Merge tag 'char-misc-4.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc

Pull char/misc driver updates from Greg KH:
 "Here is the big set of char/misc patches for 4.20-rc1.

  Loads of things here, we have new code in all of these driver
  subsystems:
   - fpga
   - stm
   - extcon
   - nvmem
   - eeprom
   - hyper-v
   - gsmi
   - coresight
   - thunderbolt
   - vmw_balloon
   - goldfish
   - soundwire
  along with lots of fixes and minor changes to other small drivers.

  All of these have been in linux-next for a while with no reported
  issues"

* tag 'char-misc-4.20-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc: (245 commits)
  Documentation/security-bugs: Clarify treatment of embargoed information
  lib: Fix ia64 bootloader linkage
  MAINTAINERS: Clarify UIO vs UIOVEC maintainer
  docs/uio: fix a grammar nitpick
  docs: fpga: document programming fpgas using regions
  fpga: add devm_fpga_region_create
  fpga: bridge: add devm_fpga_bridge_create
  fpga: mgr: add devm_fpga_mgr_create
  hv_balloon: Replace spin_is_locked() with lockdep
  sgi-xp: Replace spin_is_locked() with lockdep
  eeprom: New ee1004 driver for DDR4 memory
  eeprom: at25: remove unneeded 'at25_remove'
  w1: IAD Register is yet readable trough iad sys file. Fix snprintf (%u for unsigned, count for max size).
  misc: mic: scif: remove set but not used variables 'src_dma_addr, dst_dma_addr'
  misc: mic: fix a DMA pool free failure
  platform: goldfish: pipe: Add a blank line to separate varibles and code
  platform: goldfish: pipe: Remove redundant casting
  platform: goldfish: pipe: Call misc_deregister if init fails
  platform: goldfish: pipe: Move the file-scope goldfish_pipe_dev variable into the driver state
  platform: goldfish: pipe: Move the file-scope goldfish_pipe_miscdev variable into the driver state
  ...

146 files changed, 6032 insertions, 3006 deletions

diff --git a/drivers/android/Kconfig b/drivers/android/Kconfig
index 432e9ad77070..51e8250d113f 100644
--- a/drivers/android/Kconfig
+++ b/drivers/android/Kconfig
@@ -10,7 +10,7 @@ if ANDROID
 
 config ANDROID_BINDER_IPC
 	bool "Android Binder IPC Driver"
-	depends on MMU
+	depends on MMU && !CPU_CACHE_VIVT
 	default n
 	---help---
 	  Binder is used in Android for both communication between processes,
diff --git a/drivers/android/binder.c b/drivers/android/binder.c
index d58763b6b009..cb30a524d16d 100644
--- a/drivers/android/binder.c
+++ b/drivers/android/binder.c
@@ -71,6 +71,7 @@
 #include <linux/security.h>
 #include <linux/spinlock.h>
 #include <linux/ratelimit.h>
+#include <linux/syscalls.h>
 
 #include <uapi/linux/android/binder.h>
 
@@ -457,9 +458,8 @@ struct binder_ref {
 };
 
 enum binder_deferred_state {
-	BINDER_DEFERRED_PUT_FILES    = 0x01,
-	BINDER_DEFERRED_FLUSH        = 0x02,
-	BINDER_DEFERRED_RELEASE      = 0x04,
+	BINDER_DEFERRED_FLUSH        = 0x01,
+	BINDER_DEFERRED_RELEASE      = 0x02,
 };
 
 /**
@@ -480,9 +480,6 @@ enum binder_deferred_state {
  *                        (invariant after initialized)
  * @tsk                   task_struct for group_leader of process
  *                        (invariant after initialized)
- * @files                 files_struct for process
- *                        (protected by @files_lock)
- * @files_lock            mutex to protect @files
  * @deferred_work_node:   element for binder_deferred_list
  *                        (protected by binder_deferred_lock)
  * @deferred_work:        bitmap of deferred work to perform
@@ -527,8 +524,6 @@ struct binder_proc {
 	struct list_head waiting_threads;
 	int pid;
 	struct task_struct *tsk;
-	struct files_struct *files;
-	struct mutex files_lock;
 	struct hlist_node deferred_work_node;
 	int deferred_work;
 	bool is_dead;
@@ -611,6 +606,23 @@ struct binder_thread {
 	bool is_dead;
 };
 
+/**
+ * struct binder_txn_fd_fixup - transaction fd fixup list element
+ * @fixup_entry:          list entry
+ * @file:                 struct file to be associated with new fd
+ * @offset:               offset in buffer data to this fixup
+ *
+ * List element for fd fixups in a transaction. Since file
+ * descriptors need to be allocated in the context of the
+ * target process, we pass each fd to be processed in this
+ * struct.
+ */
+struct binder_txn_fd_fixup {
+	struct list_head fixup_entry;
+	struct file *file;
+	size_t offset;
+};
+
 struct binder_transaction {
 	int debug_id;
 	struct binder_work work;
@@ -628,6 +640,7 @@ struct binder_transaction {
 	long	priority;
 	long	saved_priority;
 	kuid_t	sender_euid;
+	struct list_head fd_fixups;
 	/**
 	 * @lock:  protects @from, @to_proc, and @to_thread
 	 *
@@ -822,6 +835,7 @@ static void
 binder_enqueue_deferred_thread_work_ilocked(struct binder_thread *thread,
 					    struct binder_work *work)
 {
+	WARN_ON(!list_empty(&thread->waiting_thread_node));
 	binder_enqueue_work_ilocked(work, &thread->todo);
 }
 
@@ -839,6 +853,7 @@ static void
 binder_enqueue_thread_work_ilocked(struct binder_thread *thread,
 				   struct binder_work *work)
 {
+	WARN_ON(!list_empty(&thread->waiting_thread_node));
 	binder_enqueue_work_ilocked(work, &thread->todo);
 	thread->process_todo = true;
 }
@@ -920,66 +935,6 @@ static void binder_free_thread(struct binder_thread *thread);
 static void binder_free_proc(struct binder_proc *proc);
 static void binder_inc_node_tmpref_ilocked(struct binder_node *node);
 
-static int task_get_unused_fd_flags(struct binder_proc *proc, int flags)
-{
-	unsigned long rlim_cur;
-	unsigned long irqs;
-	int ret;
-
-	mutex_lock(&proc->files_lock);
-	if (proc->files == NULL) {
-		ret = -ESRCH;
-		goto err;
-	}
-	if (!lock_task_sighand(proc->tsk, &irqs)) {
-		ret = -EMFILE;
-		goto err;
-	}
-	rlim_cur = task_rlimit(proc->tsk, RLIMIT_NOFILE);
-	unlock_task_sighand(proc->tsk, &irqs);
-
-	ret = __alloc_fd(proc->files, 0, rlim_cur, flags);
-err:
-	mutex_unlock(&proc->files_lock);
-	return ret;
-}
-
-/*
- * copied from fd_install
- */
-static void task_fd_install(
-	struct binder_proc *proc, unsigned int fd, struct file *file)
-{
-	mutex_lock(&proc->files_lock);
-	if (proc->files)
-		__fd_install(proc->files, fd, file);
-	mutex_unlock(&proc->files_lock);
-}
-
-/*
- * copied from sys_close
- */
-static long task_close_fd(struct binder_proc *proc, unsigned int fd)
-{
-	int retval;
-
-	mutex_lock(&proc->files_lock);
-	if (proc->files == NULL) {
-		retval = -ESRCH;
-		goto err;
-	}
-	retval = __close_fd(proc->files, fd);
-	/* can't restart close syscall because file table entry was cleared */
-	if (unlikely(retval == -ERESTARTSYS ||
-		     retval == -ERESTARTNOINTR ||
-		     retval == -ERESTARTNOHAND ||
-		     retval == -ERESTART_RESTARTBLOCK))
-		retval = -EINTR;
-err:
-	mutex_unlock(&proc->files_lock);
-	return retval;
-}
-
 static bool binder_has_work_ilocked(struct binder_thread *thread,
 				    bool do_proc_work)
 {
@@ -1270,19 +1225,12 @@ static int binder_inc_node_nilocked(struct binder_node *node, int strong,
 		} else
 			node->local_strong_refs++;
 		if (!node->has_strong_ref && target_list) {
+			struct binder_thread *thread = container_of(target_list,
+						    struct binder_thread, todo);
 			binder_dequeue_work_ilocked(&node->work);
-			/*
-			 * Note: this function is the only place where we queue
-			 * directly to a thread->todo without using the
-			 * corresponding binder_enqueue_thread_work() helper
-			 * functions; in this case it's ok to not set the
-			 * process_todo flag, since we know this node work will
-			 * always be followed by other work that starts queue
-			 * processing: in case of synchronous transactions, a
-			 * BR_REPLY or BR_ERROR; in case of oneway
-			 * transactions, a BR_TRANSACTION_COMPLETE.
-			 */
-			binder_enqueue_work_ilocked(&node->work, target_list);
+			BUG_ON(&thread->todo != target_list);
+			binder_enqueue_deferred_thread_work_ilocked(thread,
+								   &node->work);
 		}
 	} else {
 		if (!internal)
@@ -1958,10 +1906,32 @@ static struct binder_thread *binder_get_txn_from_and_acq_inner(
 	return NULL;
 }
 
+/**
+ * binder_free_txn_fixups() - free unprocessed fd fixups
+ * @t:	binder transaction for t->from
+ *
+ * If the transaction is being torn down prior to being
+ * processed by the target process, free all of the
+ * fd fixups and fput the file structs. It is safe to
+ * call this function after the fixups have been
+ * processed -- in that case, the list will be empty.
+ */
+static void binder_free_txn_fixups(struct binder_transaction *t)
+{
+	struct binder_txn_fd_fixup *fixup, *tmp;
+
+	list_for_each_entry_safe(fixup, tmp, &t->fd_fixups, fixup_entry) {
+		fput(fixup->file);
+		list_del(&fixup->fixup_entry);
+		kfree(fixup);
+	}
+}
+
 static void binder_free_transaction(struct binder_transaction *t)
 {
 	if (t->buffer)
 		t->buffer->transaction = NULL;
+	binder_free_txn_fixups(t);
 	kfree(t);
 	binder_stats_deleted(BINDER_STAT_TRANSACTION);
 }
@@ -2262,12 +2232,17 @@ static void binder_transaction_buffer_release(struct binder_proc *proc,
 		} break;
 
 		case BINDER_TYPE_FD: {
-			struct binder_fd_object *fp = to_binder_fd_object(hdr);
-
-			binder_debug(BINDER_DEBUG_TRANSACTION,
-				     "        fd %d\n", fp->fd);
-			if (failed_at)
-				task_close_fd(proc, fp->fd);
+			/*
+			 * No need to close the file here since user-space
+			 * closes it for for successfully delivered
+			 * transactions. For transactions that weren't
+			 * delivered, the new fd was never allocated so
+			 * there is no need to close and the fput on the
+			 * file is done when the transaction is torn
+			 * down.
+			 */
+			WARN_ON(failed_at &&
+				proc->tsk == current->group_leader);
 		} break;
 		case BINDER_TYPE_PTR:
 			/*
@@ -2283,6 +2258,15 @@ static void binder_transaction_buffer_release(struct binder_proc *proc,
 			size_t fd_index;
 			binder_size_t fd_buf_size;
 
+			if (proc->tsk != current->group_leader) {
+				/*
+				 * Nothing to do if running in sender context
+				 * The fd fixups have not been applied so no
+				 * fds need to be closed.
+				 */
+				continue;
+			}
+
 			fda = to_binder_fd_array_object(hdr);
 			parent = binder_validate_ptr(buffer, fda->parent,
 						     off_start,
@@ -2315,7 +2299,7 @@ static void binder_transaction_buffer_release(struct binder_proc *proc,
 			}
 			fd_array = (u32 *)(parent_buffer + (uintptr_t)fda->parent_offset);
 			for (fd_index = 0; fd_index < fda->num_fds; fd_index++)
-				task_close_fd(proc, fd_array[fd_index]);
+				ksys_close(fd_array[fd_index]);
 		} break;
 		default:
 			pr_err("transaction release %d bad object type %x\n",
@@ -2447,17 +2431,18 @@ done:
 	return ret;
 }
 
-static int binder_translate_fd(int fd,
+static int binder_translate_fd(u32 *fdp,
 			       struct binder_transaction *t,
 			       struct binder_thread *thread,
 			       struct binder_transaction *in_reply_to)
 {
 	struct binder_proc *proc = thread->proc;
 	struct binder_proc *target_proc = t->to_proc;
-	int target_fd;
+	struct binder_txn_fd_fixup *fixup;
 	struct file *file;
-	int ret;
+	int ret = 0;
 	bool target_allows_fd;
+	int fd = *fdp;
 
 	if (in_reply_to)
 		target_allows_fd = !!(in_reply_to->flags & TF_ACCEPT_FDS);
@@ -2485,19 +2470,24 @@ static int binder_translate_fd(int fd,
 		goto err_security;
 	}
 
-	target_fd = task_get_unused_fd_flags(target_proc, O_CLOEXEC);
-	if (target_fd < 0) {
+	/*
+	 * Add fixup record for this transaction. The allocation
+	 * of the fd in the target needs to be done from a
+	 * target thread.
+	 */
+	fixup = kzalloc(sizeof(*fixup), GFP_KERNEL);
+	if (!fixup) {
 		ret = -ENOMEM;
-		goto err_get_unused_fd;
+		goto err_alloc;
 	}
-	task_fd_install(target_proc, target_fd, file);
-	trace_binder_transaction_fd(t, fd, target_fd);
-	binder_debug(BINDER_DEBUG_TRANSACTION, "        fd %d -> %d\n",
-		     fd, target_fd);
+	fixup->file = file;
+	fixup->offset = (uintptr_t)fdp - (uintptr_t)t->buffer->data;
+	trace_binder_transaction_fd_send(t, fd, fixup->offset);
+	list_add_tail(&fixup->fixup_entry, &t->fd_fixups);
 
-	return target_fd;
+	return ret;
 
-err_get_unused_fd:
+err_alloc:
 err_security:
 	fput(file);
 err_fget:
@@ -2511,8 +2501,7 @@ static int binder_translate_fd_array(struct binder_fd_array_object *fda,
 				     struct binder_thread *thread,
 				     struct binder_transaction *in_reply_to)
 {
-	binder_size_t fdi, fd_buf_size, num_installed_fds;
-	int target_fd;
+	binder_size_t fdi, fd_buf_size;
 	uintptr_t parent_buffer;
 	u32 *fd_array;
 	struct binder_proc *proc = thread->proc;
@@ -2544,23 +2533,12 @@ static int binder_translate_fd_array(struct binder_fd_array_object *fda,
 		return -EINVAL;
 	}
 	for (fdi = 0; fdi < fda->num_fds; fdi++) {
-		target_fd = binder_translate_fd(fd_array[fdi], t, thread,
+		int ret = binder_translate_fd(&fd_array[fdi], t, thread,
 						in_reply_to);
-		if (target_fd < 0)
-			goto err_translate_fd_failed;
-		fd_array[fdi] = target_fd;
+		if (ret < 0)
+			return ret;
 	}
 	return 0;
-
-err_translate_fd_failed:
-	/*
-	 * Failed to allocate fd or security error, free fds
-	 * installed so far.
-	 */
-	num_installed_fds = fdi;
-	for (fdi = 0; fdi < num_installed_fds; fdi++)
-		task_close_fd(target_proc, fd_array[fdi]);
-	return target_fd;
 }
 
 static int binder_fixup_parent(struct binder_transaction *t,
@@ -2723,6 +2701,7 @@ static void binder_transaction(struct binder_proc *proc,
 {
 	int ret;
 	struct binder_transaction *t;
+	struct binder_work *w;
 	struct binder_work *tcomplete;
 	binder_size_t *offp, *off_end, *off_start;
 	binder_size_t off_min;
@@ -2864,6 +2843,29 @@ static void binder_transaction(struct binder_proc *proc,
 			goto err_invalid_target_handle;
 		}
 		binder_inner_proc_lock(proc);
+
+		w = list_first_entry_or_null(&thread->todo,
+					     struct binder_work, entry);
+		if (!(tr->flags & TF_ONE_WAY) && w &&
+		    w->type == BINDER_WORK_TRANSACTION) {
+			/*
+			 * Do not allow new outgoing transaction from a
+			 * thread that has a transaction at the head of
+			 * its todo list. Only need to check the head
+			 * because binder_select_thread_ilocked picks a
+			 * thread from proc->waiting_threads to enqueue
+			 * the transaction, and nothing is queued to the
+			 * todo list while the thread is on waiting_threads.
+			 */
+			binder_user_error("%d:%d new transaction not allowed when there is a transaction on thread todo\n",
+					  proc->pid, thread->pid);
+			binder_inner_proc_unlock(proc);
+			return_error = BR_FAILED_REPLY;
+			return_error_param = -EPROTO;
+			return_error_line = __LINE__;
+			goto err_bad_todo_list;
+		}
+
 		if (!(tr->flags & TF_ONE_WAY) && thread->transaction_stack) {
 			struct binder_transaction *tmp;
 
@@ -2911,6 +2913,7 @@ static void binder_transaction(struct binder_proc *proc,
 		return_error_line = __LINE__;
 		goto err_alloc_t_failed;
 	}
+	INIT_LIST_HEAD(&t->fd_fixups);
 	binder_stats_created(BINDER_STAT_TRANSACTION);
 	spin_lock_init(&t->lock);
 
@@ -3066,17 +3069,16 @@ static void binder_transaction(struct binder_proc *proc,
 
 		case BINDER_TYPE_FD: {
 			struct binder_fd_object *fp = to_binder_fd_object(hdr);
-			int target_fd = binder_translate_fd(fp->fd, t, thread,
-							    in_reply_to);
+			int ret = binder_translate_fd(&fp->fd, t, thread,
+						      in_reply_to);
 
-			if (target_fd < 0) {
+			if (ret < 0) {
 				return_error = BR_FAILED_REPLY;
-				return_error_param = target_fd;
+				return_error_param = ret;
 				return_error_line = __LINE__;
 				goto err_translate_failed;
 			}
 			fp->pad_binder = 0;
-			fp->fd = target_fd;
 		} break;
 		case BINDER_TYPE_FDA: {
 			struct binder_fd_array_object *fda =
@@ -3233,6 +3235,7 @@ err_bad_object_type:
 err_bad_offset:
 err_bad_parent:
 err_copy_data_failed:
+	binder_free_txn_fixups(t);
 	trace_binder_transaction_failed_buffer_release(t->buffer);
 	binder_transaction_buffer_release(target_proc, t->buffer, offp);
 	if (target_node)
@@ -3247,6 +3250,7 @@ err_alloc_tcomplete_failed:
 	kfree(t);
 	binder_stats_deleted(BINDER_STAT_TRANSACTION);
 err_alloc_t_failed:
+err_bad_todo_list:
 err_bad_call_stack:
 err_empty_call_stack:
 err_dead_binder:
@@ -3294,6 +3298,47 @@ err_invalid_target_handle:
 	}
 }
 
+/**
+ * binder_free_buf() - free the specified buffer
+ * @proc:	binder proc that owns buffer
+ * @buffer:	buffer to be freed
+ *
+ * If buffer for an async transaction, enqueue the next async
+ * transaction from the node.
+ *
+ * Cleanup buffer and free it.
+ */
+static void
+binder_free_buf(struct binder_proc *proc, struct binder_buffer *buffer)
+{
+	if (buffer->transaction) {
+		buffer->transaction->buffer = NULL;
+		buffer->transaction = NULL;
+	}
+	if (buffer->async_transaction && buffer->target_node) {
+		struct binder_node *buf_node;
+		struct binder_work *w;
+
+		buf_node = buffer->target_node;
+		binder_node_inner_lock(buf_node);
+		BUG_ON(!buf_node->has_async_transaction);
+		BUG_ON(buf_node->proc != proc);
+		w = binder_dequeue_work_head_ilocked(
+				&buf_node->async_todo);
+		if (!w) {
+			buf_node->has_async_transaction = false;
+		} else {
+			binder_enqueue_work_ilocked(
+					w, &proc->todo);
+			binder_wakeup_proc_ilocked(proc);
+		}
+		binder_node_inner_unlock(buf_node);
+	}
+	trace_binder_transaction_buffer_release(buffer);
+	binder_transaction_buffer_release(proc, buffer, NULL);
+	binder_alloc_free_buf(&proc->alloc, buffer);
+}
+
 static int binder_thread_write(struct binder_proc *proc,
 			struct binder_thread *thread,
 			binder_uintptr_t binder_buffer, size_t size,
@@ -3480,33 +3525,7 @@ static int binder_thread_write(struct binder_proc *proc,
 				     proc->pid, thread->pid, (u64)data_ptr,
 				     buffer->debug_id,
 				     buffer->transaction ? "active" : "finished");
-
-			if (buffer->transaction) {
-				buffer->transaction->buffer = NULL;
-				buffer->transaction = NULL;
-			}
-			if (buffer->async_transaction && buffer->target_node) {
-				struct binder_node *buf_node;
-				struct binder_work *w;
-
-				buf_node = buffer->target_node;
-				binder_node_inner_lock(buf_node);
-				BUG_ON(!buf_node->has_async_transaction);
-				BUG_ON(buf_node->proc != proc);
-				w = binder_dequeue_work_head_ilocked(
-						&buf_node->async_todo);
-				if (!w) {
-					buf_node->has_async_transaction = false;
-				} else {
-					binder_enqueue_work_ilocked(
-							w, &proc->todo);
-					binder_wakeup_proc_ilocked(proc);
-				}
-				binder_node_inner_unlock(buf_node);
-			}
-			trace_binder_transaction_buffer_release(buffer);
-			binder_transaction_buffer_release(proc, buffer, NULL);
-			binder_alloc_free_buf(&proc->alloc, buffer);
+			binder_free_buf(proc, buffer);
 			break;
 		}
 
@@ -3829,6 +3848,76 @@ static int binder_wait_for_work(struct binder_thread *thread,
 	return ret;
 }
 
+/**
+ * binder_apply_fd_fixups() - finish fd translation
+ * @t:	binder transaction with list of fd fixups
+ *
+ * Now that we are in the context of the transaction target
+ * process, we can allocate and install fds. Process the
+ * list of fds to translate and fixup the buffer with the
+ * new fds.
+ *
+ * If we fail to allocate an fd, then free the resources by
+ * fput'ing files that have not been processed and ksys_close'ing
+ * any fds that have already been allocated.
+ */
+static int binder_apply_fd_fixups(struct binder_transaction *t)
+{
+	struct binder_txn_fd_fixup *fixup, *tmp;
+	int ret = 0;
+
+	list_for_each_entry(fixup, &t->fd_fixups, fixup_entry) {
+		int fd = get_unused_fd_flags(O_CLOEXEC);
+		u32 *fdp;
+
+		if (fd < 0) {
+			binder_debug(BINDER_DEBUG_TRANSACTION,
+				     "failed fd fixup txn %d fd %d\n",
+				     t->debug_id, fd);
+			ret = -ENOMEM;
+			break;
+		}
+		binder_debug(BINDER_DEBUG_TRANSACTION,
+			     "fd fixup txn %d fd %d\n",
+			     t->debug_id, fd);
+		trace_binder_transaction_fd_recv(t, fd, fixup->offset);
+		fd_install(fd, fixup->file);
+		fixup->file = NULL;
+		fdp = (u32 *)(t->buffer->data + fixup->offset);
+		/*
+		 * This store can cause problems for CPUs with a
+		 * VIVT cache (eg ARMv5) since the cache cannot
+		 * detect virtual aliases to the same physical cacheline.
+		 * To support VIVT, this address and the user-space VA
+		 * would both need to be flushed. Since this kernel
+		 * VA is not constructed via page_to_virt(), we can't
+		 * use flush_dcache_page() on it, so we'd have to use
+		 * an internal function. If devices with VIVT ever
+		 * need to run Android, we'll either need to go back
+		 * to patching the translated fd from the sender side
+		 * (using the non-standard kernel functions), or rework
+		 * how the kernel uses the buffer to use page_to_virt()
+		 * addresses instead of allocating in our own vm area.
+		 *
+		 * For now, we disable compilation if CONFIG_CPU_CACHE_VIVT.
+		 */
+		*fdp = fd;
+	}
+	list_for_each_entry_safe(fixup, tmp, &t->fd_fixups, fixup_entry) {
+		if (fixup->file) {
+			fput(fixup->file);
+		} else if (ret) {
+			u32 *fdp = (u32 *)(t->buffer->data + fixup->offset);
+
+			ksys_close(*fdp);
+		}
+		list_del(&fixup->fixup_entry);
+		kfree(fixup);
+	}
+
+	return ret;
+}
+
 static int binder_thread_read(struct binder_proc *proc,
 			      struct binder_thread *thread,
 			      binder_uintptr_t binder_buffer, size_t size,
@@ -4110,6 +4199,34 @@ retry:
 			tr.sender_pid = 0;
 		}
 
+		ret = binder_apply_fd_fixups(t);
+		if (ret) {
+			struct binder_buffer *buffer = t->buffer;
+			bool oneway = !!(t->flags & TF_ONE_WAY);
+			int tid = t->debug_id;
+
+			if (t_from)
+				binder_thread_dec_tmpref(t_from);
+			buffer->transaction = NULL;
+			binder_cleanup_transaction(t, "fd fixups failed",
+						   BR_FAILED_REPLY);
+			binder_free_buf(proc, buffer);
+			binder_debug(BINDER_DEBUG_FAILED_TRANSACTION,
+				     "%d:%d %stransaction %d fd fixups failed %d/%d, line %d\n",
+				     proc->pid, thread->pid,
+				     oneway ? "async " :
+					(cmd == BR_REPLY ? "reply " : ""),
+				     tid, BR_FAILED_REPLY, ret, __LINE__);
+			if (cmd == BR_REPLY) {
+				cmd = BR_FAILED_REPLY;
+				if (put_user(cmd, (uint32_t __user *)ptr))
+					return -EFAULT;
+				ptr += sizeof(uint32_t);
+				binder_stat_br(proc, thread, cmd);
+				break;
+			}
+			continue;
+		}
 		tr.data_size = t->buffer->data_size;
 		tr.offsets_size = t->buffer->offsets_size;
 		tr.data.ptr.buffer = (binder_uintptr_t)
@@ -4544,6 +4661,42 @@ out:
 	return ret;
 }
 
+static int binder_ioctl_get_node_info_for_ref(struct binder_proc *proc,
+		struct binder_node_info_for_ref *info)
+{
+	struct binder_node *node;
+	struct binder_context *context = proc->context;
+	__u32 handle = info->handle;
+
+	if (info->strong_count || info->weak_count || info->reserved1 ||
+	    info->reserved2 || info->reserved3) {
+		binder_user_error("%d BINDER_GET_NODE_INFO_FOR_REF: only handle may be non-zero.",
+				  proc->pid);
+		return -EINVAL;
+	}
+
+	/* This ioctl may only be used by the context manager */
+	mutex_lock(&context->context_mgr_node_lock);
+	if (!context->binder_context_mgr_node ||
+		context->binder_context_mgr_node->proc != proc) {
+		mutex_unlock(&context->context_mgr_node_lock);
+		return -EPERM;
+	}
+	mutex_unlock(&context->context_mgr_node_lock);
+
+	node = binder_get_node_from_ref(proc, handle, true, NULL);
+	if (!node)
+		return -EINVAL;
+
+	info->strong_count = node->local_strong_refs +
+		node->internal_strong_refs;
+	info->weak_count = node->local_weak_refs;
+
+	binder_put_node(node);
+
+	return 0;
+}
+
 static int binder_ioctl_get_node_debug_info(struct binder_proc *proc,
 				struct binder_node_debug_info *info)
 {
@@ -4638,6 +4791,25 @@ static long binder_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
 		}
 		break;
 	}
+	case BINDER_GET_NODE_INFO_FOR_REF: {
+		struct binder_node_info_for_ref info;
+
+		if (copy_from_user(&info, ubuf, sizeof(info))) {
+			ret = -EFAULT;
+			goto err;
+		}
+
+		ret = binder_ioctl_get_node_info_for_ref(proc, &info);
+		if (ret < 0)
+			goto err;
+
+		if (copy_to_user(ubuf, &info, sizeof(info))) {
+			ret = -EFAULT;
+			goto err;
+		}
+
+		break;
+	}
 	case BINDER_GET_NODE_DEBUG_INFO: {
 		struct binder_node_debug_info info;
 
@@ -4693,7 +4865,6 @@ static void binder_vma_close(struct vm_area_struct *vma)
 		     (vma->vm_end - vma->vm_start) / SZ_1K, vma->vm_flags,
 		     (unsigned long)pgprot_val(vma->vm_page_prot));
 	binder_alloc_vma_close(&proc->alloc);
-	binder_defer_work(proc, BINDER_DEFERRED_PUT_FILES);
 }
 
 static vm_fault_t binder_vm_fault(struct vm_fault *vmf)
@@ -4739,9 +4910,6 @@ static int binder_mmap(struct file *filp, struct vm_area_struct *vma)
 	ret = binder_alloc_mmap_handler(&proc->alloc, vma);
 	if (ret)
 		return ret;
-	mutex_lock(&proc->files_lock);
-	proc->files = get_files_struct(current);
-	mutex_unlock(&proc->files_lock);
 	return 0;
 
 err_bad_arg:
@@ -4765,7 +4933,6 @@ static int binder_open(struct inode *nodp, struct file *filp)
 	spin_lock_init(&proc->outer_lock);
 	get_task_struct(current->group_leader);
 	proc->tsk = current->group_leader;
-	mutex_init(&proc->files_lock);
 	INIT_LIST_HEAD(&proc->todo);
 	proc->default_priority = task_nice(current);
 	binder_dev = container_of(filp->private_data, struct binder_device,
@@ -4915,8 +5082,6 @@ static void binder_deferred_release(struct binder_proc *proc)
 	struct rb_node *n;
 	int threads, nodes, incoming_refs, outgoing_refs, active_transactions;
 
-	BUG_ON(proc->files);
-
 	mutex_lock(&binder_procs_lock);
 	hlist_del(&proc->proc_node);
 	mutex_unlock(&binder_procs_lock);
@@ -4998,7 +5163,6 @@ static void binder_deferred_release(struct binder_proc *proc)
 static void binder_deferred_func(struct work_struct *work)
 {
 	struct binder_proc *proc;
-	struct files_struct *files;
 
 	int defer;
 
@@ -5016,23 +5180,11 @@ static void binder_deferred_func(struct work_struct *work)
 		}
 		mutex_unlock(&binder_deferred_lock);
 
-		files = NULL;
-		if (defer & BINDER_DEFERRED_PUT_FILES) {
-			mutex_lock(&proc->files_lock);
-			files = proc->files;
-			if (files)
-				proc->files = NULL;
-			mutex_unlock(&proc->files_lock);
-		}
-
 		if (defer & BINDER_DEFERRED_FLUSH)
 			binder_deferred_flush(proc);
 
 		if (defer & BINDER_DEFERRED_RELEASE)
 			binder_deferred_release(proc); /* frees proc */
-
-		if (files)
-			put_files_struct(files);
 	} while (proc);
 }
 static DECLARE_WORK(binder_deferred_work, binder_deferred_func);
@@ -5667,12 +5819,11 @@ static int __init binder_init(void)
 	 * Copy the module_parameter string, because we don't want to
 	 * tokenize it in-place.
 	 */
-	device_names = kzalloc(strlen(binder_devices_param) + 1, GFP_KERNEL);
+	device_names = kstrdup(binder_devices_param, GFP_KERNEL);
 	if (!device_names) {
 		ret = -ENOMEM;
 		goto err_alloc_device_names_failed;
 	}
-	strcpy(device_names, binder_devices_param);
 
 	device_tmp = device_names;
 	while ((device_name = strsep(&device_tmp, ","))) {
diff --git a/drivers/android/binder_trace.h b/drivers/android/binder_trace.h
index 588eb3ec3507..14de7ac57a34 100644
--- a/drivers/android/binder_trace.h
+++ b/drivers/android/binder_trace.h
@@ -223,22 +223,40 @@ TRACE_EVENT(binder_transaction_ref_to_ref,
 		  __entry->dest_ref_debug_id, __entry->dest_ref_desc)
 );
 
-TRACE_EVENT(binder_transaction_fd,
-	TP_PROTO(struct binder_transaction *t, int src_fd, int dest_fd),
-	TP_ARGS(t, src_fd, dest_fd),
+TRACE_EVENT(binder_transaction_fd_send,
+	TP_PROTO(struct binder_transaction *t, int fd, size_t offset),
+	TP_ARGS(t, fd, offset),
 
 	TP_STRUCT__entry(
 		__field(int, debug_id)
-		__field(int, src_fd)
-		__field(int, dest_fd)
+		__field(int, fd)
+		__field(size_t, offset)
+	),
+	TP_fast_assign(
+		__entry->debug_id = t->debug_id;
+		__entry->fd = fd;
+		__entry->offset = offset;
+	),
+	TP_printk("transaction=%d src_fd=%d offset=%zu",
+		  __entry->debug_id, __entry->fd, __entry->offset)
+);
+
+TRACE_EVENT(binder_transaction_fd_recv,
+	TP_PROTO(struct binder_transaction *t, int fd, size_t offset),
+	TP_ARGS(t, fd, offset),
+
+	TP_STRUCT__entry(
+		__field(int, debug_id)
+		__field(int, fd)
+		__field(size_t, offset)
 	),
 	TP_fast_assign(
 		__entry->debug_id = t->debug_id;
-		__entry->src_fd = src_fd;
-		__entry->dest_fd = dest_fd;
+		__entry->fd = fd;
+		__entry->offset = offset;
 	),
-	TP_printk("transaction=%d src_fd=%d ==> dest_fd=%d",
-		  __entry->debug_id, __entry->src_fd, __entry->dest_fd)
+	TP_printk("transaction=%d dest_fd=%d offset=%zu",
+		  __entry->debug_id, __entry->fd, __entry->offset)
 );
 
 DECLARE_EVENT_CLASS(binder_buffer_class,
diff --git a/drivers/extcon/extcon-intel-cht-wc.c b/drivers/extcon/extcon-intel-cht-wc.c
index 5e1dd2772278..5ef215297101 100644
--- a/drivers/extcon/extcon-intel-cht-wc.c
+++ b/drivers/extcon/extcon-intel-cht-wc.c
@@ -1,18 +1,10 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Extcon charger detection driver for Intel Cherrytrail Whiskey Cove PMIC
  * Copyright (C) 2017 Hans de Goede <hdegoede@redhat.com>
  *
  * Based on various non upstream patches to support the CHT Whiskey Cove PMIC:
  * Copyright (C) 2013-2015 Intel Corporation. All rights reserved.
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
  */
 
 #include <linux/extcon-provider.h>
@@ -32,10 +24,10 @@
 #define CHT_WC_CHGRCTRL0_EMRGCHREN	BIT(1)
 #define CHT_WC_CHGRCTRL0_EXTCHRDIS	BIT(2)
 #define CHT_WC_CHGRCTRL0_SWCONTROL	BIT(3)
-#define CHT_WC_CHGRCTRL0_TTLCK_MASK	BIT(4)
-#define CHT_WC_CHGRCTRL0_CCSM_OFF_MASK	BIT(5)
-#define CHT_WC_CHGRCTRL0_DBPOFF_MASK	BIT(6)
-#define CHT_WC_CHGRCTRL0_WDT_NOKICK	BIT(7)
+#define CHT_WC_CHGRCTRL0_TTLCK		BIT(4)
+#define CHT_WC_CHGRCTRL0_CCSM_OFF	BIT(5)
+#define CHT_WC_CHGRCTRL0_DBPOFF		BIT(6)
+#define CHT_WC_CHGRCTRL0_CHR_WDT_NOKICK	BIT(7)
 
 #define CHT_WC_CHGRCTRL1		0x5e17
 
@@ -52,7 +44,7 @@
 #define CHT_WC_USBSRC_TYPE_ACA		4
 #define CHT_WC_USBSRC_TYPE_SE1		5
 #define CHT_WC_USBSRC_TYPE_MHL		6
-#define CHT_WC_USBSRC_TYPE_FLOAT_DP_DN	7
+#define CHT_WC_USBSRC_TYPE_FLOATING	7
 #define CHT_WC_USBSRC_TYPE_OTHER	8
 #define CHT_WC_USBSRC_TYPE_DCP_EXTPHY	9
 
@@ -61,9 +53,12 @@
 #define CHT_WC_PWRSRC_STS		0x6e1e
 #define CHT_WC_PWRSRC_VBUS		BIT(0)
 #define CHT_WC_PWRSRC_DC		BIT(1)
-#define CHT_WC_PWRSRC_BAT		BIT(2)
-#define CHT_WC_PWRSRC_ID_GND		BIT(3)
-#define CHT_WC_PWRSRC_ID_FLOAT		BIT(4)
+#define CHT_WC_PWRSRC_BATT		BIT(2)
+#define CHT_WC_PWRSRC_USBID_MASK	GENMASK(4, 3)
+#define CHT_WC_PWRSRC_USBID_SHIFT	3
+#define CHT_WC_PWRSRC_RID_ACA		0
+#define CHT_WC_PWRSRC_RID_GND		1
+#define CHT_WC_PWRSRC_RID_FLOAT		2
 
 #define CHT_WC_VBUS_GPIO_CTLO		0x6e2d
 #define CHT_WC_VBUS_GPIO_CTLO_OUTPUT	BIT(0)
@@ -104,16 +99,20 @@ struct cht_wc_extcon_data {
 
 static int cht_wc_extcon_get_id(struct cht_wc_extcon_data *ext, int pwrsrc_sts)
 {
-	if (pwrsrc_sts & CHT_WC_PWRSRC_ID_GND)
+	switch ((pwrsrc_sts & CHT_WC_PWRSRC_USBID_MASK) >> CHT_WC_PWRSRC_USBID_SHIFT) {
+	case CHT_WC_PWRSRC_RID_GND:
 		return USB_ID_GND;
-	if (pwrsrc_sts & CHT_WC_PWRSRC_ID_FLOAT)
+	case CHT_WC_PWRSRC_RID_FLOAT:
 		return USB_ID_FLOAT;
-
-	/*
-	 * Once we have iio support for the gpadc we should read the USBID
-	 * gpadc channel here and determine ACA role based on that.
-	 */
-	return USB_ID_FLOAT;
+	case CHT_WC_PWRSRC_RID_ACA:
+	default:
+		/*
+		 * Once we have IIO support for the GPADC we should read
+		 * the USBID GPADC channel here and determine ACA role
+		 * based on that.
+		 */
+		return USB_ID_FLOAT;
+	}
 }
 
 static int cht_wc_extcon_get_charger(struct cht_wc_extcon_data *ext,
@@ -156,9 +155,9 @@ static int cht_wc_extcon_get_charger(struct cht_wc_extcon_data *ext,
 		dev_warn(ext->dev,
 			"Unhandled charger type %d, defaulting to SDP\n",
 			 ret);
-		/* Fall through, treat as SDP */
+		return EXTCON_CHG_USB_SDP;
 	case CHT_WC_USBSRC_TYPE_SDP:
-	case CHT_WC_USBSRC_TYPE_FLOAT_DP_DN:
+	case CHT_WC_USBSRC_TYPE_FLOATING:
 	case CHT_WC_USBSRC_TYPE_OTHER:
 		return EXTCON_CHG_USB_SDP;
 	case CHT_WC_USBSRC_TYPE_CDP:
@@ -279,7 +278,7 @@ static int cht_wc_extcon_sw_control(struct cht_wc_extcon_data *ext, bool enable)
 {
 	int ret, mask, val;
 
-	mask = CHT_WC_CHGRCTRL0_SWCONTROL | CHT_WC_CHGRCTRL0_CCSM_OFF_MASK;
+	mask = CHT_WC_CHGRCTRL0_SWCONTROL | CHT_WC_CHGRCTRL0_CCSM_OFF;
 	val = enable ? mask : 0;
 	ret = regmap_update_bits(ext->regmap, CHT_WC_CHGRCTRL0, mask, val);
 	if (ret)
@@ -292,6 +291,7 @@ static int cht_wc_extcon_probe(struct platform_device *pdev)
 {
 	struct intel_soc_pmic *pmic = dev_get_drvdata(pdev->dev.parent);
 	struct cht_wc_extcon_data *ext;
+	unsigned long mask = ~(CHT_WC_PWRSRC_VBUS | CHT_WC_PWRSRC_USBID_MASK);
 	int irq, ret;
 
 	irq = platform_get_irq(pdev, 0);
@@ -352,9 +352,7 @@ static int cht_wc_extcon_probe(struct platform_device *pdev)
 	}
 
 	/* Unmask irqs */
-	ret = regmap_write(ext->regmap, CHT_WC_PWRSRC_IRQ_MASK,
-			   (int)~(CHT_WC_PWRSRC_VBUS | CHT_WC_PWRSRC_ID_GND |
-				  CHT_WC_PWRSRC_ID_FLOAT));
+	ret = regmap_write(ext->regmap, CHT_WC_PWRSRC_IRQ_MASK, mask);
 	if (ret) {
 		dev_err(ext->dev, "Error writing irq-mask: %d\n", ret);
 		goto disable_sw_control;
diff --git a/drivers/extcon/extcon-intel-int3496.c b/drivers/extcon/extcon-intel-int3496.c
index fd24debe58a3..80c9abcc3f97 100644
--- a/drivers/extcon/extcon-intel-int3496.c
+++ b/drivers/extcon/extcon-intel-int3496.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Intel INT3496 ACPI device extcon driver
  *
@@ -7,15 +8,6 @@
  *
  * Copyright (c) 2014, Intel Corporation.
  * Author: David Cohen <david.a.cohen@linux.intel.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
  */
 
 #include <linux/acpi.h>
@@ -192,4 +184,4 @@ module_platform_driver(int3496_driver);
 
 MODULE_AUTHOR("Hans de Goede <hdegoede@redhat.com>");
 MODULE_DESCRIPTION("Intel INT3496 ACPI device extcon driver");
-MODULE_LICENSE("GPL");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/extcon/extcon-max14577.c b/drivers/extcon/extcon-max14577.c
index b871836da8a4..22d2feb1f8bc 100644
--- a/drivers/extcon/extcon-max14577.c
+++ b/drivers/extcon/extcon-max14577.c
@@ -1,20 +1,10 @@
-/*
- * extcon-max14577.c - MAX14577/77836 extcon driver to support MUIC
- *
- * Copyright (C) 2013,2014 Samsung Electronics
- * Chanwoo Choi <cw00.choi@samsung.com>
- * Krzysztof Kozlowski <krzk@kernel.org>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- */
+// SPDX-License-Identifier: GPL-2.0+
+//
+// extcon-max14577.c - MAX14577/77836 extcon driver to support MUIC
+//
+// Copyright (C) 2013,2014 Samsung Electronics
+// Chanwoo Choi <cw00.choi@samsung.com>
+// Krzysztof Kozlowski <krzk@kernel.org>
 
 #include <linux/kernel.h>
 #include <linux/module.h>
diff --git a/drivers/extcon/extcon-max77693.c b/drivers/extcon/extcon-max77693.c
index 227651ff9666..a79537ebb671 100644
--- a/drivers/extcon/extcon-max77693.c
+++ b/drivers/extcon/extcon-max77693.c
@@ -1,19 +1,9 @@
-/*
- * extcon-max77693.c - MAX77693 extcon driver to support MAX77693 MUIC
- *
- * Copyright (C) 2012 Samsung Electrnoics
- * Chanwoo Choi <cw00.choi@samsung.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- */
+// SPDX-License-Identifier: GPL-2.0+
+//
+// extcon-max77693.c - MAX77693 extcon driver to support MAX77693 MUIC
+//
+// Copyright (C) 2012 Samsung Electrnoics
+// Chanwoo Choi <cw00.choi@samsung.com>
 
 #include <linux/kernel.h>
 #include <linux/module.h>
diff --git a/drivers/extcon/extcon-max77843.c b/drivers/extcon/extcon-max77843.c
index c9fcd6cd41cb..b98cbd0362f5 100644
--- a/drivers/extcon/extcon-max77843.c
+++ b/drivers/extcon/extcon-max77843.c
@@ -1,15 +1,10 @@
-/*
- * extcon-max77843.c - Maxim MAX77843 extcon driver to support
- *			MUIC(Micro USB Interface Controller)
- *
- * Copyright (C) 2015 Samsung Electronics
- * Author: Jaewon Kim <jaewon02.kim@samsung.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- */
+// SPDX-License-Identifier: GPL-2.0+
+//
+// extcon-max77843.c - Maxim MAX77843 extcon driver to support
+//			MUIC(Micro USB Interface Controller)
+//
+// Copyright (C) 2015 Samsung Electronics
+// Author: Jaewon Kim <jaewon02.kim@samsung.com>
 
 #include <linux/extcon-provider.h>
 #include <linux/i2c.h>
diff --git a/drivers/extcon/extcon-max8997.c b/drivers/extcon/extcon-max8997.c
index 9f30f4929b72..bdabb2479e0d 100644
--- a/drivers/extcon/extcon-max8997.c
+++ b/drivers/extcon/extcon-max8997.c
@@ -1,19 +1,9 @@
-/*
- * extcon-max8997.c - MAX8997 extcon driver to support MAX8997 MUIC
- *
- *  Copyright (C) 2012 Samsung Electronics
- *  Donggeun Kim <dg77.kim@samsung.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- */
+// SPDX-License-Identifier: GPL-2.0+
+//
+// extcon-max8997.c - MAX8997 extcon driver to support MAX8997 MUIC
+//
+//  Copyright (C) 2012 Samsung Electronics
+//  Donggeun Kim <dg77.kim@samsung.com>
 
 #include <linux/kernel.h>
 #include <linux/module.h>
diff --git a/drivers/extcon/extcon.c b/drivers/extcon/extcon.c
index b9d27c8fe57e..5ab0498be652 100644
--- a/drivers/extcon/extcon.c
+++ b/drivers/extcon/extcon.c
@@ -628,7 +628,7 @@ int extcon_get_property(struct extcon_dev *edev, unsigned int id,
 	unsigned long flags;
 	int index, ret = 0;
 
-	*prop_val = (union extcon_property_value)(0);
+	*prop_val = (union extcon_property_value){0};
 
 	if (!edev)
 		return -EINVAL;
@@ -1123,7 +1123,6 @@ int extcon_dev_register(struct extcon_dev *edev)
 			(unsigned long)atomic_inc_return(&edev_no));
 
 	if (edev->max_supported) {
-		char buf[10];
 		char *str;
 		struct extcon_cable *cable;
 
@@ -1137,9 +1136,7 @@ int extcon_dev_register(struct extcon_dev *edev)
 		for (index = 0; index < edev->max_supported; index++) {
 			cable = &edev->cables[index];
 
-			snprintf(buf, 10, "cable.%d", index);
-			str = kzalloc(strlen(buf) + 1,
-				      GFP_KERNEL);
+			str = kasprintf(GFP_KERNEL, "cable.%d", index);
 			if (!str) {
 				for (index--; index >= 0; index--) {
 					cable = &edev->cables[index];
@@ -1149,7 +1146,6 @@ int extcon_dev_register(struct extcon_dev *edev)
 
 				goto err_alloc_cables;
 			}
-			strcpy(str, buf);
 
 			cable->edev = edev;
 			cable->cable_index = index;
@@ -1172,7 +1168,6 @@ int extcon_dev_register(struct extcon_dev *edev)
 	}
 
 	if (edev->max_supported && edev->mutually_exclusive) {
-		char buf[80];
 		char *name;
 
 		/* Count the size of mutually_exclusive array */
@@ -1197,9 +1192,8 @@ int extcon_dev_register(struct extcon_dev *edev)
 		}
 
 		for (index = 0; edev->mutually_exclusive[index]; index++) {
-			sprintf(buf, "0x%x", edev->mutually_exclusive[index]);
-			name = kzalloc(strlen(buf) + 1,
-				       GFP_KERNEL);
+			name = kasprintf(GFP_KERNEL, "0x%x",
+					 edev->mutually_exclusive[index]);
 			if (!name) {
 				for (index--; index >= 0; index--) {
 					kfree(edev->d_attrs_muex[index].attr.
@@ -1210,7 +1204,6 @@ int extcon_dev_register(struct extcon_dev *edev)
 				ret = -ENOMEM;
 				goto err_muex;
 			}
-			strcpy(name, buf);
 			sysfs_attr_init(&edev->d_attrs_muex[index].attr);
 			edev->d_attrs_muex[index].attr.name = name;
 			edev->d_attrs_muex[index].attr.mode = 0000;
diff --git a/drivers/firmware/google/Kconfig b/drivers/firmware/google/Kconfig
index a456a000048b..91a0404affe2 100644
--- a/drivers/firmware/google/Kconfig
+++ b/drivers/firmware/google/Kconfig
@@ -10,37 +10,31 @@ if GOOGLE_FIRMWARE
 
 config GOOGLE_SMI
 	tristate "SMI interface for Google platforms"
-	depends on X86 && ACPI && DMI && EFI
-	select EFI_VARS
+	depends on X86 && ACPI && DMI
 	help
 	  Say Y here if you want to enable SMI callbacks for Google
 	  platforms.  This provides an interface for writing to and
-	  clearing the EFI event log and reading and writing NVRAM
+	  clearing the event log.  If EFI_VARS is also enabled this
+	  driver provides an interface for reading and writing NVRAM
 	  variables.
 
 config GOOGLE_COREBOOT_TABLE
-	tristate
-	depends on GOOGLE_COREBOOT_TABLE_ACPI || GOOGLE_COREBOOT_TABLE_OF
-
-config GOOGLE_COREBOOT_TABLE_ACPI
-	tristate "Coreboot Table Access - ACPI"
-	depends on ACPI
-	select GOOGLE_COREBOOT_TABLE
+	tristate "Coreboot Table Access"
+	depends on ACPI || OF
 	help
 	  This option enables the coreboot_table module, which provides other
-	  firmware modules to access to the coreboot table. The coreboot table
-	  pointer is accessed through the ACPI "GOOGCB00" object.
+	  firmware modules access to the coreboot table. The coreboot table
+	  pointer is accessed through the ACPI "GOOGCB00" object or the
+	  device tree node /firmware/coreboot.
 	  If unsure say N.
 
+config GOOGLE_COREBOOT_TABLE_ACPI
+	tristate
+	select GOOGLE_COREBOOT_TABLE
+
 config GOOGLE_COREBOOT_TABLE_OF
-	tristate "Coreboot Table Access - Device Tree"
-	depends on OF
+	tristate
 	select GOOGLE_COREBOOT_TABLE
-	help
-	  This option enable the coreboot_table module, which provide other
-	  firmware modules to access coreboot table. The coreboot table pointer
-	  is accessed through the device tree node /firmware/coreboot.
-	  If unsure say N.
 
 config GOOGLE_MEMCONSOLE
 	tristate
diff --git a/drivers/firmware/google/Makefile b/drivers/firmware/google/Makefile
index d0b3fba96194..d17caded5d88 100644
--- a/drivers/firmware/google/Makefile
+++ b/drivers/firmware/google/Makefile
@@ -2,8 +2,6 @@
 
 obj-$(CONFIG_GOOGLE_SMI)		+= gsmi.o
 obj-$(CONFIG_GOOGLE_COREBOOT_TABLE)        += coreboot_table.o
-obj-$(CONFIG_GOOGLE_COREBOOT_TABLE_ACPI)   += coreboot_table-acpi.o
-obj-$(CONFIG_GOOGLE_COREBOOT_TABLE_OF)     += coreboot_table-of.o
 obj-$(CONFIG_GOOGLE_FRAMEBUFFER_COREBOOT)  += framebuffer-coreboot.o
 obj-$(CONFIG_GOOGLE_MEMCONSOLE)            += memconsole.o
 obj-$(CONFIG_GOOGLE_MEMCONSOLE_COREBOOT)   += memconsole-coreboot.o
diff --git a/drivers/firmware/google/coreboot_table-acpi.c b/drivers/firmware/google/coreboot_table-acpi.c
deleted file mode 100644
index 77197fe3d42f..000000000000
--- a/drivers/firmware/google/coreboot_table-acpi.c
+++ /dev/null
@@ -1,88 +0,0 @@
-/*
- * coreboot_table-acpi.c
- *
- * Using ACPI to locate Coreboot table and provide coreboot table access.
- *
- * Copyright 2017 Google Inc.
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License v2.0 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- */
-
-#include <linux/acpi.h>
-#include <linux/device.h>
-#include <linux/err.h>
-#include <linux/init.h>
-#include <linux/io.h>
-#include <linux/kernel.h>
-#include <linux/module.h>
-#include <linux/platform_device.h>
-
-#include "coreboot_table.h"
-
-static int coreboot_table_acpi_probe(struct platform_device *pdev)
-{
-	phys_addr_t phyaddr;
-	resource_size_t len;
-	struct coreboot_table_header __iomem *header = NULL;
-	struct resource *res;
-	void __iomem *ptr = NULL;
-
-	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (!res)
-		return -EINVAL;
-
-	len = resource_size(res);
-	if (!res->start || !len)
-		return -EINVAL;
-
-	phyaddr = res->start;
-	header = ioremap_cache(phyaddr, sizeof(*header));
-	if (header == NULL)
-		return -ENOMEM;
-
-	ptr = ioremap_cache(phyaddr,
-			    header->header_bytes + header->table_bytes);
-	iounmap(header);
-	if (!ptr)
-		return -ENOMEM;
-
-	return coreboot_table_init(&pdev->dev, ptr);
-}
-
-static int coreboot_table_acpi_remove(struct platform_device *pdev)
-{
-	return coreboot_table_exit();
-}
-
-static const struct acpi_device_id cros_coreboot_acpi_match[] = {
-	{ "GOOGCB00", 0 },
-	{ "BOOT0000", 0 },
-	{ }
-};
-MODULE_DEVICE_TABLE(acpi, cros_coreboot_acpi_match);
-
-static struct platform_driver coreboot_table_acpi_driver = {
-	.probe = coreboot_table_acpi_probe,
-	.remove = coreboot_table_acpi_remove,
-	.driver = {
-		.name = "coreboot_table_acpi",
-		.acpi_match_table = ACPI_PTR(cros_coreboot_acpi_match),
-	},
-};
-
-static int __init coreboot_table_acpi_init(void)
-{
-	return platform_driver_register(&coreboot_table_acpi_driver);
-}
-
-module_init(coreboot_table_acpi_init);
-
-MODULE_AUTHOR("Google, Inc.");
-MODULE_LICENSE("GPL");
diff --git a/drivers/firmware/google/coreboot_table-of.c b/drivers/firmware/google/coreboot_table-of.c
deleted file mode 100644
index f15bf404c579..000000000000
--- a/drivers/firmware/google/coreboot_table-of.c
+++ /dev/null
@@ -1,82 +0,0 @@
-/*
- * coreboot_table-of.c
- *
- * Coreboot table access through open firmware.
- *
- * Copyright 2017 Google Inc.
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License v2.0 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- */
-
-#include <linux/device.h>
-#include <linux/io.h>
-#include <linux/module.h>
-#include <linux/of_address.h>
-#include <linux/of_platform.h>
-#include <linux/platform_device.h>
-
-#include "coreboot_table.h"
-
-static int coreboot_table_of_probe(struct platform_device *pdev)
-{
-	struct device_node *fw_dn = pdev->dev.of_node;
-	void __iomem *ptr;
-
-	ptr = of_iomap(fw_dn, 0);
-	of_node_put(fw_dn);
-	if (!ptr)
-		return -ENOMEM;
-
-	return coreboot_table_init(&pdev->dev, ptr);
-}
-
-static int coreboot_table_of_remove(struct platform_device *pdev)
-{
-	return coreboot_table_exit();
-}
-
-static const struct of_device_id coreboot_of_match[] = {
-	{ .compatible = "coreboot" },
-	{},
-};
-
-static struct platform_driver coreboot_table_of_driver = {
-	.probe = coreboot_table_of_probe,
-	.remove = coreboot_table_of_remove,
-	.driver = {
-		.name = "coreboot_table_of",
-		.of_match_table = coreboot_of_match,
-	},
-};
-
-static int __init platform_coreboot_table_of_init(void)
-{
-	struct platform_device *pdev;
-	struct device_node *of_node;
-
-	/* Limit device creation to the presence of /firmware/coreboot node */
-	of_node = of_find_node_by_path("/firmware/coreboot");
-	if (!of_node)
-		return -ENODEV;
-
-	if (!of_match_node(coreboot_of_match, of_node))
-		return -ENODEV;
-
-	pdev = of_platform_device_create(of_node, "coreboot_table_of", NULL);
-	if (!pdev)
-		return -ENODEV;
-
-	return platform_driver_register(&coreboot_table_of_driver);
-}
-
-module_init(platform_coreboot_table_of_init);
-
-MODULE_AUTHOR("Google, Inc.");
-MODULE_LICENSE("GPL");
diff --git a/drivers/firmware/google/coreboot_table.c b/drivers/firmware/google/coreboot_table.c
index 19db5709ae28..078d3bbe632f 100644
--- a/drivers/firmware/google/coreboot_table.c
+++ b/drivers/firmware/google/coreboot_table.c
@@ -16,12 +16,15 @@
  * GNU General Public License for more details.
  */
 
+#include <linux/acpi.h>
 #include <linux/device.h>
 #include <linux/err.h>
 #include <linux/init.h>
 #include <linux/io.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
 #include <linux/slab.h>
 
 #include "coreboot_table.h"
@@ -29,8 +32,6 @@
 #define CB_DEV(d) container_of(d, struct coreboot_device, dev)
 #define CB_DRV(d) container_of(d, struct coreboot_driver, drv)
 
-static struct coreboot_table_header __iomem *ptr_header;
-
 static int coreboot_bus_match(struct device *dev, struct device_driver *drv)
 {
 	struct coreboot_device *device = CB_DEV(dev);
@@ -70,12 +71,6 @@ static struct bus_type coreboot_bus_type = {
 	.remove		= coreboot_bus_remove,
 };
 
-static int __init coreboot_bus_init(void)
-{
-	return bus_register(&coreboot_bus_type);
-}
-module_init(coreboot_bus_init);
-
 static void coreboot_device_release(struct device *dev)
 {
 	struct coreboot_device *device = CB_DEV(dev);
@@ -97,62 +92,117 @@ void coreboot_driver_unregister(struct coreboot_driver *driver)
 }
 EXPORT_SYMBOL(coreboot_driver_unregister);
 
-int coreboot_table_init(struct device *dev, void __iomem *ptr)
+static int coreboot_table_populate(struct device *dev, void *ptr)
 {
 	int i, ret;
 	void *ptr_entry;
 	struct coreboot_device *device;
-	struct coreboot_table_entry entry;
-	struct coreboot_table_header header;
-
-	ptr_header = ptr;
-	memcpy_fromio(&header, ptr_header, sizeof(header));
-
-	if (strncmp(header.signature, "LBIO", sizeof(header.signature))) {
-		pr_warn("coreboot_table: coreboot table missing or corrupt!\n");
-		return -ENODEV;
-	}
+	struct coreboot_table_entry *entry;
+	struct coreboot_table_header *header = ptr;
 
-	ptr_entry = (void *)ptr_header + header.header_bytes;
-	for (i = 0; i < header.table_entries; i++) {
-		memcpy_fromio(&entry, ptr_entry, sizeof(entry));
+	ptr_entry = ptr + header->header_bytes;
+	for (i = 0; i < header->table_entries; i++) {
+		entry = ptr_entry;
 
-		device = kzalloc(sizeof(struct device) + entry.size, GFP_KERNEL);
-		if (!device) {
-			ret = -ENOMEM;
-			break;
-		}
+		device = kzalloc(sizeof(struct device) + entry->size, GFP_KERNEL);
+		if (!device)
+			return -ENOMEM;
 
 		dev_set_name(&device->dev, "coreboot%d", i);
 		device->dev.parent = dev;
 		device->dev.bus = &coreboot_bus_type;
 		device->dev.release = coreboot_device_release;
-		memcpy_fromio(&device->entry, ptr_entry, entry.size);
+		memcpy(&device->entry, ptr_entry, entry->size);
 
 		ret = device_register(&device->dev);
 		if (ret) {
 			put_device(&device->dev);
-			break;
+			return ret;
 		}
 
-		ptr_entry += entry.size;
+		ptr_entry += entry->size;
 	}
 
-	return ret;
+	return 0;
 }
-EXPORT_SYMBOL(coreboot_table_init);
 
-int coreboot_table_exit(void)
+static int coreboot_table_probe(struct platform_device *pdev)
 {
-	if (ptr_header) {
-		bus_unregister(&coreboot_bus_type);
-		iounmap(ptr_header);
-		ptr_header = NULL;
+	resource_size_t len;
+	struct coreboot_table_header *header;
+	struct resource *res;
+	struct device *dev = &pdev->dev;
+	void *ptr;
+	int ret;
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!res)
+		return -EINVAL;
+
+	len = resource_size(res);
+	if (!res->start || !len)
+		return -EINVAL;
+
+	/* Check just the header first to make sure things are sane */
+	header = memremap(res->start, sizeof(*header), MEMREMAP_WB);
+	if (!header)
+		return -ENOMEM;
+
+	len = header->header_bytes + header->table_bytes;
+	ret = strncmp(header->signature, "LBIO", sizeof(header->signature));
+	memunmap(header);
+	if (ret) {
+		dev_warn(dev, "coreboot table missing or corrupt!\n");
+		return -ENODEV;
 	}
 
+	ptr = memremap(res->start, len, MEMREMAP_WB);
+	if (!ptr)
+		return -ENOMEM;
+
+	ret = bus_register(&coreboot_bus_type);
+	if (!ret) {
+		ret = coreboot_table_populate(dev, ptr);
+		if (ret)
+			bus_unregister(&coreboot_bus_type);
+	}
+	memunmap(ptr);
+
+	return ret;
+}
+
+static int coreboot_table_remove(struct platform_device *pdev)
+{
+	bus_unregister(&coreboot_bus_type);
 	return 0;
 }
-EXPORT_SYMBOL(coreboot_table_exit);
 
+#ifdef CONFIG_ACPI
+static const struct acpi_device_id cros_coreboot_acpi_match[] = {
+	{ "GOOGCB00", 0 },
+	{ "BOOT0000", 0 },
+	{ }
+};
+MODULE_DEVICE_TABLE(acpi, cros_coreboot_acpi_match);
+#endif
+
+#ifdef CONFIG_OF
+static const struct of_device_id coreboot_of_match[] = {
+	{ .compatible = "coreboot" },
+	{}
+};
+MODULE_DEVICE_TABLE(of, coreboot_of_match);
+#endif
+
+static struct platform_driver coreboot_table_driver = {
+	.probe = coreboot_table_probe,
+	.remove = coreboot_table_remove,
+	.driver = {
+		.name = "coreboot_table",
+		.acpi_match_table = ACPI_PTR(cros_coreboot_acpi_match),
+		.of_match_table = of_match_ptr(coreboot_of_match),
+	},
+};
+module_platform_driver(coreboot_table_driver);
 MODULE_AUTHOR("Google, Inc.");
 MODULE_LICENSE("GPL");
diff --git a/drivers/firmware/google/coreboot_table.h b/drivers/firmware/google/coreboot_table.h
index 8ad95a94481b..71a9de6b15fa 100644
--- a/drivers/firmware/google/coreboot_table.h
+++ b/drivers/firmware/google/coreboot_table.h
@@ -91,10 +91,4 @@ int coreboot_driver_register(struct coreboot_driver *driver);
 /* Unregister a driver that uses the data from a coreboot table. */
 void coreboot_driver_unregister(struct coreboot_driver *driver);
 
-/* Initialize coreboot table module given a pointer to iomem */
-int coreboot_table_init(struct device *dev, void __iomem *ptr);
-
-/* Cleanup coreboot table module */
-int coreboot_table_exit(void);
-
 #endif /* __COREBOOT_TABLE_H */
diff --git a/drivers/firmware/google/gsmi.c b/drivers/firmware/google/gsmi.c
index c8f169bf2e27..82ce1e6d261e 100644
--- a/drivers/firmware/google/gsmi.c
+++ b/drivers/firmware/google/gsmi.c
@@ -29,6 +29,7 @@
 #include <linux/efi.h>
 #include <linux/module.h>
 #include <linux/ucs2_string.h>
+#include <linux/suspend.h>
 
 #define GSMI_SHUTDOWN_CLEAN	0	/* Clean Shutdown */
 /* TODO(mikew@google.com): Tie in HARDLOCKUP_DETECTOR with NMIWDT */
@@ -70,6 +71,8 @@
 #define GSMI_CMD_SET_NVRAM_VAR		0x03
 #define GSMI_CMD_SET_EVENT_LOG		0x08
 #define GSMI_CMD_CLEAR_EVENT_LOG	0x09
+#define GSMI_CMD_LOG_S0IX_SUSPEND	0x0a
+#define GSMI_CMD_LOG_S0IX_RESUME	0x0b
 #define GSMI_CMD_CLEAR_CONFIG		0x20
 #define GSMI_CMD_HANDSHAKE_TYPE		0xC1
 
@@ -84,7 +87,7 @@ struct gsmi_buf {
 	u32 address;			/* physical address of buffer */
 };
 
-struct gsmi_device {
+static struct gsmi_device {
 	struct platform_device *pdev;	/* platform device */
 	struct gsmi_buf *name_buf;	/* variable name buffer */
 	struct gsmi_buf *data_buf;	/* generic data buffer */
@@ -122,7 +125,6 @@ struct gsmi_log_entry_type_1 {
 	u32	instance;
 } __packed;
 
-
 /*
  * Some platforms don't have explicit SMI handshake
  * and need to wait for SMI to complete.
@@ -133,6 +135,15 @@ module_param(spincount, uint, 0600);
 MODULE_PARM_DESC(spincount,
 	"The number of loop iterations to use when using the spin handshake.");
 
+/*
+ * Platforms might not support S0ix logging in their GSMI handlers. In order to
+ * avoid any side-effects of generating an SMI for S0ix logging, use the S0ix
+ * related GSMI commands only for those platforms that explicitly enable this
+ * option.
+ */
+static bool s0ix_logging_enable;
+module_param(s0ix_logging_enable, bool, 0600);
+
 static struct gsmi_buf *gsmi_buf_alloc(void)
 {
 	struct gsmi_buf *smibuf;
@@ -289,6 +300,10 @@ static int gsmi_exec(u8 func, u8 sub)
 	return rc;
 }
 
+#ifdef CONFIG_EFI_VARS
+
+static struct efivars efivars;
+
 static efi_status_t gsmi_get_variable(efi_char16_t *name,
 				      efi_guid_t *vendor, u32 *attr,
 				      unsigned long *data_size,
@@ -466,6 +481,8 @@ static const struct efivar_operations efivar_ops = {
 	.get_next_variable = gsmi_get_next_variable,
 };
 
+#endif /* CONFIG_EFI_VARS */
+
 static ssize_t eventlog_write(struct file *filp, struct kobject *kobj,
 			       struct bin_attribute *bin_attr,
 			       char *buf, loff_t pos, size_t count)
@@ -480,11 +497,10 @@ static ssize_t eventlog_write(struct file *filp, struct kobject *kobj,
 	if (count < sizeof(u32))
 		return -EINVAL;
 	param.type = *(u32 *)buf;
-	count -= sizeof(u32);
 	buf += sizeof(u32);
 
 	/* The remaining buffer is the data payload */
-	if (count > gsmi_dev.data_buf->length)
+	if ((count - sizeof(u32)) > gsmi_dev.data_buf->length)
 		return -EINVAL;
 	param.data_len = count - sizeof(u32);
 
@@ -504,7 +520,7 @@ static ssize_t eventlog_write(struct file *filp, struct kobject *kobj,
 
 	spin_unlock_irqrestore(&gsmi_dev.lock, flags);
 
-	return rc;
+	return (rc == 0) ? count : rc;
 
 }
 
@@ -716,6 +732,12 @@ static const struct dmi_system_id gsmi_dmi_table[] __initconst = {
 			DMI_MATCH(DMI_BOARD_VENDOR, "Google, Inc."),
 		},
 	},
+	{
+		.ident = "Coreboot Firmware",
+		.matches = {
+			DMI_MATCH(DMI_BIOS_VENDOR, "coreboot"),
+		},
+	},
 	{}
 };
 MODULE_DEVICE_TABLE(dmi, gsmi_dmi_table);
@@ -762,7 +784,6 @@ static __init int gsmi_system_valid(void)
 }
 
 static struct kobject *gsmi_kobj;
-static struct efivars efivars;
 
 static const struct platform_device_info gsmi_dev_info = {
 	.name		= "gsmi",
@@ -771,6 +792,78 @@ static const struct platform_device_info gsmi_dev_info = {
 	.dma_mask	= DMA_BIT_MASK(32),
 };
 
