diff --git a/clang/utils/perf-training/CMakeLists.txt b/clang/utils/perf-training/CMakeLists.txt
--- a/clang/utils/perf-training/CMakeLists.txt
+++ b/clang/utils/perf-training/CMakeLists.txt
@@ -11,7 +11,7 @@
   add_lit_testsuite(generate-profraw "Generating clang PGO data"
     ${CMAKE_CURRENT_BINARY_DIR}/pgo-data/
     EXCLUDE_FROM_CHECK_ALL
-    DEPENDS clang clear-profraw
+    DEPENDS clang clear-profraw ${CLANG_PERF_TRAINING_DEPS}
     )
 
   add_custom_target(clear-profraw
diff --git a/llvm/docs/AdvancedBuilds.rst b/llvm/docs/AdvancedBuilds.rst
--- a/llvm/docs/AdvancedBuilds.rst
+++ b/llvm/docs/AdvancedBuilds.rst
@@ -164,6 +164,14 @@
 You can feed that file into the LLVM_PROFDATA_FILE option when you build your
 optimized compiler.
 
+It may be necessary to build additional targets before running perf training, such as
+builtins and runtime libraries. You can use the :code:`CLANG_PERF_TRAINING_DEPS` CMake
+variable for that purpose:
+
+.. code-block:: cmake
+
+  set(CLANG_PERF_TRAINING_DEPS builtins runtimes CACHE STRING "")
+
 The PGO cache has a slightly different stage naming scheme than other
 multi-stage builds. It generates three stages: stage1, stage2-instrumented, and
 stage2. Both of the stage2 builds are built using the stage1 compiler.