diff --git a/headers/private/kernel/kdriver_settings.h b/headers/private/kernel/kdriver_settings.h
index aa7ecd86b0..a00cb31fb8 100644
--- a/headers/private/kernel/kdriver_settings.h
+++ b/headers/private/kernel/kdriver_settings.h
@@ -16,7 +16,6 @@ extern "C" {
 #endif
 
 status_t driver_settings_init(struct kernel_args *args);
-status_t driver_settings_init_post_sem(struct kernel_args *args);
 
 #ifdef __cplusplus
 }
diff --git a/src/system/kernel/arch/generic/generic_vm_physical_page_mapper.cpp b/src/system/kernel/arch/generic/generic_vm_physical_page_mapper.cpp
index a245d12385..2257cb84d7 100644
--- a/src/system/kernel/arch/generic/generic_vm_physical_page_mapper.cpp
+++ b/src/system/kernel/arch/generic/generic_vm_physical_page_mapper.cpp
@@ -43,7 +43,7 @@ static paddr_chunk_desc **virtual_pmappings; // will be one ptr per virtual chun
 static int first_free_vmapping;
 static int num_virtual_chunks;
 static queue mapped_paddr_lru;
-static mutex sMutex;
+static mutex sMutex = MUTEX_INITIALIZER("iospace_mutex");
 static sem_id sChunkAvailableSem;
 static int32 sChunkAvailableWaitingCounter;
 
@@ -282,7 +282,6 @@ generic_vm_physical_page_mapper_init(kernel_args *args,
 	memset(virtual_pmappings, 0, sizeof(paddr_chunk_desc *) * num_virtual_chunks);
 	first_free_vmapping = 0;
 	queue_init(&mapped_paddr_lru);
-	mutex_init(&sMutex, "iospace_mutex");
 	sChunkAvailableSem = -1;
 
 	TRACE(("generic_vm_physical_page_mapper_init: done\n"));
diff --git a/src/system/kernel/debug/frame_buffer_console.cpp b/src/system/kernel/debug/frame_buffer_console.cpp
index 0155bf9297..359ca8a90e 100644
--- a/src/system/kernel/debug/frame_buffer_console.cpp
+++ b/src/system/kernel/debug/frame_buffer_console.cpp
@@ -396,6 +396,8 @@ frame_buffer_console_init(kernel_args *args)
 	if (!args->frame_buffer.enabled)
 		return B_OK;
 
+	mutex_init(&sConsole.lock, "console_lock");
+
 	void *frameBuffer;
 	sConsole.area = map_physical_memory("vesa_fb",
 		(void *)args->frame_buffer.physical_buffer.start,
@@ -443,8 +445,6 @@ frame_buffer_console_init(kernel_args *args)
 status_t
 frame_buffer_console_init_post_modules(kernel_args *args)
 {
-	mutex_init(&sConsole.lock, "console_lock");
-
 	if (sConsole.frame_buffer == 0)
 		return B_OK;
 
diff --git a/src/system/kernel/device_manager/id_generator.cpp b/src/system/kernel/device_manager/id_generator.cpp
index 2a5bed3ffe..9fadf6e709 100644
--- a/src/system/kernel/device_manager/id_generator.cpp
+++ b/src/system/kernel/device_manager/id_generator.cpp
@@ -67,7 +67,7 @@ typedef DoublyLinkedList<id_generator> GeneratorList;
 
 
 GeneratorList sGenerators;
-static mutex sLock;
+static mutex sLock = MUTEX_INITIALIZER("id generator");
 
 
 /*!	Create new generator.
@@ -169,7 +169,6 @@ void
 dm_init_id_generator(void)
 {
 	new(&sGenerators) GeneratorList;
-	mutex_init(&sLock, "id generator");
 }
 
 
diff --git a/src/system/kernel/elf.cpp b/src/system/kernel/elf.cpp
index 157602bdf1..aa44ec43d5 100644
--- a/src/system/kernel/elf.cpp
+++ b/src/system/kernel/elf.cpp
@@ -59,9 +59,11 @@ typedef struct elf_linked_image {
 static hash_table *sImagesHash;
 
 static struct elf_image_info *sKernelImage = NULL;
-static mutex sImageMutex;		// guards sImagesHash
-static mutex sImageLoadMutex;	// serializes loading/unloading add-ons
-								// locking order sImageLoadMutex -> sImageMutex
+static mutex sImageMutex = MUTEX_INITIALIZER("kimages_lock");
+	// guards sImagesHash
+static mutex sImageLoadMutex = MUTEX_INITIALIZER("kimages_load_lock");
+	// serializes loading/unloading add-ons locking order
+	// sImageLoadMutex -> sImageMutex
 static bool sInitialized = false;
 