+#ifdef CONFIG_PM
+static void gsmi_log_s0ix_info(u8 cmd)
+{
+	unsigned long flags;
+
+	/*
+	 * If platform has not enabled S0ix logging, then no action is
+	 * necessary.
+	 */
+	if (!s0ix_logging_enable)
+		return;
+
+	spin_lock_irqsave(&gsmi_dev.lock, flags);
+
+	memset(gsmi_dev.param_buf->start, 0, gsmi_dev.param_buf->length);
+
+	gsmi_exec(GSMI_CALLBACK, cmd);
+
+	spin_unlock_irqrestore(&gsmi_dev.lock, flags);
+}
+
+static int gsmi_log_s0ix_suspend(struct device *dev)
+{
+	/*
+	 * If system is not suspending via firmware using the standard ACPI Sx
+	 * types, then make a GSMI call to log the suspend info.
+	 */
+	if (!pm_suspend_via_firmware())
+		gsmi_log_s0ix_info(GSMI_CMD_LOG_S0IX_SUSPEND);
+
+	/*
+	 * Always return success, since we do not want suspend
+	 * to fail just because of logging failure.
+	 */
+	return 0;
+}
+
+static int gsmi_log_s0ix_resume(struct device *dev)
+{
+	/*
+	 * If system did not resume via firmware, then make a GSMI call to log
+	 * the resume info and wake source.
+	 */
+	if (!pm_resume_via_firmware())
+		gsmi_log_s0ix_info(GSMI_CMD_LOG_S0IX_RESUME);
+
+	/*
+	 * Always return success, since we do not want resume
+	 * to fail just because of logging failure.
+	 */
+	return 0;
+}
+
+static const struct dev_pm_ops gsmi_pm_ops = {
+	.suspend_noirq = gsmi_log_s0ix_suspend,
+	.resume_noirq = gsmi_log_s0ix_resume,
+};
+
+static int gsmi_platform_driver_probe(struct platform_device *dev)
+{
+	return 0;
+}
+
+static struct platform_driver gsmi_driver_info = {
+	.driver = {
+		.name = "gsmi",
+		.pm = &gsmi_pm_ops,
+	},
+	.probe = gsmi_platform_driver_probe,
+};
+#endif
+
 static __init int gsmi_init(void)
 {
 	unsigned long flags;
@@ -782,6 +875,14 @@ static __init int gsmi_init(void)
 
 	gsmi_dev.smi_cmd = acpi_gbl_FADT.smi_command;
 
+#ifdef CONFIG_PM
+	ret = platform_driver_register(&gsmi_driver_info);
+	if (unlikely(ret)) {
+		printk(KERN_ERR "gsmi: unable to register platform driver\n");
+		return ret;
+	}
+#endif
+
 	/* register device */
 	gsmi_dev.pdev = platform_device_register_full(&gsmi_dev_info);
 	if (IS_ERR(gsmi_dev.pdev)) {
@@ -886,11 +987,14 @@ static __init int gsmi_init(void)
 		goto out_remove_bin_file;
 	}
 
+#ifdef CONFIG_EFI_VARS
 	ret = efivars_register(&efivars, &efivar_ops, gsmi_kobj);
 	if (ret) {
 		printk(KERN_INFO "gsmi: Failed to register efivars\n");
-		goto out_remove_sysfs_files;
+		sysfs_remove_files(gsmi_kobj, gsmi_attrs);
+		goto out_remove_bin_file;
 	}
+#endif
 
 	register_reboot_notifier(&gsmi_reboot_notifier);
 	register_die_notifier(&gsmi_die_notifier);
@@ -901,8 +1005,6 @@ static __init int gsmi_init(void)
 
 	return 0;
 
-out_remove_sysfs_files:
-	sysfs_remove_files(gsmi_kobj, gsmi_attrs);
 out_remove_bin_file:
 	sysfs_remove_bin_file(gsmi_kobj, &eventlog_bin_attr);
 out_err:
@@ -922,7 +1024,9 @@ static void __exit gsmi_exit(void)
 	unregister_die_notifier(&gsmi_die_notifier);
 	atomic_notifier_chain_unregister(&panic_notifier_list,
 					 &gsmi_panic_notifier);
+#ifdef CONFIG_EFI_VARS
 	efivars_unregister(&efivars);
+#endif
 
 	sysfs_remove_files(gsmi_kobj, gsmi_attrs);
 	sysfs_remove_bin_file(gsmi_kobj, &eventlog_bin_attr);
diff --git a/drivers/firmware/google/vpd.c b/drivers/firmware/google/vpd.c
index 1aa67bb5d8c0..c0c0b4e4e281 100644
--- a/drivers/firmware/google/vpd.c
+++ b/drivers/firmware/google/vpd.c
@@ -198,7 +198,7 @@ static int vpd_section_init(const char *name, struct vpd_section *sec,
 
 	sec->name = name;
 
-	/* We want to export the raw partion with name ${name}_raw */
+	/* We want to export the raw partition with name ${name}_raw */
 	sec->raw_name = kasprintf(GFP_KERNEL, "%s_raw", name);
 	if (!sec->raw_name) {
 		err = -ENOMEM;
diff --git a/drivers/fpga/altera-cvp.c b/drivers/fpga/altera-cvp.c
index 7fa793672a7a..610a1558e0ed 100644
--- a/drivers/fpga/altera-cvp.c
+++ b/drivers/fpga/altera-cvp.c
@@ -453,8 +453,8 @@ static int altera_cvp_probe(struct pci_dev *pdev,
 	snprintf(conf->mgr_name, sizeof(conf->mgr_name), "%s @%s",
 		 ALTERA_CVP_MGR_NAME, pci_name(pdev));
 
-	mgr = fpga_mgr_create(&pdev->dev, conf->mgr_name,
-			      &altera_cvp_ops, conf);
+	mgr = devm_fpga_mgr_create(&pdev->dev, conf->mgr_name,
+				   &altera_cvp_ops, conf);
 	if (!mgr) {
 		ret = -ENOMEM;
 		goto err_unmap;
@@ -463,10 +463,8 @@ static int altera_cvp_probe(struct pci_dev *pdev,
 	pci_set_drvdata(pdev, mgr);
 
 	ret = fpga_mgr_register(mgr);
-	if (ret) {
-		fpga_mgr_free(mgr);
+	if (ret)
 		goto err_unmap;
-	}
 
 	ret = driver_create_file(&altera_cvp_driver.driver,
 				 &driver_attr_chkcfg);
diff --git a/drivers/fpga/altera-fpga2sdram.c b/drivers/fpga/altera-fpga2sdram.c
index 23660ccd634b..a78e49c63c64 100644
--- a/drivers/fpga/altera-fpga2sdram.c
+++ b/drivers/fpga/altera-fpga2sdram.c
@@ -121,18 +121,16 @@ static int alt_fpga_bridge_probe(struct platform_device *pdev)
 	/* Get f2s bridge configuration saved in handoff register */
 	regmap_read(sysmgr, SYSMGR_ISWGRP_HANDOFF3, &priv->mask);
 
-	br = fpga_bridge_create(dev, F2S_BRIDGE_NAME,
-				&altera_fpga2sdram_br_ops, priv);
+	br = devm_fpga_bridge_create(dev, F2S_BRIDGE_NAME,
+				     &altera_fpga2sdram_br_ops, priv);
 	if (!br)
 		return -ENOMEM;
 
 	platform_set_drvdata(pdev, br);
 
 	ret = fpga_bridge_register(br);
-	if (ret) {
-		fpga_bridge_free(br);
+	if (ret)
 		return ret;
-	}
 
 	dev_info(dev, "driver initialized with handoff %08x\n", priv->mask);
 
diff --git a/drivers/fpga/altera-freeze-bridge.c b/drivers/fpga/altera-freeze-bridge.c
index ffd586c48ecf..dd58c4aea92e 100644
--- a/drivers/fpga/altera-freeze-bridge.c
+++ b/drivers/fpga/altera-freeze-bridge.c
@@ -213,7 +213,6 @@ static int altera_freeze_br_probe(struct platform_device *pdev)
 	struct fpga_bridge *br;
 	struct resource *res;
 	u32 status, revision;
-	int ret;
 
 	if (!np)
 		return -ENODEV;
@@ -245,20 +244,14 @@ static int altera_freeze_br_probe(struct platform_device *pdev)
 
 	priv->base_addr = base_addr;
 
-	br = fpga_bridge_create(dev, FREEZE_BRIDGE_NAME,
-				&altera_freeze_br_br_ops, priv);
+	br = devm_fpga_bridge_create(dev, FREEZE_BRIDGE_NAME,
+				     &altera_freeze_br_br_ops, priv);
 	if (!br)
 		return -ENOMEM;
 
 	platform_set_drvdata(pdev, br);
 
-	ret = fpga_bridge_register(br);
-	if (ret) {
-		fpga_bridge_free(br);
-		return ret;
-	}
-
-	return 0;
+	return fpga_bridge_register(br);
 }
 
 static int altera_freeze_br_remove(struct platform_device *pdev)
diff --git a/drivers/fpga/altera-hps2fpga.c b/drivers/fpga/altera-hps2fpga.c
index a974d3f60321..77b95f251821 100644
--- a/drivers/fpga/altera-hps2fpga.c
+++ b/drivers/fpga/altera-hps2fpga.c
@@ -180,7 +180,8 @@ static int alt_fpga_bridge_probe(struct platform_device *pdev)
 		}
 	}
 
-	br = fpga_bridge_create(dev, priv->name, &altera_hps2fpga_br_ops, priv);
+	br = devm_fpga_bridge_create(dev, priv->name,
+				     &altera_hps2fpga_br_ops, priv);
 	if (!br) {
 		ret = -ENOMEM;
 		goto err;
@@ -190,12 +191,10 @@ static int alt_fpga_bridge_probe(struct platform_device *pdev)
 
 	ret = fpga_bridge_register(br);
 	if (ret)
-		goto err_free;
+		goto err;
 
 	return 0;
 
-err_free:
-	fpga_bridge_free(br);
 err:
 	clk_disable_unprepare(priv->clk);
 
diff --git a/drivers/fpga/altera-pr-ip-core.c b/drivers/fpga/altera-pr-ip-core.c
index 65e0b6a2c031..a7a3bf0b5202 100644
--- a/drivers/fpga/altera-pr-ip-core.c
+++ b/drivers/fpga/altera-pr-ip-core.c
@@ -177,7 +177,6 @@ int alt_pr_register(struct device *dev, void __iomem *reg_base)
 {
 	struct alt_pr_priv *priv;
 	struct fpga_manager *mgr;
-	int ret;
 	u32 val;
 
 	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
@@ -192,17 +191,13 @@ int alt_pr_register(struct device *dev, void __iomem *reg_base)
 		(val & ALT_PR_CSR_STATUS_MSK) >> ALT_PR_CSR_STATUS_SFT,
 		(int)(val & ALT_PR_CSR_PR_START));
 
-	mgr = fpga_mgr_create(dev, dev_name(dev), &alt_pr_ops, priv);
+	mgr = devm_fpga_mgr_create(dev, dev_name(dev), &alt_pr_ops, priv);
 	if (!mgr)
 		return -ENOMEM;
 
 	dev_set_drvdata(dev, mgr);
 
-	ret = fpga_mgr_register(mgr);
-	if (ret)
-		fpga_mgr_free(mgr);
-
-	return ret;
+	return fpga_mgr_register(mgr);
 }
 EXPORT_SYMBOL_GPL(alt_pr_register);
 
diff --git a/drivers/fpga/altera-ps-spi.c b/drivers/fpga/altera-ps-spi.c
index 24b25c626036..33aafda50af5 100644
--- a/drivers/fpga/altera-ps-spi.c
+++ b/drivers/fpga/altera-ps-spi.c
@@ -239,7 +239,6 @@ static int altera_ps_probe(struct spi_device *spi)
 	struct altera_ps_conf *conf;
 	const struct of_device_id *of_id;
 	struct fpga_manager *mgr;
-	int ret;
 
 	conf = devm_kzalloc(&spi->dev, sizeof(*conf), GFP_KERNEL);
 	if (!conf)
@@ -275,18 +274,14 @@ static int altera_ps_probe(struct spi_device *spi)
 	snprintf(conf->mgr_name, sizeof(conf->mgr_name), "%s %s",
 		 dev_driver_string(&spi->dev), dev_name(&spi->dev));
 
-	mgr = fpga_mgr_create(&spi->dev, conf->mgr_name,
-			      &altera_ps_ops, conf);
+	mgr = devm_fpga_mgr_create(&spi->dev, conf->mgr_name,
+				   &altera_ps_ops, conf);
 	if (!mgr)
 		return -ENOMEM;
 
 	spi_set_drvdata(spi, mgr);
 
-	ret = fpga_mgr_register(mgr);
-	if (ret)
-		fpga_mgr_free(mgr);
-
-	return ret;
+	return fpga_mgr_register(mgr);
 }
 
 static int altera_ps_remove(struct spi_device *spi)
diff --git a/drivers/fpga/dfl-afu-dma-region.c b/drivers/fpga/dfl-afu-dma-region.c
index 0e81d33af856..025aba3ea76c 100644
--- a/drivers/fpga/dfl-afu-dma-region.c
+++ b/drivers/fpga/dfl-afu-dma-region.c
@@ -70,7 +70,7 @@ static int afu_dma_adjust_locked_vm(struct device *dev, long npages, bool incr)
 	dev_dbg(dev, "[%d] RLIMIT_MEMLOCK %c%ld %ld/%ld%s\n", current->pid,
 		incr ? '+' : '-', npages << PAGE_SHIFT,
 		current->mm->locked_vm << PAGE_SHIFT, rlimit(RLIMIT_MEMLOCK),
-		ret ? "- execeeded" : "");
+		ret ? "- exceeded" : "");
 
 	up_write(&current->mm->mmap_sem);
 
diff --git a/drivers/fpga/dfl-fme-br.c b/drivers/fpga/dfl-fme-br.c
index 7cc041def8b3..3ff9f3a687ce 100644
--- a/drivers/fpga/dfl-fme-br.c
+++ b/drivers/fpga/dfl-fme-br.c
@@ -61,7 +61,6 @@ static int fme_br_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct fme_br_priv *priv;
 	struct fpga_bridge *br;
-	int ret;
 
 	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
 	if (!priv)
@@ -69,18 +68,14 @@ static int fme_br_probe(struct platform_device *pdev)
 
 	priv->pdata = dev_get_platdata(dev);
 
-	br = fpga_bridge_create(dev, "DFL FPGA FME Bridge",
-				&fme_bridge_ops, priv);
+	br = devm_fpga_bridge_create(dev, "DFL FPGA FME Bridge",
+				     &fme_bridge_ops, priv);
 	if (!br)
 		return -ENOMEM;
 
 	platform_set_drvdata(pdev, br);
 
-	ret = fpga_bridge_register(br);
-	if (ret)
-		fpga_bridge_free(br);
-
-	return ret;
+	return fpga_bridge_register(br);
 }
 
 static int fme_br_remove(struct platform_device *pdev)
diff --git a/drivers/fpga/dfl-fme-mgr.c b/drivers/fpga/dfl-fme-mgr.c
index b5ef405b6d88..76f37709dd1a 100644
--- a/drivers/fpga/dfl-fme-mgr.c
+++ b/drivers/fpga/dfl-fme-mgr.c
@@ -201,7 +201,7 @@ static int fme_mgr_write(struct fpga_manager *mgr,
 		}
 
 		if (count < 4) {
-			dev_err(dev, "Invaild PR bitstream size\n");
+			dev_err(dev, "Invalid PR bitstream size\n");
 			return -EINVAL;
 		}
 
@@ -287,7 +287,6 @@ static int fme_mgr_probe(struct platform_device *pdev)
 	struct fme_mgr_priv *priv;
 	struct fpga_manager *mgr;
 	struct resource *res;
-	int ret;
 
 	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
 	if (!priv)
@@ -309,19 +308,15 @@ static int fme_mgr_probe(struct platform_device *pdev)
 
 	fme_mgr_get_compat_id(priv->ioaddr, compat_id);
 
-	mgr = fpga_mgr_create(dev, "DFL FME FPGA Manager",
-			      &fme_mgr_ops, priv);
+	mgr = devm_fpga_mgr_create(dev, "DFL FME FPGA Manager",
+				   &fme_mgr_ops, priv);
 	if (!mgr)
 		return -ENOMEM;
 
 	mgr->compat_id = compat_id;
 	platform_set_drvdata(pdev, mgr);
 
-	ret = fpga_mgr_register(mgr);
-	if (ret)
-		fpga_mgr_free(mgr);
-
-	return ret;
+	return fpga_mgr_register(mgr);
 }
 
 static int fme_mgr_remove(struct platform_device *pdev)
diff --git a/drivers/fpga/dfl-fme-region.c b/drivers/fpga/dfl-fme-region.c
index 51a5ac2293a7..ec134ec93f08 100644
--- a/drivers/fpga/dfl-fme-region.c
+++ b/drivers/fpga/dfl-fme-region.c
@@ -39,7 +39,7 @@ static int fme_region_probe(struct platform_device *pdev)
 	if (IS_ERR(mgr))
 		return -EPROBE_DEFER;
 
-	region = fpga_region_create(dev, mgr, fme_region_get_bridges);
+	region = devm_fpga_region_create(dev, mgr, fme_region_get_bridges);
 	if (!region) {
 		ret = -ENOMEM;
 		goto eprobe_mgr_put;
@@ -51,14 +51,12 @@ static int fme_region_probe(struct platform_device *pdev)
 
 	ret = fpga_region_register(region);
 	if (ret)
-		goto region_free;
+		goto eprobe_mgr_put;
 
 	dev_dbg(dev, "DFL FME FPGA Region probed\n");
 
 	return 0;
 
-region_free:
-	fpga_region_free(region);
 eprobe_mgr_put:
 	fpga_mgr_put(mgr);
 	return ret;
diff --git a/drivers/fpga/dfl.c b/drivers/fpga/dfl.c
index a9b521bccb06..2c09e502e721 100644
--- a/drivers/fpga/dfl.c
+++ b/drivers/fpga/dfl.c
@@ -899,7 +899,7 @@ dfl_fpga_feature_devs_enumerate(struct dfl_fpga_enum_info *info)
 	if (!cdev)
 		return ERR_PTR(-ENOMEM);
 
-	cdev->region = fpga_region_create(info->dev, NULL, NULL);
+	cdev->region = devm_fpga_region_create(info->dev, NULL, NULL);
 	if (!cdev->region) {
 		ret = -ENOMEM;
 		goto free_cdev_exit;
@@ -911,7 +911,7 @@ dfl_fpga_feature_devs_enumerate(struct dfl_fpga_enum_info *info)
 
 	ret = fpga_region_register(cdev->region);
 	if (ret)
-		goto free_region_exit;
+		goto free_cdev_exit;
 
 	/* create and init build info for enumeration */
 	binfo = devm_kzalloc(info->dev, sizeof(*binfo), GFP_KERNEL);
@@ -942,8 +942,6 @@ dfl_fpga_feature_devs_enumerate(struct dfl_fpga_enum_info *info)
 
 unregister_region_exit:
 	fpga_region_unregister(cdev->region);
-free_region_exit:
-	fpga_region_free(cdev->region);
 free_cdev_exit:
 	devm_kfree(info->dev, cdev);
 	return ERR_PTR(ret);
diff --git a/drivers/fpga/fpga-bridge.c b/drivers/fpga/fpga-bridge.c
index c983dac97501..80bd8f1b2aa6 100644
--- a/drivers/fpga/fpga-bridge.c
+++ b/drivers/fpga/fpga-bridge.c
@@ -324,6 +324,9 @@ ATTRIBUTE_GROUPS(fpga_bridge);
  * @br_ops:	pointer to structure of fpga bridge ops
  * @priv:	FPGA bridge private data
  *
+ * The caller of this function is responsible for freeing the bridge with
+ * fpga_bridge_free().  Using devm_fpga_bridge_create() instead is recommended.
+ *
  * Return: struct fpga_bridge or NULL
  */
 struct fpga_bridge *fpga_bridge_create(struct device *dev, const char *name,
@@ -378,8 +381,8 @@ error_kfree:
 EXPORT_SYMBOL_GPL(fpga_bridge_create);
 
 /**
- * fpga_bridge_free - free a fpga bridge and its id
- * @bridge:	FPGA bridge struct created by fpga_bridge_create
+ * fpga_bridge_free - free a fpga bridge created by fpga_bridge_create()
+ * @bridge:	FPGA bridge struct
  */
 void fpga_bridge_free(struct fpga_bridge *bridge)
 {
@@ -388,9 +391,56 @@ void fpga_bridge_free(struct fpga_bridge *bridge)
 }
 EXPORT_SYMBOL_GPL(fpga_bridge_free);
 
+static void devm_fpga_bridge_release(struct device *dev, void *res)
+{
+	struct fpga_bridge *bridge = *(struct fpga_bridge **)res;
+
+	fpga_bridge_free(bridge);
+}
+
 /**
- * fpga_bridge_register - register a fpga bridge
- * @bridge:	FPGA bridge struct created by fpga_bridge_create
+ * devm_fpga_bridge_create - create and init a managed struct fpga_bridge
+ * @dev:	FPGA bridge device from pdev
+ * @name:	FPGA bridge name
+ * @br_ops:	pointer to structure of fpga bridge ops
+ * @priv:	FPGA bridge private data
+ *
+ * This function is intended for use in a FPGA bridge driver's probe function.
+ * After the bridge driver creates the struct with devm_fpga_bridge_create(), it
+ * should register the bridge with fpga_bridge_register().  The bridge driver's
+ * remove function should call fpga_bridge_unregister().  The bridge struct
+ * allocated with this function will be freed automatically on driver detach.
+ * This includes the case of a probe function returning error before calling
+ * fpga_bridge_register(), the struct will still get cleaned up.
+ *
+ *  Return: struct fpga_bridge or NULL
+ */
+struct fpga_bridge
+*devm_fpga_bridge_create(struct device *dev, const char *name,
+			 const struct fpga_bridge_ops *br_ops, void *priv)
+{
+	struct fpga_bridge **ptr, *bridge;
+
+	ptr = devres_alloc(devm_fpga_bridge_release, sizeof(*ptr), GFP_KERNEL);
+	if (!ptr)
+		return NULL;
+
+	bridge = fpga_bridge_create(dev, name, br_ops, priv);
+	if (!bridge) {
+		devres_free(ptr);
+	} else {
+		*ptr = bridge;
+		devres_add(dev, ptr);
+	}
+
+	return bridge;
+}
+EXPORT_SYMBOL_GPL(devm_fpga_bridge_create);
+
+/**
+ * fpga_bridge_register - register a FPGA bridge
+ *
+ * @bridge: FPGA bridge struct
  *
  * Return: 0 for success, error code otherwise.
  */
@@ -412,8 +462,11 @@ int fpga_bridge_register(struct fpga_bridge *bridge)
 EXPORT_SYMBOL_GPL(fpga_bridge_register);
 
 /**
- * fpga_bridge_unregister - unregister and free a fpga bridge
- * @bridge:	FPGA bridge struct created by fpga_bridge_create
+ * fpga_bridge_unregister - unregister a FPGA bridge
+ *
+ * @bridge: FPGA bridge struct
+ *
+ * This function is intended for use in a FPGA bridge driver's remove function.
  */
 void fpga_bridge_unregister(struct fpga_bridge *bridge)
 {
@@ -430,9 +483,6 @@ EXPORT_SYMBOL_GPL(fpga_bridge_unregister);
 
 static void fpga_bridge_dev_release(struct device *dev)
 {
-	struct fpga_bridge *bridge = to_fpga_bridge(dev);
-
-	fpga_bridge_free(bridge);
 }
 
 static int __init fpga_bridge_dev_init(void)
diff --git a/drivers/fpga/fpga-mgr.c b/drivers/fpga/fpga-mgr.c
index a41b07e37884..c3866816456a 100644
--- a/drivers/fpga/fpga-mgr.c
+++ b/drivers/fpga/fpga-mgr.c
@@ -558,6 +558,9 @@ EXPORT_SYMBOL_GPL(fpga_mgr_unlock);
  * @mops:	pointer to structure of fpga manager ops
  * @priv:	fpga manager private data
  *
+ * The caller of this function is responsible for freeing the struct with
+ * fpga_mgr_free().  Using devm_fpga_mgr_create() instead is recommended.
+ *
  * Return: pointer to struct fpga_manager or NULL
  */
 struct fpga_manager *fpga_mgr_create(struct device *dev, const char *name,
@@ -618,8 +621,8 @@ error_kfree:
 EXPORT_SYMBOL_GPL(fpga_mgr_create);
 
 /**
- * fpga_mgr_free - deallocate a FPGA manager
- * @mgr:	fpga manager struct created by fpga_mgr_create
+ * fpga_mgr_free - free a FPGA manager created with fpga_mgr_create()
+ * @mgr:	fpga manager struct
  */
 void fpga_mgr_free(struct fpga_manager *mgr)
 {
@@ -628,9 +631,55 @@ void fpga_mgr_free(struct fpga_manager *mgr)
 }
 EXPORT_SYMBOL_GPL(fpga_mgr_free);
 
+static void devm_fpga_mgr_release(struct device *dev, void *res)
+{
+	struct fpga_manager *mgr = *(struct fpga_manager **)res;
+
+	fpga_mgr_free(mgr);
+}
+
+/**
+ * devm_fpga_mgr_create - create and initialize a managed FPGA manager struct
+ * @dev:	fpga manager device from pdev
+ * @name:	fpga manager name
+ * @mops:	pointer to structure of fpga manager ops
+ * @priv:	fpga manager private data
+ *
+ * This function is intended for use in a FPGA manager driver's probe function.
+ * After the manager driver creates the manager struct with
+ * devm_fpga_mgr_create(), it should register it with fpga_mgr_register().  The
+ * manager driver's remove function should call fpga_mgr_unregister().  The
+ * manager struct allocated with this function will be freed automatically on
+ * driver detach.  This includes the case of a probe function returning error
+ * before calling fpga_mgr_register(), the struct will still get cleaned up.
+ *
+ * Return: pointer to struct fpga_manager or NULL
+ */
+struct fpga_manager *devm_fpga_mgr_create(struct device *dev, const char *name,
+					  const struct fpga_manager_ops *mops,
+					  void *priv)
+{
+	struct fpga_manager **ptr, *mgr;
+
+	ptr = devres_alloc(devm_fpga_mgr_release, sizeof(*ptr), GFP_KERNEL);
+	if (!ptr)
+		return NULL;
+
+	mgr = fpga_mgr_create(dev, name, mops, priv);
+	if (!mgr) {
+		devres_free(ptr);
+	} else {
+		*ptr = mgr;
+		devres_add(dev, ptr);
+	}
+
+	return mgr;
+}
+EXPORT_SYMBOL_GPL(devm_fpga_mgr_create);
+
 /**
  * fpga_mgr_register - register a FPGA manager
- * @mgr:	fpga manager struct created by fpga_mgr_create
+ * @mgr: fpga manager struct
  *
  * Return: 0 on success, negative error code otherwise.
  */
@@ -661,8 +710,10 @@ error_device:
 EXPORT_SYMBOL_GPL(fpga_mgr_register);
 
 /**
- * fpga_mgr_unregister - unregister and free a FPGA manager
- * @mgr:	fpga manager struct
+ * fpga_mgr_unregister - unregister a FPGA manager
+ * @mgr: fpga manager struct
+ *
+ * This function is intended for use in a FPGA manager driver's remove function.
  */
 void fpga_mgr_unregister(struct fpga_manager *mgr)
 {
@@ -681,9 +732,6 @@ EXPORT_SYMBOL_GPL(fpga_mgr_unregister);
 
 static void fpga_mgr_dev_release(struct device *dev)
 {
-	struct fpga_manager *mgr = to_fpga_manager(dev);
-
-	fpga_mgr_free(mgr);
 }
 
 static int __init fpga_mgr_class_init(void)
diff --git a/drivers/fpga/fpga-region.c b/drivers/fpga/fpga-region.c
index 0d65220d5ec5..bde5a9d460c5 100644
--- a/drivers/fpga/fpga-region.c
+++ b/drivers/fpga/fpga-region.c
@@ -185,6 +185,10 @@ ATTRIBUTE_GROUPS(fpga_region);
  * @mgr: manager that programs this region
  * @get_bridges: optional function to get bridges to a list
  *
+ * The caller of this function is responsible for freeing the resulting region
+ * struct with fpga_region_free().  Using devm_fpga_region_create() instead is
+ * recommended.
+ *
  * Return: struct fpga_region or NULL
  */
 struct fpga_region
@@ -230,8 +234,8 @@ err_free:
 EXPORT_SYMBOL_GPL(fpga_region_create);
 
 /**
- * fpga_region_free - free a struct fpga_region
- * @region: FPGA region created by fpga_region_create
+ * fpga_region_free - free a FPGA region created by fpga_region_create()
+ * @region: FPGA region
  */
 void fpga_region_free(struct fpga_region *region)
 {
@@ -240,21 +244,69 @@ void fpga_region_free(struct fpga_region *region)
 }
 EXPORT_SYMBOL_GPL(fpga_region_free);
 
+static void devm_fpga_region_release(struct device *dev, void *res)
+{
+	struct fpga_region *region = *(struct fpga_region **)res;
+
+	fpga_region_free(region);
+}
+
+/**
+ * devm_fpga_region_create - create and initialize a managed FPGA region struct
+ * @dev: device parent
+ * @mgr: manager that programs this region
+ * @get_bridges: optional function to get bridges to a list
+ *
+ * This function is intended for use in a FPGA region driver's probe function.
+ * After the region driver creates the region struct with
+ * devm_fpga_region_create(), it should register it with fpga_region_register().
+ * The region driver's remove function should call fpga_region_unregister().
+ * The region struct allocated with this function will be freed automatically on
+ * driver detach.  This includes the case of a probe function returning error
+ * before calling fpga_region_register(), the struct will still get cleaned up.
+ *
+ * Return: struct fpga_region or NULL
+ */
+struct fpga_region
+*devm_fpga_region_create(struct device *dev,
+			 struct fpga_manager *mgr,
+			 int (*get_bridges)(struct fpga_region *))
+{
+	struct fpga_region **ptr, *region;
+
+	ptr = devres_alloc(devm_fpga_region_release, sizeof(*ptr), GFP_KERNEL);
+	if (!ptr)
+		return NULL;
+
+	region = fpga_region_create(dev, mgr, get_bridges);
+	if (!region) {
+		devres_free(ptr);
+	} else {
+		*ptr = region;
+		devres_add(dev, ptr);
+	}
+
+	return region;
+}
+EXPORT_SYMBOL_GPL(devm_fpga_region_create);
+
 /**
  * fpga_region_register - register a FPGA region
- * @region: FPGA region created by fpga_region_create
+ * @region: FPGA region
+ *
  * Return: 0 or -errno
  */
 int fpga_region_register(struct fpga_region *region)
 {
 	return device_add(&region->dev);
-
 }
 EXPORT_SYMBOL_GPL(fpga_region_register);
 
 /**
- * fpga_region_unregister - unregister and free a FPGA region
+ * fpga_region_unregister - unregister a FPGA region
  * @region: FPGA region
+ *
+ * This function is intended for use in a FPGA region driver's remove function.
  */
 void fpga_region_unregister(struct fpga_region *region)
 {
@@ -264,9 +316,6 @@ EXPORT_SYMBOL_GPL(fpga_region_unregister);
 
 static void fpga_region_dev_release(struct device *dev)
 {
-	struct fpga_region *region = to_fpga_region(dev);
-
-	fpga_region_free(region);
 }
 
 /**
diff --git a/drivers/fpga/ice40-spi.c b/drivers/fpga/ice40-spi.c
index 5981c7ee7a7d..6154661b8f76 100644
--- a/drivers/fpga/ice40-spi.c
+++ b/drivers/fpga/ice40-spi.c
@@ -175,18 +175,14 @@ static int ice40_fpga_probe(struct spi_device *spi)
 		return ret;
 	}
 
-	mgr = fpga_mgr_create(dev, "Lattice iCE40 FPGA Manager",
-			      &ice40_fpga_ops, priv);
+	mgr = devm_fpga_mgr_create(dev, "Lattice iCE40 FPGA Manager",
+				   &ice40_fpga_ops, priv);
 	if (!mgr)
 		return -ENOMEM;
 
 	spi_set_drvdata(spi, mgr);
 
-	ret = fpga_mgr_register(mgr);
-	if (ret)
-		fpga_mgr_free(mgr);
-
-	return ret;
+	return fpga_mgr_register(mgr);
 }
 
 static int ice40_fpga_remove(struct spi_device *spi)
diff --git a/drivers/fpga/machxo2-spi.c b/drivers/fpga/machxo2-spi.c
index a582e0000c97..4d8a87641587 100644
--- a/drivers/fpga/machxo2-spi.c
+++ b/drivers/fpga/machxo2-spi.c
@@ -356,25 +356,20 @@ static int machxo2_spi_probe(struct spi_device *spi)
 {
 	struct device *dev = &spi->dev;
 	struct fpga_manager *mgr;
-	int ret;
 
 	if (spi->max_speed_hz > MACHXO2_MAX_SPEED) {
 		dev_err(dev, "Speed is too high\n");
 		return -EINVAL;
 	}
 
-	mgr = fpga_mgr_create(dev, "Lattice MachXO2 SPI FPGA Manager",
-			      &machxo2_ops, spi);
+	mgr = devm_fpga_mgr_create(dev, "Lattice MachXO2 SPI FPGA Manager",
+				   &machxo2_ops, spi);
 	if (!mgr)
 		return -ENOMEM;
 
 	spi_set_drvdata(spi, mgr);
 
-	ret = fpga_mgr_register(mgr);
-	if (ret)
-		fpga_mgr_free(mgr);
-
-	return ret;
+	return fpga_mgr_register(mgr);
 }
 
 static int machxo2_spi_remove(struct spi_device *spi)
diff --git a/drivers/fpga/of-fpga-region.c b/drivers/fpga/of-fpga-region.c
index 052a1342ab7e..122286fd255a 100644
--- a/drivers/fpga/of-fpga-region.c
+++ b/drivers/fpga/of-fpga-region.c
@@ -410,7 +410,7 @@ static int of_fpga_region_probe(struct platform_device *pdev)
 	if (IS_ERR(mgr))
 		return -EPROBE_DEFER;
 
-	region = fpga_region_create(dev, mgr, of_fpga_region_get_bridges);
+	region = devm_fpga_region_create(dev, mgr, of_fpga_region_get_bridges);
 	if (!region) {
 		ret = -ENOMEM;
 		goto eprobe_mgr_put;
@@ -418,7 +418,7 @@ static int of_fpga_region_probe(struct platform_device *pdev)
 
 	ret = fpga_region_register(region);
 	if (ret)
-		goto eprobe_free;
+		goto eprobe_mgr_put;
 
 	of_platform_populate(np, fpga_region_of_match, NULL, &region->dev);
 	dev_set_drvdata(dev, region);
@@ -427,8 +427,6 @@ static int of_fpga_region_probe(struct platform_device *pdev)
 
 	return 0;
 
-eprobe_free:
-	fpga_region_free(region);
 eprobe_mgr_put:
 	fpga_mgr_put(mgr);
 	return ret;
diff --git a/drivers/fpga/socfpga-a10.c b/drivers/fpga/socfpga-a10.c
index be30c48eb6e4..573d88bdf730 100644
--- a/drivers/fpga/socfpga-a10.c
+++ b/drivers/fpga/socfpga-a10.c
@@ -508,8 +508,8 @@ static int socfpga_a10_fpga_probe(struct platform_device *pdev)
 		return -EBUSY;
 	}
 
-	mgr = fpga_mgr_create(dev, "SoCFPGA Arria10 FPGA Manager",
-			      &socfpga_a10_fpga_mgr_ops, priv);
+	mgr = devm_fpga_mgr_create(dev, "SoCFPGA Arria10 FPGA Manager",
+				   &socfpga_a10_fpga_mgr_ops, priv);
 	if (!mgr)
 		return -ENOMEM;
 
@@ -517,7 +517,6 @@ static int socfpga_a10_fpga_probe(struct platform_device *pdev)
 
 	ret = fpga_mgr_register(mgr);
 	if (ret) {
-		fpga_mgr_free(mgr);
 		clk_disable_unprepare(priv->clk);
 		return ret;
 	}
diff --git a/drivers/fpga/socfpga.c b/drivers/fpga/socfpga.c
index 959d71f26896..4a8a2fcd4e6c 100644
--- a/drivers/fpga/socfpga.c
+++ b/drivers/fpga/socfpga.c
@@ -571,18 +571,14 @@ static int socfpga_fpga_probe(struct platform_device *pdev)
 	if (ret)
 		return ret;
 
-	mgr = fpga_mgr_create(dev, "Altera SOCFPGA FPGA Manager",
-			      &socfpga_fpga_ops, priv);
+	mgr = devm_fpga_mgr_create(dev, "Altera SOCFPGA FPGA Manager",
+				   &socfpga_fpga_ops, priv);
 	if (!mgr)
 		return -ENOMEM;
 
 	platform_set_drvdata(pdev, mgr);
 
-	ret = fpga_mgr_register(mgr);
-	if (ret)
-		fpga_mgr_free(mgr);
-
-	return ret;
+	return fpga_mgr_register(mgr);
 }
 
 static int socfpga_fpga_remove(struct platform_device *pdev)
diff --git a/drivers/fpga/ts73xx-fpga.c b/drivers/fpga/ts73xx-fpga.c
index 08efd1895b1b..dc22a5842609 100644
--- a/drivers/fpga/ts73xx-fpga.c
+++ b/drivers/fpga/ts73xx-fpga.c
@@ -118,7 +118,6 @@ static int ts73xx_fpga_probe(struct platform_device *pdev)
 	struct ts73xx_fpga_priv *priv;
 	struct fpga_manager *mgr;
 	struct resource *res;
-	int ret;
 
 	priv = devm_kzalloc(kdev, sizeof(*priv), GFP_KERNEL);
 	if (!priv)
@@ -133,18 +132,14 @@ static int ts73xx_fpga_probe(struct platform_device *pdev)
 		return PTR_ERR(priv->io_base);
 	}
 
-	mgr = fpga_mgr_create(kdev, "TS-73xx FPGA Manager",
-			      &ts73xx_fpga_ops, priv);
+	mgr = devm_fpga_mgr_create(kdev, "TS-73xx FPGA Manager",
+				   &ts73xx_fpga_ops, priv);
 	if (!mgr)
 		return -ENOMEM;
 
 	platform_set_drvdata(pdev, mgr);
 
-	ret = fpga_mgr_register(mgr);
-	if (ret)
-		fpga_mgr_free(mgr);
-
-	return ret;
+	return fpga_mgr_register(mgr);
 }
 
 static int ts73xx_fpga_remove(struct platform_device *pdev)
diff --git a/drivers/fpga/xilinx-pr-decoupler.c b/drivers/fpga/xilinx-pr-decoupler.c
index 07ba1539e82c..641036135207 100644
--- a/drivers/fpga/xilinx-pr-decoupler.c
+++ b/drivers/fpga/xilinx-pr-decoupler.c
@@ -121,8 +121,8 @@ static int xlnx_pr_decoupler_probe(struct platform_device *pdev)
 
 	clk_disable(priv->clk);
 
-	br = fpga_bridge_create(&pdev->dev, "Xilinx PR Decoupler",
-				&xlnx_pr_decoupler_br_ops, priv);
+	br = devm_fpga_bridge_create(&pdev->dev, "Xilinx PR Decoupler",
+				     &xlnx_pr_decoupler_br_ops, priv);
 	if (!br) {
 		err = -ENOMEM;
 		goto err_clk;
diff --git a/drivers/fpga/xilinx-spi.c b/drivers/fpga/xilinx-spi.c
index 8d1945966533..469486be20c4 100644
--- a/drivers/fpga/xilinx-spi.c
+++ b/drivers/fpga/xilinx-spi.c
@@ -144,7 +144,6 @@ static int xilinx_spi_probe(struct spi_device *spi)
 {
 	struct xilinx_spi_conf *conf;
 	struct fpga_manager *mgr;
-	int ret;
 
 	conf = devm_kzalloc(&spi->dev, sizeof(*conf), GFP_KERNEL);
 	if (!conf)
@@ -167,18 +166,15 @@ static int xilinx_spi_probe(struct spi_device *spi)
 		return PTR_ERR(conf->done);
 	}
 
-	mgr = fpga_mgr_create(&spi->dev, "Xilinx Slave Serial FPGA Manager",
-			      &xilinx_spi_ops, conf);
+	mgr = devm_fpga_mgr_create(&spi->dev,
+				   "Xilinx Slave Serial FPGA Manager",
+				   &xilinx_spi_ops, conf);
 	if (!mgr)
 		return -ENOMEM;
 
 	spi_set_drvdata(spi, mgr);
 
-	ret = fpga_mgr_register(mgr);
-	if (ret)
-		fpga_mgr_free(mgr);
-
-	return ret;
+	return fpga_mgr_register(mgr);
 }
 
 static int xilinx_spi_remove(struct spi_device *spi)
diff --git a/drivers/fpga/zynq-fpga.c b/drivers/fpga/zynq-fpga.c
index 3110e00121ca..bb82efeebb9d 100644
--- a/drivers/fpga/zynq-fpga.c
+++ b/drivers/fpga/zynq-fpga.c
@@ -614,8 +614,8 @@ static int zynq_fpga_probe(struct platform_device *pdev)
 
 	clk_disable(priv->clk);
 
-	mgr = fpga_mgr_create(dev, "Xilinx Zynq FPGA Manager",
-			      &zynq_fpga_ops, priv);
+	mgr = devm_fpga_mgr_create(dev, "Xilinx Zynq FPGA Manager",
+				   &zynq_fpga_ops, priv);
 	if (!mgr)
 		return -ENOMEM;
 
@@ -624,7 +624,6 @@ static int zynq_fpga_probe(struct platform_device *pdev)
 	err = fpga_mgr_register(mgr);
 	if (err) {
 		dev_err(dev, "unable to register FPGA manager\n");
-		fpga_mgr_free(mgr);
 		clk_unprepare(priv->clk);
 		return err;
 	}
diff --git a/drivers/hv/channel.c b/drivers/hv/channel.c
index 741857d80da1..de8193f3b838 100644
--- a/drivers/hv/channel.c
+++ b/drivers/hv/channel.c
@@ -79,85 +79,96 @@ void vmbus_setevent(struct vmbus_channel *channel)
 }
 EXPORT_SYMBOL_GPL(vmbus_setevent);
 
-/*
- * vmbus_open - Open the specified channel.
- */
-int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size,
-		     u32 recv_ringbuffer_size, void *userdata, u32 userdatalen,
-		     void (*onchannelcallback)(void *context), void *context)
+/* vmbus_free_ring - drop mapping of ring buffer */
+void vmbus_free_ring(struct vmbus_channel *channel)
+{
+	hv_ringbuffer_cleanup(&channel->outbound);
+	hv_ringbuffer_cleanup(&channel->inbound);
+
+	if (channel->ringbuffer_page) {
+		__free_pages(channel->ringbuffer_page,
+			     get_order(channel->ringbuffer_pagecount
+				       << PAGE_SHIFT));
+		channel->ringbuffer_page = NULL;
+	}
+}
+EXPORT_SYMBOL_GPL(vmbus_free_ring);
+
+/* vmbus_alloc_ring - allocate and map pages for ring buffer */
+int vmbus_alloc_ring(struct vmbus_channel *newchannel,
+		     u32 send_size, u32 recv_size)
+{
+	struct page *page;
+	int order;
+
+	if (send_size % PAGE_SIZE || recv_size % PAGE_SIZE)
+		return -EINVAL;
+
+	/* Allocate the ring buffer */
+	order = get_order(send_size + recv_size);
+	page = alloc_pages_node(cpu_to_node(newchannel->target_cpu),
+				GFP_KERNEL|__GFP_ZERO, order);
+
+	if (!page)
+		page = alloc_pages(GFP_KERNEL|__GFP_ZERO, order);
+
+	if (!page)
+		return -ENOMEM;
+
+	newchannel->ringbuffer_page = page;
+	newchannel->ringbuffer_pagecount = (send_size + recv_size) >> PAGE_SHIFT;
+	newchannel->ringbuffer_send_offset = send_size >> PAGE_SHIFT;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(vmbus_alloc_ring);
+
+static int __vmbus_open(struct vmbus_channel *newchannel,
+		       void *userdata, u32 userdatalen,
+		       void (*onchannelcallback)(void *context), void *context)
 {
 	struct vmbus_channel_open_channel *open_msg;
 	struct vmbus_channel_msginfo *open_info = NULL;
+	struct page *page = newchannel->ringbuffer_page;
+	u32 send_pages, recv_pages;
 	unsigned long flags;
-	int ret, err = 0;
-	struct page *page;
+	int err;
 
-	if (send_ringbuffer_size % PAGE_SIZE ||
-	    recv_ringbuffer_size % PAGE_SIZE)
+	if (userdatalen > MAX_USER_DEFINED_BYTES)
 		return -EINVAL;
 
+	send_pages = newchannel->ringbuffer_send_offset;
+	recv_pages = newchannel->ringbuffer_pagecount - send_pages;
+
 	spin_lock_irqsave(&newchannel->lock, flags);
-	if (newchannel->state == CHANNEL_OPEN_STATE) {
-		newchannel->state = CHANNEL_OPENING_STATE;
-	} else {
+	if (newchannel->state != CHANNEL_OPEN_STATE) {
 		spin_unlock_irqrestore(&newchannel->lock, flags);
 		return -EINVAL;
 	}
 	spin_unlock_irqrestore(&newchannel->lock, flags);
 
+	newchannel->state = CHANNEL_OPENING_STATE;
 	newchannel->onchannel_callback = onchannelcallback;
 	newchannel->channel_callback_context = context;
 
-	/* Allocate the ring buffer */
-	page = alloc_pages_node(cpu_to_node(newchannel->target_cpu),
-				GFP_KERNEL|__GFP_ZERO,
-				get_order(send_ringbuffer_size +
-				recv_ringbuffer_size));
-
-	if (!page)
-		page = alloc_pages(GFP_KERNEL|__GFP_ZERO,
-				   get_order(send_ringbuffer_size +
-					     recv_ringbuffer_size));
-
-	if (!page) {
-		err = -ENOMEM;
-		goto error_set_chnstate;
-	}
-
-	newchannel->ringbuffer_pages = page_address(page);
-	newchannel->ringbuffer_pagecount = (send_ringbuffer_size +
-					   recv_ringbuffer_size) >> PAGE_SHIFT;
-
-	ret = hv_ringbuffer_init(&newchannel->outbound, page,
-				 send_ringbuffer_size >> PAGE_SHIFT);
-
-	if (ret != 0) {
-		err = ret;
-		goto error_free_pages;
-	}
-
-	ret = hv_ringbuffer_init(&newchannel->inbound,
-				 &page[send_ringbuffer_size >> PAGE_SHIFT],
-				 recv_ringbuffer_size >> PAGE_SHIFT);
-	if (ret != 0) {
-		err = ret;
-		goto error_free_pages;
-	}
+	err = hv_ringbuffer_init(&newchannel->outbound, page, send_pages);
+	if (err)
+		goto error_clean_ring;
 
+	err = hv_ringbuffer_init(&newchannel->inbound,
+				 &page[send_pages], recv_pages);
+	if (err)
+		goto error_clean_ring;
 
 	/* Establish the gpadl for the ring buffer */
 	newchannel->ringbuffer_gpadlhandle = 0;
 
-	ret = vmbus_establish_gpadl(newchannel,
-				    page_address(page),
-				    send_ringbuffer_size +
-				    recv_ringbuffer_size,
+	err = vmbus_establish_gpadl(newchannel,
+				    page_address(newchannel->ringbuffer_page),
+				    (send_pages + recv_pages) << PAGE_SHIFT,
 				    &newchannel->ringbuffer_gpadlhandle);
-
-	if (ret != 0) {
-		err = ret;
-		goto error_free_pages;
-	}
+	if (err)
+		goto error_clean_ring;
 
 	/* Create and init the channel open message */
 	open_info = kmalloc(sizeof(*open_info) +
@@ -176,15 +187,9 @@ int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size,
 	open_msg->openid = newchannel->offermsg.child_relid;
 	open_msg->child_relid = newchannel->offermsg.child_relid;
 	open_msg->ringbuffer_gpadlhandle = newchannel->ringbuffer_gpadlhandle;
-	open_msg->downstream_ringbuffer_pageoffset = send_ringbuffer_size >>
-						  PAGE_SHIFT;
+	open_msg->downstream_ringbuffer_pageoffset = newchannel->ringbuffer_send_offset;
 	open_msg->target_vp = newchannel->target_vp;
 
-	if (userdatalen > MAX_USER_DEFINED_BYTES) {
-		err = -EINVAL;
-		goto error_free_gpadl;
-	}
-
 	if (userdatalen)
 		memcpy(open_msg->userdata, userdata, userdatalen);
 
@@ -195,18 +200,16 @@ int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size,
 
 	if (newchannel->rescind) {
 		err = -ENODEV;
-		goto error_free_gpadl;
+		goto error_free_info;
 	}
 
-	ret = vmbus_post_msg(open_msg,
+	err = vmbus_post_msg(open_msg,
 			     sizeof(struct vmbus_channel_open_channel), true);
 
-	trace_vmbus_open(open_msg, ret);
+	trace_vmbus_open(open_msg, err);
 
-	if (ret != 0) {
-		err = ret;
+	if (err != 0)
 		goto error_clean_msglist;
-	}
 
 	wait_for_completion(&open_info->waitevent);
 
@@ -216,12 +219,12 @@ int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size,
 
 	if (newchannel->rescind) {
 		err = -ENODEV;
-		goto error_free_gpadl;
+		goto error_free_info;
 	}
 
 	if (open_info->response.open_result.status) {
 		err = -EAGAIN;
-		goto error_free_gpadl;
+		goto error_free_info;
 	}
 
 	newchannel->state = CHANNEL_OPENED_STATE;
@@ -232,19 +235,50 @@ error_clean_msglist:
 	spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags);
 	list_del(&open_info->msglistentry);
 	spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags);
-
+error_free_info:
+	kfree(open_info);
 error_free_gpadl:
 	vmbus_teardown_gpadl(newchannel, newchannel->ringbuffer_gpadlhandle);
-	kfree(open_info);
-error_free_pages:
+	newchannel->ringbuffer_gpadlhandle = 0;
+error_clean_ring:
 	hv_ringbuffer_cleanup(&newchannel->outbound);
 	hv_ringbuffer_cleanup(&newchannel->inbound);
-	__free_pages(page,
-		     get_order(send_ringbuffer_size + recv_ringbuffer_size));
-error_set_chnstate:
 	newchannel->state = CHANNEL_OPEN_STATE;
 	return err;
 }
+
+/*
+ * vmbus_connect_ring - Open the channel but reuse ring buffer
+ */
+int vmbus_connect_ring(struct vmbus_channel *newchannel,
+		       void (*onchannelcallback)(void *context), void *context)
+{
+	return  __vmbus_open(newchannel, NULL, 0, onchannelcallback, context);
+}
+EXPORT_SYMBOL_GPL(vmbus_connect_ring);
+
+/*
+ * vmbus_open - Open the specified channel.
+ */
+int vmbus_open(struct vmbus_channel *newchannel,
+	       u32 send_ringbuffer_size, u32 recv_ringbuffer_size,
+	       void *userdata, u32 userdatalen,
+	       void (*onchannelcallback)(void *context), void *context)
+{
+	int err;
+
+	err = vmbus_alloc_ring(newchannel, send_ringbuffer_size,
+			       recv_ringbuffer_size);
+	if (err)
+		return err;
+
+	err = __vmbus_open(newchannel, userdata, userdatalen,
+			   onchannelcallback, context);
+	if (err)
+		vmbus_free_ring(newchannel);
+
+	return err;
+}
 EXPORT_SYMBOL_GPL(vmbus_open);
 
 /* Used for Hyper-V Socket: a guest client's connect() to the host */
@@ -612,10 +646,8 @@ static int vmbus_close_internal(struct vmbus_channel *channel)
 	 * in Hyper-V Manager), the driver's remove() invokes vmbus_close():
 	 * here we should skip most of the below cleanup work.
 	 */
-	if (channel->state != CHANNEL_OPENED_STATE) {
-		ret = -EINVAL;
-		goto out;
-	}
+	if (channel->state != CHANNEL_OPENED_STATE)
+		return -EINVAL;
 
 	channel->state = CHANNEL_OPEN_STATE;
 
@@ -637,11 +669,10 @@ static int vmbus_close_internal(struct vmbus_channel *channel)
 		 * If we failed to post the close msg,
 		 * it is perhaps better to leak memory.
 		 */
-		goto out;
 	}
 
 	/* Tear down the gpadl for the channel's ring buffer */
-	if (channel->ringbuffer_gpadlhandle) {
+	else if (channel->ringbuffer_gpadlhandle) {
 		ret = vmbus_teardown_gpadl(channel,
 					   channel->ringbuffer_gpadlhandle);
 		if (ret) {
@@ -650,74 +681,78 @@ static int vmbus_close_internal(struct vmbus_channel *channel)
 			 * If we failed to teardown gpadl,
 			 * it is perhaps better to leak memory.
 			 */
-			goto out;
 		}
-	}
-
-	/* Cleanup the ring buffers for this channel */
-	hv_ringbuffer_cleanup(&channel->outbound);
-	hv_ringbuffer_cleanup(&channel->inbound);
 
-	free_pages((unsigned long)channel->ringbuffer_pages,
-		get_order(channel->ringbuffer_pagecount * PAGE_SIZE));
+		channel->ringbuffer_gpadlhandle = 0;
+	}
 
-out:
 	return ret;
 }
 
-/*
- * vmbus_close - Close the specified channel
- */
-void vmbus_close(struct vmbus_channel *channel)
+/* disconnect ring - close all channels */
+int vmbus_disconnect_ring(struct vmbus_channel *channel)
 {
-	struct list_head *cur, *tmp;
-	struct vmbus_channel *cur_channel;
+	struct vmbus_channel *cur_channel, *tmp;
+	unsigned long flags;
+	LIST_HEAD(list);
+	int ret;
 
-	if (channel->primary_channel != NULL) {
-		/*
-		 * We will only close sub-channels when
-		 * the primary is closed.
-		 */
-		return;
-	}
-	/*
-	 * Close all the sub-channels first and then close the
-	 * primary channel.
-	 */
-	list_for_each_safe(cur, tmp, &channel->sc_list) {
-		cur_channel = list_entry(cur, struct vmbus_channel, sc_list);
-		if (cur_channel->rescind) {
+	if (channel->primary_channel != NULL)
+		return -EINVAL;
+
+	/* Snapshot the list of subchannels */
+	spin_lock_irqsave(&channel->lock, flags);
+	list_splice_init(&channel->sc_list, &list);
+	channel->num_sc = 0;
+	spin_unlock_irqrestore(&channel->lock, flags);
+
+	list_for_each_entry_safe(cur_channel, tmp, &list, sc_list) {
+		if (cur_channel->rescind)
 			wait_for_completion(&cur_channel->rescind_event);
-			mutex_lock(&vmbus_connection.channel_mutex);
-			vmbus_close_internal(cur_channel);
-			hv_process_channel_removal(
-					   cur_channel->offermsg.child_relid);
-		} else {
-			mutex_lock(&vmbus_connection.channel_mutex);
-			vmbus_close_internal(cur_channel);
+
+		mutex_lock(&vmbus_connection.channel_mutex);
+		if (vmbus_close_internal(cur_channel) == 0) {
+			vmbus_free_ring(cur_channel);
+
+			if (cur_channel->rescind)
+				hv_process_channel_removal(cur_channel);
 		}
 		mutex_unlock(&vmbus_connection.channel_mutex);
 	}
+
 	/*
 	 * Now close the primary.
 	 */
 	mutex_lock(&vmbus_connection.channel_mutex);
-	vmbus_close_internal(channel);
+	ret = vmbus_close_internal(channel);
 	mutex_unlock(&vmbus_connection.channel_mutex);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(vmbus_disconnect_ring);
+
+/*
+ * vmbus_close - Close the specified channel
+ */
+void vmbus_close(struct vmbus_channel *channel)
+{
+	if (vmbus_disconnect_ring(channel) == 0)
+		vmbus_free_ring(channel);
 }
 EXPORT_SYMBOL_GPL(vmbus_close);
 
 /**
  * vmbus_sendpacket() - Send the specified buffer on the given channel
- * @channel: Pointer to vmbus_channel structure.
- * @buffer: Pointer to the buffer you want to receive the data into.
- * @bufferlen: Maximum size of what the the buffer will hold
+ * @channel: Pointer to vmbus_channel structure
+ * @buffer: Pointer to the buffer you want to send the data from.
+ * @bufferlen: Maximum size of what the buffer holds.
  * @requestid: Identifier of the request
- * @type: Type of packet that is being send e.g. negotiate, time
- * packet etc.
+ * @type: Type of packet that is being sent e.g. negotiate, time
+ *	  packet etc.
+ * @flags: 0 or VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED
  *
- * Sends data in @buffer directly to hyper-v via the vmbus
- * This will send the data unparsed to hyper-v.
+ * Sends data in @buffer directly to Hyper-V via the vmbus.
+ * This will send the data unparsed to Hyper-V.
  *
  * Mainly used by Hyper-V drivers.
  */
@@ -850,12 +885,13 @@ int vmbus_sendpacket_mpb_desc(struct vmbus_channel *channel,
 EXPORT_SYMBOL_GPL(vmbus_sendpacket_mpb_desc);
 
 /**
- * vmbus_recvpacket() - Retrieve the user packet on the specified channel
- * @channel: Pointer to vmbus_channel structure.
+ * __vmbus_recvpacket() - Retrieve the user packet on the specified channel
+ * @channel: Pointer to vmbus_channel structure
  * @buffer: Pointer to the buffer you want to receive the data into.
- * @bufferlen: Maximum size of what the the buffer will hold
- * @buffer_actual_len: The actual size of the data after it was received
+ * @bufferlen: Maximum size of what the buffer can hold.
+ * @buffer_actual_len: The actual size of the data after it was received.
  * @requestid: Identifier of the request
+ * @raw: true means keep the vmpacket_descriptor header in the received data.
  *
  * Receives directly from the hyper-v vmbus and puts the data it received
  * into Buffer. This will receive the data unparsed from hyper-v.
diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c
index 0f0e091c117c..6277597d3d58 100644
--- a/drivers/hv/channel_mgmt.c
+++ b/drivers/hv/channel_mgmt.c
@@ -198,24 +198,19 @@ static u16 hv_get_dev_type(const struct vmbus_channel *channel)
 }
 
 /**
- * vmbus_prep_negotiate_resp() - Create default response for Hyper-V Negotiate message
+ * vmbus_prep_negotiate_resp() - Create default response for Negotiate message
  * @icmsghdrp: Pointer to msg header structure
- * @icmsg_negotiate: Pointer to negotiate message structure
  * @buf: Raw buffer channel data
+ * @fw_version: The framework versions we can support.
+ * @fw_vercnt: The size of @fw_version.
+ * @srv_version: The service versions we can support.
+ * @srv_vercnt: The size of @srv_version.
+ * @nego_fw_version: The selected framework version.
+ * @nego_srv_version: The selected service version.
  *
- * @icmsghdrp is of type &struct icmsg_hdr.
- * Set up and fill in default negotiate response message.
- *
- * The fw_version and fw_vercnt specifies the framework version that
- * we can support.
- *
- * The srv_version and srv_vercnt specifies the service
- * versions we can support.
- *
- * Versions are given in decreasing order.
- *
- * nego_fw_version and nego_srv_version store the selected protocol versions.
+ * Note: Versions are given in decreasing order.
  *
+ * Set up and fill in default negotiate response message.
  * Mainly used by Hyper-V drivers.
  */
 bool vmbus_prep_negotiate_resp(struct icmsg_hdr *icmsghdrp,
@@ -385,21 +380,14 @@ static void vmbus_release_relid(u32 relid)
 	trace_vmbus_release_relid(&msg, ret);
 }
 
-void hv_process_channel_removal(u32 relid)
+void hv_process_channel_removal(struct vmbus_channel *channel)
 {
+	struct vmbus_channel *primary_channel;
 	unsigned long flags;
-	struct vmbus_channel *primary_channel, *channel;
 
 	BUG_ON(!mutex_is_locked(&vmbus_connection.channel_mutex));
-
-	/*
-	 * Make sure channel is valid as we may have raced.
-	 */
-	channel = relid2channel(relid);
-	if (!channel)
-		return;
-
 	BUG_ON(!channel->rescind);
+
 	if (channel->target_cpu != get_cpu()) {
 		put_cpu();
 		smp_call_function_single(channel->target_cpu,
@@ -429,7 +417,7 @@ void hv_process_channel_removal(u32 relid)
 		cpumask_clear_cpu(channel->target_cpu,
 				  &primary_channel->alloced_cpus_in_node);
 
-	vmbus_release_relid(relid);
+	vmbus_release_relid(channel->offermsg.child_relid);
 
 	free_channel(channel);
 }
@@ -606,16 +594,18 @@ static void init_vp_index(struct vmbus_channel *channel, u16 dev_type)
 	bool perf_chn = vmbus_devs[dev_type].perf_device;
 	struct vmbus_channel *primary = channel->primary_channel;
 	int next_node;
-	struct cpumask available_mask;
+	cpumask_var_t available_mask;
 	struct cpumask *alloced_mask;
 
 	if ((vmbus_proto_version == VERSION_WS2008) ||
-	    (vmbus_proto_version == VERSION_WIN7) || (!perf_chn)) {
+	    (vmbus_proto_version == VERSION_WIN7) || (!perf_chn) ||
+	    !alloc_cpumask_var(&available_mask, GFP_KERNEL)) {
 		/*
 		 * Prior to win8, all channel interrupts are
 		 * delivered on cpu 0.
 		 * Also if the channel is not a performance critical
 		 * channel, bind it to cpu 0.
+		 * In case alloc_cpumask_var() fails, bind it to cpu 0.
 		 */
 		channel->numa_node = 0;
 		channel->target_cpu = 0;
@@ -653,7 +643,7 @@ static void init_vp_index(struct vmbus_channel *channel, u16 dev_type)
 		cpumask_clear(alloced_mask);
 	}
 
-	cpumask_xor(&available_mask, alloced_mask,
+	cpumask_xor(available_mask, alloced_mask,
 		    cpumask_of_node(primary->numa_node));
 
 	cur_cpu = -1;
@@ -671,10 +661,10 @@ static void init_vp_index(struct vmbus_channel *channel, u16 dev_type)
 	}
 
 	while (true) {
-		cur_cpu = cpumask_next(cur_cpu, &available_mask);
+		cur_cpu = cpumask_next(cur_cpu, available_mask);
 		if (cur_cpu >= nr_cpu_ids) {
 			cur_cpu = -1;
-			cpumask_copy(&available_mask,
+			cpumask_copy(available_mask,
 				     cpumask_of_node(primary->numa_node));
 			continue;
 		}
@@ -704,6 +694,8 @@ static void init_vp_index(struct vmbus_channel *channel, u16 dev_type)
 
 	channel->target_cpu = cur_cpu;
 	channel->target_vp = hv_cpu_number_to_vp_number(cur_cpu);
+
+	free_cpumask_var(available_mask);
 }
 
 static void vmbus_wait_for_unload(void)
@@ -943,7 +935,7 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr)
 			 * The channel is currently not open;
 			 * it is safe for us to cleanup the channel.
 			 */
-			hv_process_channel_removal(rescind->child_relid);
+			hv_process_channel_removal(channel);
 		} else {
 			complete(&channel->rescind_event);
 		}
diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index 748a1c4172a6..332d7c34be5c 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -189,6 +189,17 @@ static void hv_init_clockevent_device(struct clock_event_device *dev, int cpu)
 int hv_synic_alloc(void)
 {
 	int cpu;
+	struct hv_per_cpu_context *hv_cpu;
+
+	/*
+	 * First, zero all per-cpu memory areas so hv_synic_free() can
+	 * detect what memory has been allocated and cleanup properly
+	 * after any failures.
+	 */
+	for_each_present_cpu(cpu) {
+		hv_cpu = per_cpu_ptr(hv_context.cpu_context, cpu);
+		memset(hv_cpu, 0, sizeof(*hv_cpu));
+	}
 
 	hv_context.hv_numa_map = kcalloc(nr_node_ids, sizeof(struct cpumask),
 					 GFP_KERNEL);
@@ -198,10 +209,8 @@ int hv_synic_alloc(void)
 	}
 
 	for_each_present_cpu(cpu) {
-		struct hv_per_cpu_context *hv_cpu
-			= per_cpu_ptr(hv_context.cpu_context, cpu);
+		hv_cpu = per_cpu_ptr(hv_context.cpu_context, cpu);
 
-		memset(hv_cpu, 0, sizeof(*hv_cpu));
 		tasklet_init(&hv_cpu->msg_dpc,
 			     vmbus_on_msg_dpc, (unsigned long) hv_cpu);
 
diff --git a/drivers/hv/hv_balloon.c b/drivers/hv/hv_balloon.c
index b1b788082793..41631512ae97 100644
--- a/drivers/hv/hv_balloon.c
+++ b/drivers/hv/hv_balloon.c
@@ -689,7 +689,7 @@ static void hv_page_online_one(struct hv_hotadd_state *has, struct page *pg)
 	__online_page_increment_counters(pg);
 	__online_page_free(pg);
 
-	WARN_ON_ONCE(!spin_is_locked(&dm_device.ha_lock));
+	lockdep_assert_held(&dm_device.ha_lock);
 	dm_device.num_pages_onlined++;
 }
 
diff --git a/drivers/hv/hv_kvp.c b/drivers/hv/hv_kvp.c
index 5eed1e7da15c..a7513a8a8e37 100644
--- a/drivers/hv/hv_kvp.c
+++ b/drivers/hv/hv_kvp.c
@@ -353,7 +353,6 @@ static void process_ib_ipinfo(void *in_msg, void *out_msg, int op)
 
 		out->body.kvp_ip_val.dhcp_enabled = in->kvp_ip_val.dhcp_enabled;
 
-	default:
 		utf16s_to_utf8s((wchar_t *)in->kvp_ip_val.adapter_id,
 				MAX_ADAPTER_ID_SIZE,
 				UTF16_LITTLE_ENDIAN,
@@ -406,7 +405,7 @@ kvp_send_key(struct work_struct *dummy)
 		process_ib_ipinfo(in_msg, message, KVP_OP_SET_IP_INFO);
 		break;
 	case KVP_OP_GET_IP_INFO:
-		process_ib_ipinfo(in_msg, message, KVP_OP_GET_IP_INFO);
+		/* We only need to pass on message->kvp_hdr.operation.  */
 		break;
 	case KVP_OP_SET:
 		switch (in_msg->body.kvp_set.data.value_type) {
@@ -421,7 +420,7 @@ kvp_send_key(struct work_struct *dummy)
 				UTF16_LITTLE_ENDIAN,
 				message->body.kvp_set.data.value,
 				HV_KVP_EXCHANGE_MAX_VALUE_SIZE - 1) + 1;
-				break;
+			break;
 
 		case REG_U32:
 			/*
@@ -446,6 +445,9 @@ kvp_send_key(struct work_struct *dummy)
 			break;
 
 		}
+
+		break;
+
 	case KVP_OP_GET:
 		message->body.kvp_set.data.key_size =
 			utf16s_to_utf8s(
@@ -454,7 +456,7 @@ kvp_send_key(struct work_struct *dummy)
 			UTF16_LITTLE_ENDIAN,
 			message->body.kvp_set.data.key,
 			HV_KVP_EXCHANGE_MAX_KEY_SIZE - 1) + 1;
-			break;
+		break;
 
 	case KVP_OP_DELETE:
 		message->body.kvp_delete.key_size =
@@ -464,12 +466,12 @@ kvp_send_key(struct work_struct *dummy)
 			UTF16_LITTLE_ENDIAN,
 			message->body.kvp_delete.key,
 			HV_KVP_EXCHANGE_MAX_KEY_SIZE - 1) + 1;
-			break;
+		break;
 
 	case KVP_OP_ENUMERATE:
 		message->body.kvp_enum_data.index =
 			in_msg->body.kvp_enum_data.index;
-			break;
+		break;
 	}
 
 	kvp_transaction.state = HVUTIL_USERSPACE_REQ;
diff --git a/drivers/hv/ring_buffer.c b/drivers/hv/ring_buffer.c
index 3e90eb91db45..64d0c85d5161 100644
--- a/drivers/hv/ring_buffer.c
+++ b/drivers/hv/ring_buffer.c
@@ -241,6 +241,7 @@ int hv_ringbuffer_init(struct hv_ring_buffer_info *ring_info,
 void hv_ringbuffer_cleanup(struct hv_ring_buffer_info *ring_info)
 {
 	vunmap(ring_info->ring_buffer);
+	ring_info->ring_buffer = NULL;
 }
 
