : overriding '/EHs' with '/EHs-'
cl : Command line warning D9025 : overriding '/EHc' with '/EHc-'

aengelke · 2026-01-02T10:58:03Z

Thanks for noticing, hopefully fixed with 2b903df.

RKSimon · 2026-01-02T11:38:43Z

Thanks for noticing, hopefully fixed with 2b903df.

Thanks - that fixed it

ronlieb · 2026-01-02T14:57:51Z

Buildbot failures seem unrelated: first is just a timeout, second is spurious (tests pass in next build). @ronlieb you noted that this breaks your build -- what's happening and is this a pure downstream thing?

seems to be purely downstream, i think i am seeing an additionaly flag "-fno-rtti" during our build of offload

kuhar · 2026-01-02T16:27:19Z

Same, I can see failures downstream even with the commit above, which causes:

 Found erroneous configuration for source file PassPlugin.cpp

 LLVM's build system enforces that all source files are added to a build
 target, that exactly one build target exists in each directory, and that
 this target lists all files in that directory.  If you want multiple
 targets in the same directory, add PARTIAL_SOURCES_INTENDED to the target
 specification, though it is discouraged.

Fix for llvm#173869. If there's no strong reason, we should get rid of per-target RTTI later.

aengelke · 2026-01-02T16:34:06Z

@ronlieb that's unexpected, that should've been there all the time? offload/CMakeLists.txt makes its decision purely on LLVM_ENABLE_RTTI, which was unchanged?

@kuhar but... PassPlugin.cpp is referenced in llvm/lib/Plugins/CMakeLists.txt?

kuhar · 2026-01-02T17:09:45Z

@kuhar but... PassPlugin.cpp is referenced in llvm/lib/Plugins/CMakeLists.txt?

to me this indicates something's off with cmake, let me try a clean build, maybe these rtti changes messed up the cache etc.

…)" This reverts commit d1b88ca.

New reverts: * llvm/llvm-project#174084 -- causes linker errors related to rtti in stablehlo code * llvm/llvm-project@2b903df -- follow up fix to the above (doesn't fix it for us) Existing local reverts carried forward: * Local revert of llvm/llvm-project#169614 due to llvm/llvm-project#172932.

…)" This reverts commit d1b88ca.

Existing local reverts carried forward: * Local revert of llvm/llvm-project#169614 due to llvm/llvm-project#172932. * llvm/llvm-project#174084 -- causes linker errors related to rtti in stablehlo code * llvm/llvm-project@2b903df -- follow up fix to the above (doesn't fix it for us) Signed-off-by: Abhishek Varma <[email protected]>

Fix for llvm#173869. If there's no strong reason, we should get rid of per-target RTTI later.

dcaballe · 2026-01-06T18:26:18Z

I'm also hitting some issues with this commit. It looks like a tablegen component in my project that was built with RTTI disabled by default is now built with RTTI enabled so probably some cmake var is not cached or propagated as expected?

aengelke · 2026-01-06T18:55:32Z

Is this reproducible with a clean cmake configuration? Is LLVM_OPTIMIZED_TABLEGEN set or is this a cross compilation? Can you share the cmake configuration and the compilation commands that lead to the error?

Existing local reverts carried forward: * Local revert of llvm/llvm-project#169614 due to llvm/llvm-project#172932. * llvm/llvm-project#174084 -- causes linker errors related to rtti in stablehlo code * llvm/llvm-project@2b903df -- follow up fix to the above (doesn't fix it for us) Signed-off-by: Abhishek Varma <[email protected]> --------- Signed-off-by: Abhishek Varma <[email protected]> Co-authored-by: Jakub Kuderski <[email protected]>

Existing local reverts carried forward: * Local revert of llvm/llvm-project#169614 due to llvm/llvm-project#172932. * llvm/llvm-project#174084 -- causes linker errors related to rtti in stablehlo code * llvm/llvm-project@2b903df -- follow up fix to the above (doesn't fix it for us) Signed-off-by: Abhishek Varma <[email protected]>

…)" This reverts commit d1b88ca.

Abhishek-Varma · 2026-01-07T11:59:15Z

Hi.

