Merge pull request #8353 from opensourcerouting/llvm-20210327

5 files changed, 505 insertions, 8 deletions

diff --git a/tools/.gitignore b/tools/.gitignore
index 63a5b61c35..1cc343a11a 100644
--- a/tools/.gitignore
+++ b/tools/.gitignore
@@ -8,3 +8,4 @@
 /frrcommon.sh
 /frr.service
 /frr@.service
+/frr-llvm-cg
diff --git a/tools/frr-llvm-cg.c b/tools/frr-llvm-cg.c
index 84a756a376..fbcf9222d2 100644
--- a/tools/frr-llvm-cg.c
+++ b/tools/frr-llvm-cg.c
@@ -50,11 +50,15 @@
 
 #include <json-c/json.h>
 
+#include "frr-llvm-debuginfo.h"
+
 /* if you want to use this without the special FRRouting defines,
  * remove the following #define
  */
 #define FRR_SPECIFIC
 
+static struct dbginfo *dbginfo;
+
 static void dbgloc_add(struct json_object *jsobj, LLVMValueRef obj)
 {
 	unsigned file_len = 0;
@@ -85,6 +89,70 @@ static struct json_object *js_get_or_make(struct json_object *parent,
 	return ret;
 }
 
+static bool try_struct_fptr(struct json_object *js_call, LLVMValueRef gep,
+			    const char *prefix)
+{
+	unsigned long long val = 0;
+	bool ret = false;
+	LLVMTypeRef ptrtype = LLVMTypeOf(LLVMGetOperand(gep, 0));
+	LLVMValueRef idx;
+
+	/* middle steps like struct a -> struct b a_member; -> fptr */
+	for (int i = 1; ptrtype && i < LLVMGetNumOperands(gep) - 1; i++) {
+		if (LLVMGetTypeKind(ptrtype) == LLVMPointerTypeKind
+		    || LLVMGetTypeKind(ptrtype) == LLVMArrayTypeKind
+		    || LLVMGetTypeKind(ptrtype) == LLVMVectorTypeKind) {
+			ptrtype = LLVMGetElementType(ptrtype);
+			continue;
+		}
+
+		if (LLVMGetTypeKind(ptrtype) != LLVMStructTypeKind)
+			return false;
+
+		idx = LLVMGetOperand(gep, i);
+		if (!LLVMIsConstant(idx))
+			return false;
+		val = LLVMConstIntGetZExtValue(idx);
+
+		unsigned n = LLVMGetNumContainedTypes(ptrtype);
+		LLVMTypeRef arr[n];
+
+		if (val > n)
+			return false;
+
+		LLVMGetSubtypes(ptrtype, arr);
+		ptrtype = arr[val];
+	}
+
+	if (!ptrtype)
+		return false;
+
+	idx = LLVMGetOperand(gep, LLVMGetNumOperands(gep) - 1);
+	if (!LLVMIsConstant(idx))
+		return false;
+
+	val = LLVMConstIntGetZExtValue(idx);
+
+	char *sname = NULL, *mname = NULL;
+
+	if (dbginfo_struct_member(dbginfo, ptrtype, val, &sname, &mname)) {
+		fprintf(stderr, "%s: call to struct %s->%s\n", prefix, sname,
+			mname);
+
+		json_object_object_add(js_call, "type",
+				       json_object_new_string("struct_memb"));
+		json_object_object_add(js_call, "struct",
+				       json_object_new_string(sname));
+		json_object_object_add(js_call, "member",
+				       json_object_new_string(mname));
+		ret = true;
+	}
+	free(sname);
+	free(mname);
+
+	return ret;
+}
+
 static bool details_fptr_vars = false;
 static bool details_fptr_consts = true;
 
@@ -175,6 +243,34 @@ static void walk_const_fptrs(struct json_object *js_call, LLVMValueRef value,
 					 prefix, hdr_written);
 		return;
 