 /* Write to the ring buffer. */
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
index c71cc857b649..283d184280af 100644
--- a/drivers/hv/vmbus_drv.c
+++ b/drivers/hv/vmbus_drv.c
@@ -498,6 +498,54 @@ static ssize_t device_show(struct device *dev,
 }
 static DEVICE_ATTR_RO(device);
 
+static ssize_t driver_override_store(struct device *dev,
+				     struct device_attribute *attr,
+				     const char *buf, size_t count)
+{
+	struct hv_device *hv_dev = device_to_hv_device(dev);
+	char *driver_override, *old, *cp;
+
+	/* We need to keep extra room for a newline */
+	if (count >= (PAGE_SIZE - 1))
+		return -EINVAL;
+
+	driver_override = kstrndup(buf, count, GFP_KERNEL);
+	if (!driver_override)
+		return -ENOMEM;
+
+	cp = strchr(driver_override, '\n');
+	if (cp)
+		*cp = '\0';
+
+	device_lock(dev);
+	old = hv_dev->driver_override;
+	if (strlen(driver_override)) {
+		hv_dev->driver_override = driver_override;
+	} else {
+		kfree(driver_override);
+		hv_dev->driver_override = NULL;
+	}
+	device_unlock(dev);
+
+	kfree(old);
+
+	return count;
+}
+
+static ssize_t driver_override_show(struct device *dev,
+				    struct device_attribute *attr, char *buf)
+{
+	struct hv_device *hv_dev = device_to_hv_device(dev);
+	ssize_t len;
+
+	device_lock(dev);
+	len = snprintf(buf, PAGE_SIZE, "%s\n", hv_dev->driver_override);
+	device_unlock(dev);
+
+	return len;
+}
+static DEVICE_ATTR_RW(driver_override);
+
 /* Set up per device attributes in /sys/bus/vmbus/devices/<bus device> */
 static struct attribute *vmbus_dev_attrs[] = {
 	&dev_attr_id.attr,
@@ -528,6 +576,7 @@ static struct attribute *vmbus_dev_attrs[] = {
 	&dev_attr_channel_vp_mapping.attr,
 	&dev_attr_vendor.attr,
 	&dev_attr_device.attr,
+	&dev_attr_driver_override.attr,
 	NULL,
 };
 ATTRIBUTE_GROUPS(vmbus_dev);
@@ -563,17 +612,26 @@ static inline bool is_null_guid(const uuid_le *guid)
 	return true;
 }
 
-/*
- * Return a matching hv_vmbus_device_id pointer.
- * If there is no match, return NULL.
- */
-static const struct hv_vmbus_device_id *hv_vmbus_get_id(struct hv_driver *drv,
-					const uuid_le *guid)
+static const struct hv_vmbus_device_id *
+hv_vmbus_dev_match(const struct hv_vmbus_device_id *id, const uuid_le *guid)
+
+{
+	if (id == NULL)
+		return NULL; /* empty device table */
+
+	for (; !is_null_guid(&id->guid); id++)
+		if (!uuid_le_cmp(id->guid, *guid))
+			return id;
+
+	return NULL;
+}
+
+static const struct hv_vmbus_device_id *
+hv_vmbus_dynid_match(struct hv_driver *drv, const uuid_le *guid)
 {
 	const struct hv_vmbus_device_id *id = NULL;
 	struct vmbus_dynid *dynid;
 
-	/* Look at the dynamic ids first, before the static ones */
 	spin_lock(&drv->dynids.lock);
 	list_for_each_entry(dynid, &drv->dynids.list, node) {
 		if (!uuid_le_cmp(dynid->id.guid, *guid)) {
@@ -583,18 +641,37 @@ static const struct hv_vmbus_device_id *hv_vmbus_get_id(struct hv_driver *drv,
 	}
 	spin_unlock(&drv->dynids.lock);
 
-	if (id)
-		return id;
+	return id;
+}
 
-	id = drv->id_table;
-	if (id == NULL)
-		return NULL; /* empty device table */
+static const struct hv_vmbus_device_id vmbus_device_null = {
+	.guid = NULL_UUID_LE,
+};
 
-	for (; !is_null_guid(&id->guid); id++)
-		if (!uuid_le_cmp(id->guid, *guid))
-			return id;
+/*
+ * Return a matching hv_vmbus_device_id pointer.
+ * If there is no match, return NULL.
+ */
+static const struct hv_vmbus_device_id *hv_vmbus_get_id(struct hv_driver *drv,
+							struct hv_device *dev)
+{
+	const uuid_le *guid = &dev->dev_type;
+	const struct hv_vmbus_device_id *id;
 
-	return NULL;
+	/* When driver_override is set, only bind to the matching driver */
+	if (dev->driver_override && strcmp(dev->driver_override, drv->name))
+		return NULL;
+
+	/* Look at the dynamic ids first, before the static ones */
+	id = hv_vmbus_dynid_match(drv, guid);
+	if (!id)
+		id = hv_vmbus_dev_match(drv->id_table, guid);
+
+	/* driver_override will always match, send a dummy id */
+	if (!id && dev->driver_override)
+		id = &vmbus_device_null;
+
+	return id;
 }
 
 /* vmbus_add_dynid - add a new device ID to this driver and re-probe devices */
@@ -643,7 +720,7 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf,
 	if (retval)
 		return retval;
 
-	if (hv_vmbus_get_id(drv, &guid))
+	if (hv_vmbus_dynid_match(drv, &guid))
 		return -EEXIST;
 
 	retval = vmbus_add_dynid(drv, &guid);
@@ -708,7 +785,7 @@ static int vmbus_match(struct device *device, struct device_driver *driver)
 	if (is_hvsock_channel(hv_dev->channel))
 		return drv->hvsock;
 
-	if (hv_vmbus_get_id(drv, &hv_dev->dev_type))
+	if (hv_vmbus_get_id(drv, hv_dev))
 		return 1;
 
 	return 0;
@@ -725,7 +802,7 @@ static int vmbus_probe(struct device *child_device)
 	struct hv_device *dev = device_to_hv_device(child_device);
 	const struct hv_vmbus_device_id *dev_id;
 
-	dev_id = hv_vmbus_get_id(drv, &dev->dev_type);
+	dev_id = hv_vmbus_get_id(drv, dev);
 	if (drv->probe) {
 		ret = drv->probe(dev, dev_id);
 		if (ret != 0)
@@ -787,10 +864,9 @@ static void vmbus_device_release(struct device *device)
 	struct vmbus_channel *channel = hv_dev->channel;
 
 	mutex_lock(&vmbus_connection.channel_mutex);
-	hv_process_channel_removal(channel->offermsg.child_relid);
+	hv_process_channel_removal(channel);
 	mutex_unlock(&vmbus_connection.channel_mutex);
 	kfree(hv_dev);
-
 }
 
 /* The one and only one */
diff --git a/drivers/hwtracing/coresight/coresight-catu.c b/drivers/hwtracing/coresight/coresight-catu.c
index ff94e58845b7..170fbb66bda2 100644
--- a/drivers/hwtracing/coresight/coresight-catu.c
+++ b/drivers/hwtracing/coresight/coresight-catu.c
@@ -406,6 +406,7 @@ static inline int catu_wait_for_ready(struct catu_drvdata *drvdata)
 
 static int catu_enable_hw(struct catu_drvdata *drvdata, void *data)
 {
+	int rc;
 	u32 control, mode;
 	struct etr_buf *etr_buf = data;
 
@@ -418,6 +419,10 @@ static int catu_enable_hw(struct catu_drvdata *drvdata, void *data)
 		return -EBUSY;
 	}
 
+	rc = coresight_claim_device_unlocked(drvdata->base);
+	if (rc)
+		return rc;
+
 	control |= BIT(CATU_CONTROL_ENABLE);
 
 	if (etr_buf && etr_buf->mode == ETR_MODE_CATU) {
@@ -459,6 +464,7 @@ static int catu_disable_hw(struct catu_drvdata *drvdata)
 	int rc = 0;
 
 	catu_write_control(drvdata, 0);
+	coresight_disclaim_device_unlocked(drvdata->base);
 	if (catu_wait_for_ready(drvdata)) {
 		dev_info(drvdata->dev, "Timeout while waiting for READY\n");
 		rc = -EAGAIN;
diff --git a/drivers/hwtracing/coresight/coresight-dynamic-replicator.c b/drivers/hwtracing/coresight/coresight-dynamic-replicator.c
index f6d0571ab9dd..299667b887fc 100644
--- a/drivers/hwtracing/coresight/coresight-dynamic-replicator.c
+++ b/drivers/hwtracing/coresight/coresight-dynamic-replicator.c
@@ -34,48 +34,87 @@ struct replicator_state {
 	struct coresight_device	*csdev;
 };
 
+/*
+ * replicator_reset : Reset the replicator configuration to sane values.
+ */
+static void replicator_reset(struct replicator_state *drvdata)
+{
+	CS_UNLOCK(drvdata->base);
+
+	if (!coresight_claim_device_unlocked(drvdata->base)) {
+		writel_relaxed(0xff, drvdata->base + REPLICATOR_IDFILTER0);
+		writel_relaxed(0xff, drvdata->base + REPLICATOR_IDFILTER1);
+		coresight_disclaim_device_unlocked(drvdata->base);
+	}
+
+	CS_LOCK(drvdata->base);
+}
+
 static int replicator_enable(struct coresight_device *csdev, int inport,
 			      int outport)
 {
+	int rc = 0;
+	u32 reg;
 	struct replicator_state *drvdata = dev_get_drvdata(csdev->dev.parent);
 
+	switch (outport) {
+	case 0:
+		reg = REPLICATOR_IDFILTER0;
+		break;
+	case 1:
+		reg = REPLICATOR_IDFILTER1;
+		break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+
 	CS_UNLOCK(drvdata->base);
 
-	/*
-	 * Ensure that the other port is disabled
-	 * 0x00 - passing through the replicator unimpeded
-	 * 0xff - disable (or impede) the flow of ATB data
-	 */
-	if (outport == 0) {
-		writel_relaxed(0x00, drvdata->base + REPLICATOR_IDFILTER0);
-		writel_relaxed(0xff, drvdata->base + REPLICATOR_IDFILTER1);
-	} else {
-		writel_relaxed(0x00, drvdata->base + REPLICATOR_IDFILTER1);
-		writel_relaxed(0xff, drvdata->base + REPLICATOR_IDFILTER0);
+	if ((readl_relaxed(drvdata->base + REPLICATOR_IDFILTER0) == 0xff) &&
+	    (readl_relaxed(drvdata->base + REPLICATOR_IDFILTER1) == 0xff))
+		rc = coresight_claim_device_unlocked(drvdata->base);
+
+	/* Ensure that the outport is enabled. */
+	if (!rc) {
+		writel_relaxed(0x00, drvdata->base + reg);
+		dev_dbg(drvdata->dev, "REPLICATOR enabled\n");
 	}
 
 	CS_LOCK(drvdata->base);
 
-	dev_info(drvdata->dev, "REPLICATOR enabled\n");
-	return 0;
+	return rc;
 }
 
 static void replicator_disable(struct coresight_device *csdev, int inport,
 				int outport)
 {
+	u32 reg;
 	struct replicator_state *drvdata = dev_get_drvdata(csdev->dev.parent);
 
+	switch (outport) {
+	case 0:
+		reg = REPLICATOR_IDFILTER0;
+		break;
+	case 1:
+		reg = REPLICATOR_IDFILTER1;
+		break;
+	default:
+		WARN_ON(1);
+		return;
+	}
+
 	CS_UNLOCK(drvdata->base);
 
 	/* disable the flow of ATB data through port */
-	if (outport == 0)
-		writel_relaxed(0xff, drvdata->base + REPLICATOR_IDFILTER0);
-	else
-		writel_relaxed(0xff, drvdata->base + REPLICATOR_IDFILTER1);
+	writel_relaxed(0xff, drvdata->base + reg);
 
+	if ((readl_relaxed(drvdata->base + REPLICATOR_IDFILTER0) == 0xff) &&
+	    (readl_relaxed(drvdata->base + REPLICATOR_IDFILTER1) == 0xff))
+		coresight_disclaim_device_unlocked(drvdata->base);
 	CS_LOCK(drvdata->base);
 
-	dev_info(drvdata->dev, "REPLICATOR disabled\n");
+	dev_dbg(drvdata->dev, "REPLICATOR disabled\n");
 }
 
 static const struct coresight_ops_link replicator_link_ops = {
@@ -156,7 +195,11 @@ static int replicator_probe(struct amba_device *adev, const struct amba_id *id)
 	desc.groups = replicator_groups;
 	drvdata->csdev = coresight_register(&desc);
 
-	return PTR_ERR_OR_ZERO(drvdata->csdev);
+	if (!IS_ERR(drvdata->csdev)) {
+		replicator_reset(drvdata);
+		return 0;
+	}
+	return PTR_ERR(drvdata->csdev);
 }
 
 #ifdef CONFIG_PM
diff --git a/drivers/hwtracing/coresight/coresight-etb10.c b/drivers/hwtracing/coresight/coresight-etb10.c
index 306119eaf16a..824be0c5f592 100644
--- a/drivers/hwtracing/coresight/coresight-etb10.c
+++ b/drivers/hwtracing/coresight/coresight-etb10.c
@@ -5,7 +5,6 @@
  * Description: CoreSight Embedded Trace Buffer driver
  */
 
-#include <asm/local.h>
 #include <linux/kernel.h>
 #include <linux/init.h>
 #include <linux/types.h>
@@ -28,6 +27,7 @@
 
 
 #include "coresight-priv.h"
+#include "coresight-etm-perf.h"
 
 #define ETB_RAM_DEPTH_REG	0x004
 #define ETB_STATUS_REG		0x00c
@@ -71,8 +71,8 @@
  * @miscdev:	specifics to handle "/dev/xyz.etb" entry.
  * @spinlock:	only one at a time pls.
  * @reading:	synchronise user space access to etb buffer.
- * @mode:	this ETB is being used.
  * @buf:	area of memory where ETB buffer content gets sent.
+ * @mode:	this ETB is being used.
  * @buffer_depth: size of @buf.
  * @trigger_cntr: amount of words to store after a trigger.
  */
@@ -84,12 +84,15 @@ struct etb_drvdata {
 	struct miscdevice	miscdev;
 	spinlock_t		spinlock;
 	local_t			reading;
-	local_t			mode;
 	u8			*buf;
+	u32			mode;
 	u32			buffer_depth;
 	u32			trigger_cntr;
 };
 
+static int etb_set_buffer(struct coresight_device *csdev,
+			  struct perf_output_handle *handle);
+
 static unsigned int etb_get_buffer_depth(struct etb_drvdata *drvdata)
 {
 	u32 depth = 0;
@@ -103,7 +106,7 @@ static unsigned int etb_get_buffer_depth(struct etb_drvdata *drvdata)
 	return depth;
 }
 
-static void etb_enable_hw(struct etb_drvdata *drvdata)
+static void __etb_enable_hw(struct etb_drvdata *drvdata)
 {
 	int i;
 	u32 depth;
@@ -131,32 +134,92 @@ static void etb_enable_hw(struct etb_drvdata *drvdata)
 	CS_LOCK(drvdata->base);
 }
 
-static int etb_enable(struct coresight_device *csdev, u32 mode)
+static int etb_enable_hw(struct etb_drvdata *drvdata)
+{
+	__etb_enable_hw(drvdata);
+	return 0;
+}
+
+static int etb_enable_sysfs(struct coresight_device *csdev)
 {
-	u32 val;
+	int ret = 0;
 	unsigned long flags;
 	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
-	val = local_cmpxchg(&drvdata->mode,
-			    CS_MODE_DISABLED, mode);
-	/*
-	 * When accessing from Perf, a HW buffer can be handled
-	 * by a single trace entity.  In sysFS mode many tracers
-	 * can be logging to the same HW buffer.
-	 */
-	if (val == CS_MODE_PERF)
-		return -EBUSY;
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+
+	/* Don't messup with perf sessions. */
+	if (drvdata->mode == CS_MODE_PERF) {
+		ret = -EBUSY;
+		goto out;
+	}
 
 	/* Nothing to do, the tracer is already enabled. */
-	if (val == CS_MODE_SYSFS)
+	if (drvdata->mode == CS_MODE_SYSFS)
 		goto out;
 
-	spin_lock_irqsave(&drvdata->spinlock, flags);
-	etb_enable_hw(drvdata);
+	ret = etb_enable_hw(drvdata);
+	if (!ret)
+		drvdata->mode = CS_MODE_SYSFS;
+
+out:
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	return ret;
+}
+
+static int etb_enable_perf(struct coresight_device *csdev, void *data)
+{
+	int ret = 0;
+	unsigned long flags;
+	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+
+	/* No need to continue if the component is already in use. */
+	if (drvdata->mode != CS_MODE_DISABLED) {
+		ret = -EBUSY;
+		goto out;
+	}
+
+	/*
+	 * We don't have an internal state to clean up if we fail to setup
+	 * the perf buffer. So we can perform the step before we turn the
+	 * ETB on and leave without cleaning up.
+	 */
+	ret = etb_set_buffer(csdev, (struct perf_output_handle *)data);
+	if (ret)
+		goto out;
+
+	ret = etb_enable_hw(drvdata);
+	if (!ret)
+		drvdata->mode = CS_MODE_PERF;
 
 out:
-	dev_info(drvdata->dev, "ETB enabled\n");
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	return ret;
+}
+
+static int etb_enable(struct coresight_device *csdev, u32 mode, void *data)
+{
+	int ret;
+	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	switch (mode) {
+	case CS_MODE_SYSFS:
+		ret = etb_enable_sysfs(csdev);
+		break;
+	case CS_MODE_PERF:
+		ret = etb_enable_perf(csdev, data);
+		break;
+	default:
+		ret = -EINVAL;
+		break;
+	}
+
+	if (ret)
+		return ret;
+
+	dev_dbg(drvdata->dev, "ETB enabled\n");
 	return 0;
 }
 
@@ -256,13 +319,16 @@ static void etb_disable(struct coresight_device *csdev)
 	unsigned long flags;
 
 	spin_lock_irqsave(&drvdata->spinlock, flags);
-	etb_disable_hw(drvdata);
-	etb_dump_hw(drvdata);
-	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
-	local_set(&drvdata->mode, CS_MODE_DISABLED);
+	/* Disable the ETB only if it needs to */
+	if (drvdata->mode != CS_MODE_DISABLED) {
+		etb_disable_hw(drvdata);
+		etb_dump_hw(drvdata);
+		drvdata->mode = CS_MODE_DISABLED;
+	}
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
-	dev_info(drvdata->dev, "ETB disabled\n");
+	dev_dbg(drvdata->dev, "ETB disabled\n");
 }
 
 static void *etb_alloc_buffer(struct coresight_device *csdev, int cpu,
@@ -294,12 +360,14 @@ static void etb_free_buffer(void *config)
 }
 
 static int etb_set_buffer(struct coresight_device *csdev,
-			  struct perf_output_handle *handle,
-			  void *sink_config)
+			  struct perf_output_handle *handle)
 {
 	int ret = 0;
 	unsigned long head;
-	struct cs_buffers *buf = sink_config;
+	struct cs_buffers *buf = etm_perf_sink_config(handle);
+
+	if (!buf)
+		return -EINVAL;
 
 	/* wrap head around to the amount of space we have */
 	head = handle->head & ((buf->nr_pages << PAGE_SHIFT) - 1);
@@ -315,37 +383,7 @@ static int etb_set_buffer(struct coresight_device *csdev,
 	return ret;
 }
 
-static unsigned long etb_reset_buffer(struct coresight_device *csdev,
-				      struct perf_output_handle *handle,
-				      void *sink_config)
-{
-	unsigned long size = 0;
-	struct cs_buffers *buf = sink_config;
-
-	if (buf) {
-		/*
-		 * In snapshot mode ->data_size holds the new address of the
-		 * ring buffer's head.  The size itself is the whole address
-		 * range since we want the latest information.
-		 */
-		if (buf->snapshot)
-			handle->head = local_xchg(&buf->data_size,
-						  buf->nr_pages << PAGE_SHIFT);
-
-		/*
-		 * Tell the tracer PMU how much we got in this run and if
-		 * something went wrong along the way.  Nobody else can use
-		 * this cs_buffers instance until we are done.  As such
-		 * resetting parameters here and squaring off with the ring
-		 * buffer API in the tracer PMU is fine.
-		 */
-		size = local_xchg(&buf->data_size, 0);
-	}
-
-	return size;
-}
-
-static void etb_update_buffer(struct coresight_device *csdev,
+static unsigned long etb_update_buffer(struct coresight_device *csdev,
 			      struct perf_output_handle *handle,
 			      void *sink_config)
 {
@@ -354,13 +392,13 @@ static void etb_update_buffer(struct coresight_device *csdev,
 	u8 *buf_ptr;
 	const u32 *barrier;
 	u32 read_ptr, write_ptr, capacity;
-	u32 status, read_data, to_read;
-	unsigned long offset;
+	u32 status, read_data;
+	unsigned long offset, to_read;
 	struct cs_buffers *buf = sink_config;
 	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
 	if (!buf)
-		return;
+		return 0;
 
 	capacity = drvdata->buffer_depth * ETB_FRAME_SIZE_WORDS;
 
@@ -465,18 +503,17 @@ static void etb_update_buffer(struct coresight_device *csdev,
 	writel_relaxed(0x0, drvdata->base + ETB_RAM_WRITE_POINTER);
 
 	/*
-	 * In snapshot mode all we have to do is communicate to
-	 * perf_aux_output_end() the address of the current head.  In full
-	 * trace mode the same function expects a size to move rb->aux_head
-	 * forward.
+	 * In snapshot mode we have to update the handle->head to point
+	 * to the new location.
 	 */
-	if (buf->snapshot)
-		local_set(&buf->data_size, (cur * PAGE_SIZE) + offset);
-	else
-		local_add(to_read, &buf->data_size);
-
+	if (buf->snapshot) {
+		handle->head = (cur * PAGE_SIZE) + offset;
+		to_read = buf->nr_pages << PAGE_SHIFT;
+	}
 	etb_enable_hw(drvdata);
 	CS_LOCK(drvdata->base);
+
+	return to_read;
 }
 
 static const struct coresight_ops_sink etb_sink_ops = {
@@ -484,8 +521,6 @@ static const struct coresight_ops_sink etb_sink_ops = {
 	.disable	= etb_disable,
 	.alloc_buffer	= etb_alloc_buffer,
 	.free_buffer	= etb_free_buffer,
-	.set_buffer	= etb_set_buffer,
-	.reset_buffer	= etb_reset_buffer,
 	.update_buffer	= etb_update_buffer,
 };
 
@@ -498,14 +533,14 @@ static void etb_dump(struct etb_drvdata *drvdata)
 	unsigned long flags;
 
 	spin_lock_irqsave(&drvdata->spinlock, flags);
-	if (local_read(&drvdata->mode) == CS_MODE_SYSFS) {
+	if (drvdata->mode == CS_MODE_SYSFS) {
 		etb_disable_hw(drvdata);
 		etb_dump_hw(drvdata);
 		etb_enable_hw(drvdata);
 	}
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
-	dev_info(drvdata->dev, "ETB dumped\n");
+	dev_dbg(drvdata->dev, "ETB dumped\n");
 }
 
 static int etb_open(struct inode *inode, struct file *file)
diff --git a/drivers/hwtracing/coresight/coresight-etm-perf.c b/drivers/hwtracing/coresight/coresight-etm-perf.c
index 677695635211..abe8249b893b 100644
--- a/drivers/hwtracing/coresight/coresight-etm-perf.c
+++ b/drivers/hwtracing/coresight/coresight-etm-perf.c
@@ -12,6 +12,7 @@
 #include <linux/mm.h>
 #include <linux/init.h>
 #include <linux/perf_event.h>
+#include <linux/percpu-defs.h>
 #include <linux/slab.h>
 #include <linux/types.h>
 #include <linux/workqueue.h>
@@ -22,20 +23,6 @@
 static struct pmu etm_pmu;
 static bool etm_perf_up;
 
-/**
- * struct etm_event_data - Coresight specifics associated to an event
- * @work:		Handle to free allocated memory outside IRQ context.
- * @mask:		Hold the CPU(s) this event was set for.
- * @snk_config:		The sink configuration.
- * @path:		An array of path, each slot for one CPU.
- */
-struct etm_event_data {
-	struct work_struct work;
-	cpumask_t mask;
-	void *snk_config;
-	struct list_head **path;
-};
-
 static DEFINE_PER_CPU(struct perf_output_handle, ctx_handle);
 static DEFINE_PER_CPU(struct coresight_device *, csdev_src);
 
@@ -61,6 +48,18 @@ static const struct attribute_group *etm_pmu_attr_groups[] = {
 	NULL,
 };
 
+static inline struct list_head **
+etm_event_cpu_path_ptr(struct etm_event_data *data, int cpu)
+{
+	return per_cpu_ptr(data->path, cpu);
+}
+
+static inline struct list_head *
+etm_event_cpu_path(struct etm_event_data *data, int cpu)
+{
+	return *etm_event_cpu_path_ptr(data, cpu);
+}
+
 static void etm_event_read(struct perf_event *event) {}
 
 static int etm_addr_filters_alloc(struct perf_event *event)
@@ -114,29 +113,30 @@ static void free_event_data(struct work_struct *work)
 
 	event_data = container_of(work, struct etm_event_data, work);
 	mask = &event_data->mask;
-	/*
-	 * First deal with the sink configuration.  See comment in
-	 * etm_setup_aux() about why we take the first available path.
-	 */
-	if (event_data->snk_config) {
+
+	/* Free the sink buffers, if there are any */
+	if (event_data->snk_config && !WARN_ON(cpumask_empty(mask))) {
 		cpu = cpumask_first(mask);
-		sink = coresight_get_sink(event_data->path[cpu]);
+		sink = coresight_get_sink(etm_event_cpu_path(event_data, cpu));
 		if (sink_ops(sink)->free_buffer)
 			sink_ops(sink)->free_buffer(event_data->snk_config);
 	}
 
 	for_each_cpu(cpu, mask) {
-		if (!(IS_ERR_OR_NULL(event_data->path[cpu])))
-			coresight_release_path(event_data->path[cpu]);
+		struct list_head **ppath;
+
+		ppath = etm_event_cpu_path_ptr(event_data, cpu);
+		if (!(IS_ERR_OR_NULL(*ppath)))
+			coresight_release_path(*ppath);
+		*ppath = NULL;
 	}
 
-	kfree(event_data->path);
+	free_percpu(event_data->path);
 	kfree(event_data);
 }
 
 static void *alloc_event_data(int cpu)
 {
-	int size;
 	cpumask_t *mask;
 	struct etm_event_data *event_data;
 
@@ -145,16 +145,12 @@ static void *alloc_event_data(int cpu)
 	if (!event_data)
 		return NULL;
 
-	/* Make sure nothing disappears under us */
-	get_online_cpus();
-	size = num_online_cpus();
 
 	mask = &event_data->mask;
 	if (cpu != -1)
 		cpumask_set_cpu(cpu, mask);
 	else
-		cpumask_copy(mask, cpu_online_mask);
-	put_online_cpus();
+		cpumask_copy(mask, cpu_present_mask);
 
 	/*
 	 * Each CPU has a single path between source and destination.  As such
@@ -164,8 +160,8 @@ static void *alloc_event_data(int cpu)
 	 * unused memory when dealing with single CPU trace scenarios is small
 	 * compared to the cost of searching through an optimized array.
 	 */
-	event_data->path = kcalloc(size,
-				   sizeof(struct list_head *), GFP_KERNEL);
+	event_data->path = alloc_percpu(struct list_head *);
+
 	if (!event_data->path) {
 		kfree(event_data);
 		return NULL;
@@ -206,34 +202,53 @@ static void *etm_setup_aux(int event_cpu, void **pages,
 	 * on the cmd line.  As such the "enable_sink" flag in sysFS is reset.
 	 */
 	sink = coresight_get_enabled_sink(true);
-	if (!sink)
+	if (!sink || !sink_ops(sink)->alloc_buffer)
 		goto err;
 
 	mask = &event_data->mask;
 
-	/* Setup the path for each CPU in a trace session */
+	/*
+	 * Setup the path for each CPU in a trace session. We try to build
+	 * trace path for each CPU in the mask. If we don't find an ETM
+	 * for the CPU or fail to build a path, we clear the CPU from the
+	 * mask and continue with the rest. If ever we try to trace on those
+	 * CPUs, we can handle it and fail the session.
+	 */
 	for_each_cpu(cpu, mask) {
+		struct list_head *path;
 		struct coresight_device *csdev;
 
 		csdev = per_cpu(csdev_src, cpu);
-		if (!csdev)
-			goto err;
+		/*
+		 * If there is no ETM associated with this CPU clear it from
+		 * the mask and continue with the rest. If ever we try to trace
+		 * on this CPU, we handle it accordingly.
+		 */
+		if (!csdev) {
+			cpumask_clear_cpu(cpu, mask);
+			continue;
+		}
 
 		/*
 		 * Building a path doesn't enable it, it simply builds a
 		 * list of devices from source to sink that can be
 		 * referenced later when the path is actually needed.
 		 */
-		event_data->path[cpu] = coresight_build_path(csdev, sink);
-		if (IS_ERR(event_data->path[cpu]))
-			goto err;
+		path = coresight_build_path(csdev, sink);
+		if (IS_ERR(path)) {
+			cpumask_clear_cpu(cpu, mask);
+			continue;
+		}
+
+		*etm_event_cpu_path_ptr(event_data, cpu) = path;
 	}
 
-	if (!sink_ops(sink)->alloc_buffer)
+	/* If we don't have any CPUs ready for tracing, abort */
+	cpu = cpumask_first(mask);
+	if (cpu >= nr_cpu_ids)
 		goto err;
 
-	cpu = cpumask_first(mask);
-	/* Get the AUX specific data from the sink buffer */
+	/* Allocate the sink buffer for this session */
 	event_data->snk_config =
 			sink_ops(sink)->alloc_buffer(sink, cpu, pages,
 						     nr_pages, overwrite);
@@ -255,6 +270,7 @@ static void etm_event_start(struct perf_event *event, int flags)
 	struct etm_event_data *event_data;
 	struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle);
 	struct coresight_device *sink, *csdev = per_cpu(csdev_src, cpu);
+	struct list_head *path;
 
 	if (!csdev)
 		goto fail;
@@ -267,18 +283,14 @@ static void etm_event_start(struct perf_event *event, int flags)
 	if (!event_data)
 		goto fail;
 
+	path = etm_event_cpu_path(event_data, cpu);
 	/* We need a sink, no need to continue without one */
-	sink = coresight_get_sink(event_data->path[cpu]);
-	if (WARN_ON_ONCE(!sink || !sink_ops(sink)->set_buffer))
-		goto fail_end_stop;
-
-	/* Configure the sink */
-	if (sink_ops(sink)->set_buffer(sink, handle,
-				       event_data->snk_config))
+	sink = coresight_get_sink(path);
+	if (WARN_ON_ONCE(!sink))
 		goto fail_end_stop;
 
 	/* Nothing will happen without a path */
-	if (coresight_enable_path(event_data->path[cpu], CS_MODE_PERF))
+	if (coresight_enable_path(path, CS_MODE_PERF, handle))
 		goto fail_end_stop;
 
 	/* Tell the perf core the event is alive */
@@ -286,11 +298,13 @@ static void etm_event_start(struct perf_event *event, int flags)
 
 	/* Finally enable the tracer */
 	if (source_ops(csdev)->enable(csdev, event, CS_MODE_PERF))
-		goto fail_end_stop;
+		goto fail_disable_path;
 
 out:
 	return;
 
+fail_disable_path:
+	coresight_disable_path(path);
 fail_end_stop:
 	perf_aux_output_flag(handle, PERF_AUX_FLAG_TRUNCATED);
 	perf_aux_output_end(handle, 0);
@@ -306,6 +320,7 @@ static void etm_event_stop(struct perf_event *event, int mode)
 	struct coresight_device *sink, *csdev = per_cpu(csdev_src, cpu);
 	struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle);
 	struct etm_event_data *event_data = perf_get_aux(handle);
+	struct list_head *path;
 
 	if (event->hw.state == PERF_HES_STOPPED)
 		return;
@@ -313,7 +328,11 @@ static void etm_event_stop(struct perf_event *event, int mode)
 	if (!csdev)
 		return;
 
-	sink = coresight_get_sink(event_data->path[cpu]);
+	path = etm_event_cpu_path(event_data, cpu);
+	if (!path)
+		return;
+
+	sink = coresight_get_sink(path);
 	if (!sink)
 		return;
 
@@ -331,20 +350,13 @@ static void etm_event_stop(struct perf_event *event, int mode)
 		if (!sink_ops(sink)->update_buffer)
 			return;
 
-		sink_ops(sink)->update_buffer(sink, handle,
+		size = sink_ops(sink)->update_buffer(sink, handle,
 					      event_data->snk_config);
-
-		if (!sink_ops(sink)->reset_buffer)
-			return;
-
-		size = sink_ops(sink)->reset_buffer(sink, handle,
-						    event_data->snk_config);
-
 		perf_aux_output_end(handle, size);
 	}
 
 	/* Disabling the path make its elements available to other sessions */
-	coresight_disable_path(event_data->path[cpu]);
+	coresight_disable_path(path);
 }
 
 static int etm_event_add(struct perf_event *event, int mode)
diff --git a/drivers/hwtracing/coresight/coresight-etm-perf.h b/drivers/hwtracing/coresight/coresight-etm-perf.h
index 4197df4faf5e..da7d9336a15c 100644
--- a/drivers/hwtracing/coresight/coresight-etm-perf.h
+++ b/drivers/hwtracing/coresight/coresight-etm-perf.h
@@ -7,6 +7,7 @@
 #ifndef _CORESIGHT_ETM_PERF_H
 #define _CORESIGHT_ETM_PERF_H
 
+#include <linux/percpu-defs.h>
 #include "coresight-priv.h"
 
 struct coresight_device;
@@ -42,14 +43,39 @@ struct etm_filters {
 	bool			ssstatus;
 };
 
+/**
+ * struct etm_event_data - Coresight specifics associated to an event
+ * @work:		Handle to free allocated memory outside IRQ context.
+ * @mask:		Hold the CPU(s) this event was set for.
+ * @snk_config:		The sink configuration.
+ * @path:		An array of path, each slot for one CPU.
+ */
+struct etm_event_data {
+	struct work_struct work;
+	cpumask_t mask;
+	void *snk_config;
+	struct list_head * __percpu *path;
+};
 
 #ifdef CONFIG_CORESIGHT
 int etm_perf_symlink(struct coresight_device *csdev, bool link);
+static inline void *etm_perf_sink_config(struct perf_output_handle *handle)
+{
+	struct etm_event_data *data = perf_get_aux(handle);
 
+	if (data)
+		return data->snk_config;
+	return NULL;
+}
 #else
 static inline int etm_perf_symlink(struct coresight_device *csdev, bool link)
 { return -EINVAL; }
 
+static inline void *etm_perf_sink_config(struct perf_output_handle *handle)
+{
+	return NULL;
+}
+
 #endif /* CONFIG_CORESIGHT */
 
 #endif
diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index 7c74263c333d..fd5c4cca7db5 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -355,11 +355,10 @@ static int etm_parse_event_config(struct etm_drvdata *drvdata,
 	return 0;
 }
 
-static void etm_enable_hw(void *info)
+static int etm_enable_hw(struct etm_drvdata *drvdata)
 {
-	int i;
+	int i, rc;
 	u32 etmcr;
-	struct etm_drvdata *drvdata = info;
 	struct etm_config *config = &drvdata->config;
 
 	CS_UNLOCK(drvdata->base);
@@ -370,6 +369,9 @@ static void etm_enable_hw(void *info)
 	etm_set_pwrup(drvdata);
 	/* Make sure all registers are accessible */
 	etm_os_unlock(drvdata);
+	rc = coresight_claim_device_unlocked(drvdata->base);
+	if (rc)
+		goto done;
 
 	etm_set_prog(drvdata);
 
@@ -418,9 +420,29 @@ static void etm_enable_hw(void *info)
 	etm_writel(drvdata, 0x0, ETMVMIDCVR);
 
 	etm_clr_prog(drvdata);
+
+done:
+	if (rc)
+		etm_set_pwrdwn(drvdata);
 	CS_LOCK(drvdata->base);
 
-	dev_dbg(drvdata->dev, "cpu: %d enable smp call done\n", drvdata->cpu);
+	dev_dbg(drvdata->dev, "cpu: %d enable smp call done: %d\n",
+		drvdata->cpu, rc);
+	return rc;
+}
+
+struct etm_enable_arg {
+	struct etm_drvdata *drvdata;
+	int rc;
+};
+
+static void etm_enable_hw_smp_call(void *info)
+{
+	struct etm_enable_arg *arg = info;
+
+	if (WARN_ON(!arg))
+		return;
+	arg->rc = etm_enable_hw(arg->drvdata);
 }
 
 static int etm_cpu_id(struct coresight_device *csdev)
@@ -475,14 +497,13 @@ static int etm_enable_perf(struct coresight_device *csdev,
 	/* Configure the tracer based on the session's specifics */
 	etm_parse_event_config(drvdata, event);
 	/* And enable it */
-	etm_enable_hw(drvdata);
-
-	return 0;
+	return etm_enable_hw(drvdata);
 }
 
 static int etm_enable_sysfs(struct coresight_device *csdev)
 {
 	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	struct etm_enable_arg arg = { 0 };
 	int ret;
 
 	spin_lock(&drvdata->spinlock);
@@ -492,20 +513,21 @@ static int etm_enable_sysfs(struct coresight_device *csdev)
 	 * hw configuration will take place on the local CPU during bring up.
 	 */
 	if (cpu_online(drvdata->cpu)) {
+		arg.drvdata = drvdata;
 		ret = smp_call_function_single(drvdata->cpu,
-					       etm_enable_hw, drvdata, 1);
-		if (ret)
-			goto err;
+					       etm_enable_hw_smp_call, &arg, 1);
+		if (!ret)
+			ret = arg.rc;
+		if (!ret)
+			drvdata->sticky_enable = true;
+	} else {
+		ret = -ENODEV;
 	}
 
-	drvdata->sticky_enable = true;
 	spin_unlock(&drvdata->spinlock);
 
-	dev_info(drvdata->dev, "ETM tracing enabled\n");
-	return 0;
-
-err:
-	spin_unlock(&drvdata->spinlock);
+	if (!ret)
+		dev_dbg(drvdata->dev, "ETM tracing enabled\n");
 	return ret;
 }
 
@@ -555,6 +577,8 @@ static void etm_disable_hw(void *info)
 	for (i = 0; i < drvdata->nr_cntr; i++)
 		config->cntr_val[i] = etm_readl(drvdata, ETMCNTVRn(i));
 
+	coresight_disclaim_device_unlocked(drvdata->base);
+
 	etm_set_pwrdwn(drvdata);
 	CS_LOCK(drvdata->base);
 
@@ -604,7 +628,7 @@ static void etm_disable_sysfs(struct coresight_device *csdev)
 	spin_unlock(&drvdata->spinlock);
 	cpus_read_unlock();
 
-	dev_info(drvdata->dev, "ETM tracing disabled\n");
+	dev_dbg(drvdata->dev, "ETM tracing disabled\n");
 }
 
 static void etm_disable(struct coresight_device *csdev,
diff --git a/drivers/hwtracing/coresight/coresight-etm4x.c b/drivers/hwtracing/coresight/coresight-etm4x.c
index 1d94ebec027b..53e2fb6e86f6 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x.c
@@ -28,6 +28,7 @@
 #include <linux/pm_runtime.h>
 #include <asm/sections.h>
 #include <asm/local.h>
+#include <asm/virt.h>
 
 #include "coresight-etm4x.h"
 #include "coresight-etm-perf.h"
@@ -77,16 +78,24 @@ static int etm4_trace_id(struct coresight_device *csdev)
 	return drvdata->trcid;
 }
 
-static void etm4_enable_hw(void *info)
+struct etm4_enable_arg {
+	struct etmv4_drvdata *drvdata;
+	int rc;
+};
+
+static int etm4_enable_hw(struct etmv4_drvdata *drvdata)
 {
-	int i;
-	struct etmv4_drvdata *drvdata = info;
+	int i, rc;
 	struct etmv4_config *config = &drvdata->config;
 
 	CS_UNLOCK(drvdata->base);
 
 	etm4_os_unlock(drvdata);
 
+	rc = coresight_claim_device_unlocked(drvdata->base);
+	if (rc)
+		goto done;
+
 	/* Disable the trace unit before programming trace registers */
 	writel_relaxed(0, drvdata->base + TRCPRGCTLR);
 
@@ -174,9 +183,21 @@ static void etm4_enable_hw(void *info)
 		dev_err(drvdata->dev,
 			"timeout while waiting for Idle Trace Status\n");
 
+done:
 	CS_LOCK(drvdata->base);
 
-	dev_dbg(drvdata->dev, "cpu: %d enable smp call done\n", drvdata->cpu);
+	dev_dbg(drvdata->dev, "cpu: %d enable smp call done: %d\n",
+		drvdata->cpu, rc);
+	return rc;
+}
+
+static void etm4_enable_hw_smp_call(void *info)
+{
+	struct etm4_enable_arg *arg = info;
+
+	if (WARN_ON(!arg))
+		return;
+	arg->rc = etm4_enable_hw(arg->drvdata);
 }
 
 static int etm4_parse_event_config(struct etmv4_drvdata *drvdata,
@@ -242,7 +263,7 @@ static int etm4_enable_perf(struct coresight_device *csdev,
 	if (ret)
 		goto out;
 	/* And enable it */
-	etm4_enable_hw(drvdata);
+	ret = etm4_enable_hw(drvdata);
 
 out:
 	return ret;
@@ -251,6 +272,7 @@ out:
 static int etm4_enable_sysfs(struct coresight_device *csdev)
 {
 	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	struct etm4_enable_arg arg = { 0 };
 	int ret;
 
 	spin_lock(&drvdata->spinlock);
@@ -259,19 +281,17 @@ static int etm4_enable_sysfs(struct coresight_device *csdev)
 	 * Executing etm4_enable_hw on the cpu whose ETM is being enabled
 	 * ensures that register writes occur when cpu is powered.
 	 */
+	arg.drvdata = drvdata;
 	ret = smp_call_function_single(drvdata->cpu,
-				       etm4_enable_hw, drvdata, 1);
-	if (ret)
-		goto err;
-
-	drvdata->sticky_enable = true;
+				       etm4_enable_hw_smp_call, &arg, 1);
+	if (!ret)
+		ret = arg.rc;
+	if (!ret)
+		drvdata->sticky_enable = true;
 	spin_unlock(&drvdata->spinlock);
 
-	dev_info(drvdata->dev, "ETM tracing enabled\n");
-	return 0;
-
-err:
-	spin_unlock(&drvdata->spinlock);
+	if (!ret)
+		dev_dbg(drvdata->dev, "ETM tracing enabled\n");
 	return ret;
 }
 
@@ -328,6 +348,8 @@ static void etm4_disable_hw(void *info)
 	isb();
 	writel_relaxed(control, drvdata->base + TRCPRGCTLR);
 
+	coresight_disclaim_device_unlocked(drvdata->base);
+
 	CS_LOCK(drvdata->base);
 
 	dev_dbg(drvdata->dev, "cpu: %d disable smp call done\n", drvdata->cpu);
@@ -380,7 +402,7 @@ static void etm4_disable_sysfs(struct coresight_device *csdev)
 	spin_unlock(&drvdata->spinlock);
 	cpus_read_unlock();
 
-	dev_info(drvdata->dev, "ETM tracing disabled\n");
+	dev_dbg(drvdata->dev, "ETM tracing disabled\n");
 }
 
 static void etm4_disable(struct coresight_device *csdev,
@@ -605,7 +627,7 @@ static void etm4_set_default_config(struct etmv4_config *config)
 	config->vinst_ctrl |= BIT(0);
 }
 
-static u64 etm4_get_access_type(struct etmv4_config *config)
+static u64 etm4_get_ns_access_type(struct etmv4_config *config)
 {
 	u64 access_type = 0;
 
@@ -616,17 +638,26 @@ static u64 etm4_get_access_type(struct etmv4_config *config)
 	 *   Bit[13] Exception level 1 - OS
 	 *   Bit[14] Exception level 2 - Hypervisor
 	 *   Bit[15] Never implemented
-	 *
-	 * Always stay away from hypervisor mode.
 	 */
-	access_type = ETM_EXLEVEL_NS_HYP;
-
-	if (config->mode & ETM_MODE_EXCL_KERN)
-		access_type |= ETM_EXLEVEL_NS_OS;
+	if (!is_kernel_in_hyp_mode()) {
+		/* Stay away from hypervisor mode for non-VHE */
+		access_type =  ETM_EXLEVEL_NS_HYP;
+		if (config->mode & ETM_MODE_EXCL_KERN)
+			access_type |= ETM_EXLEVEL_NS_OS;
+	} else if (config->mode & ETM_MODE_EXCL_KERN) {
+		access_type = ETM_EXLEVEL_NS_HYP;
+	}
 
 	if (config->mode & ETM_MODE_EXCL_USER)
 		access_type |= ETM_EXLEVEL_NS_APP;
 
+	return access_type;
+}
+
+static u64 etm4_get_access_type(struct etmv4_config *config)
+{
+	u64 access_type = etm4_get_ns_access_type(config);
+
 	/*
 	 * EXLEVEL_S, bits[11:8], don't trace anything happening
 	 * in secure state.
@@ -880,20 +911,10 @@ void etm4_config_trace_mode(struct etmv4_config *config)
 
 	addr_acc = config->addr_acc[ETM_DEFAULT_ADDR_COMP];
 	/* clear default config */
-	addr_acc &= ~(ETM_EXLEVEL_NS_APP | ETM_EXLEVEL_NS_OS);
+	addr_acc &= ~(ETM_EXLEVEL_NS_APP | ETM_EXLEVEL_NS_OS |
+		      ETM_EXLEVEL_NS_HYP);
 
-	/*
-	 * EXLEVEL_NS, bits[15:12]
-	 * The Exception levels are:
-	 *   Bit[12] Exception level 0 - Application
-	 *   Bit[13] Exception level 1 - OS
-	 *   Bit[14] Exception level 2 - Hypervisor
-	 *   Bit[15] Never implemented
-	 */
-	if (mode & ETM_MODE_EXCL_KERN)
-		addr_acc |= ETM_EXLEVEL_NS_OS;
-	else
-		addr_acc |= ETM_EXLEVEL_NS_APP;
+	addr_acc |= etm4_get_ns_access_type(config);
 
 	config->addr_acc[ETM_DEFAULT_ADDR_COMP] = addr_acc;
 	config->addr_acc[ETM_DEFAULT_ADDR_COMP + 1] = addr_acc;
diff --git a/drivers/hwtracing/coresight/coresight-funnel.c b/drivers/hwtracing/coresight/coresight-funnel.c
index 448145a36675..927925151509 100644
--- a/drivers/hwtracing/coresight/coresight-funnel.c
+++ b/drivers/hwtracing/coresight/coresight-funnel.c
@@ -25,6 +25,7 @@
 #define FUNNEL_HOLDTIME_MASK	0xf00
 #define FUNNEL_HOLDTIME_SHFT	0x8
 #define FUNNEL_HOLDTIME		(0x7 << FUNNEL_HOLDTIME_SHFT)
+#define FUNNEL_ENSx_MASK	0xff
 
 /**
  * struct funnel_drvdata - specifics associated to a funnel component
@@ -42,31 +43,42 @@ struct funnel_drvdata {
 	unsigned long		priority;
 };
 
-static void funnel_enable_hw(struct funnel_drvdata *drvdata, int port)
+static int funnel_enable_hw(struct funnel_drvdata *drvdata, int port)
 {
 	u32 functl;
+	int rc = 0;
 
 	CS_UNLOCK(drvdata->base);
 
 	functl = readl_relaxed(drvdata->base + FUNNEL_FUNCTL);
+	/* Claim the device only when we enable the first slave */
+	if (!(functl & FUNNEL_ENSx_MASK)) {
+		rc = coresight_claim_device_unlocked(drvdata->base);
+		if (rc)
+			goto done;
+	}
+
 	functl &= ~FUNNEL_HOLDTIME_MASK;
 	functl |= FUNNEL_HOLDTIME;
 	functl |= (1 << port);
 	writel_relaxed(functl, drvdata->base + FUNNEL_FUNCTL);
 	writel_relaxed(drvdata->priority, drvdata->base + FUNNEL_PRICTL);
-
+done:
 	CS_LOCK(drvdata->base);
+	return rc;
 }
 
 static int funnel_enable(struct coresight_device *csdev, int inport,
 			 int outport)
 {
+	int rc;
 	struct funnel_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
-	funnel_enable_hw(drvdata, inport);
+	rc = funnel_enable_hw(drvdata, inport);
 
-	dev_info(drvdata->dev, "FUNNEL inport %d enabled\n", inport);
-	return 0;
+	if (!rc)
+		dev_dbg(drvdata->dev, "FUNNEL inport %d enabled\n", inport);
+	return rc;
 }
 
 static void funnel_disable_hw(struct funnel_drvdata *drvdata, int inport)
@@ -79,6 +91,10 @@ static void funnel_disable_hw(struct funnel_drvdata *drvdata, int inport)
 	functl &= ~(1 << inport);
 	writel_relaxed(functl, drvdata->base + FUNNEL_FUNCTL);
 
+	/* Disclaim the device if none of the slaves are now active */
+	if (!(functl & FUNNEL_ENSx_MASK))
+		coresight_disclaim_device_unlocked(drvdata->base);
+
 	CS_LOCK(drvdata->base);
 }
 
@@ -89,7 +105,7 @@ static void funnel_disable(struct coresight_device *csdev, int inport,
 
 	funnel_disable_hw(drvdata, inport);
 
-	dev_info(drvdata->dev, "FUNNEL inport %d disabled\n", inport);
+	dev_dbg(drvdata->dev, "FUNNEL inport %d disabled\n", inport);
 }
 
 static const struct coresight_ops_link funnel_link_ops = {
diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h
index 1a6cf3589866..579f34943bf1 100644
--- a/drivers/hwtracing/coresight/coresight-priv.h
+++ b/drivers/hwtracing/coresight/coresight-priv.h
@@ -25,6 +25,13 @@
 #define CORESIGHT_DEVID		0xfc8
 #define CORESIGHT_DEVTYPE	0xfcc
 
+
+/*
+ * Coresight device CLAIM protocol.
+ * See PSCI - ARM DEN 0022D, Section: 6.8.1 Debug and Trace save and restore.
+ */
+#define CORESIGHT_CLAIM_SELF_HOSTED	BIT(1)
+
 #define TIMEOUT_US		100
 #define BMVAL(val, lsb, msb)	((val & GENMASK(msb, lsb)) >> lsb)
 
@@ -137,7 +144,7 @@ static inline void coresight_write_reg_pair(void __iomem *addr, u64 val,
 }
 
 void coresight_disable_path(struct list_head *path);
-int coresight_enable_path(struct list_head *path, u32 mode);
+int coresight_enable_path(struct list_head *path, u32 mode, void *sink_data);
 struct coresight_device *coresight_get_sink(struct list_head *path);
 struct coresight_device *coresight_get_enabled_sink(bool reset);
 struct list_head *coresight_build_path(struct coresight_device *csdev,
diff --git a/drivers/hwtracing/coresight/coresight-replicator.c b/drivers/hwtracing/coresight/coresight-replicator.c
index 8d2eaaab6c2f..feac98315471 100644
--- a/drivers/hwtracing/coresight/coresight-replicator.c
+++ b/drivers/hwtracing/coresight/coresight-replicator.c
@@ -35,7 +35,7 @@ static int replicator_enable(struct coresight_device *csdev, int inport,
 {
 	struct replicator_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
-	dev_info(drvdata->dev, "REPLICATOR enabled\n");
+	dev_dbg(drvdata->dev, "REPLICATOR enabled\n");
 	return 0;
 }
 
@@ -44,7 +44,7 @@ static void replicator_disable(struct coresight_device *csdev, int inport,
 {
 	struct replicator_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
-	dev_info(drvdata->dev, "REPLICATOR disabled\n");
+	dev_dbg(drvdata->dev, "REPLICATOR disabled\n");
 }
 
 static const struct coresight_ops_link replicator_link_ops = {
diff --git a/drivers/hwtracing/coresight/coresight-stm.c b/drivers/hwtracing/coresight/coresight-stm.c
index c46c70aec1d5..35d6f9709274 100644
--- a/drivers/hwtracing/coresight/coresight-stm.c
+++ b/drivers/hwtracing/coresight/coresight-stm.c
@@ -211,7 +211,7 @@ static int stm_enable(struct coresight_device *csdev,
 	stm_enable_hw(drvdata);
 	spin_unlock(&drvdata->spinlock);
 
-	dev_info(drvdata->dev, "STM tracing enabled\n");
+	dev_dbg(drvdata->dev, "STM tracing enabled\n");
 	return 0;
 }
 
@@ -274,7 +274,7 @@ static void stm_disable(struct coresight_device *csdev,
 		pm_runtime_put(drvdata->dev);
 
 		local_set(&drvdata->mode, CS_MODE_DISABLED);
-		dev_info(drvdata->dev, "STM tracing disabled\n");
+		dev_dbg(drvdata->dev, "STM tracing disabled\n");
 	}
 }
 
diff --git a/drivers/hwtracing/coresight/coresight-tmc-etf.c b/drivers/hwtracing/coresight/coresight-tmc-etf.c
index 0549249f4b39..53fc83b72a49 100644
--- a/drivers/hwtracing/coresight/coresight-tmc-etf.c
+++ b/drivers/hwtracing/coresight/coresight-tmc-etf.c
@@ -10,8 +10,12 @@
 #include <linux/slab.h>
 #include "coresight-priv.h"
 #include "coresight-tmc.h"
+#include "coresight-etm-perf.h"
 
-static void tmc_etb_enable_hw(struct tmc_drvdata *drvdata)
+static int tmc_set_etf_buffer(struct coresight_device *csdev,
+			      struct perf_output_handle *handle);
+
+static void __tmc_etb_enable_hw(struct tmc_drvdata *drvdata)
 {
 	CS_UNLOCK(drvdata->base);
 
@@ -30,33 +34,41 @@ static void tmc_etb_enable_hw(struct tmc_drvdata *drvdata)
 	CS_LOCK(drvdata->base);
 }
 
+static int tmc_etb_enable_hw(struct tmc_drvdata *drvdata)
+{
+	int rc = coresight_claim_device(drvdata->base);
+
+	if (rc)
+		return rc;
+
+	__tmc_etb_enable_hw(drvdata);
+	return 0;
+}
+
 static void tmc_etb_dump_hw(struct tmc_drvdata *drvdata)
 {
 	char *bufp;
 	u32 read_data, lost;
-	int i;
 
 	/* Check if the buffer wrapped around. */
 	lost = readl_relaxed(drvdata->base + TMC_STS) & TMC_STS_FULL;
 	bufp = drvdata->buf;
 	drvdata->len = 0;
 	while (1) {
-		for (i = 0; i < drvdata->memwidth; i++) {
-			read_data = readl_relaxed(drvdata->base + TMC_RRD);
-			if (read_data == 0xFFFFFFFF)
-				goto done;
-			memcpy(bufp, &read_data, 4);
-			bufp += 4;
-			drvdata->len += 4;
-		}
+		read_data = readl_relaxed(drvdata->base + TMC_RRD);
+		if (read_data == 0xFFFFFFFF)
+			break;
+		memcpy(bufp, &read_data, 4);
+		bufp += 4;
+		drvdata->len += 4;
 	}
-done:
+
 	if (lost)
 		coresight_insert_barrier_packet(drvdata->buf);
 	return;
 }
 
-static void tmc_etb_disable_hw(struct tmc_drvdata *drvdata)
+static void __tmc_etb_disable_hw(struct tmc_drvdata *drvdata)
 {
 	CS_UNLOCK(drvdata->base);
 
@@ -72,7 +84,13 @@ static void tmc_etb_disable_hw(struct tmc_drvdata *drvdata)
 	CS_LOCK(drvdata->base);
 }
 
-static void tmc_etf_enable_hw(struct tmc_drvdata *drvdata)
+static void tmc_etb_disable_hw(struct tmc_drvdata *drvdata)
+{
+	coresight_disclaim_device(drvdata);
+	__tmc_etb_disable_hw(drvdata);
+}
+
+static void __tmc_etf_enable_hw(struct tmc_drvdata *drvdata)
 {
 	CS_UNLOCK(drvdata->base);
 
@@ -88,13 +106,24 @@ static void tmc_etf_enable_hw(struct tmc_drvdata *drvdata)
 	CS_LOCK(drvdata->base);
 }
 
+static int tmc_etf_enable_hw(struct tmc_drvdata *drvdata)
+{
+	int rc = coresight_claim_device(drvdata->base);
+
+	if (rc)
+		return rc;
+
+	__tmc_etf_enable_hw(drvdata);
+	return 0;
+}
+
 static void tmc_etf_disable_hw(struct tmc_drvdata *drvdata)
 {
 	CS_UNLOCK(drvdata->base);
 
 	tmc_flush_and_stop(drvdata);
 	tmc_disable_hw(drvdata);
-
+	coresight_disclaim_device_unlocked(drvdata->base);
 	CS_LOCK(drvdata->base);
 }
 
@@ -170,8 +199,12 @@ static int tmc_enable_etf_sink_sysfs(struct coresight_device *csdev)
 		drvdata->buf = buf;
 	}
 
-	drvdata->mode = CS_MODE_SYSFS;
-	tmc_etb_enable_hw(drvdata);
+	ret = tmc_etb_enable_hw(drvdata);
+	if (!ret)
+		drvdata->mode = CS_MODE_SYSFS;
+	else
+		/* Free up the buffer if we failed to enable */
+		used = false;
 out:
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
@@ -182,37 +215,40 @@ out:
 	return ret;
 }
 
-static int tmc_enable_etf_sink_perf(struct coresight_device *csdev)
+static int tmc_enable_etf_sink_perf(struct coresight_device *csdev, void *data)
 {
 	int ret = 0;
 	unsigned long flags;
 	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	struct perf_output_handle *handle = data;
 
 	spin_lock_irqsave(&drvdata->spinlock, flags);
-	if (drvdata->reading) {
+	do {
 		ret = -EINVAL;
-		goto out;
-	}
-
-	/*
-	 * In Perf mode there can be only one writer per sink.  There
-	 * is also no need to continue if the ETB/ETR is already operated
-	 * from sysFS.
-	 */
-	if (drvdata->mode != CS_MODE_DISABLED) {
-		ret = -EINVAL;
-		goto out;
-	}
+		if (drvdata->reading)
+			break;
+		/*
+		 * In Perf mode there can be only one writer per sink.  There
+		 * is also no need to continue if the ETB/ETF is already
+		 * operated from sysFS.
+		 */
+		if (drvdata->mode != CS_MODE_DISABLED)
+			break;
 
-	drvdata->mode = CS_MODE_PERF;
-	tmc_etb_enable_hw(drvdata);
-out:
+		ret = tmc_set_etf_buffer(csdev, handle);
+		if (ret)
+			break;
+		ret  = tmc_etb_enable_hw(drvdata);
+		if (!ret)
+			drvdata->mode = CS_MODE_PERF;
+	} while (0);
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
 	return ret;
 }
 
-static int tmc_enable_etf_sink(struct coresight_device *csdev, u32 mode)
+static int tmc_enable_etf_sink(struct coresight_device *csdev,
+			       u32 mode, void *data)
 {
 	int ret;
 	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
@@ -222,7 +258,7 @@ static int tmc_enable_etf_sink(struct coresight_device *csdev, u32 mode)
 		ret = tmc_enable_etf_sink_sysfs(csdev);
 		break;
 	case CS_MODE_PERF:
-		ret = tmc_enable_etf_sink_perf(csdev);
+		ret = tmc_enable_etf_sink_perf(csdev, data);
 		break;
 	/* We shouldn't be here */
 	default:
@@ -233,7 +269,7 @@ static int tmc_enable_etf_sink(struct coresight_device *csdev, u32 mode)
 	if (ret)
 		return ret;
 
-	dev_info(drvdata->dev, "TMC-ETB/ETF enabled\n");
+	dev_dbg(drvdata->dev, "TMC-ETB/ETF enabled\n");
 	return 0;
 }
 
@@ -256,12 +292,13 @@ static void tmc_disable_etf_sink(struct coresight_device *csdev)
 
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
-	dev_info(drvdata->dev, "TMC-ETB/ETF disabled\n");
+	dev_dbg(drvdata->dev, "TMC-ETB/ETF disabled\n");
 }
 
 static int tmc_enable_etf_link(struct coresight_device *csdev,
 			       int inport, int outport)
 {
+	int ret;
 	unsigned long flags;
 	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
@@ -271,12 +308,14 @@ static int tmc_enable_etf_link(struct coresight_device *csdev,
 		return -EBUSY;
 	}
 
-	tmc_etf_enable_hw(drvdata);
-	drvdata->mode = CS_MODE_SYSFS;
+	ret = tmc_etf_enable_hw(drvdata);
+	if (!ret)
+		drvdata->mode = CS_MODE_SYSFS;
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
-	dev_info(drvdata->dev, "TMC-ETF enabled\n");
-	return 0;
+	if (!ret)
+		dev_dbg(drvdata->dev, "TMC-ETF enabled\n");
+	return ret;
 }
 
 static void tmc_disable_etf_link(struct coresight_device *csdev,
@@ -295,7 +334,7 @@ static void tmc_disable_etf_link(struct coresight_device *csdev,
 	drvdata->mode = CS_MODE_DISABLED;
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
-	dev_info(drvdata->dev, "TMC-ETF disabled\n");
+	dev_dbg(drvdata->dev, "TMC-ETF disabled\n");
 }
 
 static void *tmc_alloc_etf_buffer(struct coresight_device *csdev, int cpu,
@@ -328,12 +367,14 @@ static void tmc_free_etf_buffer(void *config)
 }
 
 static int tmc_set_etf_buffer(struct coresight_device *csdev,
-			      struct perf_output_handle *handle,
-			      void *sink_config)
+			      struct perf_output_handle *handle)
 {
 	int ret = 0;
 	unsigned long head;
-	struct cs_buffers *buf = sink_config;
+	struct cs_buffers *buf = etm_perf_sink_config(handle);
+
+	if (!buf)
+		return -EINVAL;
 
 	/* wrap head around to the amount of space we have */
 	head = handle->head & ((buf->nr_pages << PAGE_SHIFT) - 1);
@@ -349,36 +390,7 @@ static int tmc_set_etf_buffer(struct coresight_device *csdev,
 	return ret;
 }
 
-static unsigned long tmc_reset_etf_buffer(struct coresight_device *csdev,
-					  struct perf_output_handle *handle,
-					  void *sink_config)
-{
-	long size = 0;
-	struct cs_buffers *buf = sink_config;
-
-	if (buf) {
-		/*
-		 * In snapshot mode ->data_size holds the new address of the
-		 * ring buffer's head.  The size itself is the whole address
-		 * range since we want the latest information.
-		 */
-		if (buf->snapshot)
-			handle->head = local_xchg(&buf->data_size,
-						  buf->nr_pages << PAGE_SHIFT);
-		/*
-		 * Tell the tracer PMU how much we got in this run and if
-		 * something went wrong along the way.  Nobody else can use
-		 * this cs_buffers instance until we are done.  As such
-		 * resetting parameters here and squaring off with the ring
-		 * buffer API in the tracer PMU is fine.
-		 */
-		size = local_xchg(&buf->data_size, 0);
-	}
-
-	return size;
-}
-
-static void tmc_update_etf_buffer(struct coresight_device *csdev,
+static unsigned long tmc_update_etf_buffer(struct coresight_device *csdev,
 				  struct perf_output_handle *handle,
 				  void *sink_config)
 {
@@ -387,17 +399,17 @@ static void tmc_update_etf_buffer(struct coresight_device *csdev,
 	const u32 *barrier;
 	u32 *buf_ptr;
 	u64 read_ptr, write_ptr;
-	u32 status, to_read;
-	unsigned long offset;
+	u32 status;
+	unsigned long offset, to_read;
 	struct cs_buffers *buf = sink_config;
 	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
 	if (!buf)
-		return;
+		return 0;
 
 	/* This shouldn't happen */
 	if (WARN_ON_ONCE(drvdata->mode != CS_MODE_PERF))
-		return;
+		return 0;
 
 	CS_UNLOCK(drvdata->base);
 
@@ -438,10 +450,10 @@ static void tmc_update_etf_buffer(struct coresight_device *csdev,
 		case TMC_MEM_INTF_WIDTH_32BITS:
 		case TMC_MEM_INTF_WIDTH_64BITS:
 		case TMC_MEM_INTF_WIDTH_128BITS:
-			mask = GENMASK(31, 5);
+			mask = GENMASK(31, 4);
 			break;
 		case TMC_MEM_INTF_WIDTH_256BITS:
-			mask = GENMASK(31, 6);
+			mask = GENMASK(31, 5);
 			break;
 		}
 
@@ -486,18 +498,14 @@ static void tmc_update_etf_buffer(struct coresight_device *csdev,
 		}
 	}
 
-	/*
-	 * In snapshot mode all we have to do is communicate to
-	 * perf_aux_output_end() the address of the current head.  In full
-	 * trace mode the same function expects a size to move rb->aux_head
-	 * forward.
-	 */
-	if (buf->snapshot)
-		local_set(&buf->data_size, (cur * PAGE_SIZE) + offset);
-	else
-		local_add(to_read, &buf->data_size);
-
+	/* In snapshot mode we have to update the head */
+	if (buf->snapshot) {
+		handle->head = (cur * PAGE_SIZE) + offset;
+		to_read = buf->nr_pages << PAGE_SHIFT;
+	}
 	CS_LOCK(drvdata->base);
+
+	return to_read;
 }
 
 static const struct coresight_ops_sink tmc_etf_sink_ops = {
@@ -505,8 +513,6 @@ static const struct coresight_ops_sink tmc_etf_sink_ops = {
 	.disable	= tmc_disable_etf_sink,
 	.alloc_buffer	= tmc_alloc_etf_buffer,
 	.free_buffer	= tmc_free_etf_buffer,
-	.set_buffer	= tmc_set_etf_buffer,
-	.reset_buffer	= tmc_reset_etf_buffer,
 	.update_buffer	= tmc_update_etf_buffer,
 };
 
@@ -563,7 +569,7 @@ int tmc_read_prepare_etb(struct tmc_drvdata *drvdata)
 
 	/* Disable the TMC if need be */
 	if (drvdata->mode == CS_MODE_SYSFS)
-		tmc_etb_disable_hw(drvdata);
+		__tmc_etb_disable_hw(drvdata);
 
 	drvdata->reading = true;
 out:
@@ -603,7 +609,7 @@ int tmc_read_unprepare_etb(struct tmc_drvdata *drvdata)
 		 * can't be NULL.
 		 */
 		memset(drvdata->buf, 0, drvdata->size);
-		tmc_etb_enable_hw(drvdata);
+		__tmc_etb_enable_hw(drvdata);
 	} else {
 		/*
 		 * The ETB/ETF is not tracing and the buffer was just read.
diff --git a/drivers/hwtracing/coresight/coresight-tmc-etr.c b/drivers/hwtracing/coresight/coresight-tmc-etr.c
index 2eda5de304c2..f684283890d3 100644
--- a/drivers/hwtracing/coresight/coresight-tmc-etr.c
+++ b/drivers/hwtracing/coresight/coresight-tmc-etr.c
@@ -10,6 +10,7 @@
 #include <linux/slab.h>
 #include <linux/vmalloc.h>
 #include "coresight-catu.h"
+#include "coresight-etm-perf.h"
 #include "coresight-priv.h"
 #include "coresight-tmc.h"
 
@@ -21,6 +22,28 @@ struct etr_flat_buf {
 };
 