I'm getting the following error (which goes away on reverting this PR's commit) :-

[876/4230] Linking CXX executable llvm-project/bin/mlir_builder_tblgen
FAILED: llvm-project/bin/mlir_builder_tblgen
: && /usr/bin/c++ -Wno-deprecated-declarations -g -Wl,-rpath-link,/home/abhvarma/iree/build/llvm-project/lib  -Wl,--gc-sections llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o -o llvm-project/bin/mlir_builder_tblgen  -Wl,-rpath,"\$ORIGIN/../lib"  llvm-project/lib/libLLVMSupport.a  llvm-project/lib/libLLVMTableGen.a  llvm-project/lib/libLLVMTableGen.a  llvm-project/lib/libMLIRTblgenLib.a  llvm-project/lib/libMLIRSupport.a  llvm-project/lib/libLLVMTableGen.a  llvm-project/lib/libMLIRTableGen.a  llvm-project/lib/libLLVMSupport.a  -lrt  -ldl  -lm  llvm-project/lib/libLLVMDemangle.a && :
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro+0x200): undefined reference to `typeinfo for llvm::cl::Option'
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro+0x230): undefined reference to `typeinfo for llvm::cl::generic_parser_base'
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro+0x278): undefined reference to `typeinfo for llvm::cl::GenericOptionValue'
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro._ZTIN4llvm7support6detail23provider_format_adapterIRmEE[_ZTIN4llvm7support6detail23provider_format_adapterIRmEE]+0x10): undefined reference to `typeinfo for llvm::support::detail::format_adapter'
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro._ZTIN4llvm7support6detail23provider_format_adapterIRNS_5TwineEEE[_ZTIN4llvm7support6detail23provider_format_adapterIRNS_5TwineEEE]+0x10): undefined reference to `typeinfo for llvm::support::detail::format_adapter'
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro._ZTIN4llvm7support6detail23provider_format_adapterIRKNS_9StringRefEEE[_ZTIN4llvm7support6detail23provider_format_adapterIRKNS_9StringRefEEE]+0x10): undefined reference to `typeinfo for llvm::support::detail::format_adapter'
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro._ZTIN4llvm7support6detail23provider_format_adapterINS_9StringRefEEE[_ZTIN4llvm7support6detail23provider_format_adapterINS_9StringRefEEE]+0x10): undefined reference to `typeinfo for llvm::support::detail::format_adapter'
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro._ZTIN4llvm7support6detail23provider_format_adapterIRNS_9StringRefEEE[_ZTIN4llvm7support6detail23provider_format_adapterIRNS_9StringRefEEE]+0x10): undefined reference to `typeinfo for llvm::support::detail::format_adapter'
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro._ZTIN4llvm7support6detail23provider_format_adapterIRNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEEEE[_ZTIN4llvm7support6detail23provider_format_adapterIRNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEEEE]+0x10): more undefined references to `typeinfo for llvm::support::detail::format_adapter' follow
/usr/bin/ld: llvm-external-projects/stablehlo/stablehlo/integrations/cpp/builder/CMakeFiles/mlir_builder_tblgen.dir/MlirBuilderTblgen.cpp.o:(.data.rel.ro._ZTIN4mlir20raw_indented_ostreamE[_ZTIN4mlir20raw_indented_ostreamE]+0x10): undefined reference to `typeinfo for llvm::raw_ostream'
collect2: error: ld returned 1 exit status

aengelke · 2026-01-07T12:40:01Z

@Abhishek-Varma Looks like a downstream problem in iree_llvm. It's not possible to just set some CMake values and expect things to work; you also need to run the code in HandleLLVMOptions.cmake. I don't think we support direct access to functions in AddLLVM (and transitively though AddMLIR) without running the outer skeleton.

dcaballe · 2026-01-07T22:15:52Z

I don't think we can expect external projects to mandatorily include HandleLLVMOptions.cmake, at least not in the way it is today. That file adds many flags beyond just -fno-rtti, including warning-related flags that are not needed to integrate with LLVM.