+	case LLVMConstantExprValueKind:
+		switch (LLVMGetConstOpcode(value)) {
+		case LLVMGetElementPtr:
+			if (try_struct_fptr(js_call, value, prefix)) {
+				*hdr_written = true;
+				return;
+			}
+
+			fprintf(stderr,
+				"%s: calls function pointer from unhandled const GEP\n",
+				prefix);
+			*hdr_written = true;
+			/* fallthru */
+		default:
+			/* to help the user / development */
+			if (!*hdr_written) {
+				fprintf(stderr,
+					"%s: calls function pointer from constexpr\n",
+					prefix);
+				*hdr_written = true;
+			}
+			dump = LLVMPrintValueToString(value);
+			fprintf(stderr, "%s-   [opcode=%d] %s\n", prefix,
+				LLVMGetConstOpcode(value), dump);
+			LLVMDisposeMessage(dump);
+		}
+		return;
+
 	default:
 		/* to help the user / development */
 		if (!*hdr_written) {
@@ -197,7 +293,7 @@ static void walk_const_fptrs(struct json_object *js_call, LLVMValueRef value,
 #ifdef FRR_SPECIFIC
 static bool is_thread_sched(const char *name, size_t len)
 {
-#define thread_prefix "funcname_"
+#define thread_prefix "_"
 	static const char *const names[] = {
 		thread_prefix "thread_add_read_write",
 		thread_prefix "thread_add_timer",
@@ -218,6 +314,225 @@ static bool is_thread_sched(const char *name, size_t len)
 }
 #endif
 
