diff --git a/.github/workflows/check_format.yml b/.github/workflows/check_format.yml
index 0c6ad0e045..f3e8388726 100644
--- a/.github/workflows/check_format.yml
+++ b/.github/workflows/check_format.yml
@@ -23,7 +23,7 @@ jobs:
 
       - name: Install tools
         run: |
-          # Need non buggy version of clang-format-18 (See #1176).
+          # Need non buggy version of clang-format-18 (See #1174).
           wget -qO llvm.key https://apt.llvm.org/llvm-snapshot.gpg.key
           echo "ce6eee4130298f79b0e0f09a89f93c1bc711cd68e7e3182d37c8e96c5227e2f0  llvm.key" > check.sha256
           shasum -c check.sha256
diff --git a/.gitignore b/.gitignore
index abbced7712..edb9730dbe 100644
--- a/.gitignore
+++ b/.gitignore
@@ -9,6 +9,8 @@
 .clangd
 .cache
 /doc/html/
+.envrc
+*.png
 
 # ------------------- C++
 
diff --git a/ci/ctest_to_gitlab.sh b/ci/ctest_to_gitlab.sh
index 1500afa73c..456dc87764 100755
--- a/ci/ctest_to_gitlab.sh
+++ b/ci/ctest_to_gitlab.sh
@@ -63,8 +63,10 @@ JOB_TEMPLATE="
     DLAF_HDF5_TEST_OUTPUT_PATH: \$CI_PROJECT_DIR
   script: mpi-ctest -L {{CATEGORY_LABEL}} -L {{RANK_LABEL}}
   artifacts:
+    when: always
     paths:
-      - codecov-reports/"
+      - codecov-reports/
+      - output/"
 else
 # CRAY_CUDA_MPS set to 0 to avoid test hanging on daint (See PR #1197)
 BASE_TEMPLATE="
@@ -101,6 +103,10 @@ JOB_TEMPLATE="
     DISABLE_AFTER_SCRIPT: 'YES'
     DLAF_HDF5_TEST_OUTPUT_PATH: \$CI_PROJECT_DIR
   script: mpi-ctest -L {{CATEGORY_LABEL}} -L {{RANK_LABEL}}"
+  artifacts:
+    when: always
+    paths:
+      - output/"
 fi
 
 JOBS=""
diff --git a/ci/mpi-ctest b/ci/mpi-ctest
index d753fbcb18..7eefed90f3 100755
--- a/ci/mpi-ctest
+++ b/ci/mpi-ctest
@@ -7,6 +7,8 @@ if [[ "$ENABLE_COVERAGE" == "YES" ]]; then
     mkdir -p "$SHARED_REPORTS"
 fi;
 
+CTEST_OUTPUT="$CI_PROJECT_DIR/output/ctest.$SLURM_PROCID.txt"
+
 pushd /DLA-Future-build > /dev/null
 
 export CUDA_MPS_PIPE_DIRECTORY=/tmp/nvidia-mps
@@ -25,10 +27,10 @@ sleep 5
 # Run the tests, only output on the first rank
 if [[ $SLURM_PROCID == "0" ]]; then
     TZ=CET date +"Run started at: %H:%M:%S %z"
-    ctest -V $@
+    ctest --output-log "$CTEST_OUTPUT" -V $@
     TZ=CET date +"Run finished at: %H:%M:%S %z"
 else
-    ctest -Q $@
+    ctest --output-log "$CTEST_OUTPUT" -V -Q $@
 fi
 
 if [ $START_MPS -eq 1 ]; then
diff --git a/include/dlaf/common/eti.h b/include/dlaf/common/eti.h
index b2dd13af7b..f98b79df40 100644
--- a/include/dlaf/common/eti.h
+++ b/include/dlaf/common/eti.h
@@ -25,7 +25,7 @@
   ETI_MACRO(KWORD, float, Device::GPU);                \
   ETI_MACRO(KWORD, double, Device::GPU);               \
   ETI_MACRO(KWORD, std::complex<float>, Device::GPU);  \
-  ETI_MACRO(KWORD, std::complex<double>, Device::GPU);
+  ETI_MACRO(KWORD, std::complex<double>, Device::GPU)
 #define DLAF_EXPAND_ETI_SDCZ_DEVICE_VA_ARGS(ETI_MACRO, KWORD, ...)  \
   ETI_MACRO(KWORD, float, Device::CPU, __VA_ARGS__);                \
   ETI_MACRO(KWORD, double, Device::CPU, __VA_ARGS__);               \
@@ -34,16 +34,16 @@
   ETI_MACRO(KWORD, float, Device::GPU, __VA_ARGS__);                \
   ETI_MACRO(KWORD, double, Device::GPU, __VA_ARGS__);               \
   ETI_MACRO(KWORD, std::complex<float>, Device::GPU, __VA_ARGS__);  \
-  ETI_MACRO(KWORD, std::complex<double>, Device::GPU, __VA_ARGS__);
+  ETI_MACRO(KWORD, std::complex<double>, Device::GPU, __VA_ARGS__)
 #else
 #define DLAF_EXPAND_ETI_SDCZ_DEVICE(ETI_MACRO, KWORD) \
   ETI_MACRO(KWORD, float, Device::CPU);               \
   ETI_MACRO(KWORD, double, Device::CPU);              \
   ETI_MACRO(KWORD, std::complex<float>, Device::CPU); \
-  ETI_MACRO(KWORD, std::complex<double>, Device::CPU);
+  ETI_MACRO(KWORD, std::complex<double>, Device::CPU)
 #define DLAF_EXPAND_ETI_SDCZ_DEVICE_VA_ARGS(ETI_MACRO, KWORD, ...) \
   ETI_MACRO(KWORD, float, Device::CPU, __VA_ARGS__);               \
   ETI_MACRO(KWORD, double, Device::CPU, __VA_ARGS__);              \
   ETI_MACRO(KWORD, std::complex<float>, Device::CPU, __VA_ARGS__); \
-  ETI_MACRO(KWORD, std::complex<double>, Device::CPU, __VA_ARGS__);
+  ETI_MACRO(KWORD, std::complex<double>, Device::CPU, __VA_ARGS__)
 #endif
diff --git a/test/unit/multiplication/test_multiplication_hermitian.cpp b/test/unit/multiplication/test_multiplication_hermitian.cpp
index 0f21eab622..452d047433 100644
--- a/test/unit/multiplication/test_multiplication_hermitian.cpp
+++ b/test/unit/multiplication/test_multiplication_hermitian.cpp
@@ -96,9 +96,6 @@ void testHermitianMultiplication(const blas::Side side, const blas::Uplo uplo, c
     hermitian_multiplication<B>(side, uplo, alpha, mat_a.get(), mat_b.get(), beta, mat_c.get());
   }
 
-  // SCOPED_TRACE cannot yield.
-  mat_ch.waitLocalTiles();
-  SCOPED_TRACE(::testing::Message() << "m " << m << "n " << n << ", mb " << mb << ", nb " << nb);
   CHECK_MATRIX_NEAR(res_c, mat_ch, 10 * (m + 1) * TypeUtilities<T>::error,
                     10 * (m + 1) * TypeUtilities<T>::error);
 }