I think this PR fixes the problem: #174873
The issue is that LLVM_CXXFLAGS_RTTI_DISABLE and LLVM_CXXFLAGS_RTTI_ENABLE are now undefined for external projects after this commit, since they are defined at LLVM config time in HandleLLVMOptions.cmake but not exported via LLVMConfig.cmake.in. So, any external project calling llvm_update_compile_flags() will result in an empty string instead of the expected -fno-rtti flag (See:

llvm-project/llvm/cmake/modules/AddLLVM.cmake

Line 58 in d1b88ca

list(APPEND LLVM_COMPILE_CXXFLAGS ${LLVM_CXXFLAGS_RTTI_DISABLE})

)

I also found a typo in another CMake variable.

I noticed that this function in upstream LLVM now depends on also including HandleLLVMOptions (llvm/llvm-project#174084). We could pull that in but it enables a lot of warnings globally, and seems likely to create additional churn when updating LLVM. Instead, we can just disable exceptions and rtti ourselves, and remove calls to the function.

Existing local reverts carried forward: * Local revert of llvm/llvm-project#169614 due to llvm/llvm-project#172932. * llvm/llvm-project#174084 -- causes linker errors related to rtti in stablehlo code * llvm/llvm-project@2b903df -- follow up fix to the above (doesn't fix it for us) Signed-off-by: Abhishek Varma <[email protected]> --------- Signed-off-by: Abhishek Varma <[email protected]>

…)" This reverts commit d1b88ca.

Existing local reverts carried forward: * Local revert of llvm/llvm-project#169614 due to llvm/llvm-project#172932. * llvm/llvm-project#174084 -- causes linker errors related to rtti in stablehlo code * llvm/llvm-project@2b903df -- follow up fix to the above (doesn't fix it for us) -- Along with the above, IREE header fixes were done to accomodate changes introduced via llvm/llvm-project#82190 for `Affine/Transforms/Passes.h`. NOTE: [Twine fix commit](iree-org/llvm-project@2eaa29c) was dropped from our fork since llvm/llvm-project#175077 got merged upstream. Signed-off-by: Abhishek Varma <[email protected]>

Existing local reverts carried forward: * Local revert of llvm/llvm-project#169614 due to llvm/llvm-project#172932. * llvm/llvm-project#174084 -- causes linker errors related to rtti in stablehlo code * llvm/llvm-project@2b903df -- follow up fix to the above (doesn't fix it for us) -- Along with the above, IREE header fixes were done to accomodate changes introduced via llvm/llvm-project#82190 for `Affine/Transforms/Passes.h`. NOTE: [Twine fix commit](iree-org/llvm-project@2eaa29c) was dropped from our fork since llvm/llvm-project#175077 got merged upstream. Signed-off-by: Abhishek Varma <[email protected]> Signed-off-by: Abhishek Varma <[email protected]>

I noticed that `llvm_update_compile_flags ` in upstream LLVM now depends on also including HandleLLVMOptions (llvm/llvm-project#174084). We could pull that in but it enables a lot of warnings globally, and seems likely to create additional churn when updating LLVM. Instead, we can just disable exceptions and rtti ourselves, and remove calls to the LLVM provided helpers. This makes the library builds easier to follow, and reduces the chance of further upstream LLVM changes breaking the build.

[CMake] Re-add -fno-rtti to llvm-config --cxxflags

dd55b7d

Fix for LLVM_CONFIG_HAS_RTTI.

aengelke requested a review from nikic December 31, 2025 11:25

llvmbot added the cmake Build system in general and CMake in particular label Dec 31, 2025

nikic reviewed Jan 1, 2026

View reviewed changes

different, less hacky approach

f8864b7

This comment was marked as outdated.