 /*
+ * etr_perf_buffer - Perf buffer used for ETR
+ * @etr_buf		- Actual buffer used by the ETR
+ * @snaphost		- Perf session mode
+ * @head		- handle->head at the beginning of the session.
+ * @nr_pages		- Number of pages in the ring buffer.
+ * @pages		- Array of Pages in the ring buffer.
+ */
+struct etr_perf_buffer {
+	struct etr_buf		*etr_buf;
+	bool			snapshot;
+	unsigned long		head;
+	int			nr_pages;
+	void			**pages;
+};
+
+/* Convert the perf index to an offset within the ETR buffer */
+#define PERF_IDX2OFF(idx, buf)	((idx) % ((buf)->nr_pages << PAGE_SHIFT))
+
+/* Lower limit for ETR hardware buffer */
+#define TMC_ETR_PERF_MIN_BUF_SIZE	SZ_1M
+
+/*
  * The TMC ETR SG has a page size of 4K. The SG table contains pointers
  * to 4KB buffers. However, the OS may use a PAGE_SIZE different from
  * 4K (i.e, 16KB or 64KB). This implies that a single OS page could
@@ -536,7 +559,7 @@ tmc_init_etr_sg_table(struct device *dev, int node,
 	sg_table = tmc_alloc_sg_table(dev, node, nr_tpages, nr_dpages, pages);
 	if (IS_ERR(sg_table)) {
 		kfree(etr_table);
-		return ERR_PTR(PTR_ERR(sg_table));
+		return ERR_CAST(sg_table);
 	}
 
 	etr_table->sg_table = sg_table;
@@ -728,12 +751,14 @@ tmc_etr_get_catu_device(struct tmc_drvdata *drvdata)
 	return NULL;
 }
 
-static inline void tmc_etr_enable_catu(struct tmc_drvdata *drvdata)
+static inline int tmc_etr_enable_catu(struct tmc_drvdata *drvdata,
+				      struct etr_buf *etr_buf)
 {
 	struct coresight_device *catu = tmc_etr_get_catu_device(drvdata);
 
 	if (catu && helper_ops(catu)->enable)
-		helper_ops(catu)->enable(catu, drvdata->etr_buf);
+		return helper_ops(catu)->enable(catu, etr_buf);
+	return 0;
 }
 
 static inline void tmc_etr_disable_catu(struct tmc_drvdata *drvdata)
@@ -895,17 +920,11 @@ static void tmc_sync_etr_buf(struct tmc_drvdata *drvdata)
 		tmc_etr_buf_insert_barrier_packet(etr_buf, etr_buf->offset);
 }
 
-static void tmc_etr_enable_hw(struct tmc_drvdata *drvdata)
+static void __tmc_etr_enable_hw(struct tmc_drvdata *drvdata)
 {
 	u32 axictl, sts;
 	struct etr_buf *etr_buf = drvdata->etr_buf;
 
-	/*
-	 * If this ETR is connected to a CATU, enable it before we turn
-	 * this on
-	 */
-	tmc_etr_enable_catu(drvdata);
-
 	CS_UNLOCK(drvdata->base);
 
 	/* Wait for TMCSReady bit to be set */
@@ -924,11 +943,8 @@ static void tmc_etr_enable_hw(struct tmc_drvdata *drvdata)
 		axictl |= TMC_AXICTL_ARCACHE_OS;
 	}
 
-	if (etr_buf->mode == ETR_MODE_ETR_SG) {
-		if (WARN_ON(!tmc_etr_has_cap(drvdata, TMC_ETR_SG)))
-			return;
+	if (etr_buf->mode == ETR_MODE_ETR_SG)
 		axictl |= TMC_AXICTL_SCT_GAT_MODE;
-	}
 
 	writel_relaxed(axictl, drvdata->base + TMC_AXICTL);
 	tmc_write_dba(drvdata, etr_buf->hwaddr);
@@ -954,19 +970,54 @@ static void tmc_etr_enable_hw(struct tmc_drvdata *drvdata)
 	CS_LOCK(drvdata->base);
 }
 
+static int tmc_etr_enable_hw(struct tmc_drvdata *drvdata,
+			     struct etr_buf *etr_buf)
+{
+	int rc;
+
+	/* Callers should provide an appropriate buffer for use */
+	if (WARN_ON(!etr_buf))
+		return -EINVAL;
+
+	if ((etr_buf->mode == ETR_MODE_ETR_SG) &&
+	    WARN_ON(!tmc_etr_has_cap(drvdata, TMC_ETR_SG)))
+		return -EINVAL;
+
+	if (WARN_ON(drvdata->etr_buf))
+		return -EBUSY;
+
+	/*
+	 * If this ETR is connected to a CATU, enable it before we turn
+	 * this on.
+	 */
+	rc = tmc_etr_enable_catu(drvdata, etr_buf);
+	if (rc)
+		return rc;
+	rc = coresight_claim_device(drvdata->base);
+	if (!rc) {
+		drvdata->etr_buf = etr_buf;
+		__tmc_etr_enable_hw(drvdata);
+	}
+
+	return rc;
+}
+
 /*
  * Return the available trace data in the buffer (starts at etr_buf->offset,
  * limited by etr_buf->len) from @pos, with a maximum limit of @len,
  * also updating the @bufpp on where to find it. Since the trace data
  * starts at anywhere in the buffer, depending on the RRP, we adjust the
  * @len returned to handle buffer wrapping around.
+ *
+ * We are protected here by drvdata->reading != 0, which ensures the
+ * sysfs_buf stays alive.
  */
 ssize_t tmc_etr_get_sysfs_trace(struct tmc_drvdata *drvdata,
 				loff_t pos, size_t len, char **bufpp)
 {
 	s64 offset;
 	ssize_t actual = len;
-	struct etr_buf *etr_buf = drvdata->etr_buf;
+	struct etr_buf *etr_buf = drvdata->sysfs_buf;
 
 	if (pos + actual > etr_buf->len)
 		actual = etr_buf->len - pos;
@@ -996,10 +1047,17 @@ tmc_etr_free_sysfs_buf(struct etr_buf *buf)
 
 static void tmc_etr_sync_sysfs_buf(struct tmc_drvdata *drvdata)
 {
-	tmc_sync_etr_buf(drvdata);
+	struct etr_buf *etr_buf = drvdata->etr_buf;
+
+	if (WARN_ON(drvdata->sysfs_buf != etr_buf)) {
+		tmc_etr_free_sysfs_buf(drvdata->sysfs_buf);
+		drvdata->sysfs_buf = NULL;
+	} else {
+		tmc_sync_etr_buf(drvdata);
+	}
 }
 
-static void tmc_etr_disable_hw(struct tmc_drvdata *drvdata)
+static void __tmc_etr_disable_hw(struct tmc_drvdata *drvdata)
 {
 	CS_UNLOCK(drvdata->base);
 
@@ -1015,8 +1073,16 @@ static void tmc_etr_disable_hw(struct tmc_drvdata *drvdata)
 
 	CS_LOCK(drvdata->base);
 
+}
+
+static void tmc_etr_disable_hw(struct tmc_drvdata *drvdata)
+{
+	__tmc_etr_disable_hw(drvdata);
 	/* Disable CATU device if this ETR is connected to one */
 	tmc_etr_disable_catu(drvdata);
+	coresight_disclaim_device(drvdata->base);
+	/* Reset the ETR buf used by hardware */
+	drvdata->etr_buf = NULL;
 }
 
 static int tmc_enable_etr_sink_sysfs(struct coresight_device *csdev)
@@ -1024,7 +1090,7 @@ static int tmc_enable_etr_sink_sysfs(struct coresight_device *csdev)
 	int ret = 0;
 	unsigned long flags;
 	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
-	struct etr_buf *new_buf = NULL, *free_buf = NULL;
+	struct etr_buf *sysfs_buf = NULL, *new_buf = NULL, *free_buf = NULL;
 
 	/*
 	 * If we are enabling the ETR from disabled state, we need to make
@@ -1035,7 +1101,8 @@ static int tmc_enable_etr_sink_sysfs(struct coresight_device *csdev)
 	 * with the lock released.
 	 */
 	spin_lock_irqsave(&drvdata->spinlock, flags);
-	if (!drvdata->etr_buf || (drvdata->etr_buf->size != drvdata->size)) {
+	sysfs_buf = READ_ONCE(drvdata->sysfs_buf);
+	if (!sysfs_buf || (sysfs_buf->size != drvdata->size)) {
 		spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
 		/* Allocate memory with the locks released */
@@ -1064,14 +1131,15 @@ static int tmc_enable_etr_sink_sysfs(struct coresight_device *csdev)
 	 * If we don't have a buffer or it doesn't match the requested size,
 	 * use the buffer allocated above. Otherwise reuse the existing buffer.
 	 */
-	if (!drvdata->etr_buf ||
-	    (new_buf && drvdata->etr_buf->size != new_buf->size)) {
-		free_buf = drvdata->etr_buf;
-		drvdata->etr_buf = new_buf;
+	sysfs_buf = READ_ONCE(drvdata->sysfs_buf);
+	if (!sysfs_buf || (new_buf && sysfs_buf->size != new_buf->size)) {
+		free_buf = sysfs_buf;
+		drvdata->sysfs_buf = new_buf;
 	}
 
-	drvdata->mode = CS_MODE_SYSFS;
-	tmc_etr_enable_hw(drvdata);
+	ret = tmc_etr_enable_hw(drvdata, drvdata->sysfs_buf);
+	if (!ret)
+		drvdata->mode = CS_MODE_SYSFS;
 out:
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
@@ -1080,24 +1148,244 @@ out:
 		tmc_etr_free_sysfs_buf(free_buf);
 
 	if (!ret)
-		dev_info(drvdata->dev, "TMC-ETR enabled\n");
+		dev_dbg(drvdata->dev, "TMC-ETR enabled\n");
 
 	return ret;
 }
 
-static int tmc_enable_etr_sink_perf(struct coresight_device *csdev)
+/*
+ * tmc_etr_setup_perf_buf: Allocate ETR buffer for use by perf.
+ * The size of the hardware buffer is dependent on the size configured
+ * via sysfs and the perf ring buffer size. We prefer to allocate the
+ * largest possible size, scaling down the size by half until it
+ * reaches a minimum limit (1M), beyond which we give up.
+ */
+static struct etr_perf_buffer *
+tmc_etr_setup_perf_buf(struct tmc_drvdata *drvdata, int node, int nr_pages,
+		       void **pages, bool snapshot)
 {
-	/* We don't support perf mode yet ! */
-	return -EINVAL;
+	struct etr_buf *etr_buf;
+	struct etr_perf_buffer *etr_perf;
+	unsigned long size;
+
+	etr_perf = kzalloc_node(sizeof(*etr_perf), GFP_KERNEL, node);
+	if (!etr_perf)
+		return ERR_PTR(-ENOMEM);
+
+	/*
+	 * Try to match the perf ring buffer size if it is larger
+	 * than the size requested via sysfs.
+	 */
+	if ((nr_pages << PAGE_SHIFT) > drvdata->size) {
+		etr_buf = tmc_alloc_etr_buf(drvdata, (nr_pages << PAGE_SHIFT),
+					    0, node, NULL);
+		if (!IS_ERR(etr_buf))
+			goto done;
+	}
+
+	/*
+	 * Else switch to configured size for this ETR
+	 * and scale down until we hit the minimum limit.
+	 */
+	size = drvdata->size;
+	do {
+		etr_buf = tmc_alloc_etr_buf(drvdata, size, 0, node, NULL);
+		if (!IS_ERR(etr_buf))
+			goto done;
+		size /= 2;
+	} while (size >= TMC_ETR_PERF_MIN_BUF_SIZE);
+
+	kfree(etr_perf);
+	return ERR_PTR(-ENOMEM);
+
+done:
+	etr_perf->etr_buf = etr_buf;
+	return etr_perf;
 }
 
-static int tmc_enable_etr_sink(struct coresight_device *csdev, u32 mode)
+
+static void *tmc_alloc_etr_buffer(struct coresight_device *csdev,
+				  int cpu, void **pages, int nr_pages,
+				  bool snapshot)
+{
+	struct etr_perf_buffer *etr_perf;
+	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	if (cpu == -1)
+		cpu = smp_processor_id();
+
+	etr_perf = tmc_etr_setup_perf_buf(drvdata, cpu_to_node(cpu),
+					  nr_pages, pages, snapshot);
+	if (IS_ERR(etr_perf)) {
+		dev_dbg(drvdata->dev, "Unable to allocate ETR buffer\n");
+		return NULL;
+	}
+
+	etr_perf->snapshot = snapshot;
+	etr_perf->nr_pages = nr_pages;
+	etr_perf->pages = pages;
+
+	return etr_perf;
+}
+
+static void tmc_free_etr_buffer(void *config)
+{
+	struct etr_perf_buffer *etr_perf = config;
+
+	if (etr_perf->etr_buf)
+		tmc_free_etr_buf(etr_perf->etr_buf);
+	kfree(etr_perf);
+}
+
+/*
+ * tmc_etr_sync_perf_buffer: Copy the actual trace data from the hardware
+ * buffer to the perf ring buffer.
+ */
+static void tmc_etr_sync_perf_buffer(struct etr_perf_buffer *etr_perf)
+{
+	long bytes, to_copy;
+	long pg_idx, pg_offset, src_offset;
+	unsigned long head = etr_perf->head;
+	char **dst_pages, *src_buf;
+	struct etr_buf *etr_buf = etr_perf->etr_buf;
+
+	head = etr_perf->head;
+	pg_idx = head >> PAGE_SHIFT;
+	pg_offset = head & (PAGE_SIZE - 1);
+	dst_pages = (char **)etr_perf->pages;
+	src_offset = etr_buf->offset;
+	to_copy = etr_buf->len;
+
+	while (to_copy > 0) {
+		/*
+		 * In one iteration, we can copy minimum of :
+		 *  1) what is available in the source buffer,
+		 *  2) what is available in the source buffer, before it
+		 *     wraps around.
+		 *  3) what is available in the destination page.
+		 * in one iteration.
+		 */
+		bytes = tmc_etr_buf_get_data(etr_buf, src_offset, to_copy,
+					     &src_buf);
+		if (WARN_ON_ONCE(bytes <= 0))
+			break;
+		bytes = min(bytes, (long)(PAGE_SIZE - pg_offset));
+
+		memcpy(dst_pages[pg_idx] + pg_offset, src_buf, bytes);
+
+		to_copy -= bytes;
+
+		/* Move destination pointers */
+		pg_offset += bytes;
+		if (pg_offset == PAGE_SIZE) {
+			pg_offset = 0;
+			if (++pg_idx == etr_perf->nr_pages)
+				pg_idx = 0;
+		}
+
+		/* Move source pointers */
+		src_offset += bytes;
+		if (src_offset >= etr_buf->size)
+			src_offset -= etr_buf->size;
+	}
+}
+
+/*
+ * tmc_update_etr_buffer : Update the perf ring buffer with the
+ * available trace data. We use software double buffering at the moment.
+ *
+ * TODO: Add support for reusing the perf ring buffer.
+ */
+static unsigned long
+tmc_update_etr_buffer(struct coresight_device *csdev,
+		      struct perf_output_handle *handle,
+		      void *config)
+{
+	bool lost = false;
+	unsigned long flags, size = 0;
+	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	struct etr_perf_buffer *etr_perf = config;
+	struct etr_buf *etr_buf = etr_perf->etr_buf;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	if (WARN_ON(drvdata->perf_data != etr_perf)) {
+		lost = true;
+		spin_unlock_irqrestore(&drvdata->spinlock, flags);
+		goto out;
+	}
+
+	CS_UNLOCK(drvdata->base);
+
+	tmc_flush_and_stop(drvdata);
+	tmc_sync_etr_buf(drvdata);
+
+	CS_LOCK(drvdata->base);
+	/* Reset perf specific data */
+	drvdata->perf_data = NULL;
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+
+	size = etr_buf->len;
+	tmc_etr_sync_perf_buffer(etr_perf);
+
+	/*
+	 * Update handle->head in snapshot mode. Also update the size to the
+	 * hardware buffer size if there was an overflow.
+	 */
+	if (etr_perf->snapshot) {
+		handle->head += size;
+		if (etr_buf->full)
+			size = etr_buf->size;
+	}
+
+	lost |= etr_buf->full;
+out:
+	if (lost)
+		perf_aux_output_flag(handle, PERF_AUX_FLAG_TRUNCATED);
+	return size;
+}
+
+static int tmc_enable_etr_sink_perf(struct coresight_device *csdev, void *data)
+{
+	int rc = 0;
+	unsigned long flags;
+	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	struct perf_output_handle *handle = data;
+	struct etr_perf_buffer *etr_perf = etm_perf_sink_config(handle);
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+	/*
+	 * There can be only one writer per sink in perf mode. If the sink
+	 * is already open in SYSFS mode, we can't use it.
+	 */
+	if (drvdata->mode != CS_MODE_DISABLED || WARN_ON(drvdata->perf_data)) {
+		rc = -EBUSY;
+		goto unlock_out;
+	}
+
+	if (WARN_ON(!etr_perf || !etr_perf->etr_buf)) {
+		rc = -EINVAL;
+		goto unlock_out;
+	}
+
+	etr_perf->head = PERF_IDX2OFF(handle->head, etr_perf);
+	drvdata->perf_data = etr_perf;
+	rc = tmc_etr_enable_hw(drvdata, etr_perf->etr_buf);
+	if (!rc)
+		drvdata->mode = CS_MODE_PERF;
+
+unlock_out:
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+	return rc;
+}
+
+static int tmc_enable_etr_sink(struct coresight_device *csdev,
+			       u32 mode, void *data)
 {
 	switch (mode) {
 	case CS_MODE_SYSFS:
 		return tmc_enable_etr_sink_sysfs(csdev);
 	case CS_MODE_PERF:
-		return tmc_enable_etr_sink_perf(csdev);
+		return tmc_enable_etr_sink_perf(csdev, data);
 	}
 
 	/* We shouldn't be here */
@@ -1123,12 +1411,15 @@ static void tmc_disable_etr_sink(struct coresight_device *csdev)
 
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
-	dev_info(drvdata->dev, "TMC-ETR disabled\n");
+	dev_dbg(drvdata->dev, "TMC-ETR disabled\n");
 }
 
 static const struct coresight_ops_sink tmc_etr_sink_ops = {
 	.enable		= tmc_enable_etr_sink,
 	.disable	= tmc_disable_etr_sink,
+	.alloc_buffer	= tmc_alloc_etr_buffer,
+	.update_buffer	= tmc_update_etr_buffer,
+	.free_buffer	= tmc_free_etr_buffer,
 };
 
 const struct coresight_ops tmc_etr_cs_ops = {
@@ -1150,21 +1441,19 @@ int tmc_read_prepare_etr(struct tmc_drvdata *drvdata)
 		goto out;
 	}
 
-	/* Don't interfere if operated from Perf */
-	if (drvdata->mode == CS_MODE_PERF) {
-		ret = -EINVAL;
-		goto out;
-	}
-
-	/* If drvdata::etr_buf is NULL the trace data has been read already */
-	if (drvdata->etr_buf == NULL) {
+	/*
+	 * We can safely allow reads even if the ETR is operating in PERF mode,
+	 * since the sysfs session is captured in mode specific data.
+	 * If drvdata::sysfs_data is NULL the trace data has been read already.
+	 */
+	if (!drvdata->sysfs_buf) {
 		ret = -EINVAL;
 		goto out;
 	}
 
-	/* Disable the TMC if need be */
+	/* Disable the TMC if we are trying to read from a running session. */
 	if (drvdata->mode == CS_MODE_SYSFS)
-		tmc_etr_disable_hw(drvdata);
+		__tmc_etr_disable_hw(drvdata);
 
 	drvdata->reading = true;
 out:
@@ -1176,7 +1465,7 @@ out:
 int tmc_read_unprepare_etr(struct tmc_drvdata *drvdata)
 {
 	unsigned long flags;
-	struct etr_buf *etr_buf = NULL;
+	struct etr_buf *sysfs_buf = NULL;
 
 	/* config types are set a boot time and never change */
 	if (WARN_ON_ONCE(drvdata->config_type != TMC_CONFIG_TYPE_ETR))
@@ -1191,22 +1480,22 @@ int tmc_read_unprepare_etr(struct tmc_drvdata *drvdata)
 		 * buffer. Since the tracer is still enabled drvdata::buf can't
 		 * be NULL.
 		 */
-		tmc_etr_enable_hw(drvdata);
+		__tmc_etr_enable_hw(drvdata);
 	} else {
 		/*
 		 * The ETR is not tracing and the buffer was just read.
 		 * As such prepare to free the trace buffer.
 		 */
-		etr_buf =  drvdata->etr_buf;
-		drvdata->etr_buf = NULL;
+		sysfs_buf = drvdata->sysfs_buf;
+		drvdata->sysfs_buf = NULL;
 	}
 
 	drvdata->reading = false;
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
 	/* Free allocated memory out side of the spinlock */
-	if (etr_buf)
-		tmc_free_etr_buf(etr_buf);
+	if (sysfs_buf)
+		tmc_etr_free_sysfs_buf(sysfs_buf);
 
 	return 0;
 }
diff --git a/drivers/hwtracing/coresight/coresight-tmc.c b/drivers/hwtracing/coresight/coresight-tmc.c
index 1b817ec1192c..ea249f0bcd73 100644
--- a/drivers/hwtracing/coresight/coresight-tmc.c
+++ b/drivers/hwtracing/coresight/coresight-tmc.c
@@ -81,7 +81,7 @@ static int tmc_read_prepare(struct tmc_drvdata *drvdata)
 	}
 
 	if (!ret)
-		dev_info(drvdata->dev, "TMC read start\n");
+		dev_dbg(drvdata->dev, "TMC read start\n");
 
 	return ret;
 }
@@ -103,7 +103,7 @@ static int tmc_read_unprepare(struct tmc_drvdata *drvdata)
 	}
 
 	if (!ret)
-		dev_info(drvdata->dev, "TMC read end\n");
+		dev_dbg(drvdata->dev, "TMC read end\n");
 
 	return ret;
 }
diff --git a/drivers/hwtracing/coresight/coresight-tmc.h b/drivers/hwtracing/coresight/coresight-tmc.h
index 7027bd60c4cc..487c53701e9c 100644
--- a/drivers/hwtracing/coresight/coresight-tmc.h
+++ b/drivers/hwtracing/coresight/coresight-tmc.h
@@ -170,6 +170,8 @@ struct etr_buf {
  * @trigger_cntr: amount of words to store after a trigger.
  * @etr_caps:	Bitmask of capabilities of the TMC ETR, inferred from the
  *		device configuration register (DEVID)
+ * @perf_data:	PERF buffer for ETR.
+ * @sysfs_data:	SYSFS buffer for ETR.
  */
 struct tmc_drvdata {
 	void __iomem		*base;
@@ -189,6 +191,8 @@ struct tmc_drvdata {
 	enum tmc_mem_intf_width	memwidth;
 	u32			trigger_cntr;
 	u32			etr_caps;
+	struct etr_buf		*sysfs_buf;
+	void			*perf_data;
 };
 
 struct etr_buf_operations {
diff --git a/drivers/hwtracing/coresight/coresight-tpiu.c b/drivers/hwtracing/coresight/coresight-tpiu.c
index 459ef930d98c..b2f72a1fa402 100644
--- a/drivers/hwtracing/coresight/coresight-tpiu.c
+++ b/drivers/hwtracing/coresight/coresight-tpiu.c
@@ -68,13 +68,13 @@ static void tpiu_enable_hw(struct tpiu_drvdata *drvdata)
 	CS_LOCK(drvdata->base);
 }
 
-static int tpiu_enable(struct coresight_device *csdev, u32 mode)
+static int tpiu_enable(struct coresight_device *csdev, u32 mode, void *__unused)
 {
 	struct tpiu_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
 	tpiu_enable_hw(drvdata);
 
-	dev_info(drvdata->dev, "TPIU enabled\n");
+	dev_dbg(drvdata->dev, "TPIU enabled\n");
 	return 0;
 }
 
@@ -100,7 +100,7 @@ static void tpiu_disable(struct coresight_device *csdev)
 
 	tpiu_disable_hw(drvdata);
 
-	dev_info(drvdata->dev, "TPIU disabled\n");
+	dev_dbg(drvdata->dev, "TPIU disabled\n");
 }
 
 static const struct coresight_ops_sink tpiu_sink_ops = {
diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c
index 3e07fd335f8c..2b0df1a0a8df 100644
--- a/drivers/hwtracing/coresight/coresight.c
+++ b/drivers/hwtracing/coresight/coresight.c
@@ -128,16 +128,105 @@ static int coresight_find_link_outport(struct coresight_device *csdev,
 	return -ENODEV;
 }
 
-static int coresight_enable_sink(struct coresight_device *csdev, u32 mode)
+static inline u32 coresight_read_claim_tags(void __iomem *base)
+{
+	return readl_relaxed(base + CORESIGHT_CLAIMCLR);
+}
+
+static inline bool coresight_is_claimed_self_hosted(void __iomem *base)
+{
+	return coresight_read_claim_tags(base) == CORESIGHT_CLAIM_SELF_HOSTED;
+}
+
+static inline bool coresight_is_claimed_any(void __iomem *base)
+{
+	return coresight_read_claim_tags(base) != 0;
+}
+
+static inline void coresight_set_claim_tags(void __iomem *base)
+{
+	writel_relaxed(CORESIGHT_CLAIM_SELF_HOSTED, base + CORESIGHT_CLAIMSET);
+	isb();
+}
+
+static inline void coresight_clear_claim_tags(void __iomem *base)
+{
+	writel_relaxed(CORESIGHT_CLAIM_SELF_HOSTED, base + CORESIGHT_CLAIMCLR);
+	isb();
+}
+
+/*
+ * coresight_claim_device_unlocked : Claim the device for self-hosted usage
+ * to prevent an external tool from touching this device. As per PSCI
+ * standards, section "Preserving the execution context" => "Debug and Trace
+ * save and Restore", DBGCLAIM[1] is reserved for Self-hosted debug/trace and
+ * DBGCLAIM[0] is reserved for external tools.
+ *
+ * Called with CS_UNLOCKed for the component.
+ * Returns : 0 on success
+ */
+int coresight_claim_device_unlocked(void __iomem *base)
+{
+	if (coresight_is_claimed_any(base))
+		return -EBUSY;
+
+	coresight_set_claim_tags(base);
+	if (coresight_is_claimed_self_hosted(base))
+		return 0;
+	/* There was a race setting the tags, clean up and fail */
+	coresight_clear_claim_tags(base);
+	return -EBUSY;
+}
+
+int coresight_claim_device(void __iomem *base)
+{
+	int rc;
+
+	CS_UNLOCK(base);
+	rc = coresight_claim_device_unlocked(base);
+	CS_LOCK(base);
+
+	return rc;
+}
+
+/*
+ * coresight_disclaim_device_unlocked : Clear the claim tags for the device.
+ * Called with CS_UNLOCKed for the component.
+ */
+void coresight_disclaim_device_unlocked(void __iomem *base)
+{
+
+	if (coresight_is_claimed_self_hosted(base))
+		coresight_clear_claim_tags(base);
+	else
+		/*
+		 * The external agent may have not honoured our claim
+		 * and has manipulated it. Or something else has seriously
+		 * gone wrong in our driver.
+		 */
+		WARN_ON_ONCE(1);
+}
+
+void coresight_disclaim_device(void __iomem *base)
+{
+	CS_UNLOCK(base);
+	coresight_disclaim_device_unlocked(base);
+	CS_LOCK(base);
+}
+
+static int coresight_enable_sink(struct coresight_device *csdev,
+				 u32 mode, void *data)
 {
 	int ret;
 
-	if (!csdev->enable) {
-		if (sink_ops(csdev)->enable) {
-			ret = sink_ops(csdev)->enable(csdev, mode);
-			if (ret)
-				return ret;
-		}
+	/*
+	 * We need to make sure the "new" session is compatible with the
+	 * existing "mode" of operation.
+	 */
+	if (sink_ops(csdev)->enable) {
+		ret = sink_ops(csdev)->enable(csdev, mode, data);
+		if (ret)
+			return ret;
 		csdev->enable = true;
 	}
 
@@ -184,8 +273,10 @@ static int coresight_enable_link(struct coresight_device *csdev,
 	if (atomic_inc_return(&csdev->refcnt[refport]) == 1) {
 		if (link_ops(csdev)->enable) {
 			ret = link_ops(csdev)->enable(csdev, inport, outport);
-			if (ret)
+			if (ret) {
+				atomic_dec(&csdev->refcnt[refport]);
 				return ret;
+			}
 		}
 	}
 
@@ -274,13 +365,21 @@ static bool coresight_disable_source(struct coresight_device *csdev)
 	return !csdev->enable;
 }
 
-void coresight_disable_path(struct list_head *path)
+/*
+ * coresight_disable_path_from : Disable components in the given path beyond
+ * @nd in the list. If @nd is NULL, all the components, except the SOURCE are
+ * disabled.
+ */
+static void coresight_disable_path_from(struct list_head *path,
+					struct coresight_node *nd)
 {
 	u32 type;
-	struct coresight_node *nd;
 	struct coresight_device *csdev, *parent, *child;
 
-	list_for_each_entry(nd, path, link) {
+	if (!nd)
+		nd = list_first_entry(path, struct coresight_node, link);
+
+	list_for_each_entry_continue(nd, path, link) {
 		csdev = nd->csdev;
 		type = csdev->type;
 
@@ -300,7 +399,12 @@ void coresight_disable_path(struct list_head *path)
 			coresight_disable_sink(csdev);
 			break;
 		case CORESIGHT_DEV_TYPE_SOURCE:
-			/* sources are disabled from either sysFS or Perf */
+			/*
+			 * We skip the first node in the path assuming that it
+			 * is the source. So we don't expect a source device in
+			 * the middle of a path.
+			 */
+			WARN_ON(1);
 			break;
 		case CORESIGHT_DEV_TYPE_LINK:
 			parent = list_prev_entry(nd, link)->csdev;
@@ -313,7 +417,12 @@ void coresight_disable_path(struct list_head *path)
 	}
 }
 
-int coresight_enable_path(struct list_head *path, u32 mode)
+void coresight_disable_path(struct list_head *path)
+{
+	coresight_disable_path_from(path, NULL);
+}
+
+int coresight_enable_path(struct list_head *path, u32 mode, void *sink_data)
 {
 
 	int ret = 0;
@@ -338,9 +447,15 @@ int coresight_enable_path(struct list_head *path, u32 mode)
 
 		switch (type) {
 		case CORESIGHT_DEV_TYPE_SINK:
-			ret = coresight_enable_sink(csdev, mode);
+			ret = coresight_enable_sink(csdev, mode, sink_data);
+			/*
+			 * Sink is the first component turned on. If we
+			 * failed to enable the sink, there are no components
+			 * that need disabling. Disabling the path here
+			 * would mean we could disrupt an existing session.
+			 */
 			if (ret)
-				goto err;
+				goto out;
 			break;
 		case CORESIGHT_DEV_TYPE_SOURCE:
 			/* sources are enabled from either sysFS or Perf */
@@ -360,7 +475,7 @@ int coresight_enable_path(struct list_head *path, u32 mode)
 out:
 	return ret;
 err:
-	coresight_disable_path(path);
+	coresight_disable_path_from(path, nd);
 	goto out;
 }
 
@@ -635,7 +750,7 @@ int coresight_enable(struct coresight_device *csdev)
 		goto out;
 	}
 
-	ret = coresight_enable_path(path, CS_MODE_SYSFS);
+	ret = coresight_enable_path(path, CS_MODE_SYSFS, NULL);
 	if (ret)
 		goto err_path;
 
@@ -995,18 +1110,16 @@ postcore_initcall(coresight_init);
 
 struct coresight_device *coresight_register(struct coresight_desc *desc)
 {
-	int i;
 	int ret;
 	int link_subtype;
 	int nr_refcnts = 1;
 	atomic_t *refcnts = NULL;
 	struct coresight_device *csdev;
-	struct coresight_connection *conns = NULL;
 
 	csdev = kzalloc(sizeof(*csdev), GFP_KERNEL);
 	if (!csdev) {
 		ret = -ENOMEM;
-		goto err_kzalloc_csdev;
+		goto err_out;
 	}
 
 	if (desc->type == CORESIGHT_DEV_TYPE_LINK ||
@@ -1022,7 +1135,7 @@ struct coresight_device *coresight_register(struct coresight_desc *desc)
 	refcnts = kcalloc(nr_refcnts, sizeof(*refcnts), GFP_KERNEL);
 	if (!refcnts) {
 		ret = -ENOMEM;
-		goto err_kzalloc_refcnts;
+		goto err_free_csdev;
 	}
 
 	csdev->refcnt = refcnts;
@@ -1030,22 +1143,7 @@ struct coresight_device *coresight_register(struct coresight_desc *desc)
 	csdev->nr_inport = desc->pdata->nr_inport;
 	csdev->nr_outport = desc->pdata->nr_outport;
 
-	/* Initialise connections if there is at least one outport */
-	if (csdev->nr_outport) {
-		conns = kcalloc(csdev->nr_outport, sizeof(*conns), GFP_KERNEL);
-		if (!conns) {
-			ret = -ENOMEM;
-			goto err_kzalloc_conns;
-		}
-
-		for (i = 0; i < csdev->nr_outport; i++) {
-			conns[i].outport = desc->pdata->outports[i];
-			conns[i].child_name = desc->pdata->child_names[i];
-			conns[i].child_port = desc->pdata->child_ports[i];
-		}
-	}
-
-	csdev->conns = conns;
+	csdev->conns = desc->pdata->conns;
 
 	csdev->type = desc->type;
 	csdev->subtype = desc->subtype;
@@ -1062,7 +1160,11 @@ struct coresight_device *coresight_register(struct coresight_desc *desc)
 	ret = device_register(&csdev->dev);
 	if (ret) {
 		put_device(&csdev->dev);
-		goto err_kzalloc_csdev;
+		/*
+		 * All resources are free'd explicitly via
+		 * coresight_device_release(), triggered from put_device().
+		 */
+		goto err_out;
 	}
 
 	mutex_lock(&coresight_mutex);
@@ -1074,11 +1176,9 @@ struct coresight_device *coresight_register(struct coresight_desc *desc)
 
 	return csdev;
 
-err_kzalloc_conns:
-	kfree(refcnts);
-err_kzalloc_refcnts:
+err_free_csdev:
 	kfree(csdev);
-err_kzalloc_csdev:
+err_out:
 	return ERR_PTR(ret);
 }
 EXPORT_SYMBOL_GPL(coresight_register);
diff --git a/drivers/hwtracing/coresight/of_coresight.c b/drivers/hwtracing/coresight/of_coresight.c
index 6880bee195c8..89092f83567e 100644
--- a/drivers/hwtracing/coresight/of_coresight.c
+++ b/drivers/hwtracing/coresight/of_coresight.c
@@ -45,8 +45,13 @@ of_coresight_get_endpoint_device(struct device_node *endpoint)
 			       endpoint, of_dev_node_match);
 }
 
-static void of_coresight_get_ports(const struct device_node *node,
-				   int *nr_inport, int *nr_outport)
+static inline bool of_coresight_legacy_ep_is_input(struct device_node *ep)
+{
+	return of_property_read_bool(ep, "slave-mode");
+}
+
+static void of_coresight_get_ports_legacy(const struct device_node *node,
+					  int *nr_inport, int *nr_outport)
 {
 	struct device_node *ep = NULL;
 	int in = 0, out = 0;
@@ -56,7 +61,7 @@ static void of_coresight_get_ports(const struct device_node *node,
 		if (!ep)
 			break;
 
-		if (of_property_read_bool(ep, "slave-mode"))
+		if (of_coresight_legacy_ep_is_input(ep))
 			in++;
 		else
 			out++;
@@ -67,32 +72,77 @@ static void of_coresight_get_ports(const struct device_node *node,
 	*nr_outport = out;
 }
 
+static struct device_node *of_coresight_get_port_parent(struct device_node *ep)
+{
+	struct device_node *parent = of_graph_get_port_parent(ep);
+
+	/*
+	 * Skip one-level up to the real device node, if we
+	 * are using the new bindings.
+	 */
+	if (!of_node_cmp(parent->name, "in-ports") ||
+	    !of_node_cmp(parent->name, "out-ports"))
+		parent = of_get_next_parent(parent);
+
+	return parent;
+}
+
+static inline struct device_node *
+of_coresight_get_input_ports_node(const struct device_node *node)
+{
+	return of_get_child_by_name(node, "in-ports");
+}
+
+static inline struct device_node *
+of_coresight_get_output_ports_node(const struct device_node *node)
+{
+	return of_get_child_by_name(node, "out-ports");
+}
+
+static inline int
+of_coresight_count_ports(struct device_node *port_parent)
+{
+	int i = 0;
+	struct device_node *ep = NULL;
+
+	while ((ep = of_graph_get_next_endpoint(port_parent, ep)))
+		i++;
+	return i;
+}
+
+static void of_coresight_get_ports(const struct device_node *node,
+				   int *nr_inport, int *nr_outport)
+{
+	struct device_node *input_ports = NULL, *output_ports = NULL;
+
+	input_ports = of_coresight_get_input_ports_node(node);
+	output_ports = of_coresight_get_output_ports_node(node);
+
+	if (input_ports || output_ports) {
+		if (input_ports) {
+			*nr_inport = of_coresight_count_ports(input_ports);
+			of_node_put(input_ports);
+		}
+		if (output_ports) {
+			*nr_outport = of_coresight_count_ports(output_ports);
+			of_node_put(output_ports);
+		}
+	} else {
+		/* Fall back to legacy DT bindings parsing */
+		of_coresight_get_ports_legacy(node, nr_inport, nr_outport);
+	}
+}
+
 static int of_coresight_alloc_memory(struct device *dev,
 			struct coresight_platform_data *pdata)
 {
-	/* List of output port on this component */
-	pdata->outports = devm_kcalloc(dev,
-				       pdata->nr_outport,
-				       sizeof(*pdata->outports),
-				       GFP_KERNEL);
-	if (!pdata->outports)
-		return -ENOMEM;
-
-	/* Children connected to this component via @outports */
-	pdata->child_names = devm_kcalloc(dev,
-					  pdata->nr_outport,
-					  sizeof(*pdata->child_names),
-					  GFP_KERNEL);
-	if (!pdata->child_names)
-		return -ENOMEM;
-
-	/* Port number on the child this component is connected to */
-	pdata->child_ports = devm_kcalloc(dev,
-					  pdata->nr_outport,
-					  sizeof(*pdata->child_ports),
-					  GFP_KERNEL);
-	if (!pdata->child_ports)
-		return -ENOMEM;
+	if (pdata->nr_outport) {
+		pdata->conns = devm_kzalloc(dev, pdata->nr_outport *
+					    sizeof(*pdata->conns),
+					    GFP_KERNEL);
+		if (!pdata->conns)
+			return -ENOMEM;
+	}
 
 	return 0;
 }
@@ -114,17 +164,78 @@ int of_coresight_get_cpu(const struct device_node *node)
 }
 EXPORT_SYMBOL_GPL(of_coresight_get_cpu);
 
+/*
+ * of_coresight_parse_endpoint : Parse the given output endpoint @ep
+ * and fill the connection information in @conn
+ *
+ * Parses the local port, remote device name and the remote port.
+ *
+ * Returns :
+ *	 1	- If the parsing is successful and a connection record
+ *		  was created for an output connection.
+ *	 0	- If the parsing completed without any fatal errors.
+ *	-Errno	- Fatal error, abort the scanning.
+ */
+static int of_coresight_parse_endpoint(struct device *dev,
+				       struct device_node *ep,
+				       struct coresight_connection *conn)
+{
+	int ret = 0;
+	struct of_endpoint endpoint, rendpoint;
+	struct device_node *rparent = NULL;
+	struct device_node *rep = NULL;
+	struct device *rdev = NULL;
+
+	do {
+		/* Parse the local port details */
+		if (of_graph_parse_endpoint(ep, &endpoint))
+			break;
+		/*
+		 * Get a handle on the remote endpoint and the device it is
+		 * attached to.
+		 */
+		rep = of_graph_get_remote_endpoint(ep);
+		if (!rep)
+			break;
+		rparent = of_coresight_get_port_parent(rep);
+		if (!rparent)
+			break;
+		if (of_graph_parse_endpoint(rep, &rendpoint))
+			break;
+
+		/* If the remote device is not available, defer probing */
+		rdev = of_coresight_get_endpoint_device(rparent);
+		if (!rdev) {
+			ret = -EPROBE_DEFER;
+			break;
+		}
+
+		conn->outport = endpoint.port;
+		conn->child_name = devm_kstrdup(dev,
+						dev_name(rdev),
+						GFP_KERNEL);
+		conn->child_port = rendpoint.port;
+		/* Connection record updated */
+		ret = 1;
+	} while (0);
+
+	of_node_put(rparent);
+	of_node_put(rep);
+	put_device(rdev);
+
+	return ret;
+}
+
 struct coresight_platform_data *
 of_get_coresight_platform_data(struct device *dev,
 			       const struct device_node *node)
 {
-	int i = 0, ret = 0;
+	int ret = 0;
 	struct coresight_platform_data *pdata;
-	struct of_endpoint endpoint, rendpoint;
-	struct device *rdev;
+	struct coresight_connection *conn;
 	struct device_node *ep = NULL;
-	struct device_node *rparent = NULL;
-	struct device_node *rport = NULL;
+	const struct device_node *parent = NULL;
+	bool legacy_binding = false;
 
 	pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
 	if (!pdata)
@@ -132,63 +243,54 @@ of_get_coresight_platform_data(struct device *dev,
 
 	/* Use device name as sysfs handle */
 	pdata->name = dev_name(dev);
+	pdata->cpu = of_coresight_get_cpu(node);
 
 	/* Get the number of input and output port for this component */
 	of_coresight_get_ports(node, &pdata->nr_inport, &pdata->nr_outport);
 
-	if (pdata->nr_outport) {
-		ret = of_coresight_alloc_memory(dev, pdata);
-		if (ret)
-			return ERR_PTR(ret);
-
-		/* Iterate through each port to discover topology */
-		do {
-			/* Get a handle on a port */
-			ep = of_graph_get_next_endpoint(node, ep);
-			if (!ep)
-				break;
-
-			/*
-			 * No need to deal with input ports, processing for as
-			 * processing for output ports will deal with them.
-			 */
-			if (of_find_property(ep, "slave-mode", NULL))
-				continue;
-
-			/* Get a handle on the local endpoint */
-			ret = of_graph_parse_endpoint(ep, &endpoint);
-
-			if (ret)
-				continue;
-
-			/* The local out port number */
-			pdata->outports[i] = endpoint.port;
-
-			/*
-			 * Get a handle on the remote port and parent
-			 * attached to it.
-			 */
-			rparent = of_graph_get_remote_port_parent(ep);
-			rport = of_graph_get_remote_port(ep);
-
-			if (!rparent || !rport)
-				continue;
+	/* If there are no output connections, we are done */
+	if (!pdata->nr_outport)
+		return pdata;
 
-			if (of_graph_parse_endpoint(rport, &rendpoint))
-				continue;
+	ret = of_coresight_alloc_memory(dev, pdata);
+	if (ret)
+		return ERR_PTR(ret);
 
-			rdev = of_coresight_get_endpoint_device(rparent);
-			if (!rdev)
-				return ERR_PTR(-EPROBE_DEFER);
-
-			pdata->child_names[i] = dev_name(rdev);
-			pdata->child_ports[i] = rendpoint.id;
-
-			i++;
-		} while (ep);
+	parent = of_coresight_get_output_ports_node(node);
+	/*
+	 * If the DT uses obsoleted bindings, the ports are listed
+	 * under the device and we need to filter out the input
+	 * ports.
+	 */
+	if (!parent) {
+		legacy_binding = true;
+		parent = node;
+		dev_warn_once(dev, "Uses obsolete Coresight DT bindings\n");
 	}
 
-	pdata->cpu = of_coresight_get_cpu(node);
+	conn = pdata->conns;
+
+	/* Iterate through each output port to discover topology */
+	while ((ep = of_graph_get_next_endpoint(parent, ep))) {
+		/*
+		 * Legacy binding mixes input/output ports under the
+		 * same parent. So, skip the input ports if we are dealing
+		 * with legacy binding, as they processed with their
+		 * connected output ports.
+		 */
+		if (legacy_binding && of_coresight_legacy_ep_is_input(ep))
+			continue;
+
+		ret = of_coresight_parse_endpoint(dev, ep, conn);
+		switch (ret) {
+		case 1:
+			conn++;		/* Fall through */
+		case 0:
+			break;
+		default:
+			return ERR_PTR(ret);
+		}
+	}
 
 	return pdata;
 }
diff --git a/drivers/hwtracing/stm/Kconfig b/drivers/hwtracing/stm/Kconfig
index 723e2d90083d..752dd66742bf 100644
--- a/drivers/hwtracing/stm/Kconfig
+++ b/drivers/hwtracing/stm/Kconfig
@@ -11,6 +11,35 @@ config STM
 
 if STM
 
+config STM_PROTO_BASIC
+	tristate "Basic STM framing protocol driver"
+	default CONFIG_STM
+	help
+	  This is a simple framing protocol for sending data over STM
+	  devices. This was the protocol that the STM framework used
+	  exclusively until the MIPI SyS-T support was added. Use this
+	  driver for compatibility with your existing STM setup.
+
+	  The receiving side only needs to be able to decode the MIPI
+	  STP protocol in order to extract the data.
+
+	  If you want to be able to use the basic protocol or want the
+	  backwards compatibility for your existing setup, say Y.
+
+config STM_PROTO_SYS_T
+	tristate "MIPI SyS-T STM framing protocol driver"
+	default CONFIG_STM
+	help
+	  This is an implementation of MIPI SyS-T protocol to be used
+	  over the STP transport. In addition to the data payload, it
+	  also carries additional metadata for time correlation, better
+	  means of trace source identification, etc.
+
+	  The receiving side must be able to decode this protocol in
+	  addition to the MIPI STP, in order to extract the data.
+
+	  If you don't know what this is, say N.
+
 config STM_DUMMY
 	tristate "Dummy STM driver"
 	help
diff --git a/drivers/hwtracing/stm/Makefile b/drivers/hwtracing/stm/Makefile
index effc19e5190f..1692fcd29277 100644
--- a/drivers/hwtracing/stm/Makefile
+++ b/drivers/hwtracing/stm/Makefile
@@ -3,6 +3,12 @@ obj-$(CONFIG_STM)	+= stm_core.o
 
 stm_core-y		:= core.o policy.o
 
+obj-$(CONFIG_STM_PROTO_BASIC) += stm_p_basic.o
+obj-$(CONFIG_STM_PROTO_SYS_T) += stm_p_sys-t.o
+
+stm_p_basic-y		:= p_basic.o
+stm_p_sys-t-y		:= p_sys-t.o
+
 obj-$(CONFIG_STM_DUMMY)	+= dummy_stm.o
 
 obj-$(CONFIG_STM_SOURCE_CONSOLE)	+= stm_console.o
diff --git a/drivers/hwtracing/stm/core.c b/drivers/hwtracing/stm/core.c
index 10bcb5d73f90..93ce3aa740a9 100644
--- a/drivers/hwtracing/stm/core.c
+++ b/drivers/hwtracing/stm/core.c
@@ -293,15 +293,15 @@ static int stm_output_assign(struct stm_device *stm, unsigned int width,
 	if (width > stm->data->sw_nchannels)
 		return -EINVAL;
 
-	if (policy_node) {
-		stp_policy_node_get_ranges(policy_node,
-					   &midx, &mend, &cidx, &cend);
-	} else {
-		midx = stm->data->sw_start;
-		cidx = 0;
-		mend = stm->data->sw_end;
-		cend = stm->data->sw_nchannels - 1;
-	}
+	/* We no longer accept policy_node==NULL here */
+	if (WARN_ON_ONCE(!policy_node))
+		return -EINVAL;
+
+	/*
+	 * Also, the caller holds reference to policy_node, so it won't
+	 * disappear on us.
+	 */
+	stp_policy_node_get_ranges(policy_node, &midx, &mend, &cidx, &cend);
 
 	spin_lock(&stm->mc_lock);
 	spin_lock(&output->lock);
@@ -316,11 +316,26 @@ static int stm_output_assign(struct stm_device *stm, unsigned int width,
 	output->master = midx;
 	output->channel = cidx;
 	output->nr_chans = width;
+	if (stm->pdrv->output_open) {
+		void *priv = stp_policy_node_priv(policy_node);
+
+		if (WARN_ON_ONCE(!priv))
+			goto unlock;
+
+		/* configfs subsys mutex is held by the caller */
+		ret = stm->pdrv->output_open(priv, output);
+		if (ret)
+			goto unlock;
+	}
+
 	stm_output_claim(stm, output);
 	dev_dbg(&stm->dev, "assigned %u:%u (+%u)\n", midx, cidx, width);
 
 	ret = 0;
 unlock:
+	if (ret)
+		output->nr_chans = 0;
+
 	spin_unlock(&output->lock);
 	spin_unlock(&stm->mc_lock);
 
@@ -333,6 +348,8 @@ static void stm_output_free(struct stm_device *stm, struct stm_output *output)
 	spin_lock(&output->lock);
 	if (output->nr_chans)
 		stm_output_disclaim(stm, output);
+	if (stm->pdrv && stm->pdrv->output_close)
+		stm->pdrv->output_close(output);
 	spin_unlock(&output->lock);
 	spin_unlock(&stm->mc_lock);
 }
@@ -349,6 +366,127 @@ static int major_match(struct device *dev, const void *data)
 	return MAJOR(dev->devt) == major;
 }
 
+/*
+ * Framing protocol management
+ * Modules can implement STM protocol drivers and (un-)register them
+ * with the STM class framework.
+ */
+static struct list_head stm_pdrv_head;
+static struct mutex stm_pdrv_mutex;
+
+struct stm_pdrv_entry {
+	struct list_head			entry;
+	const struct stm_protocol_driver	*pdrv;
+	const struct config_item_type		*node_type;
+};
+
+static const struct stm_pdrv_entry *
+__stm_lookup_protocol(const char *name)
+{
+	struct stm_pdrv_entry *pe;
+
+	/*
+	 * If no name is given (NULL or ""), fall back to "p_basic".
+	 */
+	if (!name || !*name)
+		name = "p_basic";
+
+	list_for_each_entry(pe, &stm_pdrv_head, entry) {
+		if (!strcmp(name, pe->pdrv->name))
+			return pe;
+	}
+
+	return NULL;
+}
+
+int stm_register_protocol(const struct stm_protocol_driver *pdrv)
+{
+	struct stm_pdrv_entry *pe = NULL;
+	int ret = -ENOMEM;
+
+	mutex_lock(&stm_pdrv_mutex);
+
+	if (__stm_lookup_protocol(pdrv->name)) {
+		ret = -EEXIST;
+		goto unlock;
+	}
+
+	pe = kzalloc(sizeof(*pe), GFP_KERNEL);
+	if (!pe)
+		goto unlock;
+
+	if (pdrv->policy_attr) {
+		pe->node_type = get_policy_node_type(pdrv->policy_attr);
+		if (!pe->node_type)
+			goto unlock;
+	}
+
+	list_add_tail(&pe->entry, &stm_pdrv_head);
+	pe->pdrv = pdrv;
+
+	ret = 0;
+unlock:
+	mutex_unlock(&stm_pdrv_mutex);
+
+	if (ret)
+		kfree(pe);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(stm_register_protocol);
+
+void stm_unregister_protocol(const struct stm_protocol_driver *pdrv)
+{
+	struct stm_pdrv_entry *pe, *iter;
+
+	mutex_lock(&stm_pdrv_mutex);
+
+	list_for_each_entry_safe(pe, iter, &stm_pdrv_head, entry) {
+		if (pe->pdrv == pdrv) {
+			list_del(&pe->entry);
+
+			if (pe->node_type) {
+				kfree(pe->node_type->ct_attrs);
+				kfree(pe->node_type);
+			}
+			kfree(pe);
+			break;
+		}
+	}
+
+	mutex_unlock(&stm_pdrv_mutex);
+}
+EXPORT_SYMBOL_GPL(stm_unregister_protocol);
+
+static bool stm_get_protocol(const struct stm_protocol_driver *pdrv)
+{
+	return try_module_get(pdrv->owner);
+}
+
+void stm_put_protocol(const struct stm_protocol_driver *pdrv)
+{
+	module_put(pdrv->owner);
+}
+
+int stm_lookup_protocol(const char *name,
+			const struct stm_protocol_driver **pdrv,
+			const struct config_item_type **node_type)
+{
+	const struct stm_pdrv_entry *pe;
+
+	mutex_lock(&stm_pdrv_mutex);
+
+	pe = __stm_lookup_protocol(name);
+	if (pe && pe->pdrv && stm_get_protocol(pe->pdrv)) {
+		*pdrv = pe->pdrv;
+		*node_type = pe->node_type;
+	}
+
+	mutex_unlock(&stm_pdrv_mutex);
+
+	return pe ? 0 : -ENOENT;
+}
+
 static int stm_char_open(struct inode *inode, struct file *file)
 {
 	struct stm_file *stmf;
@@ -405,42 +543,81 @@ static int stm_char_release(struct inode *inode, struct file *file)
 	return 0;
 }
 
-static int stm_file_assign(struct stm_file *stmf, char *id, unsigned int width)
+static int
+stm_assign_first_policy(struct stm_device *stm, struct stm_output *output,
+			char **ids, unsigned int width)
 {
-	struct stm_device *stm = stmf->stm;
-	int ret;
+	struct stp_policy_node *pn;
+	int err, n;
 
-	stmf->policy_node = stp_policy_node_lookup(stm, id);
+	/*
+	 * On success, stp_policy_node_lookup() will return holding the
+	 * configfs subsystem mutex, which is then released in
+	 * stp_policy_node_put(). This allows the pdrv->output_open() in
+	 * stm_output_assign() to serialize against the attribute accessors.
+	 */
+	for (n = 0, pn = NULL; ids[n] && !pn; n++)
+		pn = stp_policy_node_lookup(stm, ids[n]);
 
-	ret = stm_output_assign(stm, width, stmf->policy_node, &stmf->output);
+	if (!pn)
+		return -EINVAL;
 
-	if (stmf->policy_node)
-		stp_policy_node_put(stmf->policy_node);
+	err = stm_output_assign(stm, width, pn, output);
 
-	return ret;
+	stp_policy_node_put(pn);
+
+	return err;
 }
 
-static ssize_t notrace stm_write(struct stm_data *data, unsigned int master,
-			  unsigned int channel, const char *buf, size_t count)
+/**
+ * stm_data_write() - send the given payload as data packets
+ * @data:	stm driver's data
+ * @m:		STP master
+ * @c:		STP channel
+ * @ts_first:	timestamp the first packet
+ * @buf:	data payload buffer
+ * @count:	data payload size
+ */
+ssize_t notrace stm_data_write(struct stm_data *data, unsigned int m,
+			       unsigned int c, bool ts_first, const void *buf,
+			       size_t count)
 {
-	unsigned int flags = STP_PACKET_TIMESTAMPED;
-	const unsigned char *p = buf, nil = 0;
-	size_t pos;
+	unsigned int flags = ts_first ? STP_PACKET_TIMESTAMPED : 0;
 	ssize_t sz;
+	size_t pos;
 
-	for (pos = 0, p = buf; count > pos; pos += sz, p += sz) {
+	for (pos = 0, sz = 0; pos < count; pos += sz) {
 		sz = min_t(unsigned int, count - pos, 8);
-		sz = data->packet(data, master, channel, STP_PACKET_DATA, flags,
-				  sz, p);
-		flags = 0;
-
-		if (sz < 0)
+		sz = data->packet(data, m, c, STP_PACKET_DATA, flags, sz,
+				  &((u8 *)buf)[pos]);
+		if (sz <= 0)
 			break;
+
+		if (ts_first) {
+			flags = 0;
+			ts_first = false;
+		}
 	}
 
-	data->packet(data, master, channel, STP_PACKET_FLAG, 0, 0, &nil);
+	return sz < 0 ? sz : pos;
+}
+EXPORT_SYMBOL_GPL(stm_data_write);
+
+static ssize_t notrace
+stm_write(struct stm_device *stm, struct stm_output *output,
+	  unsigned int chan, const char *buf, size_t count)
+{
+	int err;
+
+	/* stm->pdrv is serialized against policy_mutex */
+	if (!stm->pdrv)
+		return -ENODEV;
+
+	err = stm->pdrv->write(stm->data, output, chan, buf, count);
+	if (err < 0)
+		return err;
 
-	return pos;
+	return err;
 }
 
 static ssize_t stm_char_write(struct file *file, const char __user *buf,
@@ -455,16 +632,21 @@ static ssize_t stm_char_write(struct file *file, const char __user *buf,
 		count = PAGE_SIZE - 1;
 
 	/*
-	 * if no m/c have been assigned to this writer up to this
-	 * point, use "default" policy entry
+	 * If no m/c have been assigned to this writer up to this
+	 * point, try to use the task name and "default" policy entries.
 	 */
 	if (!stmf->output.nr_chans) {
-		err = stm_file_assign(stmf, "default", 1);
+		char comm[sizeof(current->comm)];
+		char *ids[] = { comm, "default", NULL };
+
+		get_task_comm(comm, current);
+
+		err = stm_assign_first_policy(stmf->stm, &stmf->output, ids, 1);
 		/*
 		 * EBUSY means that somebody else just assigned this
 		 * output, which is just fine for write()
 		 */
-		if (err && err != -EBUSY)
+		if (err)
 			return err;
 	}
 
@@ -480,8 +662,7 @@ static ssize_t stm_char_write(struct file *file, const char __user *buf,
 
 	pm_runtime_get_sync(&stm->dev);
 
-	count = stm_write(stm->data, stmf->output.master, stmf->output.channel,
-			  kbuf, count);
+	count = stm_write(stm, &stmf->output, 0, kbuf, count);
 
 	pm_runtime_mark_last_busy(&stm->dev);
 	pm_runtime_put_autosuspend(&stm->dev);
@@ -550,6 +731,7 @@ static int stm_char_policy_set_ioctl(struct stm_file *stmf, void __user *arg)
 {
 	struct stm_device *stm = stmf->stm;
 	struct stp_policy_id *id;
+	char *ids[] = { NULL, NULL };
 	int ret = -EINVAL;
 	u32 size;
 
@@ -582,7 +764,9 @@ static int stm_char_policy_set_ioctl(struct stm_file *stmf, void __user *arg)
 	    id->width > PAGE_SIZE / stm->data->sw_mmiosz)
 		goto err_free;
 
-	ret = stm_file_assign(stmf, id->id, id->width);
+	ids[0] = id->id;
+	ret = stm_assign_first_policy(stmf->stm, &stmf->output, ids,
+				      id->width);
 	if (ret)
 		goto err_free;
 
@@ -818,8 +1002,8 @@ EXPORT_SYMBOL_GPL(stm_unregister_device);
 static int stm_source_link_add(struct stm_source_device *src,
 			       struct stm_device *stm)
 {
-	char *id;
-	int err;
+	char *ids[] = { NULL, "default", NULL };
+	int err = -ENOMEM;
 
 	mutex_lock(&stm->link_mutex);
 	spin_lock(&stm->link_lock);
@@ -833,19 +1017,13 @@ static int stm_source_link_add(struct stm_source_device *src,
 	spin_unlock(&stm->link_lock);
 	mutex_unlock(&stm->link_mutex);
 
-	id = kstrdup(src->data->name, GFP_KERNEL);
-	if (id) {
-		src->policy_node =
-			stp_policy_node_lookup(stm, id);
-
-		kfree(id);
-	}
-
-	err = stm_output_assign(stm, src->data->nr_chans,
-				src->policy_node, &src->output);
+	ids[0] = kstrdup(src->data->name, GFP_KERNEL);
+	if (!ids[0])
+		goto fail_detach;
 
-	if (src->policy_node)
-		stp_policy_node_put(src->policy_node);
+	err = stm_assign_first_policy(stm, &src->output, ids,
+				      src->data->nr_chans);
+	kfree(ids[0]);
 
 	if (err)
 		goto fail_detach;
@@ -1134,9 +1312,7 @@ int notrace stm_source_write(struct stm_source_data *data,
 
 	stm = srcu_dereference(src->link, &stm_source_srcu);
 	if (stm)
-		count = stm_write(stm->data, src->output.master,
-				  src->output.channel + chan,
-				  buf, count);
+		count = stm_write(stm, &src->output, chan, buf, count);
 	else
 		count = -ENODEV;
 
@@ -1163,7 +1339,15 @@ static int __init stm_core_init(void)
 		goto err_src;
 
 	init_srcu_struct(&stm_source_srcu);
+	INIT_LIST_HEAD(&stm_pdrv_head);
+	mutex_init(&stm_pdrv_mutex);
 
+	/*
+	 * So as to not confuse existing users with a requirement
+	 * to load yet another module, do it here.
+	 */
+	if (IS_ENABLED(CONFIG_STM_PROTO_BASIC))
+		(void)request_module_nowait("stm_p_basic");
 	stm_core_up++;
 
 	return 0;
diff --git a/drivers/hwtracing/stm/heartbeat.c b/drivers/hwtracing/stm/heartbeat.c
index 7db42395e131..3e7df1c0477f 100644
--- a/drivers/hwtracing/stm/heartbeat.c
+++ b/drivers/hwtracing/stm/heartbeat.c
@@ -76,7 +76,7 @@ static int stm_heartbeat_init(void)
 			goto fail_unregister;
 
 		stm_heartbeat[i].data.nr_chans	= 1;
-		stm_heartbeat[i].data.link		= stm_heartbeat_link;
+		stm_heartbeat[i].data.link	= stm_heartbeat_link;
 		stm_heartbeat[i].data.unlink	= stm_heartbeat_unlink;
 		hrtimer_init(&stm_heartbeat[i].hrtimer, CLOCK_MONOTONIC,
 			     HRTIMER_MODE_ABS);
diff --git a/drivers/hwtracing/stm/p_basic.c b/drivers/hwtracing/stm/p_basic.c
new file mode 100644
index 000000000000..8980a6a5fd6c
--- /dev/null
+++ b/drivers/hwtracing/stm/p_basic.c
@@ -0,0 +1,48 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Basic framing protocol for STM devices.
+ * Copyright (c) 2018, Intel Corporation.
+ */
+
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/stm.h>
+#include "stm.h"
+
+static ssize_t basic_write(struct stm_data *data, struct stm_output *output,
+			   unsigned int chan, const char *buf, size_t count)
+{
+	unsigned int c = output->channel + chan;
+	unsigned int m = output->master;
+	const unsigned char nil = 0;
+	ssize_t sz;
+
+	sz = stm_data_write(data, m, c, true, buf, count);
+	if (sz > 0)
+		data->packet(data, m, c, STP_PACKET_FLAG, 0, 0, &nil);
+
+	return sz;
+}
+
+static const struct stm_protocol_driver basic_pdrv = {
+	.owner	= THIS_MODULE,
+	.name	= "p_basic",
+	.write	= basic_write,
+};
+
+static int basic_stm_init(void)
+{
+	return stm_register_protocol(&basic_pdrv);
+}
+
+static void basic_stm_exit(void)
+{
+	stm_unregister_protocol(&basic_pdrv);
+}
+
+module_init(basic_stm_init);
+module_exit(basic_stm_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("Basic STM framing protocol driver");
+MODULE_AUTHOR("Alexander Shishkin <alexander.shishkin@linux.intel.com>");
diff --git a/drivers/hwtracing/stm/p_sys-t.c b/drivers/hwtracing/stm/p_sys-t.c
new file mode 100644
index 000000000000..b178a5495b67
--- /dev/null
+++ b/drivers/hwtracing/stm/p_sys-t.c
@@ -0,0 +1,382 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * MIPI SyS-T framing protocol for STM devices.
+ * Copyright (c) 2018, Intel Corporation.
+ */
+
+#include <linux/configfs.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/slab.h>
+#include <linux/uuid.h>
+#include <linux/stm.h>
+#include "stm.h"
+
+enum sys_t_message_type {
+	MIPI_SYST_TYPE_BUILD	= 0,
+	MIPI_SYST_TYPE_SHORT32,
+	MIPI_SYST_TYPE_STRING,
+	MIPI_SYST_TYPE_CATALOG,
+	MIPI_SYST_TYPE_RAW	= 6,
+	MIPI_SYST_TYPE_SHORT64,
+	MIPI_SYST_TYPE_CLOCK,
+};
+
+enum sys_t_message_severity {
+	MIPI_SYST_SEVERITY_MAX	= 0,
+	MIPI_SYST_SEVERITY_FATAL,
+	MIPI_SYST_SEVERITY_ERROR,
+	MIPI_SYST_SEVERITY_WARNING,
+	MIPI_SYST_SEVERITY_INFO,
+	MIPI_SYST_SEVERITY_USER1,
+	MIPI_SYST_SEVERITY_USER2,
+	MIPI_SYST_SEVERITY_DEBUG,
+};
+
+enum sys_t_message_build_subtype {
+	MIPI_SYST_BUILD_ID_COMPACT32 = 0,
+	MIPI_SYST_BUILD_ID_COMPACT64,
+	MIPI_SYST_BUILD_ID_LONG,
+};
+
+enum sys_t_message_clock_subtype {
+	MIPI_SYST_CLOCK_TRANSPORT_SYNC = 1,
+};
+
+enum sys_t_message_string_subtype {
+	MIPI_SYST_STRING_GENERIC	= 1,
+	MIPI_SYST_STRING_FUNCTIONENTER,
+	MIPI_SYST_STRING_FUNCTIONEXIT,
+	MIPI_SYST_STRING_INVALIDPARAM	= 5,
+	MIPI_SYST_STRING_ASSERT		= 7,
+	MIPI_SYST_STRING_PRINTF_32	= 11,
+	MIPI_SYST_STRING_PRINTF_64	= 12,
+};
+
+#define MIPI_SYST_TYPE(t)		((u32)(MIPI_SYST_TYPE_ ## t))
+#define MIPI_SYST_SEVERITY(s)		((u32)(MIPI_SYST_SEVERITY_ ## s) << 4)
+#define MIPI_SYST_OPT_LOC		BIT(8)
+#define MIPI_SYST_OPT_LEN		BIT(9)
+#define MIPI_SYST_OPT_CHK		BIT(10)
+#define MIPI_SYST_OPT_TS		BIT(11)
+#define MIPI_SYST_UNIT(u)		((u32)(u) << 12)
+#define MIPI_SYST_ORIGIN(o)		((u32)(o) << 16)
+#define MIPI_SYST_OPT_GUID		BIT(23)
+#define MIPI_SYST_SUBTYPE(s)		((u32)(MIPI_SYST_ ## s) << 24)
+#define MIPI_SYST_UNITLARGE(u)		(MIPI_SYST_UNIT(u & 0xf) | \
+					 MIPI_SYST_ORIGIN(u >> 4))
+#define MIPI_SYST_TYPES(t, s)		(MIPI_SYST_TYPE(t) | \
+					 MIPI_SYST_SUBTYPE(t ## _ ## s))
+
+#define DATA_HEADER	(MIPI_SYST_TYPES(STRING, GENERIC)	| \
+			 MIPI_SYST_SEVERITY(INFO)		| \
+			 MIPI_SYST_OPT_GUID)
+
+#define CLOCK_SYNC_HEADER	(MIPI_SYST_TYPES(CLOCK, TRANSPORT_SYNC)	| \
+				 MIPI_SYST_SEVERITY(MAX))
+
+struct sys_t_policy_node {
+	uuid_t		uuid;
+	bool		do_len;
+	unsigned long	ts_interval;
+	unsigned long	clocksync_interval;
+};
+
+struct sys_t_output {
+	struct sys_t_policy_node	node;
+	unsigned long	ts_jiffies;
+	unsigned long	clocksync_jiffies;
+};
+
+static void sys_t_policy_node_init(void *priv)
+{
+	struct sys_t_policy_node *pn = priv;
+
+	generate_random_uuid(pn->uuid.b);
+}
+
+static int sys_t_output_open(void *priv, struct stm_output *output)
+{
+	struct sys_t_policy_node *pn = priv;
+	struct sys_t_output *opriv;
+
+	opriv = kzalloc(sizeof(*opriv), GFP_ATOMIC);
+	if (!opriv)
+		return -ENOMEM;
+
+	memcpy(&opriv->node, pn, sizeof(opriv->node));
+	output->pdrv_private = opriv;
+
+	return 0;
+}
+
+static void sys_t_output_close(struct stm_output *output)
+{
+	kfree(output->pdrv_private);
+}
+
+static ssize_t sys_t_policy_uuid_show(struct config_item *item,
+				      char *page)
+{
+	struct sys_t_policy_node *pn = to_pdrv_policy_node(item);
+
+	return sprintf(page, "%pU\n", &pn->uuid);
+}
+
+static ssize_t
+sys_t_policy_uuid_store(struct config_item *item, const char *page,
+			size_t count)
+{
+	struct mutex *mutexp = &item->ci_group->cg_subsys->su_mutex;
+	struct sys_t_policy_node *pn = to_pdrv_policy_node(item);
+	int ret;
+
+	mutex_lock(mutexp);
+	ret = uuid_parse(page, &pn->uuid);
+	mutex_unlock(mutexp);
+
+	return ret < 0 ? ret : count;
+}
+
+CONFIGFS_ATTR(sys_t_policy_, uuid);
+
+static ssize_t sys_t_policy_do_len_show(struct config_item *item,
+				      char *page)
+{
+	struct sys_t_policy_node *pn = to_pdrv_policy_node(item);
+
+	return sprintf(page, "%d\n", pn->do_len);
+}
+
+static ssize_t
+sys_t_policy_do_len_store(struct config_item *item, const char *page,
+			size_t count)
+{
+	struct mutex *mutexp = &item->ci_group->cg_subsys->su_mutex;
+	struct sys_t_policy_node *pn = to_pdrv_policy_node(item);
+	int ret;
+
+	mutex_lock(mutexp);
+	ret = kstrtobool(page, &pn->do_len);
+	mutex_unlock(mutexp);
+
+	return ret ? ret : count;
+}
+
+CONFIGFS_ATTR(sys_t_policy_, do_len);
+
+static ssize_t sys_t_policy_ts_interval_show(struct config_item *item,
+					     char *page)
+{
+	struct sys_t_policy_node *pn = to_pdrv_policy_node(item);
+
+	return sprintf(page, "%u\n", jiffies_to_msecs(pn->ts_interval));
+}
+
+static ssize_t
+sys_t_policy_ts_interval_store(struct config_item *item, const char *page,
+			       size_t count)
+{
+	struct mutex *mutexp = &item->ci_group->cg_subsys->su_mutex;
+	struct sys_t_policy_node *pn = to_pdrv_policy_node(item);
+	unsigned int ms;
+	int ret;
+
+	mutex_lock(mutexp);
+	ret = kstrtouint(page, 10, &ms);
+	mutex_unlock(mutexp);
+
+	if (!ret) {
+		pn->ts_interval = msecs_to_jiffies(ms);
+		return count;
+	}
+
+	return ret;
+}
+
+CONFIGFS_ATTR(sys_t_policy_, ts_interval);
+
+static ssize_t sys_t_policy_clocksync_interval_show(struct config_item *item,
+						    char *page)
+{
+	struct sys_t_policy_node *pn = to_pdrv_policy_node(item);
+
+	return sprintf(page, "%u\n", jiffies_to_msecs(pn->clocksync_interval));
+}
+
+static ssize_t
+sys_t_policy_clocksync_interval_store(struct config_item *item,
+				      const char *page, size_t count)
+{
+	struct mutex *mutexp = &item->ci_group->cg_subsys->su_mutex;
+	struct sys_t_policy_node *pn = to_pdrv_policy_node(item);
+	unsigned int ms;
+	int ret;
+
+	mutex_lock(mutexp);
+	ret = kstrtouint(page, 10, &ms);
+	mutex_unlock(mutexp);
+
+	if (!ret) {
+		pn->clocksync_interval = msecs_to_jiffies(ms);
+		return count;
+	}
+
+	return ret;
+}
+
+CONFIGFS_ATTR(sys_t_policy_, clocksync_interval);
+
+static struct configfs_attribute *sys_t_policy_attrs[] = {
+	&sys_t_policy_attr_uuid,
+	&sys_t_policy_attr_do_len,
+	&sys_t_policy_attr_ts_interval,
+	&sys_t_policy_attr_clocksync_interval,
+	NULL,
+};
+
+static inline bool sys_t_need_ts(struct sys_t_output *op)
+{
+	if (op->node.ts_interval &&
+	    time_after(op->ts_jiffies + op->node.ts_interval, jiffies)) {
+		op->ts_jiffies = jiffies;
+
+		return true;
+	}
+
+	return false;
+}
+
+static bool sys_t_need_clock_sync(struct sys_t_output *op)
+{
+	if (op->node.clocksync_interval &&
+	    time_after(op->clocksync_jiffies + op->node.clocksync_interval,
+		       jiffies)) {
+		op->clocksync_jiffies = jiffies;
+
+		return true;
+	}
+
+	return false;
+}
+
+static ssize_t
+sys_t_clock_sync(struct stm_data *data, unsigned int m, unsigned int c)
+{
+	u32 header = CLOCK_SYNC_HEADER;
+	const unsigned char nil = 0;
+	u64 payload[2]; /* Clock value and frequency */
+	ssize_t sz;
+
+	sz = data->packet(data, m, c, STP_PACKET_DATA, STP_PACKET_TIMESTAMPED,
+			  4, (u8 *)&header);
+	if (sz <= 0)
+		return sz;
+
+	payload[0] = ktime_get_real_ns();
+	payload[1] = NSEC_PER_SEC;
+	sz = stm_data_write(data, m, c, false, &payload, sizeof(payload));
+	if (sz <= 0)
+		return sz;
+
+	data->packet(data, m, c, STP_PACKET_FLAG, 0, 0, &nil);
+
+	return sizeof(header) + sizeof(payload);
+}
+
+static ssize_t sys_t_write(struct stm_data *data, struct stm_output *output,
+			   unsigned int chan, const char *buf, size_t count)
+{
+	struct sys_t_output *op = output->pdrv_private;
+	unsigned int c = output->channel + chan;
+	unsigned int m = output->master;
+	const unsigned char nil = 0;
+	u32 header = DATA_HEADER;
+	ssize_t sz;
+
+	/* We require an existing policy node to proceed */
+	if (!op)
+		return -EINVAL;
+
+	if (sys_t_need_clock_sync(op)) {
+		sz = sys_t_clock_sync(data, m, c);
+		if (sz <= 0)
+			return sz;
+	}
+
+	if (op->node.do_len)
+		header |= MIPI_SYST_OPT_LEN;
+	if (sys_t_need_ts(op))
+		header |= MIPI_SYST_OPT_TS;
+
+	/*
+	 * STP framing rules for SyS-T frames:
+	 *   * the first packet of the SyS-T frame is timestamped;
+	 *   * the last packet is a FLAG.
+	 */
+	/* Message layout: HEADER / GUID / [LENGTH /][TIMESTAMP /] DATA */
+	/* HEADER */
+	sz = data->packet(data, m, c, STP_PACKET_DATA, STP_PACKET_TIMESTAMPED,
+			  4, (u8 *)&header);
+	if (sz <= 0)
+		return sz;
+
+	/* GUID */
+	sz = stm_data_write(data, m, c, false, op->node.uuid.b, UUID_SIZE);
+	if (sz <= 0)
+		return sz;
+
+	/* [LENGTH] */
+	if (op->node.do_len) {
+		u16 length = count;
+
+		sz = data->packet(data, m, c, STP_PACKET_DATA, 0, 2,
+				  (u8 *)&length);
+		if (sz <= 0)
+			return sz;
+	}
+
+	/* [TIMESTAMP] */
+	if (header & MIPI_SYST_OPT_TS) {
+		u64 ts = ktime_get_real_ns();
+
+		sz = stm_data_write(data, m, c, false, &ts, sizeof(ts));
+		if (sz <= 0)
+			return sz;
+	}
+
+	/* DATA */
+	sz = stm_data_write(data, m, c, false, buf, count);
+	if (sz > 0)
+		data->packet(data, m, c, STP_PACKET_FLAG, 0, 0, &nil);
+
+	return sz;
+}
+
+static const struct stm_protocol_driver sys_t_pdrv = {
+	.owner			= THIS_MODULE,
+	.name			= "p_sys-t",
+	.priv_sz		= sizeof(struct sys_t_policy_node),
+	.write			= sys_t_write,
+	.policy_attr		= sys_t_policy_attrs,
+	.policy_node_init	= sys_t_policy_node_init,
+	.output_open		= sys_t_output_open,
+	.output_close		= sys_t_output_close,
+};
+
+static int sys_t_stm_init(void)
+{
+	return stm_register_protocol(&sys_t_pdrv);
+}
+
+static void sys_t_stm_exit(void)
+{
+	stm_unregister_protocol(&sys_t_pdrv);
+}
+
+module_init(sys_t_stm_init);
+module_exit(sys_t_stm_exit);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("MIPI SyS-T STM framing protocol driver");
+MODULE_AUTHOR("Alexander Shishkin <alexander.shishkin@linux.intel.com>");
diff --git a/drivers/hwtracing/stm/policy.c b/drivers/hwtracing/stm/policy.c
index 3fd07e275b34..0910ec807187 100644
--- a/drivers/hwtracing/stm/policy.c
+++ b/drivers/hwtracing/stm/policy.c
@@ -33,8 +33,18 @@ struct stp_policy_node {
 	unsigned int		last_master;
 	unsigned int		first_channel;
 	unsigned int		last_channel;
+	/* this is the one that's exposed to the attributes */
+	unsigned char		priv[0];
 };
 