+static bool _check_val(bool cond, const char *text, LLVMValueRef dumpval)
+{
+	if (cond)
+		return true;
+
+	char *dump = LLVMPrintValueToString(dumpval);
+	fprintf(stderr, "check failed: %s\ndump:\n\t%s\n", text, dump);
+	LLVMDisposeMessage(dump);
+	return false;
+}
+
+#define check_val(cond, dump)                                                  \
+	if (!_check_val(cond, #cond, dump))                                    \
+		return;
+
+static char *get_string(LLVMValueRef value)
+{
+	if (!LLVMIsAConstant(value))
+		return strdup("!NOT-A-CONST");
+
+	if (LLVMGetValueKind(value) == LLVMConstantExprValueKind
+	    && LLVMGetConstOpcode(value) == LLVMGetElementPtr) {
+		value = LLVMGetOperand(value, 0);
+
+		if (!LLVMIsAConstant(value))
+			return strdup("!NOT-A-CONST-2");
+	}
+
+	if (LLVMIsAGlobalVariable(value))
+		value = LLVMGetInitializer(value);
+
+	size_t len = 0;
+	const char *sval = LLVMGetAsString(value, &len);
+
+	return strndup(sval, len);
+}
+
+static void handle_yang_module(struct json_object *js_special,
+			       LLVMValueRef yang_mod)
+{
+	check_val(LLVMIsAGlobalVariable(yang_mod), yang_mod);
+
+	LLVMValueRef value;
+
+	value = LLVMGetInitializer(yang_mod);
+	LLVMValueKind kind = LLVMGetValueKind(value);
+
+	check_val(kind == LLVMConstantStructValueKind, value);
+
+	size_t var_len = 0;
+	const char *var_name = LLVMGetValueName2(yang_mod, &var_len);
+	char buf_name[var_len + 1];
+
+	memcpy(buf_name, var_name, var_len);
+	buf_name[var_len] = '\0';
+
+	struct json_object *js_yang, *js_yangmod, *js_items;
+
+	js_yang = js_get_or_make(js_special, "yang", json_object_new_object);
+	js_yangmod = js_get_or_make(js_yang, buf_name, json_object_new_object);
+	js_items = js_get_or_make(js_yangmod, "items", json_object_new_array);
+
+	char *mod_name = get_string(LLVMGetOperand(value, 0));
+	json_object_object_add(js_yangmod, "name",
+			       json_object_new_string(mod_name));
+	free(mod_name);
+
+	value = LLVMGetOperand(value, 1);
+	kind = LLVMGetValueKind(value);
+	check_val(kind == LLVMConstantArrayValueKind, value);
+
+	unsigned len = LLVMGetArrayLength(LLVMTypeOf(value));
+
+	for (unsigned i = 0; i < len - 1; i++) {
+		struct json_object *js_item, *js_cbs;
+		LLVMValueRef item = LLVMGetOperand(value, i);
+		char *xpath = get_string(LLVMGetOperand(item, 0));
+
+		js_item = json_object_new_object();
+		json_object_array_add(js_items, js_item);
+
+		json_object_object_add(js_item, "xpath",
+				       json_object_new_string(xpath));
+		js_cbs = js_get_or_make(js_item, "cbs", json_object_new_object);
+
+		free(xpath);
+
+		LLVMValueRef cbs = LLVMGetOperand(item, 1);
+
+		check_val(LLVMGetValueKind(cbs) == LLVMConstantStructValueKind,
+			  value);
+
+		LLVMTypeRef cbs_type = LLVMTypeOf(cbs);
+		unsigned cblen = LLVMCountStructElementTypes(cbs_type);
+
+		for (unsigned i = 0; i < cblen; i++) {
+			LLVMValueRef cb = LLVMGetOperand(cbs, i);
+
+			char *sname = NULL;
+			char *mname = NULL;
+
+			if (dbginfo_struct_member(dbginfo, cbs_type, i, &sname,
+						  &mname)) {
+				(void)0;
+			}
+
+			if (LLVMIsAFunction(cb)) {
+				size_t fn_len;
+				const char *fn_name;
+
+				fn_name = LLVMGetValueName2(cb, &fn_len);
+
+				json_object_object_add(
+					js_cbs, mname,
+					json_object_new_string_len(fn_name,
+								   fn_len));
+			}
+
+			free(sname);
+			free(mname);
+		}
+	}
+}
+
+static void handle_daemoninfo(struct json_object *js_special,
+			      LLVMValueRef daemoninfo)
+{
+	check_val(LLVMIsAGlobalVariable(daemoninfo), daemoninfo);
+
+	LLVMTypeRef type;
+	LLVMValueRef value;
+	unsigned len;
+
+	type = LLVMGlobalGetValueType(daemoninfo);
+	value = LLVMGetInitializer(daemoninfo);
+	LLVMValueKind kind = LLVMGetValueKind(value);
+
+	check_val(kind == LLVMConstantStructValueKind, value);
+
+	int yang_idx = -1;
+
+	len = LLVMCountStructElementTypes(type);
+
+	LLVMTypeRef fieldtypes[len];
+	LLVMGetSubtypes(type, fieldtypes);
+
+	for (unsigned i = 0; i < len; i++) {
+		LLVMTypeRef t = fieldtypes[i];
+
+		if (LLVMGetTypeKind(t) != LLVMPointerTypeKind)
+			continue;
+		t = LLVMGetElementType(t);
+		if (LLVMGetTypeKind(t) != LLVMPointerTypeKind)
+			continue;
+		t = LLVMGetElementType(t);
+		if (LLVMGetTypeKind(t) != LLVMStructTypeKind)
+			continue;
+
+		const char *name = LLVMGetStructName(t);
+		if (!strcmp(name, "struct.frr_yang_module_info"))
+			yang_idx = i;
+	}
+
+	if (yang_idx == -1)
+		return;
+
+	LLVMValueRef yang_mods = LLVMGetOperand(value, yang_idx);
+	LLVMValueRef yang_size = LLVMGetOperand(value, yang_idx + 1);
+
+	check_val(LLVMIsConstant(yang_size), yang_size);
+
+	unsigned long long ival = LLVMConstIntGetZExtValue(yang_size);
+
+	check_val(LLVMGetValueKind(yang_mods) == LLVMConstantExprValueKind
+			  && LLVMGetConstOpcode(yang_mods) == LLVMGetElementPtr,
+		  yang_mods);
+
+	yang_mods = LLVMGetOperand(yang_mods, 0);
+
+	check_val(LLVMIsAGlobalVariable(yang_mods), yang_mods);
+
+	yang_mods = LLVMGetInitializer(yang_mods);
+
+	check_val(LLVMGetValueKind(yang_mods) == LLVMConstantArrayValueKind,
+		  yang_mods);
+
+	len = LLVMGetArrayLength(LLVMTypeOf(yang_mods));
+
+	if (len != ival)
+		fprintf(stderr, "length mismatch - %llu vs. %u\n", ival, len);
+
+	for (unsigned i = 0; i < len; i++) {
+		char *dump;
+
+		LLVMValueRef item = LLVMGetOperand(yang_mods, i);
+		LLVMValueKind kind = LLVMGetValueKind(item);
+
+		check_val(kind == LLVMGlobalVariableValueKind
+				  || kind == LLVMConstantExprValueKind,
+			  item);
+
+		if (kind == LLVMGlobalVariableValueKind)
+			continue;
+
+		LLVMOpcode opcode = LLVMGetConstOpcode(item);
+		switch (opcode) {
+		case LLVMBitCast:
+			item = LLVMGetOperand(item, 0);
+			handle_yang_module(js_special, item);
+			break;
+
+		default:
+			dump = LLVMPrintValueToString(item);
+			printf("[%u] = [opcode=%u] %s\n", i, opcode, dump);
+			LLVMDisposeMessage(dump);
+		}
+	}
+}
+
 static void process_call(struct json_object *js_calls,
 			 struct json_object *js_special,
 			 LLVMValueRef instr,
@@ -227,6 +542,9 @@ static void process_call(struct json_object *js_calls,
 