 
@@ -1785,9 +1787,6 @@ elf_init(kernel_args *args)
 
 	image_init();
 
-	mutex_init(&sImageMutex, "kimages_lock");
-	mutex_init(&sImageLoadMutex, "kimages_load_lock");
-
 	sImagesHash = hash_init(IMAGE_HASH_SIZE, 0, image_compare, image_hash);
 	if (sImagesHash == NULL)
 		return B_NO_MEMORY;
diff --git a/src/system/kernel/fs/vfs.cpp b/src/system/kernel/fs/vfs.cpp
index 638ea660c0..cd1cb0bf90 100644
--- a/src/system/kernel/fs/vfs.cpp
+++ b/src/system/kernel/fs/vfs.cpp
@@ -169,7 +169,7 @@ struct advisory_locking {
 	}
 };
 
-static mutex sFileSystemsMutex;
+static mutex sFileSystemsMutex = MUTEX_INITIALIZER("vfs_lock");
 
 /*!	\brief Guards sMountsTable.
 
@@ -177,7 +177,7 @@ static mutex sFileSystemsMutex;
 	Manipulation of the fs_mount structures themselves
 	(and their destruction) requires different locks though.
 */
-static mutex sMountMutex;
+static mutex sMountMutex = MUTEX_INITIALIZER("vfs_mount_lock");
 
 /*!	\brief Guards mount/unmount operations.
 
@@ -201,7 +201,8 @@ static recursive_lock sMountOpLock;
 
 	The thread trying to lock the must not hold sVnodeMutex.
 */
-static mutex sVnodeCoveredByMutex;
+static mutex sVnodeCoveredByMutex
+	= MUTEX_INITIALIZER("vfs_vnode_covered_by_lock");
 
 /*!	\brief Guards sVnodeTable.
 
@@ -215,7 +216,7 @@ static mutex sVnodeCoveredByMutex;
 	You must not have this mutex held when calling create_sem(), as this
 	might call vfs_free_unused_vnodes().
 */
-static mutex sVnodeMutex;
+static mutex sVnodeMutex = MUTEX_INITIALIZER("vfs_vnode_lock");
 
 /*!	\brief Guards io_context::root.
 
@@ -223,7 +224,7 @@ static mutex sVnodeMutex;
 	The only operation allowed while holding this lock besides getting or
 	setting the field is inc_vnode_ref_count() on io_context::root.
 */
-static mutex sIOContextRootLock;
+static mutex sIOContextRootLock = MUTEX_INITIALIZER("io_context::root lock");
 
 #define VNODE_HASH_TABLE_SIZE 1024
 static hash_table *sVnodeTable;
@@ -4414,13 +4415,7 @@ vfs_init(kernel_args *args)
 
 	sRoot = NULL;
 
-	mutex_init(&sFileSystemsMutex, "vfs_lock");
-
 	recursive_lock_init(&sMountOpLock, "vfs_mount_op_lock");
-	mutex_init(&sMountMutex, "vfs_mount_lock");
-	mutex_init(&sVnodeCoveredByMutex, "vfs_vnode_covered_by_lock");
-	mutex_init(&sVnodeMutex, "vfs_vnode_lock");
-	mutex_init(&sIOContextRootLock, "io_context::root lock");
 
 	if (block_cache_init() != B_OK)
 		return B_ERROR;
diff --git a/src/system/kernel/image.c b/src/system/kernel/image.c
index 20f4ff36a6..4b78476a61 100644
--- a/src/system/kernel/image.c
+++ b/src/system/kernel/image.c
@@ -38,7 +38,7 @@ struct image {
 
 
 static image_id sNextImageID = 1;
-static mutex sImageMutex;
+static mutex sImageMutex = MUTEX_INITIALIZER("image");
 
 
 /*!	Registers an image with the specified team.
@@ -274,7 +274,6 @@ image_init(void)
 	add_debugger_command("team_images", &dump_images_list, "Dump all registered images from the current team");
 #endif
 
-	mutex_init(&sImageMutex, "image");
 	return B_OK;
 }
 
diff --git a/src/system/kernel/kernel_daemon.cpp b/src/system/kernel/kernel_daemon.cpp
index 6ce1712cbc..f3cddbfc55 100644
--- a/src/system/kernel/kernel_daemon.cpp
+++ b/src/system/kernel/kernel_daemon.cpp
@@ -33,7 +33,7 @@ struct daemon : DoublyLinkedListLinkImpl<struct daemon> {
 
 typedef DoublyLinkedList<struct daemon> DaemonList;
 
-static mutex sDaemonMutex;
+static mutex sDaemonMutex = MUTEX_INITIALIZER("kernel daemon");
 static DaemonList sDaemons;
 