+void *stp_policy_node_priv(struct stp_policy_node *pn)
+{
+	if (!pn)
+		return NULL;
+
+	return pn->priv;
+}
+
 static struct configfs_subsystem stp_policy_subsys;
 
 void stp_policy_node_get_ranges(struct stp_policy_node *policy_node,
@@ -68,6 +78,14 @@ to_stp_policy_node(struct config_item *item)
 		NULL;
 }
 
+void *to_pdrv_policy_node(struct config_item *item)
+{
+	struct stp_policy_node *node = to_stp_policy_node(item);
+
+	return stp_policy_node_priv(node);
+}
+EXPORT_SYMBOL_GPL(to_pdrv_policy_node);
+
 static ssize_t
 stp_policy_node_masters_show(struct config_item *item, char *page)
 {
@@ -163,7 +181,9 @@ unlock:
 
 static void stp_policy_node_release(struct config_item *item)
 {
-	kfree(to_stp_policy_node(item));
+	struct stp_policy_node *node = to_stp_policy_node(item);
+
+	kfree(node);
 }
 
 static struct configfs_item_operations stp_policy_node_item_ops = {
@@ -182,10 +202,34 @@ static struct configfs_attribute *stp_policy_node_attrs[] = {
 static const struct config_item_type stp_policy_type;
 static const struct config_item_type stp_policy_node_type;
 
+const struct config_item_type *
+get_policy_node_type(struct configfs_attribute **attrs)
+{
+	struct config_item_type *type;
+	struct configfs_attribute **merged;
+
+	type = kmemdup(&stp_policy_node_type, sizeof(stp_policy_node_type),
+		       GFP_KERNEL);
+	if (!type)
+		return NULL;
+
+	merged = memcat_p(stp_policy_node_attrs, attrs);
+	if (!merged) {
+		kfree(type);
+		return NULL;
+	}
+
+	type->ct_attrs = merged;
+
+	return type;
+}
+
 static struct config_group *
 stp_policy_node_make(struct config_group *group, const char *name)
 {
+	const struct config_item_type *type = &stp_policy_node_type;
 	struct stp_policy_node *policy_node, *parent_node;
+	const struct stm_protocol_driver *pdrv;
 	struct stp_policy *policy;
 
 	if (group->cg_item.ci_type == &stp_policy_type) {
@@ -199,12 +243,20 @@ stp_policy_node_make(struct config_group *group, const char *name)
 	if (!policy->stm)
 		return ERR_PTR(-ENODEV);
 
-	policy_node = kzalloc(sizeof(struct stp_policy_node), GFP_KERNEL);
+	pdrv = policy->stm->pdrv;
+	policy_node =
+		kzalloc(offsetof(struct stp_policy_node, priv[pdrv->priv_sz]),
+			GFP_KERNEL);
 	if (!policy_node)
 		return ERR_PTR(-ENOMEM);
 
-	config_group_init_type_name(&policy_node->group, name,
-				    &stp_policy_node_type);
+	if (pdrv->policy_node_init)
+		pdrv->policy_node_init((void *)policy_node->priv);
+
+	if (policy->stm->pdrv_node_type)
+		type = policy->stm->pdrv_node_type;
+
+	config_group_init_type_name(&policy_node->group, name, type);
 
 	policy_node->policy = policy;
 
@@ -254,8 +306,25 @@ static ssize_t stp_policy_device_show(struct config_item *item,
 
 CONFIGFS_ATTR_RO(stp_policy_, device);
 
+static ssize_t stp_policy_protocol_show(struct config_item *item,
+					char *page)
+{
+	struct stp_policy *policy = to_stp_policy(item);
+	ssize_t count;
+
+	count = sprintf(page, "%s\n",
+			(policy && policy->stm) ?
+			policy->stm->pdrv->name :
+			"<none>");
+
+	return count;
+}
+
+CONFIGFS_ATTR_RO(stp_policy_, protocol);
+
 static struct configfs_attribute *stp_policy_attrs[] = {
 	&stp_policy_attr_device,
+	&stp_policy_attr_protocol,
 	NULL,
 };
 
@@ -276,6 +345,7 @@ void stp_policy_unbind(struct stp_policy *policy)
 	stm->policy = NULL;
 	policy->stm = NULL;
 
+	stm_put_protocol(stm->pdrv);
 	stm_put_device(stm);
 }
 
@@ -311,11 +381,14 @@ static const struct config_item_type stp_policy_type = {
 };
 
 static struct config_group *
-stp_policies_make(struct config_group *group, const char *name)
+stp_policy_make(struct config_group *group, const char *name)
 {
+	const struct config_item_type *pdrv_node_type;
+	const struct stm_protocol_driver *pdrv;
+	char *devname, *proto, *p;
 	struct config_group *ret;
 	struct stm_device *stm;
-	char *devname, *p;
+	int err;
 
 	devname = kasprintf(GFP_KERNEL, "%s", name);
 	if (!devname)
@@ -326,6 +399,7 @@ stp_policies_make(struct config_group *group, const char *name)
 	 * <device_name> is the name of an existing stm device; may
 	 *               contain dots;
 	 * <policy_name> is an arbitrary string; may not contain dots
+	 * <device_name>:<protocol_name>.<policy_name>
 	 */
 	p = strrchr(devname, '.');
 	if (!p) {
@@ -335,11 +409,28 @@ stp_policies_make(struct config_group *group, const char *name)
 
 	*p = '\0';
 
+	/*
+	 * look for ":<protocol_name>":
+	 *  + no protocol suffix: fall back to whatever is available;
+	 *  + unknown protocol: fail the whole thing
+	 */
+	proto = strrchr(devname, ':');
+	if (proto)
+		*proto++ = '\0';
+
 	stm = stm_find_device(devname);
+	if (!stm) {
+		kfree(devname);
+		return ERR_PTR(-ENODEV);
+	}
+
+	err = stm_lookup_protocol(proto, &pdrv, &pdrv_node_type);
 	kfree(devname);
 
-	if (!stm)
+	if (err) {
+		stm_put_device(stm);
 		return ERR_PTR(-ENODEV);
+	}
 
 	mutex_lock(&stm->policy_mutex);
 	if (stm->policy) {
@@ -349,31 +440,37 @@ stp_policies_make(struct config_group *group, const char *name)
 
 	stm->policy = kzalloc(sizeof(*stm->policy), GFP_KERNEL);
 	if (!stm->policy) {
-		ret = ERR_PTR(-ENOMEM);
-		goto unlock_policy;
+		mutex_unlock(&stm->policy_mutex);
+		stm_put_protocol(pdrv);
+		stm_put_device(stm);
+		return ERR_PTR(-ENOMEM);
 	}
 
 	config_group_init_type_name(&stm->policy->group, name,
 				    &stp_policy_type);
-	stm->policy->stm = stm;
 
+	stm->pdrv = pdrv;
+	stm->pdrv_node_type = pdrv_node_type;
+	stm->policy->stm = stm;
 	ret = &stm->policy->group;
 
 unlock_policy:
 	mutex_unlock(&stm->policy_mutex);
 
-	if (IS_ERR(ret))
+	if (IS_ERR(ret)) {
+		stm_put_protocol(stm->pdrv);
 		stm_put_device(stm);
+	}
 
 	return ret;
 }
 
-static struct configfs_group_operations stp_policies_group_ops = {
-	.make_group	= stp_policies_make,
+static struct configfs_group_operations stp_policy_root_group_ops = {
+	.make_group	= stp_policy_make,
 };
 
-static const struct config_item_type stp_policies_type = {
-	.ct_group_ops	= &stp_policies_group_ops,
+static const struct config_item_type stp_policy_root_type = {
+	.ct_group_ops	= &stp_policy_root_group_ops,
 	.ct_owner	= THIS_MODULE,
 };
 
@@ -381,7 +478,7 @@ static struct configfs_subsystem stp_policy_subsys = {
 	.su_group = {
 		.cg_item = {
 			.ci_namebuf	= "stp-policy",
-			.ci_type	= &stp_policies_type,
+			.ci_type	= &stp_policy_root_type,
 		},
 	},
 };
@@ -392,7 +489,7 @@ static struct configfs_subsystem stp_policy_subsys = {
 static struct stp_policy_node *
 __stp_policy_node_lookup(struct stp_policy *policy, char *s)
 {
-	struct stp_policy_node *policy_node, *ret;
+	struct stp_policy_node *policy_node, *ret = NULL;
 	struct list_head *head = &policy->group.cg_children;
 	struct config_item *item;
 	char *start, *end = s;
@@ -400,10 +497,6 @@ __stp_policy_node_lookup(struct stp_policy *policy, char *s)
 	if (list_empty(head))
 		return NULL;
 
-	/* return the first entry if everything else fails */
-	item = list_entry(head->next, struct config_item, ci_entry);
-	ret = to_stp_policy_node(item);
-
 next:
 	for (;;) {
 		start = strsep(&end, "/");
@@ -449,25 +542,25 @@ stp_policy_node_lookup(struct stm_device *stm, char *s)
 
 	if (policy_node)
 		config_item_get(&policy_node->group.cg_item);
-	mutex_unlock(&stp_policy_subsys.su_mutex);
+	else
+		mutex_unlock(&stp_policy_subsys.su_mutex);
 
 	return policy_node;
 }
 
 void stp_policy_node_put(struct stp_policy_node *policy_node)
 {
+	lockdep_assert_held(&stp_policy_subsys.su_mutex);
+
+	mutex_unlock(&stp_policy_subsys.su_mutex);
 	config_item_put(&policy_node->group.cg_item);
 }
 
 int __init stp_configfs_init(void)
 {
-	int err;
-
 	config_group_init(&stp_policy_subsys.su_group);
 	mutex_init(&stp_policy_subsys.su_mutex);
-	err = configfs_register_subsystem(&stp_policy_subsys);
-
-	return err;
+	return configfs_register_subsystem(&stp_policy_subsys);
 }
 
 void __exit stp_configfs_exit(void)
diff --git a/drivers/hwtracing/stm/stm.h b/drivers/hwtracing/stm/stm.h
index 923571adc6f4..3569439d53bb 100644
--- a/drivers/hwtracing/stm/stm.h
+++ b/drivers/hwtracing/stm/stm.h
@@ -10,20 +10,17 @@
 #ifndef _STM_STM_H_
 #define _STM_STM_H_
 
+#include <linux/configfs.h>
+
 struct stp_policy;
 struct stp_policy_node;
+struct stm_protocol_driver;
 
-struct stp_policy_node *
-stp_policy_node_lookup(struct stm_device *stm, char *s);
-void stp_policy_node_put(struct stp_policy_node *policy_node);
-void stp_policy_unbind(struct stp_policy *policy);
-
-void stp_policy_node_get_ranges(struct stp_policy_node *policy_node,
-				unsigned int *mstart, unsigned int *mend,
-				unsigned int *cstart, unsigned int *cend);
 int stp_configfs_init(void);
 void stp_configfs_exit(void);
 
+void *stp_policy_node_priv(struct stp_policy_node *pn);
+
 struct stp_master {
 	unsigned int	nr_free;
 	unsigned long	chan_map[0];
@@ -40,6 +37,9 @@ struct stm_device {
 	struct mutex		link_mutex;
 	spinlock_t		link_lock;
 	struct list_head	link_list;
+	/* framing protocol in use */
+	const struct stm_protocol_driver	*pdrv;
+	const struct config_item_type		*pdrv_node_type;
 	/* master allocation */
 	spinlock_t		mc_lock;
 	struct stp_master	*masters[0];
@@ -48,16 +48,28 @@ struct stm_device {
 #define to_stm_device(_d)				\
 	container_of((_d), struct stm_device, dev)
 
+struct stp_policy_node *
+stp_policy_node_lookup(struct stm_device *stm, char *s);
+void stp_policy_node_put(struct stp_policy_node *policy_node);
+void stp_policy_unbind(struct stp_policy *policy);
+
+void stp_policy_node_get_ranges(struct stp_policy_node *policy_node,
+				unsigned int *mstart, unsigned int *mend,
+				unsigned int *cstart, unsigned int *cend);
+
+const struct config_item_type *
+get_policy_node_type(struct configfs_attribute **attrs);
+
 struct stm_output {
 	spinlock_t		lock;
 	unsigned int		master;
 	unsigned int		channel;
 	unsigned int		nr_chans;
+	void			*pdrv_private;
 };
 
 struct stm_file {
 	struct stm_device	*stm;
-	struct stp_policy_node	*policy_node;
 	struct stm_output	output;
 };
 
@@ -71,11 +83,35 @@ struct stm_source_device {
 	struct stm_device __rcu	*link;
 	struct list_head	link_entry;
 	/* one output per stm_source device */
-	struct stp_policy_node	*policy_node;
 	struct stm_output	output;
 };
 
 #define to_stm_source_device(_d)				\
 	container_of((_d), struct stm_source_device, dev)
 
+void *to_pdrv_policy_node(struct config_item *item);
+
+struct stm_protocol_driver {
+	struct module	*owner;
+	const char	*name;
+	ssize_t		(*write)(struct stm_data *data,
+				 struct stm_output *output, unsigned int chan,
+				 const char *buf, size_t count);
+	void		(*policy_node_init)(void *arg);
+	int		(*output_open)(void *priv, struct stm_output *output);
+	void		(*output_close)(struct stm_output *output);
+	ssize_t		priv_sz;
+	struct configfs_attribute	**policy_attr;
+};
+
+int stm_register_protocol(const struct stm_protocol_driver *pdrv);
+void stm_unregister_protocol(const struct stm_protocol_driver *pdrv);
+int stm_lookup_protocol(const char *name,
+			const struct stm_protocol_driver **pdrv,
+			const struct config_item_type **type);
+void stm_put_protocol(const struct stm_protocol_driver *pdrv);
+ssize_t stm_data_write(struct stm_data *data, unsigned int m,
+		       unsigned int c, bool ts_first, const void *buf,
+		       size_t count);
+
 #endif /* _STM_STM_H_ */
diff --git a/drivers/misc/ad525x_dpot-i2c.c b/drivers/misc/ad525x_dpot-i2c.c
index 4f832002d116..1827c69959fb 100644
--- a/drivers/misc/ad525x_dpot-i2c.c
+++ b/drivers/misc/ad525x_dpot-i2c.c
@@ -114,6 +114,6 @@ static struct i2c_driver ad_dpot_i2c_driver = {
 
 module_i2c_driver(ad_dpot_i2c_driver);
 
-MODULE_AUTHOR("Michael Hennerich <hennerich@blackfin.uclinux.org>");
+MODULE_AUTHOR("Michael Hennerich <michael.hennerich@analog.com>");
 MODULE_DESCRIPTION("digital potentiometer I2C bus driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/misc/ad525x_dpot-spi.c b/drivers/misc/ad525x_dpot-spi.c
index 39a7f517ee7e..0383ec153725 100644
--- a/drivers/misc/ad525x_dpot-spi.c
+++ b/drivers/misc/ad525x_dpot-spi.c
@@ -140,7 +140,7 @@ static struct spi_driver ad_dpot_spi_driver = {
 
 module_spi_driver(ad_dpot_spi_driver);
 
-MODULE_AUTHOR("Michael Hennerich <hennerich@blackfin.uclinux.org>");
+MODULE_AUTHOR("Michael Hennerich <michael.hennerich@analog.com>");
 MODULE_DESCRIPTION("digital potentiometer SPI bus driver");
 MODULE_LICENSE("GPL");
 MODULE_ALIAS("spi:ad_dpot");
diff --git a/drivers/misc/ad525x_dpot.c b/drivers/misc/ad525x_dpot.c
index bc591b7168db..a0afadefcc49 100644
--- a/drivers/misc/ad525x_dpot.c
+++ b/drivers/misc/ad525x_dpot.c
@@ -1,7 +1,7 @@
 /*
  * ad525x_dpot: Driver for the Analog Devices digital potentiometers
  * Copyright (c) 2009-2010 Analog Devices, Inc.
- * Author: Michael Hennerich <hennerich@blackfin.uclinux.org>
+ * Author: Michael Hennerich <michael.hennerich@analog.com>
  *
  * DEVID		#Wipers		#Positions	Resistor Options (kOhm)
  * AD5258		1		64		1, 10, 50, 100
@@ -64,7 +64,7 @@
  * Author: Chris Verges <chrisv@cyberswitching.com>
  *
  * derived from ad5252.c
- * Copyright (c) 2006-2011 Michael Hennerich <hennerich@blackfin.uclinux.org>
+ * Copyright (c) 2006-2011 Michael Hennerich <michael.hennerich@analog.com>
  *
  * Licensed under the GPL-2 or later.
  */
@@ -760,6 +760,6 @@ EXPORT_SYMBOL(ad_dpot_remove);
 
 
 MODULE_AUTHOR("Chris Verges <chrisv@cyberswitching.com>, "
-	      "Michael Hennerich <hennerich@blackfin.uclinux.org>");
+	      "Michael Hennerich <michael.hennerich@analog.com>");
 MODULE_DESCRIPTION("Digital potentiometer driver");
 MODULE_LICENSE("GPL");
diff --git a/drivers/misc/apds990x.c b/drivers/misc/apds990x.c
index ed9412d750b7..24876c615c3c 100644
--- a/drivers/misc/apds990x.c
+++ b/drivers/misc/apds990x.c
@@ -188,7 +188,6 @@ struct apds990x_chip {
 #define APDS_LUX_DEFAULT_RATE		200
 
 static const u8 again[]	= {1, 8, 16, 120}; /* ALS gain steps */
-static const u8 ir_currents[]	= {100, 50, 25, 12}; /* IRled currents in mA */
 
 /* Following two tables must match i.e 10Hz rate means 1 as persistence value */
 static const u16 arates_hz[] = {10, 5, 2, 1};
diff --git a/drivers/misc/bh1770glc.c b/drivers/misc/bh1770glc.c
index 9c62bf064f77..17e81ce9925b 100644
--- a/drivers/misc/bh1770glc.c
+++ b/drivers/misc/bh1770glc.c
@@ -180,9 +180,6 @@ static const char reg_vleds[] = "Vleds";
 static const s16 prox_rates_hz[] = {100, 50, 33, 25, 14, 10, 5, 2};
 static const s16 prox_rates_ms[] = {10, 20, 30, 40, 70, 100, 200, 500};
 
-/* Supported IR-led currents in mA */
-static const u8 prox_curr_ma[] = {5, 10, 20, 50, 100, 150, 200};
-
 /*
  * Supported stand alone rates in ms from chip data sheet
  * {100, 200, 500, 1000, 2000};
diff --git a/drivers/misc/cxl/flash.c b/drivers/misc/cxl/flash.c
index 43917898fb9a..4d6836f19489 100644
--- a/drivers/misc/cxl/flash.c
+++ b/drivers/misc/cxl/flash.c
@@ -92,8 +92,8 @@ static int update_property(struct device_node *dn, const char *name,
 
 	val = (u32 *)new_prop->value;
 	rc = cxl_update_properties(dn, new_prop);
-	pr_devel("%s: update property (%s, length: %i, value: %#x)\n",
-		  dn->name, name, vd, be32_to_cpu(*val));
+	pr_devel("%pOFn: update property (%s, length: %i, value: %#x)\n",
+		  dn, name, vd, be32_to_cpu(*val));
 
 	if (rc) {
 		kfree(new_prop->name);
diff --git a/drivers/misc/cxl/guest.c b/drivers/misc/cxl/guest.c
index 3bc0c15d4d85..5d28d9e454f5 100644
--- a/drivers/misc/cxl/guest.c
+++ b/drivers/misc/cxl/guest.c
@@ -1018,8 +1018,6 @@ err1:
 
 void cxl_guest_remove_afu(struct cxl_afu *afu)
 {
-	pr_devel("in %s - AFU(%d)\n", __func__, afu->slice);
-
 	if (!afu)
 		return;
 
diff --git a/drivers/misc/echo/echo.c b/drivers/misc/echo/echo.c
index 8a5adc0d2e88..3ebe5d75ad6a 100644
--- a/drivers/misc/echo/echo.c
+++ b/drivers/misc/echo/echo.c
@@ -381,7 +381,7 @@ int16_t oslec_update(struct oslec_state *ec, int16_t tx, int16_t rx)
 	 */
 	ec->factor = 0;
 	ec->shift = 0;
-	if ((ec->nonupdate_dwell == 0)) {
+	if (!ec->nonupdate_dwell) {
 		int p, logp, shift;
 
 		/* Determine:
diff --git a/drivers/misc/eeprom/Kconfig b/drivers/misc/eeprom/Kconfig
index 68a1ac929917..fe7a1d27a017 100644
--- a/drivers/misc/eeprom/Kconfig
+++ b/drivers/misc/eeprom/Kconfig
@@ -111,4 +111,15 @@ config EEPROM_IDT_89HPESX
 	  This driver can also be built as a module. If so, the module
 	  will be called idt_89hpesx.
 
+config EEPROM_EE1004
+	tristate "SPD EEPROMs on DDR4 memory modules"
+	depends on I2C && SYSFS
+	help
+	  Enable this driver to get read support to SPD EEPROMs following
+	  the JEDEC EE1004 standard. These are typically found on DDR4
+	  SDRAM memory modules.
+
+	  This driver can also be built as a module.  If so, the module
+	  will be called ee1004.
+
 endmenu
diff --git a/drivers/misc/eeprom/Makefile b/drivers/misc/eeprom/Makefile
index 2aab60ef3e3e..a9b4b6579b75 100644
--- a/drivers/misc/eeprom/Makefile
+++ b/drivers/misc/eeprom/Makefile
@@ -7,3 +7,4 @@ obj-$(CONFIG_EEPROM_93CX6)	+= eeprom_93cx6.o
 obj-$(CONFIG_EEPROM_93XX46)	+= eeprom_93xx46.o
 obj-$(CONFIG_EEPROM_DIGSY_MTC_CFG) += digsy_mtc_eeprom.o
 obj-$(CONFIG_EEPROM_IDT_89HPESX) += idt_89hpesx.o
+obj-$(CONFIG_EEPROM_EE1004)	+= ee1004.o
diff --git a/drivers/misc/eeprom/at25.c b/drivers/misc/eeprom/at25.c
index 840afb398f9e..99de6939cd5a 100644
--- a/drivers/misc/eeprom/at25.c
+++ b/drivers/misc/eeprom/at25.c
@@ -366,7 +366,7 @@ static int at25_probe(struct spi_device *spi)
 	at25->nvmem_config.word_size = 1;
 	at25->nvmem_config.size = chip.byte_len;
 
-	at25->nvmem = nvmem_register(&at25->nvmem_config);
+	at25->nvmem = devm_nvmem_register(&spi->dev, &at25->nvmem_config);
 	if (IS_ERR(at25->nvmem))
 		return PTR_ERR(at25->nvmem);
 
@@ -379,16 +379,6 @@ static int at25_probe(struct spi_device *spi)
 	return 0;
 }
 
-static int at25_remove(struct spi_device *spi)
-{
-	struct at25_data	*at25;
-
-	at25 = spi_get_drvdata(spi);
-	nvmem_unregister(at25->nvmem);
-
-	return 0;
-}
-
 /*-------------------------------------------------------------------------*/
 
 static const struct of_device_id at25_of_match[] = {
@@ -403,7 +393,6 @@ static struct spi_driver at25_driver = {
 		.of_match_table = at25_of_match,
 	},
 	.probe		= at25_probe,
-	.remove		= at25_remove,
 };
 
 module_spi_driver(at25_driver);
diff --git a/drivers/misc/eeprom/ee1004.c b/drivers/misc/eeprom/ee1004.c
new file mode 100644
index 000000000000..276c1690ea1b
--- /dev/null
+++ b/drivers/misc/eeprom/ee1004.c
@@ -0,0 +1,281 @@
+/*
+ * ee1004 - driver for DDR4 SPD EEPROMs
+ *
+ * Copyright (C) 2017 Jean Delvare
+ *
+ * Based on the at24 driver:
+ * Copyright (C) 2005-2007 David Brownell
+ * Copyright (C) 2008 Wolfram Sang, Pengutronix
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/i2c.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/mod_devicetable.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+
+/*
+ * DDR4 memory modules use special EEPROMs following the Jedec EE1004
+ * specification. These are 512-byte EEPROMs using a single I2C address
+ * in the 0x50-0x57 range for data. One of two 256-byte page is selected
+ * by writing a command to I2C address 0x36 or 0x37 on the same I2C bus.
+ *
+ * Therefore we need to request these 2 additional addresses, and serialize
+ * access to all such EEPROMs with a single mutex.
+ *
+ * We assume it is safe to read up to 32 bytes at once from these EEPROMs.
+ * We use SMBus access even if I2C is available, these EEPROMs are small
+ * enough, and reading from them infrequent enough, that we favor simplicity
+ * over performance.
+ */
+
+#define EE1004_ADDR_SET_PAGE		0x36
+#define EE1004_EEPROM_SIZE		512
+#define EE1004_PAGE_SIZE		256
+#define EE1004_PAGE_SHIFT		8
+
+/*
+ * Mutex protects ee1004_set_page and ee1004_dev_count, and must be held
+ * from page selection to end of read.
+ */
+static DEFINE_MUTEX(ee1004_bus_lock);
+static struct i2c_client *ee1004_set_page[2];
+static unsigned int ee1004_dev_count;
+static int ee1004_current_page;
+
+static const struct i2c_device_id ee1004_ids[] = {
+	{ "ee1004", 0 },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, ee1004_ids);
+
+/*-------------------------------------------------------------------------*/
+
+static ssize_t ee1004_eeprom_read(struct i2c_client *client, char *buf,
+				  unsigned int offset, size_t count)
+{
+	int status;
+
+	if (count > I2C_SMBUS_BLOCK_MAX)
+		count = I2C_SMBUS_BLOCK_MAX;
+	/* Can't cross page boundaries */
+	if (unlikely(offset + count > EE1004_PAGE_SIZE))
+		count = EE1004_PAGE_SIZE - offset;
+
+	status = i2c_smbus_read_i2c_block_data_or_emulated(client, offset,
+							   count, buf);
+	dev_dbg(&client->dev, "read %zu@%d --> %d\n", count, offset, status);
+
+	return status;
+}
+
+static ssize_t ee1004_read(struct file *filp, struct kobject *kobj,
+			   struct bin_attribute *bin_attr,
+			   char *buf, loff_t off, size_t count)
+{
+	struct device *dev = kobj_to_dev(kobj);
+	struct i2c_client *client = to_i2c_client(dev);
+	size_t requested = count;
+	int page;
+
+	if (unlikely(!count))
+		return count;
+
+	page = off >> EE1004_PAGE_SHIFT;
+	if (unlikely(page > 1))
+		return 0;
+	off &= (1 << EE1004_PAGE_SHIFT) - 1;
+
+	/*
+	 * Read data from chip, protecting against concurrent access to
+	 * other EE1004 SPD EEPROMs on the same adapter.
+	 */
+	mutex_lock(&ee1004_bus_lock);
+
+	while (count) {
+		int status;
+
+		/* Select page */
+		if (page != ee1004_current_page) {
+			/* Data is ignored */
+			status = i2c_smbus_write_byte(ee1004_set_page[page],
+						      0x00);
+			if (status < 0) {
+				dev_err(dev, "Failed to select page %d (%d)\n",
+					page, status);
+				mutex_unlock(&ee1004_bus_lock);
+				return status;
+			}
+			dev_dbg(dev, "Selected page %d\n", page);
+			ee1004_current_page = page;
+		}
+
+		status = ee1004_eeprom_read(client, buf, off, count);
+		if (status < 0) {
+			mutex_unlock(&ee1004_bus_lock);
+			return status;
+		}
+		buf += status;
+		off += status;
+		count -= status;
+
+		if (off == EE1004_PAGE_SIZE) {
+			page++;
+			off = 0;
+		}
+	}
+
+	mutex_unlock(&ee1004_bus_lock);
+
+	return requested;
+}
+
+static const struct bin_attribute eeprom_attr = {
+	.attr = {
+		.name = "eeprom",
+		.mode = 0444,
+	},
+	.size = EE1004_EEPROM_SIZE,
+	.read = ee1004_read,
+};
+
+static int ee1004_probe(struct i2c_client *client,
+			const struct i2c_device_id *id)
+{
+	int err, cnr = 0;
+	const char *slow = NULL;
+
+	/* Make sure we can operate on this adapter */
+	if (!i2c_check_functionality(client->adapter,
+				     I2C_FUNC_SMBUS_READ_BYTE |
+				     I2C_FUNC_SMBUS_READ_I2C_BLOCK)) {
+		if (i2c_check_functionality(client->adapter,
+				     I2C_FUNC_SMBUS_READ_BYTE |
+				     I2C_FUNC_SMBUS_READ_WORD_DATA))
+			slow = "word";
+		else if (i2c_check_functionality(client->adapter,
+				     I2C_FUNC_SMBUS_READ_BYTE |
+				     I2C_FUNC_SMBUS_READ_BYTE_DATA))
+			slow = "byte";
+		else
+			return -EPFNOSUPPORT;
+	}
+
+	/* Use 2 dummy devices for page select command */
+	mutex_lock(&ee1004_bus_lock);
+	if (++ee1004_dev_count == 1) {
+		for (cnr = 0; cnr < 2; cnr++) {
+			ee1004_set_page[cnr] = i2c_new_dummy(client->adapter,
+						EE1004_ADDR_SET_PAGE + cnr);
+			if (!ee1004_set_page[cnr]) {
+				dev_err(&client->dev,
+					"address 0x%02x unavailable\n",
+					EE1004_ADDR_SET_PAGE + cnr);
+				err = -EADDRINUSE;
+				goto err_clients;
+			}
+		}
+	} else if (i2c_adapter_id(client->adapter) !=
+		   i2c_adapter_id(ee1004_set_page[0]->adapter)) {
+		dev_err(&client->dev,
+			"Driver only supports devices on a single I2C bus\n");
+		err = -EOPNOTSUPP;
+		goto err_clients;
+	}
+
+	/* Remember current page to avoid unneeded page select */
+	err = i2c_smbus_read_byte(ee1004_set_page[0]);
+	if (err == -ENXIO) {
+		/* Nack means page 1 is selected */
+		ee1004_current_page = 1;
+	} else if (err < 0) {
+		/* Anything else is a real error, bail out */
+		goto err_clients;
+	} else {
+		/* Ack means page 0 is selected, returned value meaningless */
+		ee1004_current_page = 0;
+	}
+	dev_dbg(&client->dev, "Currently selected page: %d\n",
+		ee1004_current_page);
+	mutex_unlock(&ee1004_bus_lock);
+
+	/* Create the sysfs eeprom file */
+	err = sysfs_create_bin_file(&client->dev.kobj, &eeprom_attr);
+	if (err)
+		goto err_clients_lock;
+
+	dev_info(&client->dev,
+		 "%u byte EE1004-compliant SPD EEPROM, read-only\n",
+		 EE1004_EEPROM_SIZE);
+	if (slow)
+		dev_notice(&client->dev,
+			   "Falling back to %s reads, performance will suffer\n",
+			   slow);
+
+	return 0;
+
+ err_clients_lock:
+	mutex_lock(&ee1004_bus_lock);
+ err_clients:
+	if (--ee1004_dev_count == 0) {
+		for (cnr--; cnr >= 0; cnr--) {
+			i2c_unregister_device(ee1004_set_page[cnr]);
+			ee1004_set_page[cnr] = NULL;
+		}
+	}
+	mutex_unlock(&ee1004_bus_lock);
+
+	return err;
+}
+
+static int ee1004_remove(struct i2c_client *client)
+{
+	int i;
+
+	sysfs_remove_bin_file(&client->dev.kobj, &eeprom_attr);
+
+	/* Remove page select clients if this is the last device */
+	mutex_lock(&ee1004_bus_lock);
+	if (--ee1004_dev_count == 0) {
+		for (i = 0; i < 2; i++) {
+			i2c_unregister_device(ee1004_set_page[i]);
+			ee1004_set_page[i] = NULL;
+		}
+	}
+	mutex_unlock(&ee1004_bus_lock);
+
+	return 0;
+}
+
+/*-------------------------------------------------------------------------*/
+
+static struct i2c_driver ee1004_driver = {
+	.driver = {
+		.name = "ee1004",
+	},
+	.probe = ee1004_probe,
+	.remove = ee1004_remove,
+	.id_table = ee1004_ids,
+};
+
+static int __init ee1004_init(void)
+{
+	return i2c_add_driver(&ee1004_driver);
+}
+module_init(ee1004_init);
+
+static void __exit ee1004_exit(void)
+{
+	i2c_del_driver(&ee1004_driver);
+}
+module_exit(ee1004_exit);
+
+MODULE_DESCRIPTION("Driver for EE1004-compliant DDR4 SPD EEPROMs");
+MODULE_AUTHOR("Jean Delvare");
+MODULE_LICENSE("GPL");
diff --git a/drivers/misc/eeprom/eeprom_93xx46.c b/drivers/misc/eeprom/eeprom_93xx46.c
index 38766968bfa2..c6dd9ad9bf7b 100644
--- a/drivers/misc/eeprom/eeprom_93xx46.c
+++ b/drivers/misc/eeprom/eeprom_93xx46.c
@@ -439,7 +439,7 @@ static int eeprom_93xx46_probe(struct spi_device *spi)
 		return -ENODEV;
 	}
 
-	edev = kzalloc(sizeof(*edev), GFP_KERNEL);
+	edev = devm_kzalloc(&spi->dev, sizeof(*edev), GFP_KERNEL);
 	if (!edev)
 		return -ENOMEM;
 
@@ -449,8 +449,7 @@ static int eeprom_93xx46_probe(struct spi_device *spi)
 		edev->addrlen = 6;
 	else {
 		dev_err(&spi->dev, "unspecified address type\n");
-		err = -EINVAL;
-		goto fail;
+		return -EINVAL;
 	}
 
 	mutex_init(&edev->lock);
@@ -473,11 +472,9 @@ static int eeprom_93xx46_probe(struct spi_device *spi)
 	edev->nvmem_config.word_size = 1;
 	edev->nvmem_config.size = edev->size;
 
-	edev->nvmem = nvmem_register(&edev->nvmem_config);
-	if (IS_ERR(edev->nvmem)) {
-		err = PTR_ERR(edev->nvmem);
-		goto fail;
-	}
+	edev->nvmem = devm_nvmem_register(&spi->dev, &edev->nvmem_config);
+	if (IS_ERR(edev->nvmem))
+		return PTR_ERR(edev->nvmem);
 
 	dev_info(&spi->dev, "%d-bit eeprom %s\n",
 		(pd->flags & EE_ADDR8) ? 8 : 16,
@@ -490,21 +487,15 @@ static int eeprom_93xx46_probe(struct spi_device *spi)
 
 	spi_set_drvdata(spi, edev);
 	return 0;
-fail:
-	kfree(edev);
-	return err;
 }
 
 static int eeprom_93xx46_remove(struct spi_device *spi)
 {
 	struct eeprom_93xx46_dev *edev = spi_get_drvdata(spi);
 
-	nvmem_unregister(edev->nvmem);
-
 	if (!(edev->pdata->flags & EE_READONLY))
 		device_remove_file(&spi->dev, &dev_attr_erase);
 
-	kfree(edev);
 	return 0;
 }
 
diff --git a/drivers/misc/genwqe/card_base.c b/drivers/misc/genwqe/card_base.c
index c7cd3675bcd1..d137d0fab9bf 100644
--- a/drivers/misc/genwqe/card_base.c
+++ b/drivers/misc/genwqe/card_base.c
@@ -24,7 +24,6 @@
  * controlled from here.
  */
 
-#include <linux/module.h>
 #include <linux/types.h>
 #include <linux/pci.h>
 #include <linux/err.h>
diff --git a/drivers/misc/genwqe/card_ddcb.c b/drivers/misc/genwqe/card_ddcb.c
index 656449cb4476..9a65bd9d6152 100644
--- a/drivers/misc/genwqe/card_ddcb.c
+++ b/drivers/misc/genwqe/card_ddcb.c
@@ -27,7 +27,6 @@
  */
 
 #include <linux/types.h>
-#include <linux/module.h>
 #include <linux/sched.h>
 #include <linux/wait.h>
 #include <linux/pci.h>
diff --git a/drivers/misc/genwqe/card_utils.c b/drivers/misc/genwqe/card_utils.c
index 8679e0bd8ec2..3fcb9a2fe1c9 100644
--- a/drivers/misc/genwqe/card_utils.c
+++ b/drivers/misc/genwqe/card_utils.c
@@ -23,14 +23,12 @@
  */
 
 #include <linux/kernel.h>
-#include <linux/dma-mapping.h>
 #include <linux/sched.h>
 #include <linux/vmalloc.h>
 #include <linux/page-flags.h>
 #include <linux/scatterlist.h>
 #include <linux/hugetlb.h>
 #include <linux/iommu.h>
-#include <linux/delay.h>
 #include <linux/pci.h>
 #include <linux/dma-mapping.h>
 #include <linux/ctype.h>
@@ -298,7 +296,7 @@ static int genwqe_sgl_size(int num_pages)
 int genwqe_alloc_sync_sgl(struct genwqe_dev *cd, struct genwqe_sgl *sgl,
 			  void __user *user_addr, size_t user_size, int write)
 {
-	int rc;
+	int ret = -ENOMEM;
 	struct pci_dev *pci_dev = cd->pci_dev;
 
 	sgl->fpage_offs = offset_in_page((unsigned long)user_addr);
@@ -318,7 +316,7 @@ int genwqe_alloc_sync_sgl(struct genwqe_dev *cd, struct genwqe_sgl *sgl,
 	if (get_order(sgl->sgl_size) > MAX_ORDER) {
 		dev_err(&pci_dev->dev,
 			"[%s] err: too much memory requested!\n", __func__);
-		return -ENOMEM;
+		return ret;
 	}
 
 	sgl->sgl = __genwqe_alloc_consistent(cd, sgl->sgl_size,
@@ -326,7 +324,7 @@ int genwqe_alloc_sync_sgl(struct genwqe_dev *cd, struct genwqe_sgl *sgl,
 	if (sgl->sgl == NULL) {
 		dev_err(&pci_dev->dev,
 			"[%s] err: no memory available!\n", __func__);
-		return -ENOMEM;
+		return ret;
 	}
 
 	/* Only use buffering on incomplete pages */
@@ -339,7 +337,7 @@ int genwqe_alloc_sync_sgl(struct genwqe_dev *cd, struct genwqe_sgl *sgl,
 		/* Sync with user memory */
 		if (copy_from_user(sgl->fpage + sgl->fpage_offs,
 				   user_addr, sgl->fpage_size)) {
-			rc = -EFAULT;
+			ret = -EFAULT;
 			goto err_out;
 		}
 	}
@@ -352,7 +350,7 @@ int genwqe_alloc_sync_sgl(struct genwqe_dev *cd, struct genwqe_sgl *sgl,
 		/* Sync with user memory */
 		if (copy_from_user(sgl->lpage, user_addr + user_size -
 				   sgl->lpage_size, sgl->lpage_size)) {
-			rc = -EFAULT;
+			ret = -EFAULT;
 			goto err_out2;
 		}
 	}
@@ -374,7 +372,8 @@ int genwqe_alloc_sync_sgl(struct genwqe_dev *cd, struct genwqe_sgl *sgl,
 	sgl->sgl = NULL;
 	sgl->sgl_dma_addr = 0;
 	sgl->sgl_size = 0;
-	return -ENOMEM;
+
+	return ret;
 }
 
 int genwqe_setup_sgl(struct genwqe_dev *cd, struct genwqe_sgl *sgl,
diff --git a/drivers/misc/kgdbts.c b/drivers/misc/kgdbts.c
index 6193270e7b3d..de20bdaa148d 100644
--- a/drivers/misc/kgdbts.c
+++ b/drivers/misc/kgdbts.c
@@ -985,6 +985,12 @@ static void kgdbts_run_tests(void)
 	int nmi_sleep = 0;
 	int i;
 
+	verbose = 0;
+	if (strstr(config, "V1"))
+		verbose = 1;
+	if (strstr(config, "V2"))
+		verbose = 2;
+
 	ptr = strchr(config, 'F');
 	if (ptr)
 		fork_test = simple_strtol(ptr + 1, NULL, 10);
@@ -1068,13 +1074,6 @@ static int kgdbts_option_setup(char *opt)
 		return -ENOSPC;
 	}
 	strcpy(config, opt);
-
-	verbose = 0;
-	if (strstr(config, "V1"))
-		verbose = 1;
-	if (strstr(config, "V2"))
-		verbose = 2;
-
 	return 0;
 }
 
@@ -1086,9 +1085,6 @@ static int configure_kgdbts(void)
 
 	if (!strlen(config) || isspace(config[0]))
 		goto noconfig;
-	err = kgdbts_option_setup(config);
-	if (err)
-		goto noconfig;
 
 	final_ack = 0;
 	run_plant_and_detach_test(1);
diff --git a/drivers/misc/lkdtm/usercopy.c b/drivers/misc/lkdtm/usercopy.c
index 389475b25bb7..d5a0e7f1813b 100644
--- a/drivers/misc/lkdtm/usercopy.c
+++ b/drivers/misc/lkdtm/usercopy.c
@@ -18,7 +18,7 @@
  * hardened usercopy checks by added "unconst" to all the const copies,
  * and making sure "cache_size" isn't optimized into a const.
  */
-static volatile size_t unconst = 0;
+static volatile size_t unconst;
 static volatile size_t cache_size = 1024;
 static struct kmem_cache *whitelist_cache;
 
diff --git a/drivers/misc/mei/bus-fixup.c b/drivers/misc/mei/bus-fixup.c
index a6f41f96f2a1..80215c312f0e 100644
--- a/drivers/misc/mei/bus-fixup.c
+++ b/drivers/misc/mei/bus-fixup.c
@@ -17,7 +17,6 @@
 #include <linux/kernel.h>
 #include <linux/sched.h>
 #include <linux/module.h>
-#include <linux/moduleparam.h>
 #include <linux/device.h>
 #include <linux/slab.h>
 #include <linux/uuid.h>
diff --git a/drivers/misc/mei/main.c b/drivers/misc/mei/main.c
index 4d77a6ae183a..87281b3695e6 100644
--- a/drivers/misc/mei/main.c
+++ b/drivers/misc/mei/main.c
@@ -599,10 +599,10 @@ static __poll_t mei_poll(struct file *file, poll_table *wait)
 			mei_cl_read_start(cl, mei_cl_mtu(cl), file);
 	}
 
-	if (req_events & (POLLOUT | POLLWRNORM)) {
+	if (req_events & (EPOLLOUT | EPOLLWRNORM)) {
 		poll_wait(file, &cl->tx_wait, wait);
 		if (cl->tx_cb_queued < dev->tx_queue_limit)
-			mask |= POLLOUT | POLLWRNORM;
+			mask |= EPOLLOUT | EPOLLWRNORM;
 	}
 
 out:
diff --git a/drivers/misc/mic/scif/scif_dma.c b/drivers/misc/mic/scif/scif_dma.c
index 6369aeaa7056..18b8ed57c4ac 100644
--- a/drivers/misc/mic/scif/scif_dma.c
+++ b/drivers/misc/mic/scif/scif_dma.c
@@ -1035,8 +1035,6 @@ scif_rma_list_dma_copy_unaligned(struct scif_copy_work *work,
 			}
 			dma_async_issue_pending(chan);
 		}
-		if (ret < 0)
-			goto err;
 		offset += loop_len;
 		temp += loop_len;
 		temp_phys += loop_len;
@@ -1553,9 +1551,8 @@ static int scif_rma_list_dma_copy_wrapper(struct scif_endpt *epd,
 	int src_cache_off, dst_cache_off;
 	s64 src_offset = work->src_offset, dst_offset = work->dst_offset;
 	u8 *temp = NULL;
-	bool src_local = true, dst_local = false;
+	bool src_local = true;
 	struct scif_dma_comp_cb *comp_cb;
-	dma_addr_t src_dma_addr, dst_dma_addr;
 	int err;
 
 	if (is_dma_copy_aligned(chan->device, 1, 1, 1))
@@ -1569,12 +1566,8 @@ static int scif_rma_list_dma_copy_wrapper(struct scif_endpt *epd,
 
 	if (work->loopback)
 		return scif_rma_list_cpu_copy(work);
-	src_dma_addr = __scif_off_to_dma_addr(work->src_window, src_offset);
-	dst_dma_addr = __scif_off_to_dma_addr(work->dst_window, dst_offset);
 	src_local = work->src_window->type == SCIF_WINDOW_SELF;
-	dst_local = work->dst_window->type == SCIF_WINDOW_SELF;
 
-	dst_local = dst_local;
 	/* Allocate dma_completion cb */
 	comp_cb = kzalloc(sizeof(*comp_cb), GFP_KERNEL);
 	if (!comp_cb)
diff --git a/drivers/misc/mic/scif/scif_fence.c b/drivers/misc/mic/scif/scif_fence.c
index cac3bcc308a7..7bb929f05d85 100644
--- a/drivers/misc/mic/scif/scif_fence.c
+++ b/drivers/misc/mic/scif/scif_fence.c
@@ -272,7 +272,7 @@ static int _scif_prog_signal(scif_epd_t epd, dma_addr_t dst, u64 val)
 dma_fail:
 	if (!x100)
 		dma_pool_free(ep->remote_dev->signal_pool, status,
-			      status->src_dma_addr);
+			      src - offsetof(struct scif_status, val));
 alloc_fail:
 	return err;
 }
diff --git a/drivers/misc/sgi-gru/grukservices.c b/drivers/misc/sgi-gru/grukservices.c
index 030769018461..4b23d586fc3f 100644
--- a/drivers/misc/sgi-gru/grukservices.c
+++ b/drivers/misc/sgi-gru/grukservices.c
@@ -634,7 +634,7 @@ static int send_noop_message(void *cb, struct gru_message_queue_desc *mqd,
 			break;
 		case CBSS_PAGE_OVERFLOW:
 			STAT(mesq_noop_page_overflow);
-			/* fallthru */
+			/* fall through */
 		default:
 			BUG();
 		}
@@ -792,7 +792,7 @@ static int send_message_failure(void *cb, struct gru_message_queue_desc *mqd,
 		break;
 	case CBSS_PAGE_OVERFLOW:
 		STAT(mesq_page_overflow);
-		/* fallthru */
+		/* fall through */
 	default:
 		BUG();
 	}
diff --git a/drivers/misc/sgi-xp/xpc_channel.c b/drivers/misc/sgi-xp/xpc_channel.c
index 05a890ce2ab8..8e6607fc8a67 100644
--- a/drivers/misc/sgi-xp/xpc_channel.c
+++ b/drivers/misc/sgi-xp/xpc_channel.c
@@ -28,7 +28,7 @@ xpc_process_connect(struct xpc_channel *ch, unsigned long *irq_flags)
 {
 	enum xp_retval ret;
 
-	DBUG_ON(!spin_is_locked(&ch->lock));
+	lockdep_assert_held(&ch->lock);
 
 	if (!(ch->flags & XPC_C_OPENREQUEST) ||
 	    !(ch->flags & XPC_C_ROPENREQUEST)) {
@@ -82,7 +82,7 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags)
 	struct xpc_partition *part = &xpc_partitions[ch->partid];
 	u32 channel_was_connected = (ch->flags & XPC_C_WASCONNECTED);
 
-	DBUG_ON(!spin_is_locked(&ch->lock));
+	lockdep_assert_held(&ch->lock);
 
 	if (!(ch->flags & XPC_C_DISCONNECTING))
 		return;
@@ -755,7 +755,7 @@ xpc_disconnect_channel(const int line, struct xpc_channel *ch,
 {
 	u32 channel_was_connected = (ch->flags & XPC_C_CONNECTED);
 
-	DBUG_ON(!spin_is_locked(&ch->lock));
+	lockdep_assert_held(&ch->lock);
 
 	if (ch->flags & (XPC_C_DISCONNECTING | XPC_C_DISCONNECTED))
 		return;
diff --git a/drivers/misc/sgi-xp/xpc_partition.c b/drivers/misc/sgi-xp/xpc_partition.c
index 0c3ef6f1df54..3eba1c420cc0 100644
--- a/drivers/misc/sgi-xp/xpc_partition.c
+++ b/drivers/misc/sgi-xp/xpc_partition.c
@@ -98,8 +98,7 @@ xpc_get_rsvd_page_pa(int nasid)
 			len = L1_CACHE_ALIGN(len);
 
 		if (len > buf_len) {
-			if (buf_base != NULL)
-				kfree(buf_base);
+			kfree(buf_base);
 			buf_len = L1_CACHE_ALIGN(len);
 			buf = xpc_kmalloc_cacheline_aligned(buf_len, GFP_KERNEL,
 							    &buf_base);
diff --git a/drivers/misc/sgi-xp/xpc_sn2.c b/drivers/misc/sgi-xp/xpc_sn2.c
index 5a12d2a54049..0ae69b9390ce 100644
--- a/drivers/misc/sgi-xp/xpc_sn2.c
+++ b/drivers/misc/sgi-xp/xpc_sn2.c
@@ -1671,7 +1671,7 @@ xpc_teardown_msg_structures_sn2(struct xpc_channel *ch)
 {
 	struct xpc_channel_sn2 *ch_sn2 = &ch->sn.sn2;
 
-	DBUG_ON(!spin_is_locked(&ch->lock));
+	lockdep_assert_held(&ch->lock);
 
 	ch_sn2->remote_msgqueue_pa = 0;
 
diff --git a/drivers/misc/sgi-xp/xpc_uv.c b/drivers/misc/sgi-xp/xpc_uv.c
index 340b44d9e8cf..0441abe87880 100644
--- a/drivers/misc/sgi-xp/xpc_uv.c
+++ b/drivers/misc/sgi-xp/xpc_uv.c
@@ -1183,7 +1183,7 @@ xpc_teardown_msg_structures_uv(struct xpc_channel *ch)
 {
 	struct xpc_channel_uv *ch_uv = &ch->sn.uv;
 
-	DBUG_ON(!spin_is_locked(&ch->lock));
+	lockdep_assert_held(&ch->lock);
 
 	kfree(ch_uv->cached_notify_gru_mq_desc);
 	ch_uv->cached_notify_gru_mq_desc = NULL;
diff --git a/drivers/misc/sram.c b/drivers/misc/sram.c
index 74b183baf044..80d8cbe8c01a 100644
--- a/drivers/misc/sram.c
+++ b/drivers/misc/sram.c
@@ -323,10 +323,8 @@ static int sram_reserve_regions(struct sram_dev *sram, struct resource *res)
 		cur_start = block->start + block->size;
 	}
 
- err_chunks:
-	if (child)
-		of_node_put(child);
-
+err_chunks:
+	of_node_put(child);
 	kfree(rblocks);
 
 	return ret;
diff --git a/drivers/misc/vmw_balloon.c b/drivers/misc/vmw_balloon.c
index 2543ef1ece17..9b0b3fa4f836 100644
--- a/drivers/misc/vmw_balloon.c
+++ b/drivers/misc/vmw_balloon.c
@@ -25,6 +25,9 @@
 #include <linux/workqueue.h>
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
+#include <linux/rwsem.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
 #include <linux/vmw_vmci_defs.h>
 #include <linux/vmw_vmci_api.h>
 #include <asm/hypervisor.h>
@@ -37,20 +40,20 @@ MODULE_ALIAS("vmware_vmmemctl");
 MODULE_LICENSE("GPL");
 
 /*
- * Use __GFP_HIGHMEM to allow pages from HIGHMEM zone. We don't
- * allow wait (__GFP_RECLAIM) for NOSLEEP page allocations. Use
- * __GFP_NOWARN, to suppress page allocation failure warnings.
+ * Use __GFP_HIGHMEM to allow pages from HIGHMEM zone. We don't allow wait
+ * (__GFP_RECLAIM) for huge page allocations. Use __GFP_NOWARN, to suppress page
+ * allocation failure warnings. Disallow access to emergency low-memory pools.
  */
-#define VMW_PAGE_ALLOC_NOSLEEP		(__GFP_HIGHMEM|__GFP_NOWARN)
+#define VMW_HUGE_PAGE_ALLOC_FLAGS	(__GFP_HIGHMEM|__GFP_NOWARN|	\
+					 __GFP_NOMEMALLOC)
 
 /*
- * Use GFP_HIGHUSER when executing in a separate kernel thread
- * context and allocation can sleep.  This is less stressful to
- * the guest memory system, since it allows the thread to block
- * while memory is reclaimed, and won't take pages from emergency
- * low-memory pools.
+ * Use __GFP_HIGHMEM to allow pages from HIGHMEM zone. We allow lightweight
+ * reclamation (__GFP_NORETRY). Use __GFP_NOWARN, to suppress page allocation
+ * failure warnings. Disallow access to emergency low-memory pools.
  */
-#define VMW_PAGE_ALLOC_CANSLEEP		(GFP_HIGHUSER)
+#define VMW_PAGE_ALLOC_FLAGS		(__GFP_HIGHMEM|__GFP_NOWARN|	\
+					 __GFP_NOMEMALLOC|__GFP_NORETRY)
 
 /* Maximum number of refused pages we accumulate during inflation cycle */
 #define VMW_BALLOON_MAX_REFUSED		16
@@ -77,225 +80,420 @@ enum vmwballoon_capabilities {
 					| VMW_BALLOON_BATCHED_2M_CMDS \
 					| VMW_BALLOON_SIGNALLED_WAKEUP_CMD)
 
-#define VMW_BALLOON_2M_SHIFT		(9)
-#define VMW_BALLOON_NUM_PAGE_SIZES	(2)
+#define VMW_BALLOON_2M_ORDER		(PMD_SHIFT - PAGE_SHIFT)
 
-/*
- * Backdoor commands availability:
- *
- * START, GET_TARGET and GUEST_ID are always available,
- *
- * VMW_BALLOON_BASIC_CMDS:
- *	LOCK and UNLOCK commands,
- * VMW_BALLOON_BATCHED_CMDS:
- *	BATCHED_LOCK and BATCHED_UNLOCK commands.
- * VMW BALLOON_BATCHED_2M_CMDS:
- *	BATCHED_2M_LOCK and BATCHED_2M_UNLOCK commands,
- * VMW VMW_BALLOON_SIGNALLED_WAKEUP_CMD:
- *	VMW_BALLOON_CMD_VMCI_DOORBELL_SET command.
- */
-#define VMW_BALLOON_CMD_START			0
-#define VMW_BALLOON_CMD_GET_TARGET		1
-#define VMW_BALLOON_CMD_LOCK			2
-#define VMW_BALLOON_CMD_UNLOCK			3
-#define VMW_BALLOON_CMD_GUEST_ID		4
-#define VMW_BALLOON_CMD_BATCHED_LOCK		6
-#define VMW_BALLOON_CMD_BATCHED_UNLOCK		7
-#define VMW_BALLOON_CMD_BATCHED_2M_LOCK		8
-#define VMW_BALLOON_CMD_BATCHED_2M_UNLOCK	9
-#define VMW_BALLOON_CMD_VMCI_DOORBELL_SET	10
-
-
-/* error codes */
-#define VMW_BALLOON_SUCCESS		        0
-#define VMW_BALLOON_FAILURE		        -1
-#define VMW_BALLOON_ERROR_CMD_INVALID	        1
-#define VMW_BALLOON_ERROR_PPN_INVALID	        2
-#define VMW_BALLOON_ERROR_PPN_LOCKED	        3
-#define VMW_BALLOON_ERROR_PPN_UNLOCKED	        4
-#define VMW_BALLOON_ERROR_PPN_PINNED	        5
-#define VMW_BALLOON_ERROR_PPN_NOTNEEDED	        6
-#define VMW_BALLOON_ERROR_RESET		        7
-#define VMW_BALLOON_ERROR_BUSY		        8
+enum vmballoon_page_size_type {
+	VMW_BALLOON_4K_PAGE,
+	VMW_BALLOON_2M_PAGE,
+	VMW_BALLOON_LAST_SIZE = VMW_BALLOON_2M_PAGE
+};
 
-#define VMW_BALLOON_SUCCESS_WITH_CAPABILITIES	(0x03000000)
+#define VMW_BALLOON_NUM_PAGE_SIZES	(VMW_BALLOON_LAST_SIZE + 1)
 
-/* Batch page description */
+static const char * const vmballoon_page_size_names[] = {
+	[VMW_BALLOON_4K_PAGE]			= "4k",
+	[VMW_BALLOON_2M_PAGE]			= "2M"
+};
 
-/*
- * Layout of a page in the batch page:
+enum vmballoon_op {
+	VMW_BALLOON_INFLATE,
+	VMW_BALLOON_DEFLATE
+};
+
+enum vmballoon_op_stat_type {
+	VMW_BALLOON_OP_STAT,
+	VMW_BALLOON_OP_FAIL_STAT
+};
+
+#define VMW_BALLOON_OP_STAT_TYPES	(VMW_BALLOON_OP_FAIL_STAT + 1)
+
+/**
+ * enum vmballoon_cmd_type - backdoor commands.
  *
- * +-------------+----------+--------+
- * |             |          |        |
- * | Page number | Reserved | Status |
- * |             |          |        |
- * +-------------+----------+--------+
- * 64  PAGE_SHIFT          6         0
+ * Availability of the commands is as followed:
  *
- * The reserved field should be set to 0.
+ * %VMW_BALLOON_CMD_START, %VMW_BALLOON_CMD_GET_TARGET and
+ * %VMW_BALLOON_CMD_GUEST_ID are always available.
+ *
+ * If the host reports %VMW_BALLOON_BASIC_CMDS are supported then
+ * %VMW_BALLOON_CMD_LOCK and %VMW_BALLOON_CMD_UNLOCK commands are available.
+ *
+ * If the host reports %VMW_BALLOON_BATCHED_CMDS are supported then
+ * %VMW_BALLOON_CMD_BATCHED_LOCK and VMW_BALLOON_CMD_BATCHED_UNLOCK commands
+ * are available.
+ *
+ * If the host reports %VMW_BALLOON_BATCHED_2M_CMDS are supported then
+ * %VMW_BALLOON_CMD_BATCHED_2M_LOCK and %VMW_BALLOON_CMD_BATCHED_2M_UNLOCK
+ * are supported.
+ *
+ * If the host reports  VMW_BALLOON_SIGNALLED_WAKEUP_CMD is supported then
+ * VMW_BALLOON_CMD_VMCI_DOORBELL_SET command is supported.
+ *
+ * @VMW_BALLOON_CMD_START: Communicating supported version with the hypervisor.
+ * @VMW_BALLOON_CMD_GET_TARGET: Gets the balloon target size.
+ * @VMW_BALLOON_CMD_LOCK: Informs the hypervisor about a ballooned page.
+ * @VMW_BALLOON_CMD_UNLOCK: Informs the hypervisor about a page that is about
+ *			    to be deflated from the balloon.
+ * @VMW_BALLOON_CMD_GUEST_ID: Informs the hypervisor about the type of OS that
+ *			      runs in the VM.
+ * @VMW_BALLOON_CMD_BATCHED_LOCK: Inform the hypervisor about a batch of
+ *				  ballooned pages (up to 512).
+ * @VMW_BALLOON_CMD_BATCHED_UNLOCK: Inform the hypervisor about a batch of
+ *				  pages that are about to be deflated from the
+ *				  balloon (up to 512).
+ * @VMW_BALLOON_CMD_BATCHED_2M_LOCK: Similar to @VMW_BALLOON_CMD_BATCHED_LOCK
+ *				     for 2MB pages.
+ * @VMW_BALLOON_CMD_BATCHED_2M_UNLOCK: Similar to
+ *				       @VMW_BALLOON_CMD_BATCHED_UNLOCK for 2MB
+ *				       pages.
+ * @VMW_BALLOON_CMD_VMCI_DOORBELL_SET: A command to set doorbell notification
+ *				       that would be invoked when the balloon
+ *				       size changes.
+ * @VMW_BALLOON_CMD_LAST: Value of the last command.
  */
-#define VMW_BALLOON_BATCH_MAX_PAGES	(PAGE_SIZE / sizeof(u64))
-#define VMW_BALLOON_BATCH_STATUS_MASK	((1UL << 5) - 1)
-#define VMW_BALLOON_BATCH_PAGE_MASK	(~((1UL << PAGE_SHIFT) - 1))
-
-struct vmballoon_batch_page {
-	u64 pages[VMW_BALLOON_BATCH_MAX_PAGES];
+enum vmballoon_cmd_type {
+	VMW_BALLOON_CMD_START,
+	VMW_BALLOON_CMD_GET_TARGET,
+	VMW_BALLOON_CMD_LOCK,
+	VMW_BALLOON_CMD_UNLOCK,
+	VMW_BALLOON_CMD_GUEST_ID,
+	/* No command 5 */
+	VMW_BALLOON_CMD_BATCHED_LOCK = 6,
+	VMW_BALLOON_CMD_BATCHED_UNLOCK,
+	VMW_BALLOON_CMD_BATCHED_2M_LOCK,
+	VMW_BALLOON_CMD_BATCHED_2M_UNLOCK,
+	VMW_BALLOON_CMD_VMCI_DOORBELL_SET,
+	VMW_BALLOON_CMD_LAST = VMW_BALLOON_CMD_VMCI_DOORBELL_SET,
 };
 