 	LLVMValueRef called = LLVMGetCalledValue(instr);
 
+	if (LLVMIsAInlineAsm(called))
+		return;
+
 	if (LLVMIsAConstantExpr(called)) {
 		LLVMOpcode opcode = LLVMGetConstOpcode(called);
 
@@ -277,6 +595,12 @@ static void process_call(struct json_object *js_calls,
 		snprintf(prefix, sizeof(prefix), "%.*s:%d:%.*s()",
 			 (int)file_len, file, line, (int)name_len, name_c);
 
+		if (LLVMIsALoadInst(called)
+		    && LLVMIsAGetElementPtrInst(LLVMGetOperand(called, 0))
+		    && try_struct_fptr(js_call, LLVMGetOperand(called, 0),
+				       prefix))
+			goto out_struct_fptr;
+
 		while (LLVMIsALoadInst(last) || LLVMIsAGetElementPtrInst(last))
 			/* skipping over details for GEP here, but meh. */
 			last = LLVMGetOperand(last, 0);
@@ -324,12 +648,11 @@ static void process_call(struct json_object *js_calls,
 					prefix);
 		} else {
 			char *dump = LLVMPrintValueToString(called);
-			printf("\t%s\n", dump);
+			fprintf(stderr, "%s: ??? %s\n", prefix, dump);
 			LLVMDisposeMessage(dump);
 		}
-		return;
 #ifdef FRR_SPECIFIC
-	} else if (!strcmp(called_name, "install_element")) {
+	} else if (!strcmp(called_name, "_install_element")) {
 		called_type = FN_INSTALL_ELEMENT;
 
 		LLVMValueRef param0 = LLVMGetOperand(instr, 0);
@@ -380,10 +703,7 @@ static void process_call(struct json_object *js_calls,
 			json_object_new_string_len(called_name, called_len));
 
