jit: Do not try to shut down LLVM state in case of LLVM triggered errors.

anarazel · anarazel · commit edb4d95ddf89 · 2021-09-13T18:26:15.000-07:00
If an allocation failed within LLVM it is not safe to call back into LLVM as LLVM is not generally safe against exceptions / stack-unwinding. Thus errors while in LLVM code are promoted to FATAL. However llvm_shutdown() did call back into LLVM even in such cases, while llvm_release_context() was careful not to do so. We cannot generally skip shutting down LLVM, as that can break profiling. But it's OK to do so if there was an error from within LLVM. Reported-By: Jelte Fennema <Jelte.Fennema@microsoft.com> Author: Andres Freund <andres@anarazel.de> Author: Justin Pryzby <pryzby@telsasoft.com> Discussion: https://postgr.es/m/AM5PR83MB0178C52CCA0A8DEA0207DC14F7FF9@AM5PR83MB0178.EURPRD83.prod.outlook.com Backpatch: 11-, where jit was introduced
diff --git a/src/backend/jit/llvm/llvmjit.c b/src/backend/jit/llvm/llvmjit.c
@@ -172,8 +172,6 @@ llvm_release_context(JitContext *context)
 {
 	LLVMJitContext *llvm_context = (LLVMJitContext *) context;
 
-	llvm_enter_fatal_on_oom();
-
 	/*
 	 * When this backend is exiting, don't clean up LLVM. As an error might
 	 * have occurred from within LLVM, we do not want to risk reentering. All
@@ -182,6 +180,8 @@ llvm_release_context(JitContext *context)
 	if (proc_exit_inprogress)
 		return;
 
+	llvm_enter_fatal_on_oom();
+
 	if (llvm_context->module)
 	{
 		LLVMDisposeModule(llvm_context->module);
@@ -885,6 +885,20 @@ llvm_session_initialize(void)
 static void
 llvm_shutdown(int code, Datum arg)
 {
+	/*
+	 * If llvm_shutdown() is reached while in a fatal-on-oom section an error
+	 * has occurred in the middle of LLVM code. It is not safe to call back
+	 * into LLVM (which is why a FATAL error was thrown).
+	 *
+	 * We do need to shutdown LLVM in other shutdown cases, otherwise
+	 * e.g. profiling data won't be written out.
+	 */
+	if (llvm_in_fatal_on_oom())
+	{
+		Assert(proc_exit_inprogress);
+		return;
+	}
+
 #if LLVM_VERSION_MAJOR > 11
 	{
 		if (llvm_opt3_orc)
diff --git a/src/backend/jit/llvm/llvmjit_error.cpp b/src/backend/jit/llvm/llvmjit_error.cpp
@@ -83,6 +83,16 @@ llvm_leave_fatal_on_oom(void)
 	}
 }
 
+/*
+ * Are we currently in an fatal-on-oom section? Useful to skip cleanup in case
+ * of errors.
+ */
+bool
+llvm_in_fatal_on_oom(void)
+{
+	return fatal_new_handler_depth > 0;
+}
+
 /*
  * Reset fatal error handling. This should only be called in error recovery
  * loops like PostgresMain()'s.
diff --git a/src/include/jit/llvmjit.h b/src/include/jit/llvmjit.h
@@ -84,6 +84,7 @@ extern LLVMValueRef AttributeTemplate;
 
 extern void llvm_enter_fatal_on_oom(void);
 extern void llvm_leave_fatal_on_oom(void);
+extern bool llvm_in_fatal_on_oom(void);
 extern void llvm_reset_after_error(void);
 extern void llvm_assert_in_fatal_section(void);
 