 
@@ -130,12 +130,9 @@ register_kernel_daemon(daemon_hook function, void* arg, int frequency)
 extern "C" status_t
 kernel_daemon_init(void)
 {
-	thread_id thread;
-
-	mutex_init(&sDaemonMutex, "kernel daemon");
 	new(&sDaemons) DaemonList;
 
-	thread = spawn_kernel_thread(&kernel_daemon, "kernel daemon",
+	thread_id thread = spawn_kernel_thread(&kernel_daemon, "kernel daemon",
 		B_LOW_PRIORITY, NULL);
 	send_signal_etc(thread, SIGCONT, B_DO_NOT_RESCHEDULE);
 
diff --git a/src/system/kernel/lock.cpp b/src/system/kernel/lock.cpp
index 8ae7f72c52..6a8371dd11 100644
--- a/src/system/kernel/lock.cpp
+++ b/src/system/kernel/lock.cpp
@@ -400,15 +400,14 @@ _mutex_lock(mutex* lock, bool threadsLocked)
 	// Might have been released after we decremented the count, but before
 	// we acquired the spinlock.
 #ifdef KDEBUG
-	if (!kernel_startup && lock->holder == thread_get_current_thread_id()) {
-		panic("_mutex_lock(): double lock of %p by thread %ld", lock,
-			lock->holder);
-	}
-
-	if (lock->holder <= 0) {
+	if (lock->holder < 0) {
 		lock->holder = thread_get_current_thread_id();
 		return B_OK;
-	}
+	} else if (lock->holder == thread_get_current_thread_id()) {
+		panic("_mutex_lock(): double lock of %p by thread %ld", lock,
+			lock->holder);
+	} else if (lock->holder == 0)
+		panic("_mutex_lock(): using unitialized lock %p", lock);
 #else
 	if ((lock->flags & MUTEX_FLAG_RELEASED) != 0) {
 		lock->flags &= ~MUTEX_FLAG_RELEASED;
diff --git a/src/system/kernel/main.cpp b/src/system/kernel/main.cpp
index 957a3c1b80..d10628ee48 100644
--- a/src/system/kernel/main.cpp
+++ b/src/system/kernel/main.cpp
@@ -156,8 +156,6 @@ _start(kernel_args *bootKernelArgs, int currentCPU)
 		// now we can create and use semaphores
 		TRACE("init VM semaphores\n");
 		vm_init_post_sem(&sKernelArgs);
-		TRACE("init driver_settings\n");
-		driver_settings_init_post_sem(&sKernelArgs);
 		TRACE("init generic syscall\n");
 		generic_syscall_init();
 		smp_init_post_generic_syscalls();
diff --git a/src/system/kernel/slab/Slab.cpp b/src/system/kernel/slab/Slab.cpp
index 7338925f89..04020781b9 100644
--- a/src/system/kernel/slab/Slab.cpp
+++ b/src/system/kernel/slab/Slab.cpp
@@ -164,7 +164,7 @@ struct depot_cpu_store {
 
 
 static ObjectCacheList sObjectCaches;
-static mutex sObjectCacheListLock;
+static mutex sObjectCacheListLock = MUTEX_INITIALIZER("object cache list");
 
 static uint8 *sInitialBegin, *sInitialLimit, *sInitialPointer;
 static kernel_args *sKernelArgs;
@@ -1454,8 +1454,6 @@ slab_init(kernel_args *args, addr_t initialBase, size_t initialSize)
 void
 slab_init_post_sem()
 {
-	mutex_init(&sObjectCacheListLock, "object cache list");
-
 	ObjectCacheList::Iterator it = sObjectCaches.GetIterator();
 
 	while (it.HasNext()) {
diff --git a/src/system/kernel/syscalls.cpp b/src/system/kernel/syscalls.cpp
index e2652b182a..622cfa34b4 100644
--- a/src/system/kernel/syscalls.cpp
+++ b/src/system/kernel/syscalls.cpp
@@ -55,7 +55,7 @@ struct generic_syscall {
 	generic_syscall	*previous;
 };
 
-static struct mutex sGenericSyscallLock;
+static mutex sGenericSyscallLock = MUTEX_INITIALIZER("generic syscall");
 static struct list sGenericSyscalls;
 
 
@@ -212,7 +212,6 @@ status_t
 generic_syscall_init(void)
 {
 	list_init(&sGenericSyscalls);
-	mutex_init(&sGenericSyscallLock, "generic syscall");
 