-static u64 vmballoon_batch_get_pa(struct vmballoon_batch_page *batch, int idx)
-{
-	return batch->pages[idx] & VMW_BALLOON_BATCH_PAGE_MASK;
-}
+#define VMW_BALLOON_CMD_NUM	(VMW_BALLOON_CMD_LAST + 1)
+
+enum vmballoon_error_codes {
+	VMW_BALLOON_SUCCESS,
+	VMW_BALLOON_ERROR_CMD_INVALID,
+	VMW_BALLOON_ERROR_PPN_INVALID,
+	VMW_BALLOON_ERROR_PPN_LOCKED,
+	VMW_BALLOON_ERROR_PPN_UNLOCKED,
+	VMW_BALLOON_ERROR_PPN_PINNED,
+	VMW_BALLOON_ERROR_PPN_NOTNEEDED,
+	VMW_BALLOON_ERROR_RESET,
+	VMW_BALLOON_ERROR_BUSY
+};
 
-static int vmballoon_batch_get_status(struct vmballoon_batch_page *batch,
-				int idx)
-{
-	return (int)(batch->pages[idx] & VMW_BALLOON_BATCH_STATUS_MASK);
-}
+#define VMW_BALLOON_SUCCESS_WITH_CAPABILITIES	(0x03000000)
 
-static void vmballoon_batch_set_pa(struct vmballoon_batch_page *batch, int idx,
-				u64 pa)
-{
-	batch->pages[idx] = pa;
-}
+#define VMW_BALLOON_CMD_WITH_TARGET_MASK			\
+	((1UL << VMW_BALLOON_CMD_GET_TARGET)		|	\
+	 (1UL << VMW_BALLOON_CMD_LOCK)			|	\
+	 (1UL << VMW_BALLOON_CMD_UNLOCK)		|	\
+	 (1UL << VMW_BALLOON_CMD_BATCHED_LOCK)		|	\
+	 (1UL << VMW_BALLOON_CMD_BATCHED_UNLOCK)	|	\
+	 (1UL << VMW_BALLOON_CMD_BATCHED_2M_LOCK)	|	\
+	 (1UL << VMW_BALLOON_CMD_BATCHED_2M_UNLOCK))
+
+static const char * const vmballoon_cmd_names[] = {
+	[VMW_BALLOON_CMD_START]			= "start",
+	[VMW_BALLOON_CMD_GET_TARGET]		= "target",
+	[VMW_BALLOON_CMD_LOCK]			= "lock",
+	[VMW_BALLOON_CMD_UNLOCK]		= "unlock",
+	[VMW_BALLOON_CMD_GUEST_ID]		= "guestType",
+	[VMW_BALLOON_CMD_BATCHED_LOCK]		= "batchLock",
+	[VMW_BALLOON_CMD_BATCHED_UNLOCK]	= "batchUnlock",
+	[VMW_BALLOON_CMD_BATCHED_2M_LOCK]	= "2m-lock",
+	[VMW_BALLOON_CMD_BATCHED_2M_UNLOCK]	= "2m-unlock",
+	[VMW_BALLOON_CMD_VMCI_DOORBELL_SET]	= "doorbellSet"
+};
 
+enum vmballoon_stat_page {
+	VMW_BALLOON_PAGE_STAT_ALLOC,
+	VMW_BALLOON_PAGE_STAT_ALLOC_FAIL,
+	VMW_BALLOON_PAGE_STAT_REFUSED_ALLOC,
+	VMW_BALLOON_PAGE_STAT_REFUSED_FREE,
+	VMW_BALLOON_PAGE_STAT_FREE,
+	VMW_BALLOON_PAGE_STAT_LAST = VMW_BALLOON_PAGE_STAT_FREE
+};
 
-#define VMWARE_BALLOON_CMD(cmd, arg1, arg2, result)		\
-({								\
-	unsigned long __status, __dummy1, __dummy2, __dummy3;	\
-	__asm__ __volatile__ ("inl %%dx" :			\
-		"=a"(__status),					\
-		"=c"(__dummy1),					\
-		"=d"(__dummy2),					\
-		"=b"(result),					\
-		"=S" (__dummy3) :				\
-		"0"(VMW_BALLOON_HV_MAGIC),			\
-		"1"(VMW_BALLOON_CMD_##cmd),			\
-		"2"(VMW_BALLOON_HV_PORT),			\
-		"3"(arg1),					\
-		"4" (arg2) :					\
-		"memory");					\
-	if (VMW_BALLOON_CMD_##cmd == VMW_BALLOON_CMD_START)	\
-		result = __dummy1;				\
-	result &= -1UL;						\
-	__status & -1UL;					\
-})
+#define VMW_BALLOON_PAGE_STAT_NUM	(VMW_BALLOON_PAGE_STAT_LAST + 1)
 
-#ifdef CONFIG_DEBUG_FS
-struct vmballoon_stats {
-	unsigned int timer;
-	unsigned int doorbell;
-
-	/* allocation statistics */
-	unsigned int alloc[VMW_BALLOON_NUM_PAGE_SIZES];
-	unsigned int alloc_fail[VMW_BALLOON_NUM_PAGE_SIZES];
-	unsigned int sleep_alloc;
-	unsigned int sleep_alloc_fail;
-	unsigned int refused_alloc[VMW_BALLOON_NUM_PAGE_SIZES];
-	unsigned int refused_free[VMW_BALLOON_NUM_PAGE_SIZES];
-	unsigned int free[VMW_BALLOON_NUM_PAGE_SIZES];
-
-	/* monitor operations */
-	unsigned int lock[VMW_BALLOON_NUM_PAGE_SIZES];
-	unsigned int lock_fail[VMW_BALLOON_NUM_PAGE_SIZES];
-	unsigned int unlock[VMW_BALLOON_NUM_PAGE_SIZES];
-	unsigned int unlock_fail[VMW_BALLOON_NUM_PAGE_SIZES];
-	unsigned int target;
-	unsigned int target_fail;
-	unsigned int start;
-	unsigned int start_fail;
-	unsigned int guest_type;
-	unsigned int guest_type_fail;
-	unsigned int doorbell_set;
-	unsigned int doorbell_unset;
+enum vmballoon_stat_general {
+	VMW_BALLOON_STAT_TIMER,
+	VMW_BALLOON_STAT_DOORBELL,
+	VMW_BALLOON_STAT_RESET,
+	VMW_BALLOON_STAT_LAST = VMW_BALLOON_STAT_RESET
 };
 
-#define STATS_INC(stat) (stat)++
-#else
-#define STATS_INC(stat)
-#endif
+#define VMW_BALLOON_STAT_NUM		(VMW_BALLOON_STAT_LAST + 1)
 
-struct vmballoon;
 
-struct vmballoon_ops {
-	void (*add_page)(struct vmballoon *b, int idx, struct page *p);
-	int (*lock)(struct vmballoon *b, unsigned int num_pages,
-			bool is_2m_pages, unsigned int *target);
-	int (*unlock)(struct vmballoon *b, unsigned int num_pages,
-			bool is_2m_pages, unsigned int *target);
+static DEFINE_STATIC_KEY_TRUE(vmw_balloon_batching);
+static DEFINE_STATIC_KEY_FALSE(balloon_stat_enabled);
+
+struct vmballoon_ctl {
+	struct list_head pages;
+	struct list_head refused_pages;
+	unsigned int n_refused_pages;
+	unsigned int n_pages;
+	enum vmballoon_page_size_type page_size;
+	enum vmballoon_op op;
 };
 
 struct vmballoon_page_size {
 	/* list of reserved physical pages */
 	struct list_head pages;
-
-	/* transient list of non-balloonable pages */
-	struct list_head refused_pages;
-	unsigned int n_refused_pages;
 };
 
+/**
+ * struct vmballoon_batch_entry - a batch entry for lock or unlock.
+ *
+ * @status: the status of the operation, which is written by the hypervisor.
+ * @reserved: reserved for future use. Must be set to zero.
+ * @pfn: the physical frame number of the page to be locked or unlocked.
+ */
+struct vmballoon_batch_entry {
+	u64 status : 5;
+	u64 reserved : PAGE_SHIFT - 5;
+	u64 pfn : 52;
+} __packed;
+
 struct vmballoon {
 	struct vmballoon_page_size page_sizes[VMW_BALLOON_NUM_PAGE_SIZES];
 
-	/* supported page sizes. 1 == 4k pages only, 2 == 4k and 2m pages */
-	unsigned supported_page_sizes;
+	/**
+	 * @max_page_size: maximum supported page size for ballooning.
+	 *
+	 * Protected by @conf_sem
+	 */
+	enum vmballoon_page_size_type max_page_size;
+
+	/**
+	 * @size: balloon actual size in basic page size (frames).
+	 *
+	 * While we currently do not support size which is bigger than 32-bit,
+	 * in preparation for future support, use 64-bits.
+	 */
+	atomic64_t size;
 
-	/* balloon size in pages */
-	unsigned int size;
-	unsigned int target;
+	/**
+	 * @target: balloon target size in basic page size (frames).
+	 *
+	 * We do not protect the target under the assumption that setting the
+	 * value is always done through a single write. If this assumption ever
+	 * breaks, we would have to use X_ONCE for accesses, and suffer the less
+	 * optimized code. Although we may read stale target value if multiple
+	 * accesses happen at once, the performance impact should be minor.
+	 */
+	unsigned long target;
 
-	/* reset flag */
+	/**
+	 * @reset_required: reset flag
+	 *
+	 * Setting this flag may introduce races, but the code is expected to
+	 * handle them gracefully. In the worst case, another operation will
+	 * fail as reset did not take place. Clearing the flag is done while
+	 * holding @conf_sem for write.
+	 */
 	bool reset_required;
 
+	/**
+	 * @capabilities: hypervisor balloon capabilities.
+	 *
+	 * Protected by @conf_sem.
+	 */
 	unsigned long capabilities;
 
-	struct vmballoon_batch_page *batch_page;
+	/**
+	 * @batch_page: pointer to communication batch page.
+	 *
+	 * When batching is used, batch_page points to a page, which holds up to
+	 * %VMW_BALLOON_BATCH_MAX_PAGES entries for locking or unlocking.
+	 */
+	struct vmballoon_batch_entry *batch_page;
+
+	/**
+	 * @batch_max_pages: maximum pages that can be locked/unlocked.
+	 *
+	 * Indicates the number of pages that the hypervisor can lock or unlock
+	 * at once, according to whether batching is enabled. If batching is
+	 * disabled, only a single page can be locked/unlock on each operation.
+	 *
+	 * Protected by @conf_sem.
+	 */
 	unsigned int batch_max_pages;
-	struct page *page;
 
-	const struct vmballoon_ops *ops;
+	/**
+	 * @page: page to be locked/unlocked by the hypervisor
+	 *
+	 * @page is only used when batching is disabled and a single page is
+	 * reclaimed on each iteration.
+	 *
+	 * Protected by @comm_lock.
+	 */
+	struct page *page;
 
-#ifdef CONFIG_DEBUG_FS
 	/* statistics */
-	struct vmballoon_stats stats;
+	struct vmballoon_stats *stats;
 
+#ifdef CONFIG_DEBUG_FS
 	/* debugfs file exporting statistics */
 	struct dentry *dbg_entry;
 #endif
 
-	struct sysinfo sysinfo;
-
 	struct delayed_work dwork;
 
+	/**
+	 * @vmci_doorbell.
+	 *
+	 * Protected by @conf_sem.
+	 */
 	struct vmci_handle vmci_doorbell;
+
+	/**
+	 * @conf_sem: semaphore to protect the configuration and the statistics.
+	 */
+	struct rw_semaphore conf_sem;
+
+	/**
+	 * @comm_lock: lock to protect the communication with the host.
+	 *
+	 * Lock ordering: @conf_sem -> @comm_lock .
+	 */
+	spinlock_t comm_lock;
 };
 
 static struct vmballoon balloon;
 
+struct vmballoon_stats {
+	/* timer / doorbell operations */
+	atomic64_t general_stat[VMW_BALLOON_STAT_NUM];
+
+	/* allocation statistics for huge and small pages */
+	atomic64_t
+	       page_stat[VMW_BALLOON_PAGE_STAT_NUM][VMW_BALLOON_NUM_PAGE_SIZES];
+
+	/* Monitor operations: total operations, and failures */
+	atomic64_t ops[VMW_BALLOON_CMD_NUM][VMW_BALLOON_OP_STAT_TYPES];
+};
+
+static inline bool is_vmballoon_stats_on(void)
+{
+	return IS_ENABLED(CONFIG_DEBUG_FS) &&
+		static_branch_unlikely(&balloon_stat_enabled);
+}
+
+static inline void vmballoon_stats_op_inc(struct vmballoon *b, unsigned int op,
+					  enum vmballoon_op_stat_type type)
+{
+	if (is_vmballoon_stats_on())
+		atomic64_inc(&b->stats->ops[op][type]);
+}
+
+static inline void vmballoon_stats_gen_inc(struct vmballoon *b,
+					   enum vmballoon_stat_general stat)
+{
+	if (is_vmballoon_stats_on())
+		atomic64_inc(&b->stats->general_stat[stat]);
+}
+
+static inline void vmballoon_stats_gen_add(struct vmballoon *b,
+					   enum vmballoon_stat_general stat,
+					   unsigned int val)
+{
+	if (is_vmballoon_stats_on())
+		atomic64_add(val, &b->stats->general_stat[stat]);
+}
+
+static inline void vmballoon_stats_page_inc(struct vmballoon *b,
+					    enum vmballoon_stat_page stat,
+					    enum vmballoon_page_size_type size)
+{
+	if (is_vmballoon_stats_on())
+		atomic64_inc(&b->stats->page_stat[stat][size]);
+}
+
+static inline void vmballoon_stats_page_add(struct vmballoon *b,
+					    enum vmballoon_stat_page stat,
+					    enum vmballoon_page_size_type size,
+					    unsigned int val)
+{
+	if (is_vmballoon_stats_on())
+		atomic64_add(val, &b->stats->page_stat[stat][size]);
+}
+
+static inline unsigned long
+__vmballoon_cmd(struct vmballoon *b, unsigned long cmd, unsigned long arg1,
+		unsigned long arg2, unsigned long *result)
+{
+	unsigned long status, dummy1, dummy2, dummy3, local_result;
+
+	vmballoon_stats_op_inc(b, cmd, VMW_BALLOON_OP_STAT);
+
+	asm volatile ("inl %%dx" :
+		"=a"(status),
+		"=c"(dummy1),
+		"=d"(dummy2),
+		"=b"(local_result),
+		"=S"(dummy3) :
+		"0"(VMW_BALLOON_HV_MAGIC),
+		"1"(cmd),
+		"2"(VMW_BALLOON_HV_PORT),
+		"3"(arg1),
+		"4"(arg2) :
+		"memory");
+
+	/* update the result if needed */
+	if (result)
+		*result = (cmd == VMW_BALLOON_CMD_START) ? dummy1 :
+							   local_result;
+
+	/* update target when applicable */
+	if (status == VMW_BALLOON_SUCCESS &&
+	    ((1ul << cmd) & VMW_BALLOON_CMD_WITH_TARGET_MASK))
+		WRITE_ONCE(b->target, local_result);
+
+	if (status != VMW_BALLOON_SUCCESS &&
+	    status != VMW_BALLOON_SUCCESS_WITH_CAPABILITIES) {
+		vmballoon_stats_op_inc(b, cmd, VMW_BALLOON_OP_FAIL_STAT);
+		pr_debug("%s: %s [0x%lx,0x%lx) failed, returned %ld\n",
+			 __func__, vmballoon_cmd_names[cmd], arg1, arg2,
+			 status);
+	}
+
+	/* mark reset required accordingly */
+	if (status == VMW_BALLOON_ERROR_RESET)
+		b->reset_required = true;
+
+	return status;
+}
+
+static __always_inline unsigned long
+vmballoon_cmd(struct vmballoon *b, unsigned long cmd, unsigned long arg1,
+	      unsigned long arg2)
+{
+	unsigned long dummy;
+
+	return __vmballoon_cmd(b, cmd, arg1, arg2, &dummy);
+}
+
 /*
  * Send "start" command to the host, communicating supported version
  * of the protocol.
  */
-static bool vmballoon_send_start(struct vmballoon *b, unsigned long req_caps)
+static int vmballoon_send_start(struct vmballoon *b, unsigned long req_caps)
 {
-	unsigned long status, capabilities, dummy = 0;
-	bool success;
-
-	STATS_INC(b->stats.start);
+	unsigned long status, capabilities;
 
-	status = VMWARE_BALLOON_CMD(START, req_caps, dummy, capabilities);
+	status = __vmballoon_cmd(b, VMW_BALLOON_CMD_START, req_caps, 0,
+				 &capabilities);
 
 	switch (status) {
 	case VMW_BALLOON_SUCCESS_WITH_CAPABILITIES:
 		b->capabilities = capabilities;
-		success = true;
 		break;
 	case VMW_BALLOON_SUCCESS:
 		b->capabilities = VMW_BALLOON_BASIC_CMDS;
-		success = true;
 		break;
 	default:
-		success = false;
+		return -EIO;
 	}
 
 	/*
@@ -303,626 +501,693 @@ static bool vmballoon_send_start(struct vmballoon *b, unsigned long req_caps)
 	 * reason disabled, do not use 2MB pages, since otherwise the legacy
 	 * mechanism is used with 2MB pages, causing a failure.
 	 */
+	b->max_page_size = VMW_BALLOON_4K_PAGE;
 	if ((b->capabilities & VMW_BALLOON_BATCHED_2M_CMDS) &&
 	    (b->capabilities & VMW_BALLOON_BATCHED_CMDS))
-		b->supported_page_sizes = 2;
-	else
-		b->supported_page_sizes = 1;
-
-	if (!success) {
-		pr_debug("%s - failed, hv returns %ld\n", __func__, status);
-		STATS_INC(b->stats.start_fail);
-	}
-	return success;
-}
+		b->max_page_size = VMW_BALLOON_2M_PAGE;
 
-static bool vmballoon_check_status(struct vmballoon *b, unsigned long status)
-{
-	switch (status) {
-	case VMW_BALLOON_SUCCESS:
-		return true;
 
-	case VMW_BALLOON_ERROR_RESET:
-		b->reset_required = true;
-		/* fall through */
-
-	default:
-		return false;
-	}
+	return 0;
 }
 
-/*
+/**
+ * vmballoon_send_guest_id - communicate guest type to the host.
+ *
+ * @b: pointer to the balloon.
+ *
  * Communicate guest type to the host so that it can adjust ballooning
  * algorithm to the one most appropriate for the guest. This command
  * is normally issued after sending "start" command and is part of
  * standard reset sequence.
+ *
+ * Return: zero on success or appropriate error code.
  */
-static bool vmballoon_send_guest_id(struct vmballoon *b)
+static int vmballoon_send_guest_id(struct vmballoon *b)
 {
-	unsigned long status, dummy = 0;
-
-	status = VMWARE_BALLOON_CMD(GUEST_ID, VMW_BALLOON_GUEST_ID, dummy,
-				dummy);
-
-	STATS_INC(b->stats.guest_type);
+	unsigned long status;
 
-	if (vmballoon_check_status(b, status))
-		return true;
+	status = vmballoon_cmd(b, VMW_BALLOON_CMD_GUEST_ID,
+			       VMW_BALLOON_GUEST_ID, 0);
 
-	pr_debug("%s - failed, hv returns %ld\n", __func__, status);
-	STATS_INC(b->stats.guest_type_fail);
-	return false;
+	return status == VMW_BALLOON_SUCCESS ? 0 : -EIO;
 }
 
-static u16 vmballoon_page_size(bool is_2m_page)
+/**
+ * vmballoon_page_order() - return the order of the page
+ * @page_size: the size of the page.
+ *
+ * Return: the allocation order.
+ */
+static inline
+unsigned int vmballoon_page_order(enum vmballoon_page_size_type page_size)
 {
-	if (is_2m_page)
-		return 1 << VMW_BALLOON_2M_SHIFT;
+	return page_size == VMW_BALLOON_2M_PAGE ? VMW_BALLOON_2M_ORDER : 0;
+}
 
-	return 1;
+/**
+ * vmballoon_page_in_frames() - returns the number of frames in a page.
+ * @page_size: the size of the page.
+ *
+ * Return: the number of 4k frames.
+ */
+static inline unsigned int
+vmballoon_page_in_frames(enum vmballoon_page_size_type page_size)
+{
+	return 1 << vmballoon_page_order(page_size);
 }
 
-/*
- * Retrieve desired balloon size from the host.
+/**
+ * vmballoon_send_get_target() - Retrieve desired balloon size from the host.
+ *
+ * @b: pointer to the balloon.
+ *
+ * Return: zero on success, EINVAL if limit does not fit in 32-bit, as required
+ * by the host-guest protocol and EIO if an error occurred in communicating with
+ * the host.
  */
-static bool vmballoon_send_get_target(struct vmballoon *b, u32 *new_target)
+static int vmballoon_send_get_target(struct vmballoon *b)
 {
 	unsigned long status;
-	unsigned long target;
 	unsigned long limit;
-	unsigned long dummy = 0;
-	u32 limit32;
 
-	/*
-	 * si_meminfo() is cheap. Moreover, we want to provide dynamic
-	 * max balloon size later. So let us call si_meminfo() every
-	 * iteration.
-	 */
-	si_meminfo(&b->sysinfo);
-	limit = b->sysinfo.totalram;
+	limit = totalram_pages;
 
 	/* Ensure limit fits in 32-bits */
-	limit32 = (u32)limit;
-	if (limit != limit32)
-		return false;
-
-	/* update stats */
-	STATS_INC(b->stats.target);
+	if (limit != (u32)limit)
+		return -EINVAL;
 
-	status = VMWARE_BALLOON_CMD(GET_TARGET, limit, dummy, target);
-	if (vmballoon_check_status(b, status)) {
-		*new_target = target;
-		return true;
-	}
+	status = vmballoon_cmd(b, VMW_BALLOON_CMD_GET_TARGET, limit, 0);
 
-	pr_debug("%s - failed, hv returns %ld\n", __func__, status);
-	STATS_INC(b->stats.target_fail);
-	return false;
+	return status == VMW_BALLOON_SUCCESS ? 0 : -EIO;
 }
 
-/*
- * Notify the host about allocated page so that host can use it without
- * fear that guest will need it. Host may reject some pages, we need to
- * check the return value and maybe submit a different page.
+/**
+ * vmballoon_alloc_page_list - allocates a list of pages.
+ *
+ * @b: pointer to the balloon.
+ * @ctl: pointer for the %struct vmballoon_ctl, which defines the operation.
+ * @req_n_pages: the number of requested pages.
+ *
+ * Tries to allocate @req_n_pages. Add them to the list of balloon pages in
+ * @ctl.pages and updates @ctl.n_pages to reflect the number of pages.
+ *
+ * Return: zero on success or error code otherwise.
  */
-static int vmballoon_send_lock_page(struct vmballoon *b, unsigned long pfn,
-				unsigned int *hv_status, unsigned int *target)
+static int vmballoon_alloc_page_list(struct vmballoon *b,
+				     struct vmballoon_ctl *ctl,
+				     unsigned int req_n_pages)
 {
-	unsigned long status, dummy = 0;
-	u32 pfn32;
-
-	pfn32 = (u32)pfn;
-	if (pfn32 != pfn)
-		return -EINVAL;
-
-	STATS_INC(b->stats.lock[false]);
+	struct page *page;
+	unsigned int i;
 
-	*hv_status = status = VMWARE_BALLOON_CMD(LOCK, pfn, dummy, *target);
-	if (vmballoon_check_status(b, status))
-		return 0;
+	for (i = 0; i < req_n_pages; i++) {
+		if (ctl->page_size == VMW_BALLOON_2M_PAGE)
+			page = alloc_pages(VMW_HUGE_PAGE_ALLOC_FLAGS,
+					   VMW_BALLOON_2M_ORDER);
+		else
+			page = alloc_page(VMW_PAGE_ALLOC_FLAGS);
 
-	pr_debug("%s - ppn %lx, hv returns %ld\n", __func__, pfn, status);
-	STATS_INC(b->stats.lock_fail[false]);
-	return -EIO;
-}
+		/* Update statistics */
+		vmballoon_stats_page_inc(b, VMW_BALLOON_PAGE_STAT_ALLOC,
+					 ctl->page_size);
 
-static int vmballoon_send_batched_lock(struct vmballoon *b,
-		unsigned int num_pages, bool is_2m_pages, unsigned int *target)
-{
-	unsigned long status;
-	unsigned long pfn = PHYS_PFN(virt_to_phys(b->batch_page));
+		if (page) {
+			/* Success. Add the page to the list and continue. */
+			list_add(&page->lru, &ctl->pages);
+			continue;
+		}
 
-	STATS_INC(b->stats.lock[is_2m_pages]);
+		/* Allocation failed. Update statistics and stop. */
+		vmballoon_stats_page_inc(b, VMW_BALLOON_PAGE_STAT_ALLOC_FAIL,
+					 ctl->page_size);
+		break;
+	}
 
-	if (is_2m_pages)
-		status = VMWARE_BALLOON_CMD(BATCHED_2M_LOCK, pfn, num_pages,
-				*target);
-	else
-		status = VMWARE_BALLOON_CMD(BATCHED_LOCK, pfn, num_pages,
-				*target);
+	ctl->n_pages = i;
 
-	if (vmballoon_check_status(b, status))
-		return 0;
-
-	pr_debug("%s - batch ppn %lx, hv returns %ld\n", __func__, pfn, status);
-	STATS_INC(b->stats.lock_fail[is_2m_pages]);
-	return 1;
+	return req_n_pages == ctl->n_pages ? 0 : -ENOMEM;
 }
 
-/*
- * Notify the host that guest intends to release given page back into
- * the pool of available (to the guest) pages.
+/**
+ * vmballoon_handle_one_result - Handle lock/unlock result for a single page.
+ *
+ * @b: pointer for %struct vmballoon.
+ * @page: pointer for the page whose result should be handled.
+ * @page_size: size of the page.
+ * @status: status of the operation as provided by the hypervisor.
  */
-static bool vmballoon_send_unlock_page(struct vmballoon *b, unsigned long pfn,
-							unsigned int *target)
+static int vmballoon_handle_one_result(struct vmballoon *b, struct page *page,
+				       enum vmballoon_page_size_type page_size,
+				       unsigned long status)
 {
-	unsigned long status, dummy = 0;
-	u32 pfn32;
-
-	pfn32 = (u32)pfn;
-	if (pfn32 != pfn)
-		return false;
+	/* On success do nothing. The page is already on the balloon list. */
+	if (likely(status == VMW_BALLOON_SUCCESS))
+		return 0;
 
-	STATS_INC(b->stats.unlock[false]);
+	pr_debug("%s: failed comm pfn %lx status %lu page_size %s\n", __func__,
+		 page_to_pfn(page), status,
+		 vmballoon_page_size_names[page_size]);
 
-	status = VMWARE_BALLOON_CMD(UNLOCK, pfn, dummy, *target);
-	if (vmballoon_check_status(b, status))
-		return true;
+	/* Error occurred */
+	vmballoon_stats_page_inc(b, VMW_BALLOON_PAGE_STAT_REFUSED_ALLOC,
+				 page_size);
 
-	pr_debug("%s - ppn %lx, hv returns %ld\n", __func__, pfn, status);
-	STATS_INC(b->stats.unlock_fail[false]);
-	return false;
+	return -EIO;
 }
 
-static bool vmballoon_send_batched_unlock(struct vmballoon *b,
-		unsigned int num_pages, bool is_2m_pages, unsigned int *target)
+/**
+ * vmballoon_status_page - returns the status of (un)lock operation
+ *
+ * @b: pointer to the balloon.
+ * @idx: index for the page for which the operation is performed.
+ * @p: pointer to where the page struct is returned.
+ *
+ * Following a lock or unlock operation, returns the status of the operation for
+ * an individual page. Provides the page that the operation was performed on on
+ * the @page argument.
+ *
+ * Returns: The status of a lock or unlock operation for an individual page.
+ */
+static unsigned long vmballoon_status_page(struct vmballoon *b, int idx,
+					   struct page **p)
 {
-	unsigned long status;
-	unsigned long pfn = PHYS_PFN(virt_to_phys(b->batch_page));
-
-	STATS_INC(b->stats.unlock[is_2m_pages]);
-
-	if (is_2m_pages)
-		status = VMWARE_BALLOON_CMD(BATCHED_2M_UNLOCK, pfn, num_pages,
-				*target);
-	else
-		status = VMWARE_BALLOON_CMD(BATCHED_UNLOCK, pfn, num_pages,
-				*target);
+	if (static_branch_likely(&vmw_balloon_batching)) {
+		/* batching mode */
+		*p = pfn_to_page(b->batch_page[idx].pfn);
+		return b->batch_page[idx].status;
+	}
 
-	if (vmballoon_check_status(b, status))
-		return true;
+	/* non-batching mode */
+	*p = b->page;
 
-	pr_debug("%s - batch ppn %lx, hv returns %ld\n", __func__, pfn, status);
-	STATS_INC(b->stats.unlock_fail[is_2m_pages]);
-	return false;
+	/*
+	 * If a failure occurs, the indication will be provided in the status
+	 * of the entire operation, which is considered before the individual
+	 * page status. So for non-batching mode, the indication is always of
+	 * success.
+	 */
+	return VMW_BALLOON_SUCCESS;
 }
 
-static struct page *vmballoon_alloc_page(gfp_t flags, bool is_2m_page)
+/**
+ * vmballoon_lock_op - notifies the host about inflated/deflated pages.
+ * @b: pointer to the balloon.
+ * @num_pages: number of inflated/deflated pages.
+ * @page_size: size of the page.
+ * @op: the type of operation (lock or unlock).
+ *
+ * Notify the host about page(s) that were ballooned (or removed from the
+ * balloon) so that host can use it without fear that guest will need it (or
+ * stop using them since the VM does). Host may reject some pages, we need to
+ * check the return value and maybe submit a different page. The pages that are
+ * inflated/deflated are pointed by @b->page.
+ *
+ * Return: result as provided by the hypervisor.
+ */
+static unsigned long vmballoon_lock_op(struct vmballoon *b,
+				       unsigned int num_pages,
+				       enum vmballoon_page_size_type page_size,
+				       enum vmballoon_op op)
 {
-	if (is_2m_page)
-		return alloc_pages(flags, VMW_BALLOON_2M_SHIFT);
+	unsigned long cmd, pfn;
 
-	return alloc_page(flags);
-}
+	lockdep_assert_held(&b->comm_lock);
 
-static void vmballoon_free_page(struct page *page, bool is_2m_page)
-{
-	if (is_2m_page)
-		__free_pages(page, VMW_BALLOON_2M_SHIFT);
-	else
-		__free_page(page);
+	if (static_branch_likely(&vmw_balloon_batching)) {
+		if (op == VMW_BALLOON_INFLATE)
+			cmd = page_size == VMW_BALLOON_2M_PAGE ?
+				VMW_BALLOON_CMD_BATCHED_2M_LOCK :
+				VMW_BALLOON_CMD_BATCHED_LOCK;
+		else
+			cmd = page_size == VMW_BALLOON_2M_PAGE ?
+				VMW_BALLOON_CMD_BATCHED_2M_UNLOCK :
+				VMW_BALLOON_CMD_BATCHED_UNLOCK;
+
+		pfn = PHYS_PFN(virt_to_phys(b->batch_page));
+	} else {
+		cmd = op == VMW_BALLOON_INFLATE ? VMW_BALLOON_CMD_LOCK :
+						  VMW_BALLOON_CMD_UNLOCK;
+		pfn = page_to_pfn(b->page);
+
+		/* In non-batching mode, PFNs must fit in 32-bit */
+		if (unlikely(pfn != (u32)pfn))
+			return VMW_BALLOON_ERROR_PPN_INVALID;
+	}
+
+	return vmballoon_cmd(b, cmd, pfn, num_pages);
 }
 
-/*
- * Quickly release all pages allocated for the balloon. This function is
- * called when host decides to "reset" balloon for one reason or another.
- * Unlike normal "deflate" we do not (shall not) notify host of the pages
- * being released.
+/**
+ * vmballoon_add_page - adds a page towards lock/unlock operation.
+ *
+ * @b: pointer to the balloon.
+ * @idx: index of the page to be ballooned in this batch.
+ * @p: pointer to the page that is about to be ballooned.
+ *
+ * Adds the page to be ballooned. Must be called while holding @comm_lock.
  */
-static void vmballoon_pop(struct vmballoon *b)
+static void vmballoon_add_page(struct vmballoon *b, unsigned int idx,
+			       struct page *p)
 {
-	struct page *page, *next;
-	unsigned is_2m_pages;
-
-	for (is_2m_pages = 0; is_2m_pages < VMW_BALLOON_NUM_PAGE_SIZES;
-			is_2m_pages++) {
-		struct vmballoon_page_size *page_size =
-				&b->page_sizes[is_2m_pages];
-		u16 size_per_page = vmballoon_page_size(is_2m_pages);
-
-		list_for_each_entry_safe(page, next, &page_size->pages, lru) {
-			list_del(&page->lru);
-			vmballoon_free_page(page, is_2m_pages);
-			STATS_INC(b->stats.free[is_2m_pages]);
-			b->size -= size_per_page;
-			cond_resched();
-		}
-	}
+	lockdep_assert_held(&b->comm_lock);
 
-	/* Clearing the batch_page unconditionally has no adverse effect */
-	free_page((unsigned long)b->batch_page);
-	b->batch_page = NULL;
+	if (static_branch_likely(&vmw_balloon_batching))
+		b->batch_page[idx] = (struct vmballoon_batch_entry)
+					{ .pfn = page_to_pfn(p) };
+	else
+		b->page = p;
 }
 
-/*
- * Notify the host of a ballooned page. If host rejects the page put it on the
- * refuse list, those refused page are then released at the end of the
- * inflation cycle.
+/**
+ * vmballoon_lock - lock or unlock a batch of pages.
+ *
+ * @b: pointer to the balloon.
+ * @ctl: pointer for the %struct vmballoon_ctl, which defines the operation.
+ *
+ * Notifies the host of about ballooned pages (after inflation or deflation,
+ * according to @ctl). If the host rejects the page put it on the
+ * @ctl refuse list. These refused page are then released when moving to the
+ * next size of pages.
+ *
+ * Note that we neither free any @page here nor put them back on the ballooned
+ * pages list. Instead we queue it for later processing. We do that for several
+ * reasons. First, we do not want to free the page under the lock. Second, it
+ * allows us to unify the handling of lock and unlock. In the inflate case, the
+ * caller will check if there are too many refused pages and release them.
+ * Although it is not identical to the past behavior, it should not affect
+ * performance.
  */
-static int vmballoon_lock_page(struct vmballoon *b, unsigned int num_pages,
-				bool is_2m_pages, unsigned int *target)
+static int vmballoon_lock(struct vmballoon *b, struct vmballoon_ctl *ctl)
 {
-	int locked, hv_status;
-	struct page *page = b->page;
-	struct vmballoon_page_size *page_size = &b->page_sizes[false];
-
-	/* is_2m_pages can never happen as 2m pages support implies batching */
-
-	locked = vmballoon_send_lock_page(b, page_to_pfn(page), &hv_status,
-								target);
-	if (locked) {
-		STATS_INC(b->stats.refused_alloc[false]);
-
-		if (locked == -EIO &&
-		    (hv_status == VMW_BALLOON_ERROR_RESET ||
-		     hv_status == VMW_BALLOON_ERROR_PPN_NOTNEEDED)) {
-			vmballoon_free_page(page, false);
-			return -EIO;
-		}
-
-		/*
-		 * Place page on the list of non-balloonable pages
-		 * and retry allocation, unless we already accumulated
-		 * too many of them, in which case take a breather.
-		 */
-		if (page_size->n_refused_pages < VMW_BALLOON_MAX_REFUSED) {
-			page_size->n_refused_pages++;
-			list_add(&page->lru, &page_size->refused_pages);
-		} else {
-			vmballoon_free_page(page, false);
-		}
-		return locked;
-	}
-
-	/* track allocated page */
-	list_add(&page->lru, &page_size->pages);
+	unsigned long batch_status;
+	struct page *page;
+	unsigned int i, num_pages;
 
-	/* update balloon size */
-	b->size++;
+	num_pages = ctl->n_pages;
+	if (num_pages == 0)
+		return 0;
 
-	return 0;
-}
+	/* communication with the host is done under the communication lock */
+	spin_lock(&b->comm_lock);
 
-static int vmballoon_lock_batched_page(struct vmballoon *b,
-		unsigned int num_pages, bool is_2m_pages, unsigned int *target)
-{
-	int locked, i;
-	u16 size_per_page = vmballoon_page_size(is_2m_pages);
+	i = 0;
+	list_for_each_entry(page, &ctl->pages, lru)
+		vmballoon_add_page(b, i++, page);
 
-	locked = vmballoon_send_batched_lock(b, num_pages, is_2m_pages,
-			target);
-	if (locked > 0) {
-		for (i = 0; i < num_pages; i++) {
-			u64 pa = vmballoon_batch_get_pa(b->batch_page, i);
-			struct page *p = pfn_to_page(pa >> PAGE_SHIFT);
+	batch_status = vmballoon_lock_op(b, ctl->n_pages, ctl->page_size,
+					 ctl->op);
 
-			vmballoon_free_page(p, is_2m_pages);
-		}
+	/*
+	 * Iterate over the pages in the provided list. Since we are changing
+	 * @ctl->n_pages we are saving the original value in @num_pages and
+	 * use this value to bound the loop.
+	 */
+	for (i = 0; i < num_pages; i++) {
+		unsigned long status;
 
-		return -EIO;
-	}
+		status = vmballoon_status_page(b, i, &page);
 
-	for (i = 0; i < num_pages; i++) {
-		u64 pa = vmballoon_batch_get_pa(b->batch_page, i);
-		struct page *p = pfn_to_page(pa >> PAGE_SHIFT);
-		struct vmballoon_page_size *page_size =
-				&b->page_sizes[is_2m_pages];
+		/*
+		 * Failure of the whole batch overrides a single operation
+		 * results.
+		 */
+		if (batch_status != VMW_BALLOON_SUCCESS)
+			status = batch_status;
 
-		locked = vmballoon_batch_get_status(b->batch_page, i);
+		/* Continue if no error happened */
+		if (!vmballoon_handle_one_result(b, page, ctl->page_size,
+						 status))
+			continue;
 
-		switch (locked) {
-		case VMW_BALLOON_SUCCESS:
-			list_add(&p->lru, &page_size->pages);
-			b->size += size_per_page;
-			break;
-		case VMW_BALLOON_ERROR_PPN_PINNED:
-		case VMW_BALLOON_ERROR_PPN_INVALID:
-			if (page_size->n_refused_pages
-					< VMW_BALLOON_MAX_REFUSED) {
-				list_add(&p->lru, &page_size->refused_pages);
-				page_size->n_refused_pages++;
-				break;
-			}
-			/* Fallthrough */
-		case VMW_BALLOON_ERROR_RESET:
-		case VMW_BALLOON_ERROR_PPN_NOTNEEDED:
-			vmballoon_free_page(p, is_2m_pages);
-			break;
-		default:
-			/* This should never happen */
-			WARN_ON_ONCE(true);
-		}
+		/*
+		 * Error happened. Move the pages to the refused list and update
+		 * the pages number.
+		 */
+		list_move(&page->lru, &ctl->refused_pages);
+		ctl->n_pages--;
+		ctl->n_refused_pages++;
 	}
 
-	return 0;
+	spin_unlock(&b->comm_lock);
+
+	return batch_status == VMW_BALLOON_SUCCESS ? 0 : -EIO;
 }
 
-/*
- * Release the page allocated for the balloon. Note that we first notify
- * the host so it can make sure the page will be available for the guest
- * to use, if needed.
+/**
+ * vmballoon_release_page_list() - Releases a page list
+ *
+ * @page_list: list of pages to release.
+ * @n_pages: pointer to the number of pages.
+ * @page_size: whether the pages in the list are 2MB (or else 4KB).
+ *
+ * Releases the list of pages and zeros the number of pages.
  */
-static int vmballoon_unlock_page(struct vmballoon *b, unsigned int num_pages,
-		bool is_2m_pages, unsigned int *target)
+static void vmballoon_release_page_list(struct list_head *page_list,
+				       int *n_pages,
+				       enum vmballoon_page_size_type page_size)
 {
-	struct page *page = b->page;
-	struct vmballoon_page_size *page_size = &b->page_sizes[false];
-
-	/* is_2m_pages can never happen as 2m pages support implies batching */
+	struct page *page, *tmp;
 
-	if (!vmballoon_send_unlock_page(b, page_to_pfn(page), target)) {
-		list_add(&page->lru, &page_size->pages);
-		return -EIO;
+	list_for_each_entry_safe(page, tmp, page_list, lru) {
+		list_del(&page->lru);
+		__free_pages(page, vmballoon_page_order(page_size));
 	}
 
-	/* deallocate page */
-	vmballoon_free_page(page, false);
-	STATS_INC(b->stats.free[false]);
+	*n_pages = 0;
+}
 
-	/* update balloon size */
-	b->size--;
 
-	return 0;
+/*
+ * Release pages that were allocated while attempting to inflate the
+ * balloon but were refused by the host for one reason or another.
+ */
+static void vmballoon_release_refused_pages(struct vmballoon *b,
+					    struct vmballoon_ctl *ctl)
+{
+	vmballoon_stats_page_inc(b, VMW_BALLOON_PAGE_STAT_REFUSED_FREE,
+				 ctl->page_size);
+
+	vmballoon_release_page_list(&ctl->refused_pages, &ctl->n_refused_pages,
+				    ctl->page_size);
 }
 
-static int vmballoon_unlock_batched_page(struct vmballoon *b,
-				unsigned int num_pages, bool is_2m_pages,
-				unsigned int *target)
+/**
+ * vmballoon_change - retrieve the required balloon change
+ *
+ * @b: pointer for the balloon.
+ *
+ * Return: the required change for the balloon size. A positive number
+ * indicates inflation, a negative number indicates a deflation.
+ */
+static int64_t vmballoon_change(struct vmballoon *b)
 {
-	int locked, i, ret = 0;
-	bool hv_success;
-	u16 size_per_page = vmballoon_page_size(is_2m_pages);
+	int64_t size, target;
 
-	hv_success = vmballoon_send_batched_unlock(b, num_pages, is_2m_pages,
-			target);
-	if (!hv_success)
-		ret = -EIO;
+	size = atomic64_read(&b->size);
+	target = READ_ONCE(b->target);
 
-	for (i = 0; i < num_pages; i++) {
-		u64 pa = vmballoon_batch_get_pa(b->batch_page, i);
-		struct page *p = pfn_to_page(pa >> PAGE_SHIFT);
-		struct vmballoon_page_size *page_size =
-				&b->page_sizes[is_2m_pages];
+	/*
+	 * We must cast first because of int sizes
+	 * Otherwise we might get huge positives instead of negatives
+	 */
 
-		locked = vmballoon_batch_get_status(b->batch_page, i);
-		if (!hv_success || locked != VMW_BALLOON_SUCCESS) {
-			/*
-			 * That page wasn't successfully unlocked by the
-			 * hypervisor, re-add it to the list of pages owned by
-			 * the balloon driver.
-			 */
-			list_add(&p->lru, &page_size->pages);
-		} else {
-			/* deallocate page */
-			vmballoon_free_page(p, is_2m_pages);
-			STATS_INC(b->stats.free[is_2m_pages]);
-
-			/* update balloon size */
-			b->size -= size_per_page;
-		}
-	}
+	if (b->reset_required)
+		return 0;
+
+	/* consider a 2MB slack on deflate, unless the balloon is emptied */
+	if (target < size && target != 0 &&
+	    size - target < vmballoon_page_in_frames(VMW_BALLOON_2M_PAGE))
+		return 0;
 
-	return ret;
+	return target - size;
 }
 
-/*
- * Release pages that were allocated while attempting to inflate the
- * balloon but were refused by the host for one reason or another.
+/**
+ * vmballoon_enqueue_page_list() - Enqueues list of pages after inflation.
+ *
+ * @b: pointer to balloon.
+ * @pages: list of pages to enqueue.
+ * @n_pages: pointer to number of pages in list. The value is zeroed.
+ * @page_size: whether the pages are 2MB or 4KB pages.
+ *
+ * Enqueues the provides list of pages in the ballooned page list, clears the
+ * list and zeroes the number of pages that was provided.
  */
-static void vmballoon_release_refused_pages(struct vmballoon *b,
-		bool is_2m_pages)
+static void vmballoon_enqueue_page_list(struct vmballoon *b,
+					struct list_head *pages,
+					unsigned int *n_pages,
+					enum vmballoon_page_size_type page_size)
 {
-	struct page *page, *next;
-	struct vmballoon_page_size *page_size =
-			&b->page_sizes[is_2m_pages];
+	struct vmballoon_page_size *page_size_info = &b->page_sizes[page_size];
 
-	list_for_each_entry_safe(page, next, &page_size->refused_pages, lru) {
-		list_del(&page->lru);
-		vmballoon_free_page(page, is_2m_pages);
-		STATS_INC(b->stats.refused_free[is_2m_pages]);
-	}
-
-	page_size->n_refused_pages = 0;
+	list_splice_init(pages, &page_size_info->pages);
+	*n_pages = 0;
 }
 
-static void vmballoon_add_page(struct vmballoon *b, int idx, struct page *p)
+/**
+ * vmballoon_dequeue_page_list() - Dequeues page lists for deflation.
+ *
+ * @b: pointer to balloon.
+ * @pages: list of pages to enqueue.
+ * @n_pages: pointer to number of pages in list. The value is zeroed.
+ * @page_size: whether the pages are 2MB or 4KB pages.
+ * @n_req_pages: the number of requested pages.
+ *
+ * Dequeues the number of requested pages from the balloon for deflation. The
+ * number of dequeued pages may be lower, if not enough pages in the requested
+ * size are available.
+ */
+static void vmballoon_dequeue_page_list(struct vmballoon *b,
+					struct list_head *pages,
+					unsigned int *n_pages,
+					enum vmballoon_page_size_type page_size,
+					unsigned int n_req_pages)
 {
-	b->page = p;
-}
+	struct vmballoon_page_size *page_size_info = &b->page_sizes[page_size];
+	struct page *page, *tmp;
+	unsigned int i = 0;
 
-static void vmballoon_add_batched_page(struct vmballoon *b, int idx,
-				struct page *p)
-{
-	vmballoon_batch_set_pa(b->batch_page, idx,
-			(u64)page_to_pfn(p) << PAGE_SHIFT);
+	list_for_each_entry_safe(page, tmp, &page_size_info->pages, lru) {
+		list_move(&page->lru, pages);
+		if (++i == n_req_pages)
+			break;
+	}
+	*n_pages = i;
 }
 
-/*
- * Inflate the balloon towards its target size. Note that we try to limit
- * the rate of allocation to make sure we are not choking the rest of the
- * system.
+/**
+ * vmballoon_inflate() - Inflate the balloon towards its target size.
+ *
+ * @b: pointer to the balloon.
  */
 static void vmballoon_inflate(struct vmballoon *b)
 {
-	unsigned int num_pages = 0;
-	int error = 0;
-	gfp_t flags = VMW_PAGE_ALLOC_NOSLEEP;
-	bool is_2m_pages;
+	int64_t to_inflate_frames;
+	struct vmballoon_ctl ctl = {
+		.pages = LIST_HEAD_INIT(ctl.pages),
+		.refused_pages = LIST_HEAD_INIT(ctl.refused_pages),
+		.page_size = b->max_page_size,
+		.op = VMW_BALLOON_INFLATE
+	};
 
-	pr_debug("%s - size: %d, target %d\n", __func__, b->size, b->target);
+	while ((to_inflate_frames = vmballoon_change(b)) > 0) {
+		unsigned int to_inflate_pages, page_in_frames;
+		int alloc_error, lock_error = 0;
 
-	/*
-	 * First try NOSLEEP page allocations to inflate balloon.
-	 *
-	 * If we do not throttle nosleep allocations, we can drain all
-	 * free pages in the guest quickly (if the balloon target is high).
-	 * As a side-effect, draining free pages helps to inform (force)
-	 * the guest to start swapping if balloon target is not met yet,
-	 * which is a desired behavior. However, balloon driver can consume
-	 * all available CPU cycles if too many pages are allocated in a
-	 * second. Therefore, we throttle nosleep allocations even when
-	 * the guest is not under memory pressure. OTOH, if we have already
-	 * predicted that the guest is under memory pressure, then we
-	 * slowdown page allocations considerably.
-	 */
+		VM_BUG_ON(!list_empty(&ctl.pages));
+		VM_BUG_ON(ctl.n_pages != 0);
 
-	/*
-	 * Start with no sleep allocation rate which may be higher
-	 * than sleeping allocation rate.
-	 */
-	is_2m_pages = b->supported_page_sizes == VMW_BALLOON_NUM_PAGE_SIZES;
+		page_in_frames = vmballoon_page_in_frames(ctl.page_size);
 
-	pr_debug("%s - goal: %d",  __func__, b->target - b->size);
+		to_inflate_pages = min_t(unsigned long, b->batch_max_pages,
+					 DIV_ROUND_UP_ULL(to_inflate_frames,
+							  page_in_frames));
 
-	while (!b->reset_required &&
-		b->size + num_pages * vmballoon_page_size(is_2m_pages)
-		< b->target) {
-		struct page *page;
+		/* Start by allocating */
+		alloc_error = vmballoon_alloc_page_list(b, &ctl,
+							to_inflate_pages);
 
-		if (flags == VMW_PAGE_ALLOC_NOSLEEP)
-			STATS_INC(b->stats.alloc[is_2m_pages]);
-		else
-			STATS_INC(b->stats.sleep_alloc);
-
-		page = vmballoon_alloc_page(flags, is_2m_pages);
-		if (!page) {
-			STATS_INC(b->stats.alloc_fail[is_2m_pages]);
-
-			if (is_2m_pages) {
-				b->ops->lock(b, num_pages, true, &b->target);
-
-				/*
-				 * ignore errors from locking as we now switch
-				 * to 4k pages and we might get different
-				 * errors.
-				 */
-
-				num_pages = 0;
-				is_2m_pages = false;
-				continue;
-			}
-
-			if (flags == VMW_PAGE_ALLOC_CANSLEEP) {
-				/*
-				 * CANSLEEP page allocation failed, so guest
-				 * is under severe memory pressure. We just log
-				 * the event, but do not stop the inflation
-				 * due to its negative impact on performance.
-				 */
-				STATS_INC(b->stats.sleep_alloc_fail);
+		/* Actually lock the pages by telling the hypervisor */
+		lock_error = vmballoon_lock(b, &ctl);
+
+		/*
+		 * If an error indicates that something serious went wrong,
+		 * stop the inflation.
+		 */
+		if (lock_error)
+			break;
+
+		/* Update the balloon size */
+		atomic64_add(ctl.n_pages * page_in_frames, &b->size);
+
+		vmballoon_enqueue_page_list(b, &ctl.pages, &ctl.n_pages,
+					    ctl.page_size);
+
+		/*
+		 * If allocation failed or the number of refused pages exceeds
+		 * the maximum allowed, move to the next page size.
+		 */
+		if (alloc_error ||
+		    ctl.n_refused_pages >= VMW_BALLOON_MAX_REFUSED) {
+			if (ctl.page_size == VMW_BALLOON_4K_PAGE)
 				break;
-			}
 
 			/*
-			 * NOSLEEP page allocation failed, so the guest is
-			 * under memory pressure. Slowing down page alloctions
-			 * seems to be reasonable, but doing so might actually
-			 * cause the hypervisor to throttle us down, resulting
-			 * in degraded performance. We will count on the
-			 * scheduler and standard memory management mechanisms
-			 * for now.
+			 * Ignore errors from locking as we now switch to 4k
+			 * pages and we might get different errors.
 			 */
-			flags = VMW_PAGE_ALLOC_CANSLEEP;
-			continue;
-		}
-
-		b->ops->add_page(b, num_pages++, page);
-		if (num_pages == b->batch_max_pages) {
-			error = b->ops->lock(b, num_pages, is_2m_pages,
-					&b->target);
-			num_pages = 0;
-			if (error)
-				break;
+			vmballoon_release_refused_pages(b, &ctl);
+			ctl.page_size--;
 		}
 
 		cond_resched();
 	}
 
-	if (num_pages > 0)
-		b->ops->lock(b, num_pages, is_2m_pages, &b->target);
-
-	vmballoon_release_refused_pages(b, true);
-	vmballoon_release_refused_pages(b, false);
+	/*
+	 * Release pages that were allocated while attempting to inflate the
+	 * balloon but were refused by the host for one reason or another,
+	 * and update the statistics.
+	 */
+	if (ctl.n_refused_pages != 0)
+		vmballoon_release_refused_pages(b, &ctl);
 }
 
-/*
+/**
+ * vmballoon_deflate() - Decrease the size of the balloon.
+ *
+ * @b: pointer to the balloon
+ * @n_frames: the number of frames to deflate. If zero, automatically
+ * calculated according to the target size.
+ * @coordinated: whether to coordinate with the host
+ *
  * Decrease the size of the balloon allowing guest to use more memory.
+ *
+ * Return: The number of deflated frames (i.e., basic page size units)
  */
-static void vmballoon_deflate(struct vmballoon *b)
+static unsigned long vmballoon_deflate(struct vmballoon *b, uint64_t n_frames,
+				       bool coordinated)
 {
-	unsigned is_2m_pages;
-
-	pr_debug("%s - size: %d, target %d\n", __func__, b->size, b->target);
+	unsigned long deflated_frames = 0;
+	unsigned long tried_frames = 0;
+	struct vmballoon_ctl ctl = {
+		.pages = LIST_HEAD_INIT(ctl.pages),
+		.refused_pages = LIST_HEAD_INIT(ctl.refused_pages),
+		.page_size = VMW_BALLOON_4K_PAGE,
+		.op = VMW_BALLOON_DEFLATE
+	};
 
 	/* free pages to reach target */
-	for (is_2m_pages = 0; is_2m_pages < b->supported_page_sizes;
-			is_2m_pages++) {
-		struct page *page, *next;
-		unsigned int num_pages = 0;
-		struct vmballoon_page_size *page_size =
-				&b->page_sizes[is_2m_pages];
-
-		list_for_each_entry_safe(page, next, &page_size->pages, lru) {
-			if (b->reset_required ||
-				(b->target > 0 &&
-					b->size - num_pages
-					* vmballoon_page_size(is_2m_pages)
-				< b->target + vmballoon_page_size(true)))
-				break;
+	while (true) {
+		unsigned int to_deflate_pages, n_unlocked_frames;
+		unsigned int page_in_frames;
+		int64_t to_deflate_frames;
+		bool deflated_all;
+
+		page_in_frames = vmballoon_page_in_frames(ctl.page_size);
+
+		VM_BUG_ON(!list_empty(&ctl.pages));
+		VM_BUG_ON(ctl.n_pages);
+		VM_BUG_ON(!list_empty(&ctl.refused_pages));
+		VM_BUG_ON(ctl.n_refused_pages);
+
+		/*
+		 * If we were requested a specific number of frames, we try to
+		 * deflate this number of frames. Otherwise, deflation is
+		 * performed according to the target and balloon size.
+		 */
+		to_deflate_frames = n_frames ? n_frames - tried_frames :
+					       -vmballoon_change(b);
+
+		/* break if no work to do */
+		if (to_deflate_frames <= 0)
+			break;
+
+		/*
+		 * Calculate the number of frames based on current page size,
+		 * but limit the deflated frames to a single chunk
+		 */
+		to_deflate_pages = min_t(unsigned long, b->batch_max_pages,
+					 DIV_ROUND_UP_ULL(to_deflate_frames,
+							  page_in_frames));
+
+		/* First take the pages from the balloon pages. */
+		vmballoon_dequeue_page_list(b, &ctl.pages, &ctl.n_pages,
+					    ctl.page_size, to_deflate_pages);
+
+		/*
+		 * Before pages are moving to the refused list, count their
+		 * frames as frames that we tried to deflate.
+		 */
+		tried_frames += ctl.n_pages * page_in_frames;
+
+		/*
+		 * Unlock the pages by communicating with the hypervisor if the
+		 * communication is coordinated (i.e., not pop). We ignore the
+		 * return code. Instead we check if all the pages we manage to
+		 * unlock all the pages. If we failed, we will move to the next
+		 * page size, and would eventually try again later.
+		 */
+		if (coordinated)
+			vmballoon_lock(b, &ctl);
+
+		/*
+		 * Check if we deflated enough. We will move to the next page
+		 * size if we did not manage to do so. This calculation takes
+		 * place now, as once the pages are released, the number of
+		 * pages is zeroed.
+		 */
+		deflated_all = (ctl.n_pages == to_deflate_pages);
+
+		/* Update local and global counters */
+		n_unlocked_frames = ctl.n_pages * page_in_frames;
+		atomic64_sub(n_unlocked_frames, &b->size);
+		deflated_frames += n_unlocked_frames;
 
-			list_del(&page->lru);
-			b->ops->add_page(b, num_pages++, page);
+		vmballoon_stats_page_add(b, VMW_BALLOON_PAGE_STAT_FREE,
+					 ctl.page_size, ctl.n_pages);
 
-			if (num_pages == b->batch_max_pages) {
-				int error;
+		/* free the ballooned pages */
+		vmballoon_release_page_list(&ctl.pages, &ctl.n_pages,
+					    ctl.page_size);
 
-				error = b->ops->unlock(b, num_pages,
-						is_2m_pages, &b->target);
-				num_pages = 0;
-				if (error)
-					return;
-			}
+		/* Return the refused pages to the ballooned list. */
+		vmballoon_enqueue_page_list(b, &ctl.refused_pages,
+					    &ctl.n_refused_pages,
+					    ctl.page_size);
 
-			cond_resched();
+		/* If we failed to unlock all the pages, move to next size. */
+		if (!deflated_all) {
+			if (ctl.page_size == b->max_page_size)
+				break;
+			ctl.page_size++;
 		}
 
-		if (num_pages > 0)
-			b->ops->unlock(b, num_pages, is_2m_pages, &b->target);
+		cond_resched();
 	}
-}
 
-static const struct vmballoon_ops vmballoon_basic_ops = {
-	.add_page = vmballoon_add_page,
-	.lock = vmballoon_lock_page,
-	.unlock = vmballoon_unlock_page
-};
+	return deflated_frames;
+}
 
-static const struct vmballoon_ops vmballoon_batched_ops = {
-	.add_page = vmballoon_add_batched_page,
-	.lock = vmballoon_lock_batched_page,
-	.unlock = vmballoon_unlock_batched_page
-};
+/**
+ * vmballoon_deinit_batching - disables batching mode.
+ *
+ * @b: pointer to &struct vmballoon.
+ *
+ * Disables batching, by deallocating the page for communication with the
+ * hypervisor and disabling the static key to indicate that batching is off.
+ */
+static void vmballoon_deinit_batching(struct vmballoon *b)
+{
+	free_page((unsigned long)b->batch_page);
+	b->batch_page = NULL;
+	static_branch_disable(&vmw_balloon_batching);
+	b->batch_max_pages = 1;
+}
 
-static bool vmballoon_init_batching(struct vmballoon *b)
+/**
+ * vmballoon_init_batching - enable batching mode.
+ *
+ * @b: pointer to &struct vmballoon.
+ *
+ * Enables batching, by allocating a page for communication with the hypervisor
+ * and enabling the static_key to use batching.
+ *
+ * Return: zero on success or an appropriate error-code.
+ */
+static int vmballoon_init_batching(struct vmballoon *b)
 {
 	struct page *page;
 
 	page = alloc_page(GFP_KERNEL | __GFP_ZERO);
 	if (!page)
-		return false;
+		return -ENOMEM;
 
 	b->batch_page = page_address(page);
-	return true;
+	b->batch_max_pages = PAGE_SIZE / sizeof(struct vmballoon_batch_entry);
+
+	static_branch_enable(&vmw_balloon_batching);
+
+	return 0;
 }
 
 /*
@@ -932,7 +1197,7 @@ static void vmballoon_doorbell(void *client_data)
 {
 	struct vmballoon *b = client_data;
 
-	STATS_INC(b->stats.doorbell);
+	vmballoon_stats_gen_inc(b, VMW_BALLOON_STAT_DOORBELL);
 
 	mod_delayed_work(system_freezable_wq, &b->dwork, 0);
 }
@@ -942,11 +1207,8 @@ static void vmballoon_doorbell(void *client_data)
  */
 static void vmballoon_vmci_cleanup(struct vmballoon *b)
 {
-	int error;
-
-	VMWARE_BALLOON_CMD(VMCI_DOORBELL_SET, VMCI_INVALID_ID,
-			VMCI_INVALID_ID, error);
-	STATS_INC(b->stats.doorbell_unset);
+	vmballoon_cmd(b, VMW_BALLOON_CMD_VMCI_DOORBELL_SET,
+		      VMCI_INVALID_ID, VMCI_INVALID_ID);
 
 	if (!vmci_handle_is_invalid(b->vmci_doorbell)) {
 		vmci_doorbell_destroy(b->vmci_doorbell);
@@ -954,12 +1216,19 @@ static void vmballoon_vmci_cleanup(struct vmballoon *b)
 	}
 }
 
-/*
- * Initialize vmci doorbell, to get notified as soon as balloon changes
+/**
+ * vmballoon_vmci_init - Initialize vmci doorbell.
+ *
+ * @b: pointer to the balloon.
+ *
+ * Return: zero on success or when wakeup command not supported. Error-code
+ * otherwise.
+ *
+ * Initialize vmci doorbell, to get notified as soon as balloon changes.
  */
 static int vmballoon_vmci_init(struct vmballoon *b)
 {
-	unsigned long error, dummy;
+	unsigned long error;
 
 	if ((b->capabilities & VMW_BALLOON_SIGNALLED_WAKEUP_CMD) == 0)
 		return 0;
@@ -971,10 +1240,9 @@ static int vmballoon_vmci_init(struct vmballoon *b)
 	if (error != VMCI_SUCCESS)
 		goto fail;
 
-	error = VMWARE_BALLOON_CMD(VMCI_DOORBELL_SET, b->vmci_doorbell.context,
-				   b->vmci_doorbell.resource, dummy);
-
-	STATS_INC(b->stats.doorbell_set);
+	error =	__vmballoon_cmd(b, VMW_BALLOON_CMD_VMCI_DOORBELL_SET,
+				b->vmci_doorbell.context,
+				b->vmci_doorbell.resource, NULL);
 
 	if (error != VMW_BALLOON_SUCCESS)
 		goto fail;
@@ -985,6 +1253,23 @@ fail:
 	return -EIO;
 }
 
+/**
+ * vmballoon_pop - Quickly release all pages allocate for the balloon.
+ *
+ * @b: pointer to the balloon.
+ *
+ * This function is called when host decides to "reset" balloon for one reason
+ * or another. Unlike normal "deflate" we do not (shall not) notify host of the
+ * pages being released.
+ */
+static void vmballoon_pop(struct vmballoon *b)
+{
+	unsigned long size;
+
+	while ((size = atomic64_read(&b->size)))
+		vmballoon_deflate(b, size, false);
+}
+
 /*
  * Perform standard reset sequence by popping the balloon (in case it
  * is not  empty) and then restarting protocol. This operation normally
@@ -994,18 +1279,18 @@ static void vmballoon_reset(struct vmballoon *b)
 {
 	int error;
 
+	down_write(&b->conf_sem);
+
 	vmballoon_vmci_cleanup(b);
 
 	/* free all pages, skipping monitor unlock */
 	vmballoon_pop(b);
 
-	if (!vmballoon_send_start(b, VMW_BALLOON_CAPABILITIES))
+	if (vmballoon_send_start(b, VMW_BALLOON_CAPABILITIES))
 		return;
 
 	if ((b->capabilities & VMW_BALLOON_BATCHED_CMDS) != 0) {
-		b->ops = &vmballoon_batched_ops;
-		b->batch_max_pages = VMW_BALLOON_BATCH_MAX_PAGES;
-		if (!vmballoon_init_batching(b)) {
+		if (vmballoon_init_batching(b)) {
 			/*
 			 * We failed to initialize batching, inform the monitor
 			 * about it by sending a null capability.
@@ -1016,52 +1301,70 @@ static void vmballoon_reset(struct vmballoon *b)
 			return;
 		}
 	} else if ((b->capabilities & VMW_BALLOON_BASIC_CMDS) != 0) {
-		b->ops = &vmballoon_basic_ops;
-		b->batch_max_pages = 1;
+		vmballoon_deinit_batching(b);
 	}
 
+	vmballoon_stats_gen_inc(b, VMW_BALLOON_STAT_RESET);
 	b->reset_required = false;
 
 	error = vmballoon_vmci_init(b);
 	if (error)
 		pr_err("failed to initialize vmci doorbell\n");
 
-	if (!vmballoon_send_guest_id(b))
+	if (vmballoon_send_guest_id(b))
 		pr_err("failed to send guest ID to the host\n");
+
+	up_write(&b->conf_sem);
 }
 
-/*
- * Balloon work function: reset protocol, if needed, get the new size and
- * adjust balloon as needed. Repeat in 1 sec.
+/**
+ * vmballoon_work - periodic balloon worker for reset, inflation and deflation.
+ *
+ * @work: pointer to the &work_struct which is provided by the workqueue.
+ *
+ * Resets the protocol if needed, gets the new size and adjusts balloon as
+ * needed. Repeat in 1 sec.
  */
 static void vmballoon_work(struct work_struct *work)
 {
 	struct delayed_work *dwork = to_delayed_work(work);
 	struct vmballoon *b = container_of(dwork, struct vmballoon, dwork);
-	unsigned int target;
-
-	STATS_INC(b->stats.timer);
+	int64_t change = 0;
 
 	if (b->reset_required)
 		vmballoon_reset(b);
 
-	if (!b->reset_required && vmballoon_send_get_target(b, &target)) {
-		/* update target, adjust size */
-		b->target = target;
+	down_read(&b->conf_sem);
+
+	/*
+	 * Update the stats while holding the semaphore to ensure that
+	 * @stats_enabled is consistent with whether the stats are actually
+	 * enabled
+	 */
+	vmballoon_stats_gen_inc(b, VMW_BALLOON_STAT_TIMER);
+
+	if (!vmballoon_send_get_target(b))
+		change = vmballoon_change(b);
+
+	if (change != 0) {
+		pr_debug("%s - size: %llu, target %lu\n", __func__,
+			 atomic64_read(&b->size), READ_ONCE(b->target));
 
-		if (b->size < target)
+		if (change > 0)
 			vmballoon_inflate(b);
-		else if (target == 0 ||
-				b->size > target + vmballoon_page_size(true))
-			vmballoon_deflate(b);
+		else  /* (change < 0) */
+			vmballoon_deflate(b, 0, true);
 	}
 
+	up_read(&b->conf_sem);
+
 	/*
 	 * We are using a freezable workqueue so that balloon operations are
 	 * stopped while the system transitions to/from sleep/hibernation.
 	 */
 	queue_delayed_work(system_freezable_wq,
 			   dwork, round_jiffies_relative(HZ));
+
 }
 
 /*
@@ -1069,64 +1372,100 @@ static void vmballoon_work(struct work_struct *work)
  */
 #ifdef CONFIG_DEBUG_FS
 