 		LLVMValueRef fparam;
-		if (strstr(called_name, "_read_"))
-			fparam = LLVMGetOperand(instr, 2);
-		else
-			fparam = LLVMGetOperand(instr, 1);
+		fparam = LLVMGetOperand(instr, 2);
 		assert(fparam);
 
 		size_t target_len = 0;
@@ -434,12 +754,21 @@ static void process_call(struct json_object *js_calls,
 		 * - zclient->* ?
 		 */
 #endif /* FRR_SPECIFIC */
+	} else if (!strcmp(called_name, "frr_preinit")) {
+		LLVMValueRef daemoninfo = LLVMGetOperand(instr, 0);
+
+		handle_daemoninfo(js_special, daemoninfo);
+
+		json_object_object_add(
+			js_call, "target",
+			json_object_new_string_len(called_name, called_len));
 	} else {
 		json_object_object_add(
 			js_call, "target",
 			json_object_new_string_len(called_name, called_len));
 	}
 
+out_struct_fptr:
 	for (unsigned argno = 0; argno < n_args; argno++) {
 		LLVMValueRef param = LLVMGetOperand(instr, argno);
 		size_t target_len;
@@ -597,6 +926,8 @@ int main(int argc, char **argv)
 	// done with the memory buffer now, so dispose of it
 	LLVMDisposeMemoryBuffer(memoryBuffer);
 
+	dbginfo = dbginfo_load(module);
+
 	struct json_object *js_root, *js_funcs, *js_special;
 