 #if	SYSCALL_TRACING
 	add_debugger_command_etc("straced", &dump_syscall_tracing,
diff --git a/src/system/kernel/util/cbuf.c b/src/system/kernel/util/cbuf.c
index 55eccf32cf..9aa4fb9e32 100644
--- a/src/system/kernel/util/cbuf.c
+++ b/src/system/kernel/util/cbuf.c
@@ -44,7 +44,7 @@ struct cbuf {
 #define CBUF_BITMAP_SIZE (CBUF_REGION_SIZE / CBUF_LENGTH)
 
 static cbuf *sFreeBufferList;
-static mutex sFreeBufferListMutex;
+static mutex sFreeBufferListMutex = MUTEX_INITIALIZER("cbuf_free_list");
 static cbuf *sFreeBufferNoBlockList;
 static spinlock sNoBlockSpinlock;
 
@@ -950,8 +950,6 @@ cbuf_init(void)
 	// add the debug command
 	add_debugger_command("cbuf_freelist", &dbg_dump_cbuf_freelists, "Dumps the cbuf free lists");
 
-	mutex_init(&sFreeBufferListMutex, "cbuf_free_list");
-
 	// errors are fatal, that's why we don't clean up here
 
 	sBufferArea = create_area("cbuf region", (void **)&sBuffer, B_ANY_KERNEL_ADDRESS,
diff --git a/src/system/kernel/vm/vm.cpp b/src/system/kernel/vm/vm.cpp
index 18b390e0b2..3c2bb19daa 100644
--- a/src/system/kernel/vm/vm.cpp
+++ b/src/system/kernel/vm/vm.cpp
@@ -194,11 +194,11 @@ public:
 static area_id sNextAreaID;
 static hash_table *sAreaHash;
 static sem_id sAreaHashLock;
-static mutex sMappingLock;
-static mutex sAreaCacheLock;
+static mutex sMappingLock = MUTEX_INITIALIZER("page mappings");
+static mutex sAreaCacheLock = MUTEX_INITIALIZER("area->cache");
 
 static off_t sAvailableMemory;
-static mutex sAvailableMemoryLock;
+static mutex sAvailableMemoryLock = MUTEX_INITIALIZER("available memory lock");
 
 #if DEBUG_CACHE_LIST
 
@@ -3955,13 +3955,10 @@ vm_init_post_sem(kernel_args *args)
 	// since we're still single threaded and only the kernel address space exists,
 	// it isn't that hard to find all of the ones we need to create
 
-	mutex_init(&sAvailableMemoryLock, "available memory lock");
 	arch_vm_translation_map_init_post_sem(args);
 	vm_address_space_init_post_sem();
 
 	sAreaHashLock = create_sem(WRITE_COUNT, "area hash");
-	mutex_init(&sAreaCacheLock, "area->cache");
-	mutex_init(&sMappingLock, "page mappings");
 
 	slab_init_post_sem();
 	return heap_init_post_sem();
diff --git a/src/system/kernel/vm/vm_low_memory.cpp b/src/system/kernel/vm/vm_low_memory.cpp
index f36361d18d..0096a14eff 100644
--- a/src/system/kernel/vm/vm_low_memory.cpp
+++ b/src/system/kernel/vm/vm_low_memory.cpp
@@ -50,7 +50,7 @@ static const size_t kCriticalLimit = 32;
 static int32 sLowMemoryState = B_NO_LOW_MEMORY;
 static bigtime_t sLastMeasurement;
 
-static mutex sLowMemoryMutex;
+static mutex sLowMemoryMutex = MUTEX_INITIALIZER("low memory");
 static sem_id sLowMemoryWaitSem;
 static HandlerList sLowMemoryHandlers;
 
@@ -184,8 +184,6 @@ vm_low_memory_init(void)
 status_t
 vm_low_memory_init_post_thread(void)
 {
-	mutex_init(&sLowMemoryMutex, "low memory");
-	
 	sLowMemoryWaitSem = create_sem(0, "low memory wait");
 	if (sLowMemoryWaitSem < B_OK)
 		return sLowMemoryWaitSem;
diff --git a/src/system/libroot/os/driver_settings.c b/src/system/libroot/os/driver_settings.c
index f0c258eadd..17bd88a68c 100644
--- a/src/system/libroot/os/driver_settings.c
+++ b/src/system/libroot/os/driver_settings.c
@@ -88,7 +88,7 @@ enum assignment_mode {
 
 #ifdef _KERNEL_MODE
 static struct list sHandles;
-static mutex sLock;
+static mutex sLock = MUTEX_INITIALIZER("driver settings");
 #endif
 
 
@@ -652,14 +652,6 @@ driver_settings_init(kernel_args *args)
 
 	return B_OK;
 }
-
-
-status_t
-driver_settings_init_post_sem(kernel_args *args)
-{
-	mutex_init(&sLock, "driver settings");
-	return B_OK;
-}
 #endif