+static const char * const vmballoon_stat_page_names[] = {
+	[VMW_BALLOON_PAGE_STAT_ALLOC]		= "alloc",
+	[VMW_BALLOON_PAGE_STAT_ALLOC_FAIL]	= "allocFail",
+	[VMW_BALLOON_PAGE_STAT_REFUSED_ALLOC]	= "errAlloc",
+	[VMW_BALLOON_PAGE_STAT_REFUSED_FREE]	= "errFree",
+	[VMW_BALLOON_PAGE_STAT_FREE]		= "free"
+};
+
+static const char * const vmballoon_stat_names[] = {
+	[VMW_BALLOON_STAT_TIMER]		= "timer",
+	[VMW_BALLOON_STAT_DOORBELL]		= "doorbell",
+	[VMW_BALLOON_STAT_RESET]		= "reset",
+};
+
+static int vmballoon_enable_stats(struct vmballoon *b)
+{
+	int r = 0;
+
+	down_write(&b->conf_sem);
+
+	/* did we somehow race with another reader which enabled stats? */
+	if (b->stats)
+		goto out;
+
+	b->stats = kzalloc(sizeof(*b->stats), GFP_KERNEL);
+
+	if (!b->stats) {
+		/* allocation failed */
+		r = -ENOMEM;
+		goto out;
+	}
+	static_key_enable(&balloon_stat_enabled.key);
+out:
+	up_write(&b->conf_sem);
+	return r;
+}
+
+/**
+ * vmballoon_debug_show - shows statistics of balloon operations.
+ * @f: pointer to the &struct seq_file.
+ * @offset: ignored.
+ *
+ * Provides the statistics that can be accessed in vmmemctl in the debugfs.
+ * To avoid the overhead - mainly that of memory - of collecting the statistics,
+ * we only collect statistics after the first time the counters are read.
+ *
+ * Return: zero on success or an error code.
+ */
 static int vmballoon_debug_show(struct seq_file *f, void *offset)
 {
 	struct vmballoon *b = f->private;
-	struct vmballoon_stats *stats = &b->stats;
+	int i, j;
+
+	/* enables stats if they are disabled */
+	if (!b->stats) {
+		int r = vmballoon_enable_stats(b);
+
+		if (r)
+			return r;
+	}
 
 	/* format capabilities info */
-	seq_printf(f,
-		   "balloon capabilities:   %#4x\n"
-		   "used capabilities:      %#4lx\n"
-		   "is resetting:           %c\n",
-		   VMW_BALLOON_CAPABILITIES, b->capabilities,
-		   b->reset_required ? 'y' : 'n');
+	seq_printf(f, "%-22s: %#16x\n", "balloon capabilities",
+		   VMW_BALLOON_CAPABILITIES);
+	seq_printf(f, "%-22s: %#16lx\n", "used capabilities", b->capabilities);
+	seq_printf(f, "%-22s: %16s\n", "is resetting",
+		   b->reset_required ? "y" : "n");
 
 	/* format size info */
-	seq_printf(f,
-		   "target:             %8d pages\n"
-		   "current:            %8d pages\n",
-		   b->target, b->size);
-
-	seq_printf(f,
-		   "\n"
-		   "timer:              %8u\n"
-		   "doorbell:           %8u\n"
-		   "start:              %8u (%4u failed)\n"
-		   "guestType:          %8u (%4u failed)\n"
-		   "2m-lock:            %8u (%4u failed)\n"
-		   "lock:               %8u (%4u failed)\n"
-		   "2m-unlock:          %8u (%4u failed)\n"
-		   "unlock:             %8u (%4u failed)\n"
-		   "target:             %8u (%4u failed)\n"
-		   "prim2mAlloc:        %8u (%4u failed)\n"
-		   "primNoSleepAlloc:   %8u (%4u failed)\n"
-		   "primCanSleepAlloc:  %8u (%4u failed)\n"
-		   "prim2mFree:         %8u\n"
-		   "primFree:           %8u\n"
-		   "err2mAlloc:         %8u\n"
-		   "errAlloc:           %8u\n"
-		   "err2mFree:          %8u\n"
-		   "errFree:            %8u\n"
-		   "doorbellSet:        %8u\n"
-		   "doorbellUnset:      %8u\n",
-		   stats->timer,
-		   stats->doorbell,
-		   stats->start, stats->start_fail,
-		   stats->guest_type, stats->guest_type_fail,
-		   stats->lock[true],  stats->lock_fail[true],
-		   stats->lock[false],  stats->lock_fail[false],
-		   stats->unlock[true], stats->unlock_fail[true],
-		   stats->unlock[false], stats->unlock_fail[false],
-		   stats->target, stats->target_fail,
-		   stats->alloc[true], stats->alloc_fail[true],
-		   stats->alloc[false], stats->alloc_fail[false],
-		   stats->sleep_alloc, stats->sleep_alloc_fail,
-		   stats->free[true],
-		   stats->free[false],
-		   stats->refused_alloc[true], stats->refused_alloc[false],
-		   stats->refused_free[true], stats->refused_free[false],
-		   stats->doorbell_set, stats->doorbell_unset);
+	seq_printf(f, "%-22s: %16lu\n", "target", READ_ONCE(b->target));
+	seq_printf(f, "%-22s: %16llu\n", "current", atomic64_read(&b->size));
+
+	for (i = 0; i < VMW_BALLOON_CMD_NUM; i++) {
+		if (vmballoon_cmd_names[i] == NULL)
+			continue;
+
+		seq_printf(f, "%-22s: %16llu (%llu failed)\n",
+			   vmballoon_cmd_names[i],
+			   atomic64_read(&b->stats->ops[i][VMW_BALLOON_OP_STAT]),
+			   atomic64_read(&b->stats->ops[i][VMW_BALLOON_OP_FAIL_STAT]));
+	}
+
+	for (i = 0; i < VMW_BALLOON_STAT_NUM; i++)
+		seq_printf(f, "%-22s: %16llu\n",
+			   vmballoon_stat_names[i],
+			   atomic64_read(&b->stats->general_stat[i]));
+
+	for (i = 0; i < VMW_BALLOON_PAGE_STAT_NUM; i++) {
+		for (j = 0; j < VMW_BALLOON_NUM_PAGE_SIZES; j++)
+			seq_printf(f, "%-18s(%s): %16llu\n",
+				   vmballoon_stat_page_names[i],
+				   vmballoon_page_size_names[j],
+				   atomic64_read(&b->stats->page_stat[i][j]));
+	}
 
 	return 0;
 }
@@ -1161,7 +1500,10 @@ static int __init vmballoon_debugfs_init(struct vmballoon *b)
 
 static void __exit vmballoon_debugfs_exit(struct vmballoon *b)
 {
+	static_key_disable(&balloon_stat_enabled.key);
 	debugfs_remove(b->dbg_entry);
+	kfree(b->stats);
+	b->stats = NULL;
 }
 
 #else
@@ -1179,8 +1521,9 @@ static inline void vmballoon_debugfs_exit(struct vmballoon *b)
 
 static int __init vmballoon_init(void)
 {
+	enum vmballoon_page_size_type page_size;
 	int error;
-	unsigned is_2m_pages;
+
 	/*
 	 * Check if we are running on VMware's hypervisor and bail out
 	 * if we are not.
@@ -1188,11 +1531,10 @@ static int __init vmballoon_init(void)
 	if (x86_hyper_type != X86_HYPER_VMWARE)
 		return -ENODEV;
 
-	for (is_2m_pages = 0; is_2m_pages < VMW_BALLOON_NUM_PAGE_SIZES;
-			is_2m_pages++) {
-		INIT_LIST_HEAD(&balloon.page_sizes[is_2m_pages].pages);
-		INIT_LIST_HEAD(&balloon.page_sizes[is_2m_pages].refused_pages);
-	}
+	for (page_size = VMW_BALLOON_4K_PAGE;
+	     page_size <= VMW_BALLOON_LAST_SIZE; page_size++)
+		INIT_LIST_HEAD(&balloon.page_sizes[page_size].pages);
+
 
 	INIT_DELAYED_WORK(&balloon.dwork, vmballoon_work);
 
@@ -1200,6 +1542,8 @@ static int __init vmballoon_init(void)
 	if (error)
 		return error;
 
+	spin_lock_init(&balloon.comm_lock);
+	init_rwsem(&balloon.conf_sem);
 	balloon.vmci_doorbell = VMCI_INVALID_HANDLE;
 	balloon.batch_page = NULL;
 	balloon.page = NULL;
diff --git a/drivers/misc/vmw_vmci/vmci_driver.c b/drivers/misc/vmw_vmci/vmci_driver.c
index d7eaf1eb11e7..003bfba40758 100644
--- a/drivers/misc/vmw_vmci/vmci_driver.c
+++ b/drivers/misc/vmw_vmci/vmci_driver.c
@@ -113,5 +113,5 @@ module_exit(vmci_drv_exit);
 
 MODULE_AUTHOR("VMware, Inc.");
 MODULE_DESCRIPTION("VMware Virtual Machine Communication Interface.");
-MODULE_VERSION("1.1.5.0-k");
+MODULE_VERSION("1.1.6.0-k");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/misc/vmw_vmci/vmci_host.c b/drivers/misc/vmw_vmci/vmci_host.c
index 83e0c95d20a4..edfffc9699ba 100644
--- a/drivers/misc/vmw_vmci/vmci_host.c
+++ b/drivers/misc/vmw_vmci/vmci_host.c
@@ -15,7 +15,6 @@
 
 #include <linux/vmw_vmci_defs.h>
 #include <linux/vmw_vmci_api.h>
-#include <linux/moduleparam.h>
 #include <linux/miscdevice.h>
 #include <linux/interrupt.h>
 #include <linux/highmem.h>
@@ -448,15 +447,12 @@ static int vmci_host_do_alloc_queuepair(struct vmci_host_dev *vmci_host_dev,
 	struct vmci_handle handle;
 	int vmci_status;
 	int __user *retptr;
-	u32 cid;
 
 	if (vmci_host_dev->ct_type != VMCIOBJ_CONTEXT) {
 		vmci_ioctl_err("only valid for contexts\n");
 		return -EINVAL;
 	}
 
-	cid = vmci_ctx_get_id(vmci_host_dev->context);
-
 	if (vmci_host_dev->user_version < VMCI_VERSION_NOVMVM) {
 		struct vmci_qp_alloc_info_vmvm alloc_info;
 		struct vmci_qp_alloc_info_vmvm __user *info = uptr;
diff --git a/drivers/misc/vmw_vmci/vmci_resource.c b/drivers/misc/vmw_vmci/vmci_resource.c
index 1ab6e8737a5f..da1ee2e1ba99 100644
--- a/drivers/misc/vmw_vmci/vmci_resource.c
+++ b/drivers/misc/vmw_vmci/vmci_resource.c
@@ -57,7 +57,8 @@ static struct vmci_resource *vmci_resource_lookup(struct vmci_handle handle,
 
 		if (r->type == type &&
 		    rid == handle.resource &&
-		    (cid == handle.context || cid == VMCI_INVALID_ID)) {
+		    (cid == handle.context || cid == VMCI_INVALID_ID ||
+		     handle.context == VMCI_INVALID_ID)) {
 			resource = r;
 			break;
 		}
diff --git a/drivers/nvmem/core.c b/drivers/nvmem/core.c
index aa1657831b70..9b18ce90f907 100644
--- a/drivers/nvmem/core.c
+++ b/drivers/nvmem/core.c
@@ -1,17 +1,9 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * nvmem framework core.
  *
  * Copyright (C) 2015 Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
  * Copyright (C) 2013 Maxime Ripard <maxime.ripard@free-electrons.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 and
- * only version 2 as published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
  */
 
 #include <linux/device.h>
@@ -19,6 +11,7 @@
 #include <linux/fs.h>
 #include <linux/idr.h>
 #include <linux/init.h>
+#include <linux/kref.h>
 #include <linux/module.h>
 #include <linux/nvmem-consumer.h>
 #include <linux/nvmem-provider.h>
@@ -26,18 +19,18 @@
 #include <linux/slab.h>
 
 struct nvmem_device {
-	const char		*name;
 	struct module		*owner;
 	struct device		dev;
 	int			stride;
 	int			word_size;
 	int			id;
-	int			users;
+	struct kref		refcnt;
 	size_t			size;
 	bool			read_only;
 	int			flags;
 	struct bin_attribute	eeprom;
 	struct device		*base_dev;
+	struct list_head	cells;
 	nvmem_reg_read_t	reg_read;
 	nvmem_reg_write_t	reg_write;
 	void *priv;
@@ -58,8 +51,13 @@ struct nvmem_cell {
 static DEFINE_MUTEX(nvmem_mutex);
 static DEFINE_IDA(nvmem_ida);
 
-static LIST_HEAD(nvmem_cells);
-static DEFINE_MUTEX(nvmem_cells_mutex);
+static DEFINE_MUTEX(nvmem_cell_mutex);
+static LIST_HEAD(nvmem_cell_tables);
+
+static DEFINE_MUTEX(nvmem_lookup_mutex);
+static LIST_HEAD(nvmem_lookup_list);
+
+static BLOCKING_NOTIFIER_HEAD(nvmem_notifier);
 
 #ifdef CONFIG_DEBUG_LOCK_ALLOC
 static struct lock_class_key eeprom_lock_key;
@@ -156,7 +154,7 @@ static ssize_t bin_attr_nvmem_write(struct file *filp, struct kobject *kobj,
 static struct bin_attribute bin_attr_rw_nvmem = {
 	.attr	= {
 		.name	= "nvmem",
-		.mode	= S_IWUSR | S_IRUGO,
+		.mode	= 0644,
 	},
 	.read	= bin_attr_nvmem_read,
 	.write	= bin_attr_nvmem_write,
@@ -180,7 +178,7 @@ static const struct attribute_group *nvmem_rw_dev_groups[] = {
 static struct bin_attribute bin_attr_ro_nvmem = {
 	.attr	= {
 		.name	= "nvmem",
-		.mode	= S_IRUGO,
+		.mode	= 0444,
 	},
 	.read	= bin_attr_nvmem_read,
 };
@@ -203,7 +201,7 @@ static const struct attribute_group *nvmem_ro_dev_groups[] = {
 static struct bin_attribute bin_attr_rw_root_nvmem = {
 	.attr	= {
 		.name	= "nvmem",
-		.mode	= S_IWUSR | S_IRUSR,
+		.mode	= 0600,
 	},
 	.read	= bin_attr_nvmem_read,
 	.write	= bin_attr_nvmem_write,
@@ -227,7 +225,7 @@ static const struct attribute_group *nvmem_rw_root_dev_groups[] = {
 static struct bin_attribute bin_attr_ro_root_nvmem = {
 	.attr	= {
 		.name	= "nvmem",
-		.mode	= S_IRUSR,
+		.mode	= 0400,
 	},
 	.read	= bin_attr_nvmem_read,
 };
@@ -282,48 +280,42 @@ static struct nvmem_device *of_nvmem_find(struct device_node *nvmem_np)
 	return to_nvmem_device(d);
 }
 
-static struct nvmem_cell *nvmem_find_cell(const char *cell_id)
+static struct nvmem_device *nvmem_find(const char *name)
 {
-	struct nvmem_cell *p;
-
-	mutex_lock(&nvmem_cells_mutex);
+	struct device *d;
 
-	list_for_each_entry(p, &nvmem_cells, node)
-		if (!strcmp(p->name, cell_id)) {
-			mutex_unlock(&nvmem_cells_mutex);
-			return p;
-		}
+	d = bus_find_device_by_name(&nvmem_bus_type, NULL, name);
 
-	mutex_unlock(&nvmem_cells_mutex);
+	if (!d)
+		return NULL;
 
-	return NULL;
+	return to_nvmem_device(d);
 }
 
 static void nvmem_cell_drop(struct nvmem_cell *cell)
 {
-	mutex_lock(&nvmem_cells_mutex);
+	blocking_notifier_call_chain(&nvmem_notifier, NVMEM_CELL_REMOVE, cell);
+	mutex_lock(&nvmem_mutex);
 	list_del(&cell->node);
-	mutex_unlock(&nvmem_cells_mutex);
+	mutex_unlock(&nvmem_mutex);
+	kfree(cell->name);
 	kfree(cell);
 }
 
 static void nvmem_device_remove_all_cells(const struct nvmem_device *nvmem)
 {
-	struct nvmem_cell *cell;
-	struct list_head *p, *n;
+	struct nvmem_cell *cell, *p;
 
-	list_for_each_safe(p, n, &nvmem_cells) {
-		cell = list_entry(p, struct nvmem_cell, node);
-		if (cell->nvmem == nvmem)
-			nvmem_cell_drop(cell);
-	}
+	list_for_each_entry_safe(cell, p, &nvmem->cells, node)
+		nvmem_cell_drop(cell);
 }
 
 static void nvmem_cell_add(struct nvmem_cell *cell)
 {
-	mutex_lock(&nvmem_cells_mutex);
-	list_add_tail(&cell->node, &nvmem_cells);
-	mutex_unlock(&nvmem_cells_mutex);
+	mutex_lock(&nvmem_mutex);
+	list_add_tail(&cell->node, &cell->nvmem->cells);
+	mutex_unlock(&nvmem_mutex);
+	blocking_notifier_call_chain(&nvmem_notifier, NVMEM_CELL_ADD, cell);
 }
 
 static int nvmem_cell_info_to_nvmem_cell(struct nvmem_device *nvmem,
@@ -361,7 +353,7 @@ static int nvmem_cell_info_to_nvmem_cell(struct nvmem_device *nvmem,
  *
  * Return: 0 or negative error code on failure.
  */
-int nvmem_add_cells(struct nvmem_device *nvmem,
+static int nvmem_add_cells(struct nvmem_device *nvmem,
 		    const struct nvmem_cell_info *info,
 		    int ncells)
 {
@@ -400,7 +392,6 @@ err:
 
 	return rval;
 }
-EXPORT_SYMBOL_GPL(nvmem_add_cells);
 
 /*
  * nvmem_setup_compat() - Create an additional binary entry in
@@ -440,6 +431,136 @@ static int nvmem_setup_compat(struct nvmem_device *nvmem,
 }
 
 /**
+ * nvmem_register_notifier() - Register a notifier block for nvmem events.
+ *
+ * @nb: notifier block to be called on nvmem events.
+ *
+ * Return: 0 on success, negative error number on failure.
+ */
+int nvmem_register_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_register(&nvmem_notifier, nb);
+}
+EXPORT_SYMBOL_GPL(nvmem_register_notifier);
+
+/**
+ * nvmem_unregister_notifier() - Unregister a notifier block for nvmem events.
+ *
+ * @nb: notifier block to be unregistered.
+ *
+ * Return: 0 on success, negative error number on failure.
+ */
+int nvmem_unregister_notifier(struct notifier_block *nb)
+{
+	return blocking_notifier_chain_unregister(&nvmem_notifier, nb);
+}
+EXPORT_SYMBOL_GPL(nvmem_unregister_notifier);
+
+static int nvmem_add_cells_from_table(struct nvmem_device *nvmem)
+{
+	const struct nvmem_cell_info *info;
+	struct nvmem_cell_table *table;
+	struct nvmem_cell *cell;
+	int rval = 0, i;
+
+	mutex_lock(&nvmem_cell_mutex);
+	list_for_each_entry(table, &nvmem_cell_tables, node) {
+		if (strcmp(nvmem_dev_name(nvmem), table->nvmem_name) == 0) {
+			for (i = 0; i < table->ncells; i++) {
+				info = &table->cells[i];
+
+				cell = kzalloc(sizeof(*cell), GFP_KERNEL);
+				if (!cell) {
+					rval = -ENOMEM;
+					goto out;
+				}
+
+				rval = nvmem_cell_info_to_nvmem_cell(nvmem,
+								     info,
+								     cell);
+				if (rval) {
+					kfree(cell);
+					goto out;
+				}
+
+				nvmem_cell_add(cell);
+			}
+		}
+	}
+
+out:
+	mutex_unlock(&nvmem_cell_mutex);
+	return rval;
+}
+
+static struct nvmem_cell *
+nvmem_find_cell_by_name(struct nvmem_device *nvmem, const char *cell_id)
+{
+	struct nvmem_cell *cell = NULL;
+
+	mutex_lock(&nvmem_mutex);
+	list_for_each_entry(cell, &nvmem->cells, node) {
+		if (strcmp(cell_id, cell->name) == 0)
+			break;
+	}
+	mutex_unlock(&nvmem_mutex);
+
+	return cell;
+}
+
+static int nvmem_add_cells_from_of(struct nvmem_device *nvmem)
+{
+	struct device_node *parent, *child;
+	struct device *dev = &nvmem->dev;
+	struct nvmem_cell *cell;
+	const __be32 *addr;
+	int len;
+
+	parent = dev->of_node;
+
+	for_each_child_of_node(parent, child) {
+		addr = of_get_property(child, "reg", &len);
+		if (!addr || (len < 2 * sizeof(u32))) {
+			dev_err(dev, "nvmem: invalid reg on %pOF\n", child);
+			return -EINVAL;
+		}
+
+		cell = kzalloc(sizeof(*cell), GFP_KERNEL);
+		if (!cell)
+			return -ENOMEM;
+
+		cell->nvmem = nvmem;
+		cell->offset = be32_to_cpup(addr++);
+		cell->bytes = be32_to_cpup(addr);
+		cell->name = kasprintf(GFP_KERNEL, "%pOFn", child);
+
+		addr = of_get_property(child, "bits", &len);
+		if (addr && len == (2 * sizeof(u32))) {
+			cell->bit_offset = be32_to_cpup(addr++);
+			cell->nbits = be32_to_cpup(addr);
+		}
+
+		if (cell->nbits)
+			cell->bytes = DIV_ROUND_UP(
+					cell->nbits + cell->bit_offset,
+					BITS_PER_BYTE);
+
+		if (!IS_ALIGNED(cell->offset, nvmem->stride)) {
+			dev_err(dev, "cell %s unaligned to nvmem stride %d\n",
+				cell->name, nvmem->stride);
+			/* Cells already added will be freed later. */
+			kfree(cell->name);
+			kfree(cell);
+			return -EINVAL;
+		}
+
+		nvmem_cell_add(cell);
+	}
+
+	return 0;
+}
+
+/**
  * nvmem_register() - Register a nvmem device for given nvmem_config.
  * Also creates an binary entry in /sys/bus/nvmem/devices/dev-name/nvmem
  *
@@ -467,6 +588,9 @@ struct nvmem_device *nvmem_register(const struct nvmem_config *config)
 		return ERR_PTR(rval);
 	}
 
+	kref_init(&nvmem->refcnt);
+	INIT_LIST_HEAD(&nvmem->cells);
+
 	nvmem->id = rval;
 	nvmem->owner = config->owner;
 	if (!nvmem->owner && config->dev->driver)
@@ -516,11 +640,31 @@ struct nvmem_device *nvmem_register(const struct nvmem_config *config)
 			goto err_device_del;
 	}
 
-	if (config->cells)
-		nvmem_add_cells(nvmem, config->cells, config->ncells);
+	if (config->cells) {
+		rval = nvmem_add_cells(nvmem, config->cells, config->ncells);
+		if (rval)
+			goto err_teardown_compat;
+	}
+
+	rval = nvmem_add_cells_from_table(nvmem);
+	if (rval)
+		goto err_remove_cells;
+
+	rval = nvmem_add_cells_from_of(nvmem);
+	if (rval)
+		goto err_remove_cells;
+
+	rval = blocking_notifier_call_chain(&nvmem_notifier, NVMEM_ADD, nvmem);
+	if (rval)
+		goto err_remove_cells;
 
 	return nvmem;
 
+err_remove_cells:
+	nvmem_device_remove_all_cells(nvmem);
+err_teardown_compat:
+	if (config->compat)
+		device_remove_bin_file(nvmem->base_dev, &nvmem->eeprom);
 err_device_del:
 	device_del(&nvmem->dev);
 err_put_device:
@@ -530,21 +674,13 @@ err_put_device:
 }
 EXPORT_SYMBOL_GPL(nvmem_register);
 
-/**
- * nvmem_unregister() - Unregister previously registered nvmem device
- *
- * @nvmem: Pointer to previously registered nvmem device.
- *
- * Return: Will be an negative on error or a zero on success.
- */
-int nvmem_unregister(struct nvmem_device *nvmem)
+static void nvmem_device_release(struct kref *kref)
 {
-	mutex_lock(&nvmem_mutex);
-	if (nvmem->users) {
-		mutex_unlock(&nvmem_mutex);
-		return -EBUSY;
-	}
-	mutex_unlock(&nvmem_mutex);
+	struct nvmem_device *nvmem;
+
+	nvmem = container_of(kref, struct nvmem_device, refcnt);
+
+	blocking_notifier_call_chain(&nvmem_notifier, NVMEM_REMOVE, nvmem);
 
 	if (nvmem->flags & FLAG_COMPAT)
 		device_remove_bin_file(nvmem->base_dev, &nvmem->eeprom);
@@ -552,14 +688,22 @@ int nvmem_unregister(struct nvmem_device *nvmem)
 	nvmem_device_remove_all_cells(nvmem);
 	device_del(&nvmem->dev);
 	put_device(&nvmem->dev);
+}
 
-	return 0;
+/**
+ * nvmem_unregister() - Unregister previously registered nvmem device
+ *
+ * @nvmem: Pointer to previously registered nvmem device.
+ */
+void nvmem_unregister(struct nvmem_device *nvmem)
+{
+	kref_put(&nvmem->refcnt, nvmem_device_release);
 }
 EXPORT_SYMBOL_GPL(nvmem_unregister);
 
 static void devm_nvmem_release(struct device *dev, void *res)
 {
-	WARN_ON(nvmem_unregister(*(struct nvmem_device **)res));
+	nvmem_unregister(*(struct nvmem_device **)res);
 }
 
 /**
@@ -617,71 +761,34 @@ int devm_nvmem_unregister(struct device *dev, struct nvmem_device *nvmem)
 }
 EXPORT_SYMBOL(devm_nvmem_unregister);
 
-
 static struct nvmem_device *__nvmem_device_get(struct device_node *np,
-					       struct nvmem_cell **cellp,
-					       const char *cell_id)
+					       const char *nvmem_name)
 {
 	struct nvmem_device *nvmem = NULL;
 
 	mutex_lock(&nvmem_mutex);
-
-	if (np) {
-		nvmem = of_nvmem_find(np);
-		if (!nvmem) {
-			mutex_unlock(&nvmem_mutex);
-			return ERR_PTR(-EPROBE_DEFER);
-		}
-	} else {
-		struct nvmem_cell *cell = nvmem_find_cell(cell_id);
-
-		if (cell) {
-			nvmem = cell->nvmem;
-			*cellp = cell;
-		}
-
-		if (!nvmem) {
-			mutex_unlock(&nvmem_mutex);
-			return ERR_PTR(-ENOENT);
-		}
-	}
-
-	nvmem->users++;
+	nvmem = np ? of_nvmem_find(np) : nvmem_find(nvmem_name);
 	mutex_unlock(&nvmem_mutex);
+	if (!nvmem)
+		return ERR_PTR(-EPROBE_DEFER);
 
 	if (!try_module_get(nvmem->owner)) {
 		dev_err(&nvmem->dev,
 			"could not increase module refcount for cell %s\n",
-			nvmem->name);
-
-		mutex_lock(&nvmem_mutex);
-		nvmem->users--;
-		mutex_unlock(&nvmem_mutex);
+			nvmem_dev_name(nvmem));
 
 		return ERR_PTR(-EINVAL);
 	}
 
+	kref_get(&nvmem->refcnt);
+
 	return nvmem;
 }
 
 static void __nvmem_device_put(struct nvmem_device *nvmem)
 {
 	module_put(nvmem->owner);
-	mutex_lock(&nvmem_mutex);
-	nvmem->users--;
-	mutex_unlock(&nvmem_mutex);
-}
-
-static struct nvmem_device *nvmem_find(const char *name)
-{
-	struct device *d;
-
-	d = bus_find_device_by_name(&nvmem_bus_type, NULL, name);
-
-	if (!d)
-		return NULL;
-
-	return to_nvmem_device(d);
+	kref_put(&nvmem->refcnt, nvmem_device_release);
 }
 
 #if IS_ENABLED(CONFIG_OF)
@@ -706,7 +813,7 @@ struct nvmem_device *of_nvmem_device_get(struct device_node *np, const char *id)
 	if (!nvmem_np)
 		return ERR_PTR(-EINVAL);
 
-	return __nvmem_device_get(nvmem_np, NULL, NULL);
+	return __nvmem_device_get(nvmem_np, NULL);
 }
 EXPORT_SYMBOL_GPL(of_nvmem_device_get);
 #endif
@@ -810,44 +917,86 @@ struct nvmem_device *devm_nvmem_device_get(struct device *dev, const char *id)
 }
 EXPORT_SYMBOL_GPL(devm_nvmem_device_get);
 
-static struct nvmem_cell *nvmem_cell_get_from_list(const char *cell_id)
+static struct nvmem_cell *
+nvmem_cell_get_from_lookup(struct device *dev, const char *con_id)
 {
-	struct nvmem_cell *cell = NULL;
+	struct nvmem_cell *cell = ERR_PTR(-ENOENT);
+	struct nvmem_cell_lookup *lookup;
 	struct nvmem_device *nvmem;
+	const char *dev_id;
 
-	nvmem = __nvmem_device_get(NULL, &cell, cell_id);
-	if (IS_ERR(nvmem))
-		return ERR_CAST(nvmem);
+	if (!dev)
+		return ERR_PTR(-EINVAL);
 
+	dev_id = dev_name(dev);
+
+	mutex_lock(&nvmem_lookup_mutex);
+
+	list_for_each_entry(lookup, &nvmem_lookup_list, node) {
+		if ((strcmp(lookup->dev_id, dev_id) == 0) &&
+		    (strcmp(lookup->con_id, con_id) == 0)) {
+			/* This is the right entry. */
+			nvmem = __nvmem_device_get(NULL, lookup->nvmem_name);
+			if (IS_ERR(nvmem)) {
+				/* Provider may not be registered yet. */
+				cell = ERR_CAST(nvmem);
+				goto out;
+			}
+
+			cell = nvmem_find_cell_by_name(nvmem,
+						       lookup->cell_name);
+			if (!cell) {
+				__nvmem_device_put(nvmem);
+				cell = ERR_PTR(-ENOENT);
+				goto out;
+			}
+		}
+	}
+
+out:
+	mutex_unlock(&nvmem_lookup_mutex);
 	return cell;
 }
 
 #if IS_ENABLED(CONFIG_OF)
+static struct nvmem_cell *
+nvmem_find_cell_by_index(struct nvmem_device *nvmem, int index)
+{
+	struct nvmem_cell *cell = NULL;
+	int i = 0;
+
+	mutex_lock(&nvmem_mutex);
+	list_for_each_entry(cell, &nvmem->cells, node) {
+		if (index == i++)
+			break;
+	}
+	mutex_unlock(&nvmem_mutex);
+
+	return cell;
+}
+
 /**
  * of_nvmem_cell_get() - Get a nvmem cell from given device node and cell id
  *
  * @np: Device tree node that uses the nvmem cell.
- * @name: nvmem cell name from nvmem-cell-names property, or NULL
- *	  for the cell at index 0 (the lone cell with no accompanying
- *	  nvmem-cell-names property).
+ * @id: nvmem cell name from nvmem-cell-names property, or NULL
+ *      for the cell at index 0 (the lone cell with no accompanying
+ *      nvmem-cell-names property).
  *
  * Return: Will be an ERR_PTR() on error or a valid pointer
  * to a struct nvmem_cell.  The nvmem_cell will be freed by the
  * nvmem_cell_put().
  */
-struct nvmem_cell *of_nvmem_cell_get(struct device_node *np,
-					    const char *name)
+struct nvmem_cell *of_nvmem_cell_get(struct device_node *np, const char *id)
 {
 	struct device_node *cell_np, *nvmem_np;
-	struct nvmem_cell *cell;
 	struct nvmem_device *nvmem;
-	const __be32 *addr;
-	int rval, len;
+	struct nvmem_cell *cell;
 	int index = 0;
 
 	/* if cell name exists, find index to the name */
-	if (name)
-		index = of_property_match_string(np, "nvmem-cell-names", name);
+	if (id)
+		index = of_property_match_string(np, "nvmem-cell-names", id);
 
 	cell_np = of_parse_phandle(np, "nvmem-cells", index);
 	if (!cell_np)
@@ -857,59 +1006,18 @@ struct nvmem_cell *of_nvmem_cell_get(struct device_node *np,
 	if (!nvmem_np)
 		return ERR_PTR(-EINVAL);
 
-	nvmem = __nvmem_device_get(nvmem_np, NULL, NULL);
+	nvmem = __nvmem_device_get(nvmem_np, NULL);
 	of_node_put(nvmem_np);
 	if (IS_ERR(nvmem))
 		return ERR_CAST(nvmem);
 
-	addr = of_get_property(cell_np, "reg", &len);
-	if (!addr || (len < 2 * sizeof(u32))) {
-		dev_err(&nvmem->dev, "nvmem: invalid reg on %pOF\n",
-			cell_np);
-		rval  = -EINVAL;
-		goto err_mem;
-	}
-
-	cell = kzalloc(sizeof(*cell), GFP_KERNEL);
+	cell = nvmem_find_cell_by_index(nvmem, index);
 	if (!cell) {
-		rval = -ENOMEM;
-		goto err_mem;
-	}
-
-	cell->nvmem = nvmem;
-	cell->offset = be32_to_cpup(addr++);
-	cell->bytes = be32_to_cpup(addr);
-	cell->name = cell_np->name;
-
-	addr = of_get_property(cell_np, "bits", &len);
-	if (addr && len == (2 * sizeof(u32))) {
-		cell->bit_offset = be32_to_cpup(addr++);
-		cell->nbits = be32_to_cpup(addr);
+		__nvmem_device_put(nvmem);
+		return ERR_PTR(-ENOENT);
 	}
 
-	if (cell->nbits)
-		cell->bytes = DIV_ROUND_UP(cell->nbits + cell->bit_offset,
-					   BITS_PER_BYTE);
-
-	if (!IS_ALIGNED(cell->offset, nvmem->stride)) {
-			dev_err(&nvmem->dev,
-				"cell %s unaligned to nvmem stride %d\n",
-				cell->name, nvmem->stride);
-		rval  = -EINVAL;
-		goto err_sanity;
-	}
-
-	nvmem_cell_add(cell);
-
 	return cell;
-
-err_sanity:
-	kfree(cell);
-
-err_mem:
-	__nvmem_device_put(nvmem);
-
-	return ERR_PTR(rval);
 }
 EXPORT_SYMBOL_GPL(of_nvmem_cell_get);
 #endif
@@ -918,27 +1026,29 @@ EXPORT_SYMBOL_GPL(of_nvmem_cell_get);
  * nvmem_cell_get() - Get nvmem cell of device form a given cell name
  *
  * @dev: Device that requests the nvmem cell.
- * @cell_id: nvmem cell name to get.
+ * @id: nvmem cell name to get (this corresponds with the name from the
+ *      nvmem-cell-names property for DT systems and with the con_id from
+ *      the lookup entry for non-DT systems).
  *
  * Return: Will be an ERR_PTR() on error or a valid pointer
  * to a struct nvmem_cell.  The nvmem_cell will be freed by the
  * nvmem_cell_put().
  */
-struct nvmem_cell *nvmem_cell_get(struct device *dev, const char *cell_id)
+struct nvmem_cell *nvmem_cell_get(struct device *dev, const char *id)
 {
 	struct nvmem_cell *cell;
 
 	if (dev->of_node) { /* try dt first */
-		cell = of_nvmem_cell_get(dev->of_node, cell_id);
+		cell = of_nvmem_cell_get(dev->of_node, id);
 		if (!IS_ERR(cell) || PTR_ERR(cell) == -EPROBE_DEFER)
 			return cell;
 	}
 
-	/* NULL cell_id only allowed for device tree; invalid otherwise */
-	if (!cell_id)
+	/* NULL cell id only allowed for device tree; invalid otherwise */
+	if (!id)
 		return ERR_PTR(-EINVAL);
 
-	return nvmem_cell_get_from_list(cell_id);
+	return nvmem_cell_get_from_lookup(dev, id);
 }
 EXPORT_SYMBOL_GPL(nvmem_cell_get);
 
@@ -1015,7 +1125,6 @@ void nvmem_cell_put(struct nvmem_cell *cell)
 	struct nvmem_device *nvmem = cell->nvmem;
 
 	__nvmem_device_put(nvmem);
-	nvmem_cell_drop(cell);
 }
 EXPORT_SYMBOL_GPL(nvmem_cell_put);
 
@@ -1267,7 +1376,7 @@ EXPORT_SYMBOL_GPL(nvmem_device_cell_read);
  * @buf: buffer to be written to cell.
  *
  * Return: length of bytes written or negative error code on failure.
- * */
+ */
 int nvmem_device_cell_write(struct nvmem_device *nvmem,
 			    struct nvmem_cell_info *info, void *buf)
 {
@@ -1323,7 +1432,7 @@ EXPORT_SYMBOL_GPL(nvmem_device_read);
  * @buf: buffer to be written.
  *
  * Return: length of bytes written or negative error code on failure.
- * */
+ */
 int nvmem_device_write(struct nvmem_device *nvmem,
 		       unsigned int offset,
 		       size_t bytes, void *buf)
@@ -1343,6 +1452,80 @@ int nvmem_device_write(struct nvmem_device *nvmem,
 }
 EXPORT_SYMBOL_GPL(nvmem_device_write);
 
+/**
+ * nvmem_add_cell_table() - register a table of cell info entries
+ *
+ * @table: table of cell info entries
+ */
+void nvmem_add_cell_table(struct nvmem_cell_table *table)
+{
+	mutex_lock(&nvmem_cell_mutex);
+	list_add_tail(&table->node, &nvmem_cell_tables);
+	mutex_unlock(&nvmem_cell_mutex);
+}
+EXPORT_SYMBOL_GPL(nvmem_add_cell_table);
+
+/**
+ * nvmem_del_cell_table() - remove a previously registered cell info table
+ *
+ * @table: table of cell info entries
+ */
+void nvmem_del_cell_table(struct nvmem_cell_table *table)
+{
+	mutex_lock(&nvmem_cell_mutex);
+	list_del(&table->node);
+	mutex_unlock(&nvmem_cell_mutex);
+}
+EXPORT_SYMBOL_GPL(nvmem_del_cell_table);
+
+/**
+ * nvmem_add_cell_lookups() - register a list of cell lookup entries
+ *
+ * @entries: array of cell lookup entries
+ * @nentries: number of cell lookup entries in the array
+ */
+void nvmem_add_cell_lookups(struct nvmem_cell_lookup *entries, size_t nentries)
+{
+	int i;
+
+	mutex_lock(&nvmem_lookup_mutex);
+	for (i = 0; i < nentries; i++)
+		list_add_tail(&entries[i].node, &nvmem_lookup_list);
+	mutex_unlock(&nvmem_lookup_mutex);
+}
+EXPORT_SYMBOL_GPL(nvmem_add_cell_lookups);
+
+/**
+ * nvmem_del_cell_lookups() - remove a list of previously added cell lookup
+ *                            entries
+ *
+ * @entries: array of cell lookup entries
+ * @nentries: number of cell lookup entries in the array
+ */
+void nvmem_del_cell_lookups(struct nvmem_cell_lookup *entries, size_t nentries)
+{
+	int i;
+
+	mutex_lock(&nvmem_lookup_mutex);
+	for (i = 0; i < nentries; i++)
+		list_del(&entries[i].node);
+	mutex_unlock(&nvmem_lookup_mutex);
+}
+EXPORT_SYMBOL_GPL(nvmem_del_cell_lookups);
+
+/**
+ * nvmem_dev_name() - Get the name of a given nvmem device.
+ *
+ * @nvmem: nvmem device.
+ *
+ * Return: name of the nvmem device.
+ */
+const char *nvmem_dev_name(struct nvmem_device *nvmem)
+{
+	return dev_name(&nvmem->dev);
+}
+EXPORT_SYMBOL_GPL(nvmem_dev_name);
+
 static int __init nvmem_init(void)
 {
 	return bus_register(&nvmem_bus_type);
diff --git a/drivers/nvmem/lpc18xx_eeprom.c b/drivers/nvmem/lpc18xx_eeprom.c
index a9534a6e8636..66cff1e2147a 100644
--- a/drivers/nvmem/lpc18xx_eeprom.c
+++ b/drivers/nvmem/lpc18xx_eeprom.c
@@ -236,7 +236,7 @@ static int lpc18xx_eeprom_probe(struct platform_device *pdev)
 	lpc18xx_nvmem_config.dev = dev;
 	lpc18xx_nvmem_config.priv = eeprom;
 
-	eeprom->nvmem = nvmem_register(&lpc18xx_nvmem_config);
+	eeprom->nvmem = devm_nvmem_register(dev, &lpc18xx_nvmem_config);
 	if (IS_ERR(eeprom->nvmem)) {
 		ret = PTR_ERR(eeprom->nvmem);
 		goto err_clk;
@@ -255,11 +255,6 @@ err_clk:
 static int lpc18xx_eeprom_remove(struct platform_device *pdev)
 {
 	struct lpc18xx_eeprom_dev *eeprom = platform_get_drvdata(pdev);
-	int ret;
-
-	ret = nvmem_unregister(eeprom->nvmem);
-	if (ret < 0)
-		return ret;
 
 	clk_disable_unprepare(eeprom->clk);
 
diff --git a/drivers/nvmem/mxs-ocotp.c b/drivers/nvmem/mxs-ocotp.c
index 7018e2ef5714..53122f59c4b2 100644
--- a/drivers/nvmem/mxs-ocotp.c
+++ b/drivers/nvmem/mxs-ocotp.c
@@ -177,7 +177,7 @@ static int mxs_ocotp_probe(struct platform_device *pdev)
 	ocotp_config.size = data->size;
 	ocotp_config.priv = otp;
 	ocotp_config.dev = dev;
-	otp->nvmem = nvmem_register(&ocotp_config);
+	otp->nvmem = devm_nvmem_register(dev, &ocotp_config);
 	if (IS_ERR(otp->nvmem)) {
 		ret = PTR_ERR(otp->nvmem);
 		goto err_clk;
@@ -199,7 +199,7 @@ static int mxs_ocotp_remove(struct platform_device *pdev)
 
 	clk_unprepare(otp->clk);
 
-	return nvmem_unregister(otp->nvmem);
+	return 0;
 }
 
 static struct platform_driver mxs_ocotp_driver = {
diff --git a/drivers/nvmem/sunxi_sid.c b/drivers/nvmem/sunxi_sid.c
index d020f89248fd..570a2e354f30 100644
--- a/drivers/nvmem/sunxi_sid.c
+++ b/drivers/nvmem/sunxi_sid.c
@@ -154,7 +154,7 @@ static int sunxi_sid_probe(struct platform_device *pdev)
 	struct resource *res;
 	struct nvmem_device *nvmem;
 	struct sunxi_sid *sid;
-	int ret, i, size;
+	int i, size;
 	char *randomness;
 	const struct sunxi_sid_cfg *cfg;
 
@@ -181,15 +181,13 @@ static int sunxi_sid_probe(struct platform_device *pdev)
 	else
 		econfig.reg_read = sunxi_sid_read;
 	econfig.priv = sid;
-	nvmem = nvmem_register(&econfig);
+	nvmem = devm_nvmem_register(dev, &econfig);
 	if (IS_ERR(nvmem))
 		return PTR_ERR(nvmem);
 
 	randomness = kzalloc(size, GFP_KERNEL);
-	if (!randomness) {
-		ret = -EINVAL;
-		goto err_unreg_nvmem;
-	}
+	if (!randomness)
+		return -ENOMEM;
 
 	for (i = 0; i < size; i++)
 		econfig.reg_read(sid, i, &randomness[i], 1);
@@ -200,17 +198,6 @@ static int sunxi_sid_probe(struct platform_device *pdev)
 	platform_set_drvdata(pdev, nvmem);
 
 	return 0;
-
-err_unreg_nvmem:
-	nvmem_unregister(nvmem);
-	return ret;
-}
-
-static int sunxi_sid_remove(struct platform_device *pdev)
-{
-	struct nvmem_device *nvmem = platform_get_drvdata(pdev);
-
-	return nvmem_unregister(nvmem);
 }
 
 static const struct sunxi_sid_cfg sun4i_a10_cfg = {
@@ -243,7 +230,6 @@ MODULE_DEVICE_TABLE(of, sunxi_sid_of_match);
 
 static struct platform_driver sunxi_sid_driver = {
 	.probe = sunxi_sid_probe,
-	.remove = sunxi_sid_remove,
 	.driver = {
 		.name = "eeprom-sunxi-sid",
 		.of_match_table = sunxi_sid_of_match,
diff --git a/drivers/platform/goldfish/goldfish_pipe.c b/drivers/platform/goldfish/goldfish_pipe.c
index 2da567540c2d..7c639006252e 100644
--- a/drivers/platform/goldfish/goldfish_pipe.c
+++ b/drivers/platform/goldfish/goldfish_pipe.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Copyright (C) 2012 Intel, Inc.
  * Copyright (C) 2013 Intel, Inc.
@@ -46,7 +47,6 @@
  * exchange is properly mapped during a transfer.
  */
 
-
 #include <linux/module.h>
 #include <linux/mod_devicetable.h>
 #include <linux/interrupt.h>
@@ -59,10 +59,11 @@
 #include <linux/bitops.h>
 #include <linux/slab.h>
 #include <linux/io.h>
-#include <linux/goldfish.h>
 #include <linux/dma-mapping.h>
 #include <linux/mm.h>
 #include <linux/acpi.h>
+#include <linux/bug.h>
+#include "goldfish_pipe_qemu.h"
 
 /*
  * Update this when something changes in the driver's behavior so the host
@@ -73,71 +74,6 @@ enum {
 	PIPE_CURRENT_DEVICE_VERSION = 2
 };
 
-/*
- * IMPORTANT: The following constants must match the ones used and defined
- * in external/qemu/hw/goldfish_pipe.c in the Android source tree.
- */
-
-/* List of bitflags returned in status of CMD_POLL command */
-enum PipePollFlags {
-	PIPE_POLL_IN	= 1 << 0,
-	PIPE_POLL_OUT	= 1 << 1,
-	PIPE_POLL_HUP	= 1 << 2
-};
-
-/* Possible status values used to signal errors - see goldfish_pipe_error_convert */
-enum PipeErrors {
-	PIPE_ERROR_INVAL  = -1,
-	PIPE_ERROR_AGAIN  = -2,
-	PIPE_ERROR_NOMEM  = -3,
-	PIPE_ERROR_IO     = -4
-};
-
-/* Bit-flags used to signal events from the emulator */
-enum PipeWakeFlags {
-	PIPE_WAKE_CLOSED = 1 << 0,  /* emulator closed pipe */
-	PIPE_WAKE_READ   = 1 << 1,  /* pipe can now be read from */
-	PIPE_WAKE_WRITE  = 1 << 2  /* pipe can now be written to */
-};
-
-/* Bit flags for the 'flags' field */
-enum PipeFlagsBits {
-	BIT_CLOSED_ON_HOST = 0,  /* pipe closed by host */
-	BIT_WAKE_ON_WRITE  = 1,  /* want to be woken on writes */
-	BIT_WAKE_ON_READ   = 2,  /* want to be woken on reads */
-};
-
-enum PipeRegs {
-	PIPE_REG_CMD = 0,
-
-	PIPE_REG_SIGNAL_BUFFER_HIGH = 4,
-	PIPE_REG_SIGNAL_BUFFER = 8,
-	PIPE_REG_SIGNAL_BUFFER_COUNT = 12,
-
-	PIPE_REG_OPEN_BUFFER_HIGH = 20,
-	PIPE_REG_OPEN_BUFFER = 24,
-
-	PIPE_REG_VERSION = 36,
-
-	PIPE_REG_GET_SIGNALLED = 48,
-};
-
-enum PipeCmdCode {
-	PIPE_CMD_OPEN = 1,	/* to be used by the pipe device itself */
-	PIPE_CMD_CLOSE,
-	PIPE_CMD_POLL,
-	PIPE_CMD_WRITE,
-	PIPE_CMD_WAKE_ON_WRITE,
-	PIPE_CMD_READ,
-	PIPE_CMD_WAKE_ON_READ,
-
-	/*
-	 * TODO(zyy): implement a deferred read/write execution to allow
-	 * parallel processing of pipe operations on the host.
-	 */
-	PIPE_CMD_WAKE_ON_DONE_IO,
-};
-
 enum {
 	MAX_BUFFERS_PER_COMMAND = 336,
 	MAX_SIGNALLED_PIPES = 64,
@@ -145,14 +81,12 @@ enum {
 };
 
 struct goldfish_pipe_dev;
-struct goldfish_pipe;
-struct goldfish_pipe_command;
 
 /* A per-pipe command structure, shared with the host */
 struct goldfish_pipe_command {
-	s32 cmd;		/* PipeCmdCode, guest -> host */
-	s32 id;			/* pipe id, guest -> host */
-	s32 status;		/* command execution status, host -> guest */
+	s32 cmd;	/* PipeCmdCode, guest -> host */
+	s32 id;		/* pipe id, guest -> host */
+	s32 status;	/* command execution status, host -> guest */
 	s32 reserved;	/* to pad to 64-bit boundary */
 	union {
 		/* Parameters for PIPE_CMD_{READ,WRITE} */
@@ -184,19 +118,21 @@ struct open_command_param {
 /* Device-level set of buffers shared with the host */
 struct goldfish_pipe_dev_buffers {
 	struct open_command_param open_command_params;
-	struct signalled_pipe_buffer signalled_pipe_buffers[
-		MAX_SIGNALLED_PIPES];
+	struct signalled_pipe_buffer
+		signalled_pipe_buffers[MAX_SIGNALLED_PIPES];
 };
 
 /* This data type models a given pipe instance */
 struct goldfish_pipe {
 	/* pipe ID - index into goldfish_pipe_dev::pipes array */
 	u32 id;
+
 	/* The wake flags pipe is waiting for
 	 * Note: not protected with any lock, uses atomic operations
 	 *  and barriers to make it thread-safe.
 	 */
 	unsigned long flags;
+
 	/* wake flags host have signalled,
 	 *  - protected by goldfish_pipe_dev::lock
 	 */
@@ -220,8 +156,12 @@ struct goldfish_pipe {
 
 	/* A wake queue for sleeping until host signals an event */
 	wait_queue_head_t wake_queue;
+
 	/* Pointer to the parent goldfish_pipe_dev instance */
 	struct goldfish_pipe_dev *dev;
+
+	/* A buffer of pages, too large to fit into a stack frame */
+	struct page *pages[MAX_BUFFERS_PER_COMMAND];
 };
 
 /* The global driver data. Holds a reference to the i/o page used to
@@ -229,6 +169,9 @@ struct goldfish_pipe {
  * waiting to be awoken.
  */
 struct goldfish_pipe_dev {
+	/* A magic number to check if this is an instance of this struct */
+	void *magic;
+
 	/*
 	 * Global device spinlock. Protects the following members:
 	 *  - pipes, pipes_capacity
@@ -261,15 +204,22 @@ struct goldfish_pipe_dev {
 	/* Head of a doubly linked list of signalled pipes */
 	struct goldfish_pipe *first_signalled_pipe;
 
+	/* ptr to platform device's device struct */
+	struct device *pdev_dev;
+
 	/* Some device-specific data */
 	int irq;
 	int version;
 	unsigned char __iomem *base;
-};
 
-static struct goldfish_pipe_dev pipe_dev[1] = {};
+	/* an irq tasklet to run goldfish_interrupt_task */
+	struct tasklet_struct irq_tasklet;
 
-static int goldfish_cmd_locked(struct goldfish_pipe *pipe, enum PipeCmdCode cmd)
+	struct miscdevice miscdev;
+};
+
+static int goldfish_pipe_cmd_locked(struct goldfish_pipe *pipe,
+				    enum PipeCmdCode cmd)
 {
 	pipe->command_buffer->cmd = cmd;
 	/* failure by default */
@@ -278,13 +228,13 @@ static int goldfish_cmd_locked(struct goldfish_pipe *pipe, enum PipeCmdCode cmd)
 	return pipe->command_buffer->status;
 }
 
-static int goldfish_cmd(struct goldfish_pipe *pipe, enum PipeCmdCode cmd)
+static int goldfish_pipe_cmd(struct goldfish_pipe *pipe, enum PipeCmdCode cmd)
 {
 	int status;
 
 	if (mutex_lock_interruptible(&pipe->lock))
 		return PIPE_ERROR_IO;
-	status = goldfish_cmd_locked(pipe, cmd);
+	status = goldfish_pipe_cmd_locked(pipe, cmd);
 	mutex_unlock(&pipe->lock);
 	return status;
 }
@@ -307,10 +257,12 @@ static int goldfish_pipe_error_convert(int status)
 	}
 }
 
-static int pin_user_pages(unsigned long first_page, unsigned long last_page,
-	unsigned int last_page_size, int is_write,
-	struct page *pages[MAX_BUFFERS_PER_COMMAND],
-	unsigned int *iter_last_page_size)
+static int pin_user_pages(unsigned long first_page,
+			  unsigned long last_page,
+			  unsigned int last_page_size,
+			  int is_write,
+			  struct page *pages[MAX_BUFFERS_PER_COMMAND],
+			  unsigned int *iter_last_page_size)
 {
 	int ret;
 	int requested_pages = ((last_page - first_page) >> PAGE_SHIFT) + 1;
@@ -322,18 +274,18 @@ static int pin_user_pages(unsigned long first_page, unsigned long last_page,
 		*iter_last_page_size = last_page_size;
 	}
 
-	ret = get_user_pages_fast(
-			first_page, requested_pages, !is_write, pages);
+	ret = get_user_pages_fast(first_page, requested_pages, !is_write,
+				  pages);
 	if (ret <= 0)
 		return -EFAULT;
 	if (ret < requested_pages)
 		*iter_last_page_size = PAGE_SIZE;
-	return ret;
 
+	return ret;
 }
 
 static void release_user_pages(struct page **pages, int pages_count,
-	int is_write, s32 consumed_size)
+			       int is_write, s32 consumed_size)
 {
 	int i;
 
@@ -345,12 +297,15 @@ static void release_user_pages(struct page **pages, int pages_count,
 }
 
 /* Populate the call parameters, merging adjacent pages together */
-static void populate_rw_params(
-	struct page **pages, int pages_count,
-	unsigned long address, unsigned long address_end,
-	unsigned long first_page, unsigned long last_page,
-	unsigned int iter_last_page_size, int is_write,
-	struct goldfish_pipe_command *command)
+static void populate_rw_params(struct page **pages,
+			       int pages_count,
+			       unsigned long address,
+			       unsigned long address_end,
+			       unsigned long first_page,
+			       unsigned long last_page,
+			       unsigned int iter_last_page_size,
+			       int is_write,
+			       struct goldfish_pipe_command *command)
 {
 	/*
 	 * Process the first page separately - it's the only page that
@@ -382,55 +337,59 @@ static void populate_rw_params(
 }
 
 static int transfer_max_buffers(struct goldfish_pipe *pipe,
-	unsigned long address, unsigned long address_end, int is_write,
-	unsigned long last_page, unsigned int last_page_size,
-	s32 *consumed_size, int *status)
+				unsigned long address,
+				unsigned long address_end,
+				int is_write,
+				unsigned long last_page,
+				unsigned int last_page_size,
+				s32 *consumed_size,
+				int *status)
 {
-	static struct page *pages[MAX_BUFFERS_PER_COMMAND];
 	unsigned long first_page = address & PAGE_MASK;
 	unsigned int iter_last_page_size;
-	int pages_count = pin_user_pages(first_page, last_page,
-			last_page_size, is_write,
-			pages, &iter_last_page_size);
-
-	if (pages_count < 0)
-		return pages_count;
+	int pages_count;
 
 	/* Serialize access to the pipe command buffers */
 	if (mutex_lock_interruptible(&pipe->lock))
 		return -ERESTARTSYS;
 
-	populate_rw_params(pages, pages_count, address, address_end,
-		first_page, last_page, iter_last_page_size, is_write,
-		pipe->command_buffer);
+	pages_count = pin_user_pages(first_page, last_page,
+				     last_page_size, is_write,
+				     pipe->pages, &iter_last_page_size);
+	if (pages_count < 0) {
+		mutex_unlock(&pipe->lock);
+		return pages_count;
+	}
+
+	populate_rw_params(pipe->pages, pages_count, address, address_end,
+			   first_page, last_page, iter_last_page_size, is_write,
+			   pipe->command_buffer);
 
 	/* Transfer the data */
-	*status = goldfish_cmd_locked(pipe,
+	*status = goldfish_pipe_cmd_locked(pipe,
 				is_write ? PIPE_CMD_WRITE : PIPE_CMD_READ);
 
 	*consumed_size = pipe->command_buffer->rw_params.consumed_size;
 
-	release_user_pages(pages, pages_count, is_write, *consumed_size);
+	release_user_pages(pipe->pages, pages_count, is_write, *consumed_size);
 
 	mutex_unlock(&pipe->lock);
-
 	return 0;
 }
 
 static int wait_for_host_signal(struct goldfish_pipe *pipe, int is_write)
 {
-	u32 wakeBit = is_write ? BIT_WAKE_ON_WRITE : BIT_WAKE_ON_READ;
+	u32 wake_bit = is_write ? BIT_WAKE_ON_WRITE : BIT_WAKE_ON_READ;
 
-	set_bit(wakeBit, &pipe->flags);
+	set_bit(wake_bit, &pipe->flags);
 
 	/* Tell the emulator we're going to wait for a wake event */
-	(void)goldfish_cmd(pipe,
+	goldfish_pipe_cmd(pipe,
 		is_write ? PIPE_CMD_WAKE_ON_WRITE : PIPE_CMD_WAKE_ON_READ);
 
-	while (test_bit(wakeBit, &pipe->flags)) {
-		if (wait_event_interruptible(
-				pipe->wake_queue,
-				!test_bit(wakeBit, &pipe->flags)))
+	while (test_bit(wake_bit, &pipe->flags)) {
+		if (wait_event_interruptible(pipe->wake_queue,
+					     !test_bit(wake_bit, &pipe->flags)))
 			return -ERESTARTSYS;
 
 		if (test_bit(BIT_CLOSED_ON_HOST, &pipe->flags))
@@ -441,7 +400,9 @@ static int wait_for_host_signal(struct goldfish_pipe *pipe, int is_write)
 }
 
 static ssize_t goldfish_pipe_read_write(struct file *filp,
-	char __user *buffer, size_t bufflen, int is_write)
+					char __user *buffer,
+					size_t bufflen,
+					int is_write)
 {
 	struct goldfish_pipe *pipe = filp->private_data;
 	int count = 0, ret = -EINVAL;
@@ -456,7 +417,7 @@ static ssize_t goldfish_pipe_read_write(struct file *filp,
 		return 0;
 	/* Check the buffer range for access */
 	if (unlikely(!access_ok(is_write ? VERIFY_WRITE : VERIFY_READ,
-			buffer, bufflen)))
+				buffer, bufflen)))
 		return -EFAULT;
 
 	address = (unsigned long)buffer;
@@ -469,8 +430,8 @@ static ssize_t goldfish_pipe_read_write(struct file *filp,
 		int status;
 
 		ret = transfer_max_buffers(pipe, address, address_end, is_write,
-				last_page, last_page_size, &consumed_size,
-				&status);
+					   last_page, last_page_size,
+					   &consumed_size, &status);
 		if (ret < 0)
 			break;
 
@@ -496,7 +457,8 @@ static ssize_t goldfish_pipe_read_write(struct file *filp,
 			 * err.
 			 */
 			if (status != PIPE_ERROR_AGAIN)
-				pr_info_ratelimited("goldfish_pipe: backend error %d on %s\n",
+				dev_err_ratelimited(pipe->dev->pdev_dev,
+					"backend error %d on %s\n",
 					status, is_write ? "write" : "read");
 			break;
 		}
@@ -522,19 +484,21 @@ static ssize_t goldfish_pipe_read_write(struct file *filp,
 }
 
 static ssize_t goldfish_pipe_read(struct file *filp, char __user *buffer,
-				size_t bufflen, loff_t *ppos)
+				  size_t bufflen, loff_t *ppos)
 {
 	return goldfish_pipe_read_write(filp, buffer, bufflen,
-			/* is_write */ 0);
+					/* is_write */ 0);
 }
 
 static ssize_t goldfish_pipe_write(struct file *filp,
-				const char __user *buffer, size_t bufflen,
-				loff_t *ppos)
+				   const char __user *buffer, size_t bufflen,
+				   loff_t *ppos)
 {
-	return goldfish_pipe_read_write(filp,
-			/* cast away the const */(char __user *)buffer, bufflen,
-			/* is_write */ 1);
+	/* cast away the const */
+	char __user *no_const_buffer = (char __user *)buffer;
+
+	return goldfish_pipe_read_write(filp, no_const_buffer, bufflen,
+					/* is_write */ 1);
 }
 
 static __poll_t goldfish_pipe_poll(struct file *filp, poll_table *wait)
@@ -545,7 +509,7 @@ static __poll_t goldfish_pipe_poll(struct file *filp, poll_table *wait)
 
 	poll_wait(filp, &pipe->wake_queue, wait);
 
-	status = goldfish_cmd(pipe, PIPE_CMD_POLL);
+	status = goldfish_pipe_cmd(pipe, PIPE_CMD_POLL);
 	if (status < 0)
 		return -ERESTARTSYS;
 
@@ -562,7 +526,7 @@ static __poll_t goldfish_pipe_poll(struct file *filp, poll_table *wait)
 }
 
 static void signalled_pipes_add_locked(struct goldfish_pipe_dev *dev,
-	u32 id, u32 flags)
+				       u32 id, u32 flags)
 {
 	struct goldfish_pipe *pipe;
 
@@ -574,8 +538,8 @@ static void signalled_pipes_add_locked(struct goldfish_pipe_dev *dev,
 		return;
 	pipe->signalled_flags |= flags;
 
-	if (pipe->prev_signalled || pipe->next_signalled
-		|| dev->first_signalled_pipe == pipe)
+	if (pipe->prev_signalled || pipe->next_signalled ||
+		dev->first_signalled_pipe == pipe)
 		return;	/* already in the list */
 	pipe->next_signalled = dev->first_signalled_pipe;
 	if (dev->first_signalled_pipe)
@@ -584,7 +548,8 @@ static void signalled_pipes_add_locked(struct goldfish_pipe_dev *dev,
 }
 
 static void signalled_pipes_remove_locked(struct goldfish_pipe_dev *dev,
-	struct goldfish_pipe *pipe) {
+					  struct goldfish_pipe *pipe)
+{
 	if (pipe->prev_signalled)
 		pipe->prev_signalled->next_signalled = pipe->next_signalled;
 	if (pipe->next_signalled)
@@ -623,10 +588,10 @@ static struct goldfish_pipe *signalled_pipes_pop_front(
 	return pipe;
 }
 
-static void goldfish_interrupt_task(unsigned long unused)
+static void goldfish_interrupt_task(unsigned long dev_addr)
 {
-	struct goldfish_pipe_dev *dev = pipe_dev;
 	/* Iterate over the signalled pipes and wake them one by one */
+	struct goldfish_pipe_dev *dev = (struct goldfish_pipe_dev *)dev_addr;
 	struct goldfish_pipe *pipe;
 	int wakes;
 
@@ -646,7 +611,9 @@ static void goldfish_interrupt_task(unsigned long unused)
 		wake_up_interruptible(&pipe->wake_queue);
 	}
 }
-static DECLARE_TASKLET(goldfish_interrupt_tasklet, goldfish_interrupt_task, 0);
+
+static void goldfish_pipe_device_deinit(struct platform_device *pdev,
+					struct goldfish_pipe_dev *dev);
 
 /*
  * The general idea of the interrupt handling:
@@ -668,7 +635,7 @@ static irqreturn_t goldfish_pipe_interrupt(int irq, void *dev_id)
 	unsigned long flags;
 	struct goldfish_pipe_dev *dev = dev_id;
 
-	if (dev != pipe_dev)
+	if (dev->magic != &goldfish_pipe_device_deinit)
 		return IRQ_NONE;
 
 	/* Request the signalled pipes from the device */
@@ -689,7 +656,7 @@ static irqreturn_t goldfish_pipe_interrupt(int irq, void *dev_id)
 
 	spin_unlock_irqrestore(&dev->lock, flags);
 
-	tasklet_schedule(&goldfish_interrupt_tasklet);
+	tasklet_schedule(&dev->irq_tasklet);
 	return IRQ_HANDLED;
 }
 
@@ -702,7 +669,10 @@ static int get_free_pipe_id_locked(struct goldfish_pipe_dev *dev)
 			return id;
 
 	{
-		/* Reallocate the array */
+		/* Reallocate the array.
+		 * Since get_free_pipe_id_locked runs with interrupts disabled,
+		 * we don't want to make calls that could lead to sleep.
+		 */
 		u32 new_capacity = 2 * dev->pipes_capacity;
 		struct goldfish_pipe **pipes =
 			kcalloc(new_capacity, sizeof(*pipes), GFP_ATOMIC);
@@ -717,6 +687,14 @@ static int get_free_pipe_id_locked(struct goldfish_pipe_dev *dev)
 	return id;
 }
 
+/* A helper function to get the instance of goldfish_pipe_dev from file */
+static struct goldfish_pipe_dev *to_goldfish_pipe_dev(struct file *file)
+{
+	struct miscdevice *miscdev = file->private_data;
+
+	return container_of(miscdev, struct goldfish_pipe_dev, miscdev);
+}
+
 /**
  *	goldfish_pipe_open - open a channel to the AVD
  *	@inode: inode of device
@@ -730,14 +708,15 @@ static int get_free_pipe_id_locked(struct goldfish_pipe_dev *dev)
  */
 static int goldfish_pipe_open(struct inode *inode, struct file *file)
 {
-	struct goldfish_pipe_dev *dev = pipe_dev;
+	struct goldfish_pipe_dev *dev = to_goldfish_pipe_dev(file);
 	unsigned long flags;
 	int id;
 	int status;
 
 	/* Allocate new pipe kernel object */
 	struct goldfish_pipe *pipe = kzalloc(sizeof(*pipe), GFP_KERNEL);
-	if (pipe == NULL)
+
+	if (!pipe)
 		return -ENOMEM;
 
 	pipe->dev = dev;
@@ -748,6 +727,7 @@ static int goldfish_pipe_open(struct inode *inode, struct file *file)
 	 * Command buffer needs to be allocated on its own page to make sure
 	 * it is physically contiguous in host's address space.
 	 */
+	BUILD_BUG_ON(sizeof(struct goldfish_pipe_command) > PAGE_SIZE);
 	pipe->command_buffer =
 		(struct goldfish_pipe_command *)__get_free_page(GFP_KERNEL);
 	if (!pipe->command_buffer) {
@@ -772,7 +752,7 @@ static int goldfish_pipe_open(struct inode *inode, struct file *file)
 			MAX_BUFFERS_PER_COMMAND;
 	dev->buffers->open_command_params.command_buffer_ptr =
 			(u64)(unsigned long)__pa(pipe->command_buffer);
-	status = goldfish_cmd_locked(pipe, PIPE_CMD_OPEN);
+	status = goldfish_pipe_cmd_locked(pipe, PIPE_CMD_OPEN);
 	spin_unlock_irqrestore(&dev->lock, flags);
 	if (status < 0)
 		goto err_cmd;
@@ -798,7 +778,7 @@ static int goldfish_pipe_release(struct inode *inode, struct file *filp)
 	struct goldfish_pipe_dev *dev = pipe->dev;
 
 	/* The guest is closing the channel, so tell the emulator right now */
-	(void)goldfish_cmd(pipe, PIPE_CMD_CLOSE);
+	goldfish_pipe_cmd(pipe, PIPE_CMD_CLOSE);
 
 	spin_lock_irqsave(&dev->lock, flags);
 	dev->pipes[pipe->id] = NULL;
@@ -820,36 +800,55 @@ static const struct file_operations goldfish_pipe_fops = {
 	.release = goldfish_pipe_release,
 };
 
-static struct miscdevice goldfish_pipe_dev = {
-	.minor = MISC_DYNAMIC_MINOR,
-	.name = "goldfish_pipe",
-	.fops = &goldfish_pipe_fops,
-};
+static void init_miscdevice(struct miscdevice *miscdev)
+{
+	memset(miscdev, 0, sizeof(*miscdev));
+
+	miscdev->minor = MISC_DYNAMIC_MINOR;
+	miscdev->name = "goldfish_pipe";
+	miscdev->fops = &goldfish_pipe_fops;
+}
+
+static void write_pa_addr(void *addr, void __iomem *portl, void __iomem *porth)
+{
+	const unsigned long paddr = __pa(addr);
+
+	writel(upper_32_bits(paddr), porth);
+	writel(lower_32_bits(paddr), portl);
+}
 
-static int goldfish_pipe_device_init(struct platform_device *pdev)
+static int goldfish_pipe_device_init(struct platform_device *pdev,
+				     struct goldfish_pipe_dev *dev)
 {
-	char *page;
-	struct goldfish_pipe_dev *dev = pipe_dev;
-	int err = devm_request_irq(&pdev->dev, dev->irq,
-				goldfish_pipe_interrupt,
-				IRQF_SHARED, "goldfish_pipe", dev);
+	int err;
+
+	tasklet_init(&dev->irq_tasklet, &goldfish_interrupt_task,
+		     (unsigned long)dev);
+
+	err = devm_request_irq(&pdev->dev, dev->irq,
+			       goldfish_pipe_interrupt,
+			       IRQF_SHARED, "goldfish_pipe", dev);
 	if (err) {
 		dev_err(&pdev->dev, "unable to allocate IRQ for v2\n");
 		return err;
 	}
 