Sign in to view

remove debug print

20e8ba7

nikic approved these changes Jan 1, 2026

View reviewed changes

aengelke merged commit d1b88ca into llvm:main Jan 1, 2026
13 of 14 checks passed

aengelke deleted the fix-cmake-llvm-config branch January 1, 2026 21:24

ronlieb added a commit to ROCm/llvm-project that referenced this pull request Jan 2, 2026

Revert "[CMake] Re-add -fno-rtti to llvm-config --cxxflags (llvm#174084…

2e63367

…)" This reverts commit d1b88ca.

ronlieb added a commit to ROCm/llvm-project that referenced this pull request Jan 2, 2026

revert_patches.txt : [CMake] Re-add -fno-rtti to llvm-config --cxxfla…

0d85d2c

…gs (llvm#174084)

ronlieb pushed a commit to ROCm/llvm-project that referenced this pull request Jan 2, 2026

[CMake] Re-add -fno-rtti to llvm-config --cxxflags (llvm#174084)

52d3c62

Fix for llvm#173869. If there's no strong reason, we should get rid of per-target RTTI later.

kuhar added a commit to iree-org/llvm-project that referenced this pull request Jan 2, 2026

Revert "[CMake] Re-add -fno-rtti to llvm-config --cxxflags (llvm#174084…

e6187df

…)" This reverts commit d1b88ca.

kuhar mentioned this pull request Jan 2, 2026

Integrate llvm/llvm-project@7976ac990000 iree-org/iree#23010

Merged

Abhishek-Varma pushed a commit to iree-org/llvm-project that referenced this pull request Jan 5, 2026

Revert "[CMake] Re-add -fno-rtti to llvm-config --cxxflags (llvm#174084…

12d80f9

…)" This reverts commit d1b88ca.

Abhishek-Varma mentioned this pull request Jan 5, 2026

Integrate llvm/llvm-project@10a245bd02 iree-org/iree#23022

Merged

mahesh-attarde pushed a commit to mahesh-attarde/llvm-project that referenced this pull request Jan 6, 2026

[CMake] Re-add -fno-rtti to llvm-config --cxxflags (llvm#174084)

5f72978

Fix for llvm#173869. If there's no strong reason, we should get rid of per-target RTTI later.

Abhishek-Varma mentioned this pull request Jan 7, 2026

Integrate llvm/llvm-project@55eaa6c27b iree-org/iree#23048

Merged

Abhishek-Varma pushed a commit to iree-org/llvm-project that referenced this pull request Jan 7, 2026

Revert "[CMake] Re-add -fno-rtti to llvm-config --cxxflags (llvm#174084…

996f32c

…)" This reverts commit d1b88ca.

neildhar mentioned this pull request Jan 8, 2026

Remove calls to llvm_update_compile_flags in CMake triton-lang/triton#9167

Merged

Abhishek-Varma pushed a commit to iree-org/llvm-project that referenced this pull request Jan 9, 2026

Revert "[CMake] Re-add -fno-rtti to llvm-config --cxxflags (llvm#174084…

77dfbf0

…)" This reverts commit d1b88ca.

Abhishek-Varma mentioned this pull request Jan 9, 2026

Integrate llvm/llvm-project@eec258dcf38d iree-org/iree#23068

Merged

[CMake] Re-add -fno-rtti to llvm-config --cxxflags #174084

[CMake] Re-add -fno-rtti to llvm-config --cxxflags #174084

Uh oh!

Conversation

aengelke commented Dec 31, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

nikic left a comment

Choose a reason for hiding this comment

Uh oh!

This comment was marked as outdated.

This comment was marked as outdated.

aengelke commented Jan 1, 2026

Uh oh!

nikic left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

llvm-ci commented Jan 2, 2026

Uh oh!

llvm-ci commented Jan 2, 2026

Uh oh!

aengelke commented Jan 2, 2026

Uh oh!

llvm-ci commented Jan 2, 2026

Uh oh!

RKSimon commented Jan 2, 2026

Uh oh!

aengelke commented Jan 2, 2026

Uh oh!

RKSimon commented Jan 2, 2026

Uh oh!

ronlieb commented Jan 2, 2026

Uh oh!

kuhar commented Jan 2, 2026

Uh oh!

aengelke commented Jan 2, 2026

Uh oh!

kuhar commented Jan 2, 2026

Uh oh!

dcaballe commented Jan 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

aengelke commented Jan 6, 2026

Uh oh!

Abhishek-Varma commented Jan 7, 2026

Uh oh!

aengelke commented Jan 7, 2026

Uh oh!

dcaballe commented Jan 7, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

9 participants

aengelke commented Dec 31, 2025 •

edited

Loading

dcaballe commented Jan 6, 2026 •

edited

Loading