Key the object cache on the sources that generate the objects

CMakeLists.txt

@@ -196,6 +196,30 @@ add_library(dr_app STATIC
     src/app/pipeline.c
 )
 target_include_directories(dr_app PUBLIC ${DOLRECOMP_SRC})
+# The generated-object cache must invalidate when the code that generates those
+# objects changes. A hand-edited version string cannot do that -- it only
+# invalidates when somebody remembers to bump it. Hash the sources that decide
+# what gets emitted and fold the digest into every cache key. GLOB with
+# CONFIGURE_DEPENDS so a newly added emitter file is covered automatically
+# rather than needing to be added to a list that can rot.
+file(GLOB DOLRECOMP_CODEGEN_SOURCES CONFIGURE_DEPENDS
+    "${CMAKE_CURRENT_SOURCE_DIR}/src/backend/*.c"
+    "${CMAKE_CURRENT_SOURCE_DIR}/src/backend/*.h"
+    "${CMAKE_CURRENT_SOURCE_DIR}/src/backend/llvm/*.cpp"
+    "${CMAKE_CURRENT_SOURCE_DIR}/src/backend/llvm/*.h"
+    "${CMAKE_CURRENT_SOURCE_DIR}/src/ir/*.c"
+    "${CMAKE_CURRENT_SOURCE_DIR}/src/ir/*.h")
+list(SORT DOLRECOMP_CODEGEN_SOURCES)
+set(DOLRECOMP_CODEGEN_DIGEST "")
+foreach(_codegen_source IN LISTS DOLRECOMP_CODEGEN_SOURCES)
+    file(SHA256 "${_codegen_source}" _codegen_file_hash)
+    string(APPEND DOLRECOMP_CODEGEN_DIGEST "${_codegen_file_hash}")
+endforeach()
+string(SHA256 DOLRECOMP_CODEGEN_SOURCE_HASH "${DOLRECOMP_CODEGEN_DIGEST}")
+list(LENGTH DOLRECOMP_CODEGEN_SOURCES _codegen_count)
+message(STATUS "Codegen source digest: ${DOLRECOMP_CODEGEN_SOURCE_HASH} (${_codegen_count} files)")
+target_compile_definitions(dr_app PRIVATE
+    DOLRECOMP_CODEGEN_SOURCE_HASH="${DOLRECOMP_CODEGEN_SOURCE_HASH}")
 target_link_libraries(dr_app PUBLIC dr_backend dr_analysis dr_frontend dr_platform dr_ir)
 if(DOLRECOMP_ENABLE_LLVM)
     target_link_libraries(dr_app PUBLIC dr_llvm)

src/app/pipeline.c

@@ -250,6 +250,14 @@ static u64 hash_file_contents(u64 hash, const char* path) {
 static u64 llvm_job_hash(const LLVMChunkJob* job) {
     u64 hash = 1469598103934665603ull;
     hash = hash_bytes(hash, DOLLLVM_CACHE_VERSION, strlen(DOLLLVM_CACHE_VERSION));
+#ifdef DOLRECOMP_CODEGEN_SOURCE_HASH
+    /* A hand-edited version string cannot invalidate the cache when the code
+       that generates the objects changes -- it only invalidates when somebody
+       remembers to bump it. CMake hashes the emitter sources and passes the
+       digest in, so editing any of them changes every job key automatically. */
+    hash = hash_bytes(hash, DOLRECOMP_CODEGEN_SOURCE_HASH,
+                      strlen(DOLRECOMP_CODEGEN_SOURCE_HASH));
+#endif
     hash = hash_bytes(hash, &job->function_address, sizeof(job->function_address));
     hash = hash_bytes(hash, &job->count, sizeof(job->count));
     u32 state_size = (u32)sizeof(CPUState);