 	js_root = json_object_new_object();
diff --git a/tools/frr-llvm-debuginfo.cpp b/tools/frr-llvm-debuginfo.cpp
new file mode 100644
index 0000000000..ed3ad956b8
--- /dev/null
+++ b/tools/frr-llvm-debuginfo.cpp
@@ -0,0 +1,112 @@
+// This is free and unencumbered software released into the public domain.
+//
+// Anyone is free to copy, modify, publish, use, compile, sell, or
+// distribute this software, either in source code form or as a compiled
+// binary, for any purpose, commercial or non-commercial, and by any
+// means.
+//
+// In jurisdictions that recognize copyright laws, the author or authors
+// of this software dedicate any and all copyright interest in the
+// software to the public domain. We make this dedication for the benefit
+// of the public at large and to the detriment of our heirs and
+// successors. We intend this dedication to be an overt act of
+// relinquishment in perpetuity of all present and future rights to this
+// software under copyright law.
+//
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+// IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
+// OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+// ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+// OTHER DEALINGS IN THE SOFTWARE.
+//
+// For more information, please refer to <http://unlicense.org/>
+
+#include <llvm-c/BitReader.h>
+#include <llvm-c/BitWriter.h>
+#include <llvm-c/Core.h>
+#include <llvm-c/DebugInfo.h>
+
+#include <llvm/IR/Module.h>
+#include <llvm/IR/Value.h>
+#include <llvm/IR/Type.h>
+#include <llvm/IR/DebugInfo.h>
+#include <llvm/IR/DebugInfoMetadata.h>
+#include <llvm/Support/raw_ostream.h>
+
+#include <map>
+
+#include "frr-llvm-debuginfo.h"
+
+/* llvm::DebugInfoFinder is unfortunately not exposed in the llvm-c API... */
+
+struct dbginfo {
+	llvm::DebugInfoFinder finder;
+	std::map<std::string, llvm::DICompositeType *> tab;
+};
+
+struct dbginfo *dbginfo_load(LLVMModuleRef _mod)
+{
+	llvm::Module *mod = llvm::unwrap(_mod);
+	struct dbginfo *info = new dbginfo();
+
+	info->finder.processModule(*mod);
+
+	for (auto ty : info->finder.types()) {
+		if (ty->getMetadataID() != llvm::Metadata::DICompositeTypeKind)
+			continue;
+
+		llvm::DICompositeType *cty = (llvm::DICompositeType *)ty;
+		/* empty forward declarations aka "struct foobar;" */
+		if (cty->getElements().size() == 0)
+			continue;
+
+		info->tab.emplace(std::move(ty->getName().str()), cty);
+	}
+
+	return info;
+}
+
+bool dbginfo_struct_member(struct dbginfo *info, LLVMTypeRef _typ,
+			   unsigned long long idx, char **struct_name,
+			   char **member_name)
+{
+	*struct_name = NULL;
+	*member_name = NULL;
+
+	llvm::Type *typ = llvm::unwrap(_typ);
+
+	if (!typ->isStructTy())
+		return false;
+
+	llvm::StructType *styp = (llvm::StructType *)typ;
+	auto sname = styp->getStructName();
+
+	if (!sname.startswith("struct."))
+		return false;
+	sname = sname.drop_front(7);
+
+	size_t dot = sname.find_last_of(".");
+	if (dot != sname.npos)
+		sname = sname.take_front(dot);
+
+	auto item = info->tab.find(sname.str());
+	if (item == info->tab.end())
+		return false;
+
+	auto elements = item->second->getElements();
+	if (idx >= elements.size())
+		return false;
+
+	auto elem = elements[idx];
+
+	if (elem->getMetadataID() != llvm::Metadata::DIDerivedTypeKind)
+		return false;
+
+	llvm::DIDerivedType *dtyp = (llvm::DIDerivedType *)elem;
+
+	*struct_name = strdup(sname.str().c_str());
+	*member_name = strdup(dtyp->getName().str().c_str());
+	return true;
+}
diff --git a/tools/frr-llvm-debuginfo.h b/tools/frr-llvm-debuginfo.h
new file mode 100644
index 0000000000..fca4bc1f97
--- /dev/null
+++ b/tools/frr-llvm-debuginfo.h
@@ -0,0 +1,47 @@
+// This is free and unencumbered software released into the public domain.
+//
+// Anyone is free to copy, modify, publish, use, compile, sell, or
+// distribute this software, either in source code form or as a compiled
+// binary, for any purpose, commercial or non-commercial, and by any
+// means.
+//
+// In jurisdictions that recognize copyright laws, the author or authors
+// of this software dedicate any and all copyright interest in the
+// software to the public domain. We make this dedication for the benefit
+// of the public at large and to the detriment of our heirs and
+// successors. We intend this dedication to be an overt act of
+// relinquishment in perpetuity of all present and future rights to this
+// software under copyright law.
+//
+// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+// IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY CLAIM, DAMAGES OR
+// OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+// ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+// OTHER DEALINGS IN THE SOFTWARE.
+//
+// For more information, please refer to <http://unlicense.org/>
+
+#ifndef _FRR_LLVM_DEBUGINFO_H
+#define _FRR_LLVM_DEBUGINFO_H
+
+#include <stdbool.h>
+#include <llvm-c/Core.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+struct dbginfo;
+
+extern struct dbginfo *dbginfo_load(LLVMModuleRef mod);
+extern bool dbginfo_struct_member(struct dbginfo *di, LLVMTypeRef typ,
+				  unsigned long long idx, char **struct_name,
+				  char **member_name);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* _FRR_LLVM_DEBUGINFO_H */
diff --git a/tools/subdir.am b/tools/subdir.am
index 6a03a23baa..e4b9ecd84f 100644
--- a/tools/subdir.am
+++ b/tools/subdir.am
@@ -40,9 +40,15 @@ tools_ssd_CPPFLAGS =
 llvm_version = $(shell echo __clang_major__ | $(CC) -xc -P -E -)
 tools_frr_llvm_cg_CPPFLAGS = $(CPPFLAGS_BASE)
 tools_frr_llvm_cg_CFLAGS = $(AM_CFLAGS) `llvm-config-$(llvm_version) --cflags`
+tools_frr_llvm_cg_CXXFLAGS = $(AM_CXXFLAGS) -O0 -ggdb3 `llvm-config-$(llvm_version) --cxxflags`
 tools_frr_llvm_cg_LDFLAGS = `llvm-config-$(llvm_version) --ldflags --libs`
 tools_frr_llvm_cg_SOURCES = \
 	tools/frr-llvm-cg.c \
+	tools/frr-llvm-debuginfo.cpp \
+	# end
+
+noinst_HEADERS += \
+	tools/frr-llvm-debuginfo.h \
 	# end
 
 EXTRA_DIST += \