-	err = misc_register(&goldfish_pipe_dev);
+	init_miscdevice(&dev->miscdev);
+	err = misc_register(&dev->miscdev);
 	if (err) {
 		dev_err(&pdev->dev, "unable to register v2 device\n");
 		return err;
 	}
 
+	dev->pdev_dev = &pdev->dev;
 	dev->first_signalled_pipe = NULL;
 	dev->pipes_capacity = INITIAL_PIPES_CAPACITY;
 	dev->pipes = kcalloc(dev->pipes_capacity, sizeof(*dev->pipes),
-					GFP_KERNEL);
-	if (!dev->pipes)
+			     GFP_KERNEL);
+	if (!dev->pipes) {
+		misc_deregister(&dev->miscdev);
 		return -ENOMEM;
+	}
 
 	/*
 	 * We're going to pass two buffers, open_command_params and
@@ -857,75 +856,67 @@ static int goldfish_pipe_device_init(struct platform_device *pdev)
 	 * needs to be contained in a single physical page. The easiest choice
 	 * is to just allocate a page and place the buffers in it.
 	 */
-	if (WARN_ON(sizeof(*dev->buffers) > PAGE_SIZE))
-		return -ENOMEM;
-
-	page = (char *)__get_free_page(GFP_KERNEL);
-	if (!page) {
+	BUILD_BUG_ON(sizeof(struct goldfish_pipe_dev_buffers) > PAGE_SIZE);
+	dev->buffers = (struct goldfish_pipe_dev_buffers *)
+		__get_free_page(GFP_KERNEL);
+	if (!dev->buffers) {
 		kfree(dev->pipes);
+		misc_deregister(&dev->miscdev);
 		return -ENOMEM;
 	}
-	dev->buffers = (struct goldfish_pipe_dev_buffers *)page;
 
 	/* Send the buffer addresses to the host */
-	{
-		u64 paddr = __pa(&dev->buffers->signalled_pipe_buffers);
-
-		writel((u32)(unsigned long)(paddr >> 32),
-			dev->base + PIPE_REG_SIGNAL_BUFFER_HIGH);
-		writel((u32)(unsigned long)paddr,
-			dev->base + PIPE_REG_SIGNAL_BUFFER);
-		writel((u32)MAX_SIGNALLED_PIPES,
-			dev->base + PIPE_REG_SIGNAL_BUFFER_COUNT);
-
-		paddr = __pa(&dev->buffers->open_command_params);
-		writel((u32)(unsigned long)(paddr >> 32),
-			dev->base + PIPE_REG_OPEN_BUFFER_HIGH);
-		writel((u32)(unsigned long)paddr,
-			dev->base + PIPE_REG_OPEN_BUFFER);
-	}
+	write_pa_addr(&dev->buffers->signalled_pipe_buffers,
+		      dev->base + PIPE_REG_SIGNAL_BUFFER,
+		      dev->base + PIPE_REG_SIGNAL_BUFFER_HIGH);
+
+	writel(MAX_SIGNALLED_PIPES,
+	       dev->base + PIPE_REG_SIGNAL_BUFFER_COUNT);
+
+	write_pa_addr(&dev->buffers->open_command_params,
+		      dev->base + PIPE_REG_OPEN_BUFFER,
+		      dev->base + PIPE_REG_OPEN_BUFFER_HIGH);
+
+	platform_set_drvdata(pdev, dev);
 	return 0;
 }
 
-static void goldfish_pipe_device_deinit(struct platform_device *pdev)
+static void goldfish_pipe_device_deinit(struct platform_device *pdev,
+					struct goldfish_pipe_dev *dev)
 {
-	struct goldfish_pipe_dev *dev = pipe_dev;
-
-	misc_deregister(&goldfish_pipe_dev);
+	misc_deregister(&dev->miscdev);
+	tasklet_kill(&dev->irq_tasklet);
 	kfree(dev->pipes);
 	free_page((unsigned long)dev->buffers);
 }
 
 static int goldfish_pipe_probe(struct platform_device *pdev)
 {
-	int err;
 	struct resource *r;
-	struct goldfish_pipe_dev *dev = pipe_dev;
+	struct goldfish_pipe_dev *dev;
 
-	if (WARN_ON(sizeof(struct goldfish_pipe_command) > PAGE_SIZE))
+	dev = devm_kzalloc(&pdev->dev, sizeof(*dev), GFP_KERNEL);
+	if (!dev)
 		return -ENOMEM;
 
-	/* not thread safe, but this should not happen */
-	WARN_ON(dev->base != NULL);
-
+	dev->magic = &goldfish_pipe_device_deinit;
 	spin_lock_init(&dev->lock);
 
 	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	if (r == NULL || resource_size(r) < PAGE_SIZE) {
+	if (!r || resource_size(r) < PAGE_SIZE) {
 		dev_err(&pdev->dev, "can't allocate i/o page\n");
 		return -EINVAL;
 	}
 	dev->base = devm_ioremap(&pdev->dev, r->start, PAGE_SIZE);
-	if (dev->base == NULL) {
+	if (!dev->base) {
 		dev_err(&pdev->dev, "ioremap failed\n");
 		return -EINVAL;
 	}
 
 	r = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
-	if (r == NULL) {
-		err = -EINVAL;
-		goto error;
-	}
+	if (!r)
+		return -EINVAL;
+
 	dev->irq = r->start;
 
 	/*
@@ -935,25 +926,19 @@ static int goldfish_pipe_probe(struct platform_device *pdev)
 	 *  reading device version back: this allows the host implementation to
 	 *  detect the old driver (if there was no version write before read).
 	 */
-	writel((u32)PIPE_DRIVER_VERSION, dev->base + PIPE_REG_VERSION);
+	writel(PIPE_DRIVER_VERSION, dev->base + PIPE_REG_VERSION);
 	dev->version = readl(dev->base + PIPE_REG_VERSION);
 	if (WARN_ON(dev->version < PIPE_CURRENT_DEVICE_VERSION))
 		return -EINVAL;
 
-	err = goldfish_pipe_device_init(pdev);
-	if (!err)
-		return 0;
-
-error:
-	dev->base = NULL;
-	return err;
+	return goldfish_pipe_device_init(pdev, dev);
 }
 
 static int goldfish_pipe_remove(struct platform_device *pdev)
 {
-	struct goldfish_pipe_dev *dev = pipe_dev;
-	goldfish_pipe_device_deinit(pdev);
-	dev->base = NULL;
+	struct goldfish_pipe_dev *dev = platform_get_drvdata(pdev);
+
+	goldfish_pipe_device_deinit(pdev, dev);
 	return 0;
 }
 
@@ -981,4 +966,4 @@ static struct platform_driver goldfish_pipe_driver = {
 
 module_platform_driver(goldfish_pipe_driver);
 MODULE_AUTHOR("David Turner <digit@google.com>");
-MODULE_LICENSE("GPL");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/platform/goldfish/goldfish_pipe_qemu.h b/drivers/platform/goldfish/goldfish_pipe_qemu.h
new file mode 100644
index 000000000000..b4d78c108afd
--- /dev/null
+++ b/drivers/platform/goldfish/goldfish_pipe_qemu.h
@@ -0,0 +1,98 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * IMPORTANT: The following constants must match the ones used and defined in
+ * external/qemu/include/hw/misc/goldfish_pipe.h
+ */
+
+#ifndef GOLDFISH_PIPE_QEMU_H
+#define GOLDFISH_PIPE_QEMU_H
+
+/* List of bitflags returned in status of CMD_POLL command */
+enum PipePollFlags {
+	PIPE_POLL_IN	= 1 << 0,
+	PIPE_POLL_OUT	= 1 << 1,
+	PIPE_POLL_HUP	= 1 << 2
+};
+
+/* Possible status values used to signal errors */
+enum PipeErrors {
+	PIPE_ERROR_INVAL	= -1,
+	PIPE_ERROR_AGAIN	= -2,
+	PIPE_ERROR_NOMEM	= -3,
+	PIPE_ERROR_IO		= -4
+};
+
+/* Bit-flags used to signal events from the emulator */
+enum PipeWakeFlags {
+	/* emulator closed pipe */
+	PIPE_WAKE_CLOSED		= 1 << 0,
+
+	/* pipe can now be read from */
+	PIPE_WAKE_READ			= 1 << 1,
+
+	/* pipe can now be written to */
+	PIPE_WAKE_WRITE			= 1 << 2,
+
+	/* unlock this pipe's DMA buffer */
+	PIPE_WAKE_UNLOCK_DMA		= 1 << 3,
+
+	/* unlock DMA buffer of the pipe shared to this pipe */
+	PIPE_WAKE_UNLOCK_DMA_SHARED	= 1 << 4,
+};
+
+/* Possible pipe closing reasons */
+enum PipeCloseReason {
+	/* guest sent a close command */
+	PIPE_CLOSE_GRACEFUL		= 0,
+
+	/* guest rebooted, we're closing the pipes */
+	PIPE_CLOSE_REBOOT		= 1,
+
+	/* close old pipes on snapshot load */
+	PIPE_CLOSE_LOAD_SNAPSHOT	= 2,
+
+	/* some unrecoverable error on the pipe */
+	PIPE_CLOSE_ERROR		= 3,
+};
+
+/* Bit flags for the 'flags' field */
+enum PipeFlagsBits {
+	BIT_CLOSED_ON_HOST = 0,  /* pipe closed by host */
+	BIT_WAKE_ON_WRITE  = 1,  /* want to be woken on writes */
+	BIT_WAKE_ON_READ   = 2,  /* want to be woken on reads */
+};
+
+enum PipeRegs {
+	PIPE_REG_CMD = 0,
+
+	PIPE_REG_SIGNAL_BUFFER_HIGH = 4,
+	PIPE_REG_SIGNAL_BUFFER = 8,
+	PIPE_REG_SIGNAL_BUFFER_COUNT = 12,
+
+	PIPE_REG_OPEN_BUFFER_HIGH = 20,
+	PIPE_REG_OPEN_BUFFER = 24,
+
+	PIPE_REG_VERSION = 36,
+
+	PIPE_REG_GET_SIGNALLED = 48,
+};
+
+enum PipeCmdCode {
+	/* to be used by the pipe device itself */
+	PIPE_CMD_OPEN		= 1,
+
+	PIPE_CMD_CLOSE,
+	PIPE_CMD_POLL,
+	PIPE_CMD_WRITE,
+	PIPE_CMD_WAKE_ON_WRITE,
+	PIPE_CMD_READ,
+	PIPE_CMD_WAKE_ON_READ,
+
+	/*
+	 * TODO(zyy): implement a deferred read/write execution to allow
+	 * parallel processing of pipe operations on the host.
+	 */
+	PIPE_CMD_WAKE_ON_DONE_IO,
+};
+
+#endif /* GOLDFISH_PIPE_QEMU_H */
diff --git a/drivers/slimbus/core.c b/drivers/slimbus/core.c
index 95b00d28ad6e..55eda5863a6b 100644
--- a/drivers/slimbus/core.c
+++ b/drivers/slimbus/core.c
@@ -9,6 +9,7 @@
 #include <linux/init.h>
 #include <linux/idr.h>
 #include <linux/of.h>
+#include <linux/of_device.h>
 #include <linux/pm_runtime.h>
 #include <linux/slimbus.h>
 #include "slimbus.h"
@@ -32,6 +33,10 @@ static int slim_device_match(struct device *dev, struct device_driver *drv)
 	struct slim_device *sbdev = to_slim_device(dev);
 	struct slim_driver *sbdrv = to_slim_driver(drv);
 
+	/* Attempt an OF style match first */
+	if (of_driver_match_device(dev, drv))
+		return 1;
+
 	return !!slim_match(sbdrv->id_table, sbdev);
 }
 
@@ -39,8 +44,23 @@ static int slim_device_probe(struct device *dev)
 {
 	struct slim_device	*sbdev = to_slim_device(dev);
 	struct slim_driver	*sbdrv = to_slim_driver(dev->driver);
+	int ret;
 
-	return sbdrv->probe(sbdev);
+	ret = sbdrv->probe(sbdev);
+	if (ret)
+		return ret;
+
+	/* try getting the logical address after probe */
+	ret = slim_get_logical_addr(sbdev);
+	if (!ret) {
+		if (sbdrv->device_status)
+			sbdrv->device_status(sbdev, sbdev->status);
+	} else {
+		dev_err(&sbdev->dev, "Failed to get logical address\n");
+		ret = -EPROBE_DEFER;
+	}
+
+	return ret;
 }
 
 static int slim_device_remove(struct device *dev)
@@ -57,11 +77,24 @@ static int slim_device_remove(struct device *dev)
 	return 0;
 }
 
+static int slim_device_uevent(struct device *dev, struct kobj_uevent_env *env)
+{
+	struct slim_device *sbdev = to_slim_device(dev);
+	int ret;
+
+	ret = of_device_uevent_modalias(dev, env);
+	if (ret != -ENODEV)
+		return ret;
+
+	return add_uevent_var(env, "MODALIAS=slim:%s", dev_name(&sbdev->dev));
+}
+
 struct bus_type slimbus_bus = {
 	.name		= "slimbus",
 	.match		= slim_device_match,
 	.probe		= slim_device_probe,
 	.remove		= slim_device_remove,
+	.uevent		= slim_device_uevent,
 };
 EXPORT_SYMBOL_GPL(slimbus_bus);
 
@@ -77,7 +110,7 @@ EXPORT_SYMBOL_GPL(slimbus_bus);
 int __slim_driver_register(struct slim_driver *drv, struct module *owner)
 {
 	/* ID table and probe are mandatory */
-	if (!drv->id_table || !drv->probe)
+	if (!(drv->driver.of_match_table || drv->id_table) || !drv->probe)
 		return -EINVAL;
 
 	drv->driver.bus = &slimbus_bus;
diff --git a/drivers/slimbus/qcom-ngd-ctrl.c b/drivers/slimbus/qcom-ngd-ctrl.c
index 8be4d6786c61..7218fb963d0a 100644
--- a/drivers/slimbus/qcom-ngd-ctrl.c
+++ b/drivers/slimbus/qcom-ngd-ctrl.c
@@ -1004,6 +1004,7 @@ static int qcom_slim_ngd_get_laddr(struct slim_controller *ctrl,
 				   struct slim_eaddr *ea, u8 *laddr)
 {
 	struct slim_val_inf msg =  {0};
+	u8 failed_ea[6] = {0, 0, 0, 0, 0, 0};
 	struct slim_msg_txn txn;
 	u8 wbuf[10] = {0};
 	u8 rbuf[10] = {0};
@@ -1034,6 +1035,9 @@ static int qcom_slim_ngd_get_laddr(struct slim_controller *ctrl,
 		return ret;
 	}
 
+	if (!memcmp(rbuf, failed_ea, 6))
+		return -ENXIO;
+
 	*laddr = rbuf[6];
 
 	return ret;
@@ -1234,8 +1238,17 @@ static int qcom_slim_ngd_enable(struct qcom_slim_ngd_ctrl *ctrl, bool enable)
 			pm_runtime_resume(ctrl->dev);
 		pm_runtime_mark_last_busy(ctrl->dev);
 		pm_runtime_put(ctrl->dev);
+
+		ret = slim_register_controller(&ctrl->ctrl);
+		if (ret) {
+			dev_err(ctrl->dev, "error adding slim controller\n");
+			return ret;
+		}
+
+		dev_info(ctrl->dev, "SLIM controller Registered\n");
 	} else {
 		qcom_slim_qmi_exit(ctrl);
+		slim_unregister_controller(&ctrl->ctrl);
 	}
 
 	return 0;
@@ -1342,7 +1355,6 @@ static int of_qcom_slim_ngd_register(struct device *parent,
 		ngd->base = ctrl->base + ngd->id * data->offset +
 					(ngd->id - 1) * data->size;
 		ctrl->ngd = ngd;
-		platform_driver_register(&qcom_slim_ngd_driver);
 
 		return 0;
 	}
@@ -1357,11 +1369,6 @@ static int qcom_slim_ngd_probe(struct platform_device *pdev)
 	int ret;
 
 	ctrl->ctrl.dev = dev;
-	ret = slim_register_controller(&ctrl->ctrl);
-	if (ret) {
-		dev_err(dev, "error adding slim controller\n");
-		return ret;
-	}
 
 	pm_runtime_use_autosuspend(dev);
 	pm_runtime_set_autosuspend_delay(dev, QCOM_SLIM_NGD_AUTOSUSPEND);
@@ -1371,7 +1378,7 @@ static int qcom_slim_ngd_probe(struct platform_device *pdev)
 	ret = qcom_slim_ngd_qmi_svc_event_init(ctrl);
 	if (ret) {
 		dev_err(&pdev->dev, "QMI service registration failed:%d", ret);
-		goto err;
+		return ret;
 	}
 
 	INIT_WORK(&ctrl->m_work, qcom_slim_ngd_master_worker);
@@ -1383,14 +1390,12 @@ static int qcom_slim_ngd_probe(struct platform_device *pdev)
 	}
 
 	return 0;
-err:
-	slim_unregister_controller(&ctrl->ctrl);
 wq_err:
 	qcom_slim_ngd_qmi_svc_event_deinit(&ctrl->qmi);
 	if (ctrl->mwq)
 		destroy_workqueue(ctrl->mwq);
 
-	return 0;
+	return ret;
 }
 
 static int qcom_slim_ngd_ctrl_probe(struct platform_device *pdev)
@@ -1441,6 +1446,7 @@ static int qcom_slim_ngd_ctrl_probe(struct platform_device *pdev)
 	init_completion(&ctrl->reconf);
 	init_completion(&ctrl->qmi.qmi_comp);
 
+	platform_driver_register(&qcom_slim_ngd_driver);
 	return of_qcom_slim_ngd_register(dev, ctrl);
 }
 
@@ -1456,7 +1462,7 @@ static int qcom_slim_ngd_remove(struct platform_device *pdev)
 	struct qcom_slim_ngd_ctrl *ctrl = platform_get_drvdata(pdev);
 
 	pm_runtime_disable(&pdev->dev);
-	slim_unregister_controller(&ctrl->ctrl);
+	qcom_slim_ngd_enable(ctrl, false);
 	qcom_slim_ngd_exit_dma(ctrl);
 	qcom_slim_ngd_qmi_svc_event_deinit(&ctrl->qmi);
 	if (ctrl->mwq)
@@ -1467,7 +1473,7 @@ static int qcom_slim_ngd_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static int qcom_slim_ngd_runtime_idle(struct device *dev)
+static int __maybe_unused qcom_slim_ngd_runtime_idle(struct device *dev)
 {
 	struct qcom_slim_ngd_ctrl *ctrl = dev_get_drvdata(dev);
 
@@ -1477,8 +1483,7 @@ static int qcom_slim_ngd_runtime_idle(struct device *dev)
 	return -EAGAIN;
 }
 
-#ifdef CONFIG_PM
-static int qcom_slim_ngd_runtime_suspend(struct device *dev)
+static int __maybe_unused qcom_slim_ngd_runtime_suspend(struct device *dev)
 {
 	struct qcom_slim_ngd_ctrl *ctrl = dev_get_drvdata(dev);
 	int ret = 0;
@@ -1491,7 +1496,6 @@ static int qcom_slim_ngd_runtime_suspend(struct device *dev)
 
 	return ret;
 }
-#endif
 
 static const struct dev_pm_ops qcom_slim_ngd_dev_pm_ops = {
 	SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
diff --git a/drivers/soundwire/bus.c b/drivers/soundwire/bus.c
index dcc0ff9f0c22..1cbfedfc20ef 100644
--- a/drivers/soundwire/bus.c
+++ b/drivers/soundwire/bus.c
@@ -35,6 +35,11 @@ int sdw_add_bus_master(struct sdw_bus *bus)
 	INIT_LIST_HEAD(&bus->slaves);
 	INIT_LIST_HEAD(&bus->m_rt_list);
 
+	/*
+	 * Initialize multi_link flag
+	 * TODO: populate this flag by reading property from FW node
+	 */
+	bus->multi_link = false;
 	if (bus->ops->read_prop) {
 		ret = bus->ops->read_prop(bus);
 		if (ret < 0) {
@@ -175,6 +180,7 @@ static inline int do_transfer_defer(struct sdw_bus *bus,
 
 	defer->msg = msg;
 	defer->length = msg->len;
+	init_completion(&defer->complete);
 
 	for (i = 0; i <= retry; i++) {
 		resp = bus->ops->xfer_msg_defer(bus, msg, defer);
diff --git a/drivers/soundwire/bus.h b/drivers/soundwire/bus.h
index 3b15c4e25a3a..c77de05b8100 100644
--- a/drivers/soundwire/bus.h
+++ b/drivers/soundwire/bus.h
@@ -4,6 +4,8 @@
 #ifndef __SDW_BUS_H
 #define __SDW_BUS_H
 
+#define DEFAULT_BANK_SWITCH_TIMEOUT 3000
+
 #if IS_ENABLED(CONFIG_ACPI)
 int sdw_acpi_find_slaves(struct sdw_bus *bus);
 #else
@@ -99,6 +101,7 @@ struct sdw_slave_runtime {
  * this stream, can be zero.
  * @slave_rt_list: Slave runtime list
  * @port_list: List of Master Ports configured for this stream, can be zero.
+ * @stream_node: sdw_stream_runtime master_list node
  * @bus_node: sdw_bus m_rt_list node
  */
 struct sdw_master_runtime {
@@ -108,6 +111,7 @@ struct sdw_master_runtime {
 	unsigned int ch_count;
 	struct list_head slave_rt_list;
 	struct list_head port_list;
+	struct list_head stream_node;
 	struct list_head bus_node;
 };
 
diff --git a/drivers/soundwire/intel.c b/drivers/soundwire/intel.c
index 0a8990e758f9..c5ee97ee7886 100644
--- a/drivers/soundwire/intel.c
+++ b/drivers/soundwire/intel.c
@@ -398,6 +398,69 @@ static int intel_config_stream(struct sdw_intel *sdw,
 }
 
 /*
+ * bank switch routines
+ */
+
+static int intel_pre_bank_switch(struct sdw_bus *bus)
+{
+	struct sdw_cdns *cdns = bus_to_cdns(bus);
+	struct sdw_intel *sdw = cdns_to_intel(cdns);
+	void __iomem *shim = sdw->res->shim;
+	int sync_reg;
+
+	/* Write to register only for multi-link */
+	if (!bus->multi_link)
+		return 0;
+
+	/* Read SYNC register */
+	sync_reg = intel_readl(shim, SDW_SHIM_SYNC);
+	sync_reg |= SDW_SHIM_SYNC_CMDSYNC << sdw->instance;
+	intel_writel(shim, SDW_SHIM_SYNC, sync_reg);
+
+	return 0;
+}
+
+static int intel_post_bank_switch(struct sdw_bus *bus)
+{
+	struct sdw_cdns *cdns = bus_to_cdns(bus);
+	struct sdw_intel *sdw = cdns_to_intel(cdns);
+	void __iomem *shim = sdw->res->shim;
+	int sync_reg, ret;
+
+	/* Write to register only for multi-link */
+	if (!bus->multi_link)
+		return 0;
+
+	/* Read SYNC register */
+	sync_reg = intel_readl(shim, SDW_SHIM_SYNC);
+
+	/*
+	 * post_bank_switch() ops is called from the bus in loop for
+	 * all the Masters in the steam with the expectation that
+	 * we trigger the bankswitch for the only first Master in the list
+	 * and do nothing for the other Masters
+	 *
+	 * So, set the SYNCGO bit only if CMDSYNC bit is set for any Master.
+	 */
+	if (!(sync_reg & SDW_SHIM_SYNC_CMDSYNC_MASK))
+		return 0;
+
+	/*
+	 * Set SyncGO bit to synchronously trigger a bank switch for
+	 * all the masters. A write to SYNCGO bit clears CMDSYNC bit for all
+	 * the Masters.
+	 */
+	sync_reg |= SDW_SHIM_SYNC_SYNCGO;
+
+	ret = intel_clear_bit(shim, SDW_SHIM_SYNC, sync_reg,
+					SDW_SHIM_SYNC_SYNCGO);
+	if (ret < 0)
+		dev_err(sdw->cdns.dev, "Post bank switch failed: %d", ret);
+
+	return ret;
+}
+
+/*
  * DAI routines
  */
 
@@ -750,6 +813,8 @@ static struct sdw_master_ops sdw_intel_ops = {
 	.xfer_msg_defer = cdns_xfer_msg_defer,
 	.reset_page_addr = cdns_reset_page_addr,
 	.set_bus_conf = cdns_bus_conf,
+	.pre_bank_switch = intel_pre_bank_switch,
+	.post_bank_switch = intel_post_bank_switch,
 };
 
 /*
@@ -780,9 +845,6 @@ static int intel_probe(struct platform_device *pdev)
 	sdw_intel_ops.read_prop = intel_prop_read;
 	sdw->cdns.bus.ops = &sdw_intel_ops;
 
-	sdw_intel_ops.read_prop = intel_prop_read;
-	sdw->cdns.bus.ops = &sdw_intel_ops;
-
 	platform_set_drvdata(pdev, sdw);
 
 	ret = sdw_add_bus_master(&sdw->cdns.bus);
diff --git a/drivers/soundwire/intel_init.c b/drivers/soundwire/intel_init.c
index d1ea6b4d0ad3..5c8a20d99878 100644
--- a/drivers/soundwire/intel_init.c
+++ b/drivers/soundwire/intel_init.c
@@ -151,7 +151,7 @@ static acpi_status sdw_intel_acpi_cb(acpi_handle handle, u32 level,
 	struct acpi_device *adev;
 
 	if (acpi_bus_get_device(handle, &adev)) {
-		dev_err(&adev->dev, "Couldn't find ACPI handle\n");
+		pr_err("%s: Couldn't find ACPI handle\n", __func__);
 		return AE_NOT_FOUND;
 	}
 
diff --git a/drivers/soundwire/stream.c b/drivers/soundwire/stream.c
index e5c7e1ef6318..bd879b1a76c8 100644
--- a/drivers/soundwire/stream.c
+++ b/drivers/soundwire/stream.c
@@ -626,9 +626,10 @@ static int sdw_program_params(struct sdw_bus *bus)
 	return ret;
 }
 
-static int sdw_bank_switch(struct sdw_bus *bus)
+static int sdw_bank_switch(struct sdw_bus *bus, int m_rt_count)
 {
 	int col_index, row_index;
+	bool multi_link;
 	struct sdw_msg *wr_msg;
 	u8 *wbuf = NULL;
 	int ret = 0;
@@ -638,6 +639,8 @@ static int sdw_bank_switch(struct sdw_bus *bus)
 	if (!wr_msg)
 		return -ENOMEM;
 
+	bus->defer_msg.msg = wr_msg;
+
 	wbuf = kzalloc(sizeof(*wbuf), GFP_KERNEL);
 	if (!wbuf) {
 		ret = -ENOMEM;
@@ -658,17 +661,29 @@ static int sdw_bank_switch(struct sdw_bus *bus)
 					SDW_MSG_FLAG_WRITE, wbuf);
 	wr_msg->ssp_sync = true;
 
-	ret = sdw_transfer(bus, wr_msg);
+	/*
+	 * Set the multi_link flag only when both the hardware supports
+	 * and there is a stream handled by multiple masters
+	 */
+	multi_link = bus->multi_link && (m_rt_count > 1);
+
+	if (multi_link)
+		ret = sdw_transfer_defer(bus, wr_msg, &bus->defer_msg);
+	else
+		ret = sdw_transfer(bus, wr_msg);
+
 	if (ret < 0) {
 		dev_err(bus->dev, "Slave frame_ctrl reg write failed");
 		goto error;
 	}
 
-	kfree(wr_msg);
-	kfree(wbuf);
-	bus->defer_msg.msg = NULL;
-	bus->params.curr_bank = !bus->params.curr_bank;
-	bus->params.next_bank = !bus->params.next_bank;
+	if (!multi_link) {
+		kfree(wr_msg);
+		kfree(wbuf);
+		bus->defer_msg.msg = NULL;
+		bus->params.curr_bank = !bus->params.curr_bank;
+		bus->params.next_bank = !bus->params.next_bank;
+	}
 
 	return 0;
 
@@ -679,37 +694,138 @@ error_1:
 	return ret;
 }
 
+/**
+ * sdw_ml_sync_bank_switch: Multilink register bank switch
+ *
+ * @bus: SDW bus instance
+ *
+ * Caller function should free the buffers on error
+ */
+static int sdw_ml_sync_bank_switch(struct sdw_bus *bus)
+{
+	unsigned long time_left;
+
+	if (!bus->multi_link)
+		return 0;
+
+	/* Wait for completion of transfer */
+	time_left = wait_for_completion_timeout(&bus->defer_msg.complete,
+						bus->bank_switch_timeout);
+
+	if (!time_left) {
+		dev_err(bus->dev, "Controller Timed out on bank switch");
+		return -ETIMEDOUT;
+	}
+
+	bus->params.curr_bank = !bus->params.curr_bank;
+	bus->params.next_bank = !bus->params.next_bank;
+
+	if (bus->defer_msg.msg) {
+		kfree(bus->defer_msg.msg->buf);
+		kfree(bus->defer_msg.msg);
+	}
+
+	return 0;
+}
+
 static int do_bank_switch(struct sdw_stream_runtime *stream)
 {
-	struct sdw_master_runtime *m_rt = stream->m_rt;
+	struct sdw_master_runtime *m_rt = NULL;
 	const struct sdw_master_ops *ops;
-	struct sdw_bus *bus = m_rt->bus;
+	struct sdw_bus *bus = NULL;
+	bool multi_link = false;
 	int ret = 0;
 
-	ops = bus->ops;
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+		ops = bus->ops;
+
+		if (bus->multi_link) {
+			multi_link = true;
+			mutex_lock(&bus->msg_lock);
+		}
+
+		/* Pre-bank switch */
+		if (ops->pre_bank_switch) {
+			ret = ops->pre_bank_switch(bus);
+			if (ret < 0) {
+				dev_err(bus->dev,
+					"Pre bank switch op failed: %d", ret);
+				goto msg_unlock;
+			}
+		}
 
-	/* Pre-bank switch */
-	if (ops->pre_bank_switch) {
-		ret = ops->pre_bank_switch(bus);
+		/*
+		 * Perform Bank switch operation.
+		 * For multi link cases, the actual bank switch is
+		 * synchronized across all Masters and happens later as a
+		 * part of post_bank_switch ops.
+		 */
+		ret = sdw_bank_switch(bus, stream->m_rt_count);
 		if (ret < 0) {
-			dev_err(bus->dev, "Pre bank switch op failed: %d", ret);
-			return ret;
+			dev_err(bus->dev, "Bank switch failed: %d", ret);
+			goto error;
+
 		}
 	}
 
-	/* Bank switch */
-	ret = sdw_bank_switch(bus);
-	if (ret < 0) {
-		dev_err(bus->dev, "Bank switch failed: %d", ret);
-		return ret;
-	}
+	/*
+	 * For multi link cases, it is expected that the bank switch is
+	 * triggered by the post_bank_switch for the first Master in the list
+	 * and for the other Masters the post_bank_switch() should return doing
+	 * nothing.
+	 */
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+		ops = bus->ops;
+
+		/* Post-bank switch */
+		if (ops->post_bank_switch) {
+			ret = ops->post_bank_switch(bus);
+			if (ret < 0) {
+				dev_err(bus->dev,
+					"Post bank switch op failed: %d", ret);
+				goto error;
+			}
+		} else if (bus->multi_link && stream->m_rt_count > 1) {
+			dev_err(bus->dev,
+				"Post bank switch ops not implemented");
+			goto error;
+		}
+
+		/* Set the bank switch timeout to default, if not set */
+		if (!bus->bank_switch_timeout)
+			bus->bank_switch_timeout = DEFAULT_BANK_SWITCH_TIMEOUT;
 
-	/* Post-bank switch */
-	if (ops->post_bank_switch) {
-		ret = ops->post_bank_switch(bus);
+		/* Check if bank switch was successful */
+		ret = sdw_ml_sync_bank_switch(bus);
 		if (ret < 0) {
 			dev_err(bus->dev,
-					"Post bank switch op failed: %d", ret);
+				"multi link bank switch failed: %d", ret);
+			goto error;
+		}
+
+		mutex_unlock(&bus->msg_lock);
+	}
+
+	return ret;
+
+error:
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+
+		bus = m_rt->bus;
+
+		kfree(bus->defer_msg.msg->buf);
+		kfree(bus->defer_msg.msg);
+	}
+
+msg_unlock:
+
+	if (multi_link) {
+		list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+			bus = m_rt->bus;
+			if (mutex_is_locked(&bus->msg_lock))
+				mutex_unlock(&bus->msg_lock);
 		}
 	}
 
@@ -747,12 +863,29 @@ struct sdw_stream_runtime *sdw_alloc_stream(char *stream_name)
 		return NULL;
 
 	stream->name = stream_name;
+	INIT_LIST_HEAD(&stream->master_list);
 	stream->state = SDW_STREAM_ALLOCATED;
+	stream->m_rt_count = 0;
 
 	return stream;
 }
 EXPORT_SYMBOL(sdw_alloc_stream);
 
+static struct sdw_master_runtime
+*sdw_find_master_rt(struct sdw_bus *bus,
+			struct sdw_stream_runtime *stream)
+{
+	struct sdw_master_runtime *m_rt = NULL;
+
+	/* Retrieve Bus handle if already available */
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		if (m_rt->bus == bus)
+			return m_rt;
+	}
+
+	return NULL;
+}
+
 /**
  * sdw_alloc_master_rt() - Allocates and initialize Master runtime handle
  *
@@ -769,12 +902,11 @@ static struct sdw_master_runtime
 {
 	struct sdw_master_runtime *m_rt;
 
-	m_rt = stream->m_rt;
-
 	/*
 	 * check if Master is already allocated (as a result of Slave adding
 	 * it first), if so skip allocation and go to configure
 	 */
+	m_rt = sdw_find_master_rt(bus, stream);
 	if (m_rt)
 		goto stream_config;
 
@@ -785,7 +917,7 @@ static struct sdw_master_runtime
 	/* Initialization of Master runtime handle */
 	INIT_LIST_HEAD(&m_rt->port_list);
 	INIT_LIST_HEAD(&m_rt->slave_rt_list);
-	stream->m_rt = m_rt;
+	list_add_tail(&m_rt->stream_node, &stream->master_list);
 
 	list_add_tail(&m_rt->bus_node, &bus->m_rt_list);
 
@@ -843,17 +975,21 @@ static void sdw_slave_port_release(struct sdw_bus *bus,
 			struct sdw_stream_runtime *stream)
 {
 	struct sdw_port_runtime *p_rt, *_p_rt;
-	struct sdw_master_runtime *m_rt = stream->m_rt;
+	struct sdw_master_runtime *m_rt;
 	struct sdw_slave_runtime *s_rt;
 
-	list_for_each_entry(s_rt, &m_rt->slave_rt_list, m_rt_node) {
-		if (s_rt->slave != slave)
-			continue;
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		list_for_each_entry(s_rt, &m_rt->slave_rt_list, m_rt_node) {
 
-		list_for_each_entry_safe(p_rt, _p_rt,
-				&s_rt->port_list, port_node) {
-			list_del(&p_rt->port_node);
-			kfree(p_rt);
+			if (s_rt->slave != slave)
+				continue;
+
+			list_for_each_entry_safe(p_rt, _p_rt,
+					&s_rt->port_list, port_node) {
+
+				list_del(&p_rt->port_node);
+				kfree(p_rt);
+			}
 		}
 	}
 }
@@ -870,16 +1006,18 @@ static void sdw_release_slave_stream(struct sdw_slave *slave,
 			struct sdw_stream_runtime *stream)
 {
 	struct sdw_slave_runtime *s_rt, *_s_rt;
-	struct sdw_master_runtime *m_rt = stream->m_rt;
+	struct sdw_master_runtime *m_rt;
 
-	/* Retrieve Slave runtime handle */
-	list_for_each_entry_safe(s_rt, _s_rt,
-			&m_rt->slave_rt_list, m_rt_node) {
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		/* Retrieve Slave runtime handle */
+		list_for_each_entry_safe(s_rt, _s_rt,
+					&m_rt->slave_rt_list, m_rt_node) {
 
-		if (s_rt->slave == slave) {
-			list_del(&s_rt->m_rt_node);
-			kfree(s_rt);
-			return;
+			if (s_rt->slave == slave) {
+				list_del(&s_rt->m_rt_node);
+				kfree(s_rt);
+				return;
+			}
 		}
 	}
 }
@@ -887,6 +1025,7 @@ static void sdw_release_slave_stream(struct sdw_slave *slave,
 /**
  * sdw_release_master_stream() - Free Master runtime handle
  *
+ * @m_rt: Master runtime node
  * @stream: Stream runtime handle.
  *
  * This function is to be called with bus_lock held
@@ -894,9 +1033,9 @@ static void sdw_release_slave_stream(struct sdw_slave *slave,
  * handle. If this is called first then sdw_release_slave_stream() will have
  * no effect as Slave(s) runtime handle would already be freed up.
  */
-static void sdw_release_master_stream(struct sdw_stream_runtime *stream)
+static void sdw_release_master_stream(struct sdw_master_runtime *m_rt,
+			struct sdw_stream_runtime *stream)
 {
-	struct sdw_master_runtime *m_rt = stream->m_rt;
 	struct sdw_slave_runtime *s_rt, *_s_rt;
 
 	list_for_each_entry_safe(s_rt, _s_rt, &m_rt->slave_rt_list, m_rt_node) {
@@ -904,7 +1043,9 @@ static void sdw_release_master_stream(struct sdw_stream_runtime *stream)
 		sdw_release_slave_stream(s_rt->slave, stream);
 	}
 
+	list_del(&m_rt->stream_node);
 	list_del(&m_rt->bus_node);
+	kfree(m_rt);
 }
 
 /**
@@ -918,13 +1059,23 @@ static void sdw_release_master_stream(struct sdw_stream_runtime *stream)
 int sdw_stream_remove_master(struct sdw_bus *bus,
 		struct sdw_stream_runtime *stream)
 {
+	struct sdw_master_runtime *m_rt, *_m_rt;
+
 	mutex_lock(&bus->bus_lock);
 
-	sdw_release_master_stream(stream);
-	sdw_master_port_release(bus, stream->m_rt);
-	stream->state = SDW_STREAM_RELEASED;
-	kfree(stream->m_rt);
-	stream->m_rt = NULL;
+	list_for_each_entry_safe(m_rt, _m_rt,
+			&stream->master_list, stream_node) {
+
+		if (m_rt->bus != bus)
+			continue;
+
+		sdw_master_port_release(bus, m_rt);
+		sdw_release_master_stream(m_rt, stream);
+		stream->m_rt_count--;
+	}
+
+	if (list_empty(&stream->master_list))
+		stream->state = SDW_STREAM_RELEASED;
 
 	mutex_unlock(&bus->bus_lock);
 
@@ -1107,6 +1258,18 @@ int sdw_stream_add_master(struct sdw_bus *bus,
 
 	mutex_lock(&bus->bus_lock);
 
+	/*
+	 * For multi link streams, add the second master only if
+	 * the bus supports it.
+	 * Check if bus->multi_link is set
+	 */
+	if (!bus->multi_link && stream->m_rt_count > 0) {
+		dev_err(bus->dev,
+			"Multilink not supported, link %d", bus->link_id);
+		ret = -EINVAL;
+		goto unlock;
+	}
+
 	m_rt = sdw_alloc_master_rt(bus, stream_config, stream);
 	if (!m_rt) {
 		dev_err(bus->dev,
@@ -1124,10 +1287,12 @@ int sdw_stream_add_master(struct sdw_bus *bus,
 	if (ret)
 		goto stream_error;
 
+	stream->m_rt_count++;
+
 	goto unlock;
 
 stream_error:
-	sdw_release_master_stream(stream);
+	sdw_release_master_stream(m_rt, stream);
 unlock:
 	mutex_unlock(&bus->bus_lock);
 	return ret;
@@ -1205,7 +1370,7 @@ stream_error:
 	 * we hit error so cleanup the stream, release all Slave(s) and
 	 * Master runtime
 	 */
-	sdw_release_master_stream(stream);
+	sdw_release_master_stream(m_rt, stream);
 error:
 	mutex_unlock(&slave->bus->bus_lock);
 	return ret;
@@ -1245,33 +1410,82 @@ struct sdw_dpn_prop *sdw_get_slave_dpn_prop(struct sdw_slave *slave,
 	return NULL;
 }
 
+/**
+ * sdw_acquire_bus_lock: Acquire bus lock for all Master runtime(s)
+ *
+ * @stream: SoundWire stream
+ *
+ * Acquire bus_lock for each of the master runtime(m_rt) part of this
+ * stream to reconfigure the bus.
+ * NOTE: This function is called from SoundWire stream ops and is
+ * expected that a global lock is held before acquiring bus_lock.
+ */
+static void sdw_acquire_bus_lock(struct sdw_stream_runtime *stream)
+{
+	struct sdw_master_runtime *m_rt = NULL;
+	struct sdw_bus *bus = NULL;
+
+	/* Iterate for all Master(s) in Master list */
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+
+		mutex_lock(&bus->bus_lock);
+	}
+}
+
+/**
+ * sdw_release_bus_lock: Release bus lock for all Master runtime(s)
+ *
+ * @stream: SoundWire stream
+ *
+ * Release the previously held bus_lock after reconfiguring the bus.
+ * NOTE: This function is called from SoundWire stream ops and is
+ * expected that a global lock is held before releasing bus_lock.
+ */
+static void sdw_release_bus_lock(struct sdw_stream_runtime *stream)
+{
+	struct sdw_master_runtime *m_rt = NULL;
+	struct sdw_bus *bus = NULL;
+
+	/* Iterate for all Master(s) in Master list */
+	list_for_each_entry_reverse(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+		mutex_unlock(&bus->bus_lock);
+	}
+}
+
 static int _sdw_prepare_stream(struct sdw_stream_runtime *stream)
 {
-	struct sdw_master_runtime *m_rt = stream->m_rt;
-	struct sdw_bus *bus = m_rt->bus;
+	struct sdw_master_runtime *m_rt = NULL;
+	struct sdw_bus *bus = NULL;
 	struct sdw_master_prop *prop = NULL;
 	struct sdw_bus_params params;
 	int ret;
 
-	prop = &bus->prop;
-	memcpy(&params, &bus->params, sizeof(params));
+	/* Prepare  Master(s) and Slave(s) port(s) associated with stream */
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+		prop = &bus->prop;
+		memcpy(&params, &bus->params, sizeof(params));
 
-	/* TODO: Support Asynchronous mode */
-	if ((prop->max_freq % stream->params.rate) != 0) {
-		dev_err(bus->dev, "Async mode not supported");
-		return -EINVAL;
-	}
+		/* TODO: Support Asynchronous mode */
+		if ((prop->max_freq % stream->params.rate) != 0) {
+			dev_err(bus->dev, "Async mode not supported");
+			return -EINVAL;
+		}
 
-	/* Increment cumulative bus bandwidth */
-	/* TODO: Update this during Device-Device support */
-	bus->params.bandwidth += m_rt->stream->params.rate *
-		m_rt->ch_count * m_rt->stream->params.bps;
+		/* Increment cumulative bus bandwidth */
+		/* TODO: Update this during Device-Device support */
+		bus->params.bandwidth += m_rt->stream->params.rate *
+			m_rt->ch_count * m_rt->stream->params.bps;
+
+		/* Program params */
+		ret = sdw_program_params(bus);
+		if (ret < 0) {
+			dev_err(bus->dev, "Program params failed: %d", ret);
+			goto restore_params;
+		}
 
-	/* Program params */
-	ret = sdw_program_params(bus);
-	if (ret < 0) {
-		dev_err(bus->dev, "Program params failed: %d", ret);
-		goto restore_params;
 	}
 
 	ret = do_bank_switch(stream);
@@ -1280,12 +1494,16 @@ static int _sdw_prepare_stream(struct sdw_stream_runtime *stream)
 		goto restore_params;
 	}
 
-	/* Prepare port(s) on the new clock configuration */
-	ret = sdw_prep_deprep_ports(m_rt, true);
-	if (ret < 0) {
-		dev_err(bus->dev, "Prepare port(s) failed ret = %d",
-				ret);
-		return ret;
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+
+		/* Prepare port(s) on the new clock configuration */
+		ret = sdw_prep_deprep_ports(m_rt, true);
+		if (ret < 0) {
+			dev_err(bus->dev, "Prepare port(s) failed ret = %d",
+					ret);
+			return ret;
+		}
 	}
 
 	stream->state = SDW_STREAM_PREPARED;
@@ -1313,35 +1531,40 @@ int sdw_prepare_stream(struct sdw_stream_runtime *stream)
 		return -EINVAL;
 	}
 
-	mutex_lock(&stream->m_rt->bus->bus_lock);
+	sdw_acquire_bus_lock(stream);
 
 	ret = _sdw_prepare_stream(stream);
 	if (ret < 0)
 		pr_err("Prepare for stream:%s failed: %d", stream->name, ret);
 
-	mutex_unlock(&stream->m_rt->bus->bus_lock);
+	sdw_release_bus_lock(stream);
 	return ret;
 }
 EXPORT_SYMBOL(sdw_prepare_stream);
 
 static int _sdw_enable_stream(struct sdw_stream_runtime *stream)
 {
-	struct sdw_master_runtime *m_rt = stream->m_rt;
-	struct sdw_bus *bus = m_rt->bus;
+	struct sdw_master_runtime *m_rt = NULL;
+	struct sdw_bus *bus = NULL;
 	int ret;
 
-	/* Program params */
-	ret = sdw_program_params(bus);
-	if (ret < 0) {
-		dev_err(bus->dev, "Program params failed: %d", ret);
-		return ret;
-	}
+	/* Enable Master(s) and Slave(s) port(s) associated with stream */
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
 
-	/* Enable port(s) */
-	ret = sdw_enable_disable_ports(m_rt, true);
-	if (ret < 0) {
-		dev_err(bus->dev, "Enable port(s) failed ret: %d", ret);
-		return ret;
+		/* Program params */
+		ret = sdw_program_params(bus);
+		if (ret < 0) {
+			dev_err(bus->dev, "Program params failed: %d", ret);
+			return ret;
+		}
+
+		/* Enable port(s) */
+		ret = sdw_enable_disable_ports(m_rt, true);
+		if (ret < 0) {
+			dev_err(bus->dev, "Enable port(s) failed ret: %d", ret);
+			return ret;
+		}
 	}
 
 	ret = do_bank_switch(stream);
@@ -1370,37 +1593,42 @@ int sdw_enable_stream(struct sdw_stream_runtime *stream)
 		return -EINVAL;
 	}
 
-	mutex_lock(&stream->m_rt->bus->bus_lock);
+	sdw_acquire_bus_lock(stream);
 
 	ret = _sdw_enable_stream(stream);
 	if (ret < 0)
 		pr_err("Enable for stream:%s failed: %d", stream->name, ret);
 
-	mutex_unlock(&stream->m_rt->bus->bus_lock);
+	sdw_release_bus_lock(stream);
 	return ret;
 }
 EXPORT_SYMBOL(sdw_enable_stream);
 
 static int _sdw_disable_stream(struct sdw_stream_runtime *stream)
 {
-	struct sdw_master_runtime *m_rt = stream->m_rt;
-	struct sdw_bus *bus = m_rt->bus;
+	struct sdw_master_runtime *m_rt = NULL;
+	struct sdw_bus *bus = NULL;
 	int ret;
 
-	/* Disable port(s) */
-	ret = sdw_enable_disable_ports(m_rt, false);
-	if (ret < 0) {
-		dev_err(bus->dev, "Disable port(s) failed: %d", ret);
-		return ret;
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+		/* Disable port(s) */
+		ret = sdw_enable_disable_ports(m_rt, false);
+		if (ret < 0) {
+			dev_err(bus->dev, "Disable port(s) failed: %d", ret);
+			return ret;
+		}
 	}
-
 	stream->state = SDW_STREAM_DISABLED;
 
-	/* Program params */
-	ret = sdw_program_params(bus);
-	if (ret < 0) {
-		dev_err(bus->dev, "Program params failed: %d", ret);
-		return ret;
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+		/* Program params */
+		ret = sdw_program_params(bus);
+		if (ret < 0) {
+			dev_err(bus->dev, "Program params failed: %d", ret);
+			return ret;
+		}
 	}
 
 	return do_bank_switch(stream);
@@ -1422,43 +1650,46 @@ int sdw_disable_stream(struct sdw_stream_runtime *stream)
 		return -EINVAL;
 	}
 
-	mutex_lock(&stream->m_rt->bus->bus_lock);
+	sdw_acquire_bus_lock(stream);
 
 	ret = _sdw_disable_stream(stream);
 	if (ret < 0)
 		pr_err("Disable for stream:%s failed: %d", stream->name, ret);
 
-	mutex_unlock(&stream->m_rt->bus->bus_lock);
+	sdw_release_bus_lock(stream);
 	return ret;
 }
 EXPORT_SYMBOL(sdw_disable_stream);
 
 static int _sdw_deprepare_stream(struct sdw_stream_runtime *stream)
 {
-	struct sdw_master_runtime *m_rt = stream->m_rt;
-	struct sdw_bus *bus = m_rt->bus;
+	struct sdw_master_runtime *m_rt = NULL;
+	struct sdw_bus *bus = NULL;
 	int ret = 0;
 
-	/* De-prepare port(s) */
-	ret = sdw_prep_deprep_ports(m_rt, false);
-	if (ret < 0) {
-		dev_err(bus->dev, "De-prepare port(s) failed: %d", ret);
-		return ret;
-	}
+	list_for_each_entry(m_rt, &stream->master_list, stream_node) {
+		bus = m_rt->bus;
+		/* De-prepare port(s) */
+		ret = sdw_prep_deprep_ports(m_rt, false);
+		if (ret < 0) {
+			dev_err(bus->dev, "De-prepare port(s) failed: %d", ret);
+			return ret;
+		}
 
-	stream->state = SDW_STREAM_DEPREPARED;
+		/* TODO: Update this during Device-Device support */
+		bus->params.bandwidth -= m_rt->stream->params.rate *
+			m_rt->ch_count * m_rt->stream->params.bps;
 
-	/* TODO: Update this during Device-Device support */
-	bus->params.bandwidth -= m_rt->stream->params.rate *
-		m_rt->ch_count * m_rt->stream->params.bps;
+		/* Program params */
+		ret = sdw_program_params(bus);
+		if (ret < 0) {
+			dev_err(bus->dev, "Program params failed: %d", ret);
+			return ret;
+		}
 
-	/* Program params */
-	ret = sdw_program_params(bus);
-	if (ret < 0) {
-		dev_err(bus->dev, "Program params failed: %d", ret);
-		return ret;
 	}
 
+	stream->state = SDW_STREAM_DEPREPARED;
 	return do_bank_switch(stream);
 }
 
@@ -1478,13 +1709,12 @@ int sdw_deprepare_stream(struct sdw_stream_runtime *stream)
 		return -EINVAL;
 	}
 
-	mutex_lock(&stream->m_rt->bus->bus_lock);
-
+	sdw_acquire_bus_lock(stream);
 	ret = _sdw_deprepare_stream(stream);
 	if (ret < 0)
 		pr_err("De-prepare for stream:%d failed: %d", ret, ret);
 
-	mutex_unlock(&stream->m_rt->bus->bus_lock);
+	sdw_release_bus_lock(stream);
 	return ret;
 }
 EXPORT_SYMBOL(sdw_deprepare_stream);
diff --git a/drivers/thunderbolt/cap.c b/drivers/thunderbolt/cap.c
index c2277b8ee88d..9553305c63ea 100644
--- a/drivers/thunderbolt/cap.c
+++ b/drivers/thunderbolt/cap.c
@@ -1,8 +1,9 @@
 // SPDX-License-Identifier: GPL-2.0
 /*
- * Thunderbolt Cactus Ridge driver - capabilities lookup
+ * Thunderbolt driver - capabilities lookup
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #include <linux/slab.h>
diff --git a/drivers/thunderbolt/ctl.c b/drivers/thunderbolt/ctl.c
index 37a7f4c735d0..73b386de4d15 100644
--- a/drivers/thunderbolt/ctl.c
+++ b/drivers/thunderbolt/ctl.c
@@ -1,8 +1,9 @@
 // SPDX-License-Identifier: GPL-2.0
 /*
- * Thunderbolt Cactus Ridge driver - control channel and configuration commands
+ * Thunderbolt driver - control channel and configuration commands
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #include <linux/crc32.h>
@@ -631,7 +632,7 @@ struct tb_ctl *tb_ctl_alloc(struct tb_nhi *nhi, event_cb cb, void *cb_data)
 		ctl->rx_packets[i]->frame.callback = tb_ctl_rx_callback;
 	}
 
-	tb_ctl_info(ctl, "control channel created\n");
+	tb_ctl_dbg(ctl, "control channel created\n");
 	return ctl;
 err:
 	tb_ctl_free(ctl);
@@ -662,8 +663,7 @@ void tb_ctl_free(struct tb_ctl *ctl)
 		tb_ctl_pkg_free(ctl->rx_packets[i]);
 
 
-	if (ctl->frame_pool)
-		dma_pool_destroy(ctl->frame_pool);
+	dma_pool_destroy(ctl->frame_pool);
 	kfree(ctl);
 }
 
@@ -673,7 +673,7 @@ void tb_ctl_free(struct tb_ctl *ctl)
 void tb_ctl_start(struct tb_ctl *ctl)
 {
 	int i;
-	tb_ctl_info(ctl, "control channel starting...\n");
+	tb_ctl_dbg(ctl, "control channel starting...\n");
 	tb_ring_start(ctl->tx); /* is used to ack hotplug packets, start first */
 	tb_ring_start(ctl->rx);
 	for (i = 0; i < TB_CTL_RX_PKG_COUNT; i++)
@@ -702,7 +702,7 @@ void tb_ctl_stop(struct tb_ctl *ctl)
 	if (!list_empty(&ctl->request_queue))
 		tb_ctl_WARN(ctl, "dangling request in request_queue\n");
 	INIT_LIST_HEAD(&ctl->request_queue);
-	tb_ctl_info(ctl, "control channel stopped\n");
+	tb_ctl_dbg(ctl, "control channel stopped\n");
 }
 
 /* public interface, commands */
diff --git a/drivers/thunderbolt/ctl.h b/drivers/thunderbolt/ctl.h
index 3062e0b5f71e..2f1a1e111110 100644
--- a/drivers/thunderbolt/ctl.h
+++ b/drivers/thunderbolt/ctl.h
@@ -1,8 +1,9 @@
 /* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Thunderbolt Cactus Ridge driver - control channel and configuration commands
+ * Thunderbolt driver - control channel and configuration commands
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #ifndef _TB_CFG
diff --git a/drivers/thunderbolt/dma_port.c b/drivers/thunderbolt/dma_port.c
index f2701194f810..847dd07a7b17 100644
--- a/drivers/thunderbolt/dma_port.c
+++ b/drivers/thunderbolt/dma_port.c
@@ -1,13 +1,10 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Thunderbolt DMA configuration based mailbox support
  *
  * Copyright (C) 2017, Intel Corporation
  * Authors: Michael Jamet <michael.jamet@intel.com>
  *          Mika Westerberg <mika.westerberg@linux.intel.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
  */
 
 #include <linux/delay.h>
diff --git a/drivers/thunderbolt/dma_port.h b/drivers/thunderbolt/dma_port.h
index c4a69e0fbff7..7deadd97ce31 100644
--- a/drivers/thunderbolt/dma_port.h
+++ b/drivers/thunderbolt/dma_port.h
@@ -1,13 +1,10 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Thunderbolt DMA configuration based mailbox support
  *
  * Copyright (C) 2017, Intel Corporation
  * Authors: Michael Jamet <michael.jamet@intel.com>
  *          Mika Westerberg <mika.westerberg@linux.intel.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
  */
 
 #ifndef DMA_PORT_H_
diff --git a/drivers/thunderbolt/domain.c b/drivers/thunderbolt/domain.c
index 092381e2accf..93e562f18d40 100644
--- a/drivers/thunderbolt/domain.c
+++ b/drivers/thunderbolt/domain.c
@@ -1,12 +1,9 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Thunderbolt bus support
  *
  * Copyright (C) 2017, Intel Corporation
- * Author:  Mika Westerberg <mika.westerberg@linux.intel.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
+ * Author: Mika Westerberg <mika.westerberg@linux.intel.com>
  */
 
 #include <linux/device.h>
diff --git a/drivers/thunderbolt/eeprom.c b/drivers/thunderbolt/eeprom.c
index 3e8caf22c294..81e8ac4c5805 100644
--- a/drivers/thunderbolt/eeprom.c
+++ b/drivers/thunderbolt/eeprom.c
@@ -1,8 +1,9 @@
 // SPDX-License-Identifier: GPL-2.0
 /*
- * Thunderbolt Cactus Ridge driver - eeprom access
+ * Thunderbolt driver - eeprom access
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #include <linux/crc32.h>
@@ -540,7 +541,7 @@ int tb_drom_read(struct tb_switch *sw)
 		return res;
 	size &= 0x3ff;
 	size += TB_DROM_DATA_START;
-	tb_sw_info(sw, "reading drom (length: %#x)\n", size);
+	tb_sw_dbg(sw, "reading drom (length: %#x)\n", size);
 	if (size < sizeof(*header)) {
 		tb_sw_warn(sw, "drom too small, aborting\n");
 		return -EIO;
diff --git a/drivers/thunderbolt/icm.c b/drivers/thunderbolt/icm.c
index 28fc4ce75edb..e3fc920af682 100644
--- a/drivers/thunderbolt/icm.c
+++ b/drivers/thunderbolt/icm.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Internal Thunderbolt Connection Manager. This is a firmware running on
  * the Thunderbolt host controller performing most of the low-level
@@ -6,10 +7,6 @@
  * Copyright (C) 2017, Intel Corporation
  * Authors: Michael Jamet <michael.jamet@intel.com>
  *          Mika Westerberg <mika.westerberg@linux.intel.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
  */
 
 #include <linux/delay.h>
diff --git a/drivers/thunderbolt/nhi.c b/drivers/thunderbolt/nhi.c
index 5cd6bdfa068f..9aa44f9762a3 100644
--- a/drivers/thunderbolt/nhi.c
+++ b/drivers/thunderbolt/nhi.c
@@ -1,10 +1,11 @@
 /*
- * Thunderbolt Cactus Ridge driver - NHI driver
+ * Thunderbolt driver - NHI driver
  *
  * The NHI (native host interface) is the pci device that allows us to send and
  * receive frames from the thunderbolt bus.
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #include <linux/pm_runtime.h>
@@ -95,9 +96,9 @@ static void ring_interrupt_active(struct tb_ring *ring, bool active)
 	else
 		new = old & ~mask;
 
-	dev_info(&ring->nhi->pdev->dev,
-		 "%s interrupt at register %#x bit %d (%#x -> %#x)\n",
-		 active ? "enabling" : "disabling", reg, bit, old, new);
+	dev_dbg(&ring->nhi->pdev->dev,
+		"%s interrupt at register %#x bit %d (%#x -> %#x)\n",
+		active ? "enabling" : "disabling", reg, bit, old, new);
 
 	if (new == old)
 		dev_WARN(&ring->nhi->pdev->dev,
@@ -476,8 +477,9 @@ static struct tb_ring *tb_ring_alloc(struct tb_nhi *nhi, u32 hop, int size,
 				     void *poll_data)
 {
 	struct tb_ring *ring = NULL;
-	dev_info(&nhi->pdev->dev, "allocating %s ring %d of size %d\n",
-		 transmit ? "TX" : "RX", hop, size);
+
+	dev_dbg(&nhi->pdev->dev, "allocating %s ring %d of size %d\n",
+		transmit ? "TX" : "RX", hop, size);
 
 	/* Tx Ring 2 is reserved for E2E workaround */
 	if (transmit && hop == RING_E2E_UNUSED_HOPID)
@@ -585,8 +587,8 @@ void tb_ring_start(struct tb_ring *ring)
 		dev_WARN(&ring->nhi->pdev->dev, "ring already started\n");
 		goto err;
 	}
-	dev_info(&ring->nhi->pdev->dev, "starting %s %d\n",
-		 RING_TYPE(ring), ring->hop);
+	dev_dbg(&ring->nhi->pdev->dev, "starting %s %d\n",
+		RING_TYPE(ring), ring->hop);
 
 	if (ring->flags & RING_FLAG_FRAME) {
 		/* Means 4096 */
@@ -647,8 +649,8 @@ void tb_ring_stop(struct tb_ring *ring)
 {
 	spin_lock_irq(&ring->nhi->lock);
 	spin_lock(&ring->lock);
-	dev_info(&ring->nhi->pdev->dev, "stopping %s %d\n",
-		 RING_TYPE(ring), ring->hop);
+	dev_dbg(&ring->nhi->pdev->dev, "stopping %s %d\n",
+		RING_TYPE(ring), ring->hop);
 	if (ring->nhi->going_away)
 		goto err;
 	if (!ring->running) {
@@ -716,10 +718,8 @@ void tb_ring_free(struct tb_ring *ring)
 	ring->descriptors_dma = 0;
 
 
-	dev_info(&ring->nhi->pdev->dev,
-		 "freeing %s %d\n",
-		 RING_TYPE(ring),
-		 ring->hop);
+	dev_dbg(&ring->nhi->pdev->dev, "freeing %s %d\n", RING_TYPE(ring),
+		ring->hop);
 
 	/**
 	 * ring->work can no longer be scheduled (it is scheduled only
@@ -931,7 +931,8 @@ static int nhi_runtime_resume(struct device *dev)
 static void nhi_shutdown(struct tb_nhi *nhi)
 {
 	int i;
-	dev_info(&nhi->pdev->dev, "shutdown\n");
+
+	dev_dbg(&nhi->pdev->dev, "shutdown\n");
 
 	for (i = 0; i < nhi->hop_count; i++) {
 		if (nhi->tx_rings[i])
@@ -1059,7 +1060,7 @@ static int nhi_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		return -ENODEV;
 	}
 
-	dev_info(&nhi->pdev->dev, "NHI initialized, starting thunderbolt\n");
+	dev_dbg(&nhi->pdev->dev, "NHI initialized, starting thunderbolt\n");
 
 	res = tb_domain_add(tb);
 	if (res) {
diff --git a/drivers/thunderbolt/nhi.h b/drivers/thunderbolt/nhi.h
index 1696a4560948..1b5d47ecd3ed 100644
--- a/drivers/thunderbolt/nhi.h
+++ b/drivers/thunderbolt/nhi.h
@@ -1,8 +1,9 @@
 /* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Thunderbolt Cactus Ridge driver - NHI driver
+ * Thunderbolt driver - NHI driver
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #ifndef DSL3510_H_
diff --git a/drivers/thunderbolt/nhi_regs.h b/drivers/thunderbolt/nhi_regs.h
index b3e49d19c01e..a60bd98c1d04 100644
--- a/drivers/thunderbolt/nhi_regs.h
+++ b/drivers/thunderbolt/nhi_regs.h
@@ -3,6 +3,7 @@
  * Thunderbolt driver - NHI registers
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #ifndef NHI_REGS_H_
diff --git a/drivers/thunderbolt/path.c b/drivers/thunderbolt/path.c
index ff49ad880bfd..a11956522bac 100644
--- a/drivers/thunderbolt/path.c
+++ b/drivers/thunderbolt/path.c
@@ -13,19 +13,19 @@
 
 static void tb_dump_hop(struct tb_port *port, struct tb_regs_hop *hop)
 {
-	tb_port_info(port, " Hop through port %d to hop %d (%s)\n",
-		     hop->out_port, hop->next_hop,
-		     hop->enable ? "enabled" : "disabled");
-	tb_port_info(port, "  Weight: %d Priority: %d Credits: %d Drop: %d\n",
-		     hop->weight, hop->priority,
-		     hop->initial_credits, hop->drop_packages);
-	tb_port_info(port, "   Counter enabled: %d Counter index: %d\n",
-		     hop->counter_enable, hop->counter);
-	tb_port_info(port, "  Flow Control (In/Eg): %d/%d Shared Buffer (In/Eg): %d/%d\n",
-		     hop->ingress_fc, hop->egress_fc,
-		     hop->ingress_shared_buffer, hop->egress_shared_buffer);
-	tb_port_info(port, "  Unknown1: %#x Unknown2: %#x Unknown3: %#x\n",
-		     hop->unknown1, hop->unknown2, hop->unknown3);
+	tb_port_dbg(port, " Hop through port %d to hop %d (%s)\n",
+		    hop->out_port, hop->next_hop,
+		    hop->enable ? "enabled" : "disabled");
+	tb_port_dbg(port, "  Weight: %d Priority: %d Credits: %d Drop: %d\n",
+		    hop->weight, hop->priority,
+		    hop->initial_credits, hop->drop_packages);
+	tb_port_dbg(port, "   Counter enabled: %d Counter index: %d\n",
+		    hop->counter_enable, hop->counter);
+	tb_port_dbg(port, "  Flow Control (In/Eg): %d/%d Shared Buffer (In/Eg): %d/%d\n",
+		    hop->ingress_fc, hop->egress_fc,
+		    hop->ingress_shared_buffer, hop->egress_shared_buffer);
+	tb_port_dbg(port, "  Unknown1: %#x Unknown2: %#x Unknown3: %#x\n",
+		    hop->unknown1, hop->unknown2, hop->unknown3);
 }
 
 /**
diff --git a/drivers/thunderbolt/property.c b/drivers/thunderbolt/property.c
index 8fe913a95b4a..b2f0d6386cee 100644
--- a/drivers/thunderbolt/property.c
+++ b/drivers/thunderbolt/property.c
@@ -1,13 +1,10 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Thunderbolt XDomain property support
  *
  * Copyright (C) 2017, Intel Corporation
  * Authors: Michael Jamet <michael.jamet@intel.com>
  *          Mika Westerberg <mika.westerberg@linux.intel.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
  */
 
 #include <linux/err.h>
diff --git a/drivers/thunderbolt/switch.c b/drivers/thunderbolt/switch.c
index 7442bc4c6433..52ff854f0d6c 100644
--- a/drivers/thunderbolt/switch.c
+++ b/drivers/thunderbolt/switch.c
@@ -1,8 +1,9 @@
 // SPDX-License-Identifier: GPL-2.0
 /*
- * Thunderbolt Cactus Ridge driver - switch/port utility functions
+ * Thunderbolt driver - switch/port utility functions
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #include <linux/delay.h>
@@ -436,15 +437,15 @@ static const char *tb_port_type(struct tb_regs_port_header *port)
 
 static void tb_dump_port(struct tb *tb, struct tb_regs_port_header *port)
 {
-	tb_info(tb,
-		" Port %d: %x:%x (Revision: %d, TB Version: %d, Type: %s (%#x))\n",
-		port->port_number, port->vendor_id, port->device_id,
-		port->revision, port->thunderbolt_version, tb_port_type(port),
-		port->type);
-	tb_info(tb, "  Max hop id (in/out): %d/%d\n",
-		port->max_in_hop_id, port->max_out_hop_id);
-	tb_info(tb, "  Max counters: %d\n", port->max_counters);
-	tb_info(tb, "  NFC Credits: %#x\n", port->nfc_credits);
+	tb_dbg(tb,
+	       " Port %d: %x:%x (Revision: %d, TB Version: %d, Type: %s (%#x))\n",
+	       port->port_number, port->vendor_id, port->device_id,
+	       port->revision, port->thunderbolt_version, tb_port_type(port),
+	       port->type);
+	tb_dbg(tb, "  Max hop id (in/out): %d/%d\n",
+	       port->max_in_hop_id, port->max_out_hop_id);
+	tb_dbg(tb, "  Max counters: %d\n", port->max_counters);
+	tb_dbg(tb, "  NFC Credits: %#x\n", port->nfc_credits);
 }
 
 /**
@@ -605,20 +606,18 @@ static int tb_init_port(struct tb_port *port)
 
 static void tb_dump_switch(struct tb *tb, struct tb_regs_switch_header *sw)
 {
-	tb_info(tb,
-		" Switch: %x:%x (Revision: %d, TB Version: %d)\n",
-		sw->vendor_id, sw->device_id, sw->revision,
-		sw->thunderbolt_version);
-	tb_info(tb, "  Max Port Number: %d\n", sw->max_port_number);
-	tb_info(tb, "  Config:\n");
-	tb_info(tb,
+	tb_dbg(tb, " Switch: %x:%x (Revision: %d, TB Version: %d)\n",
+	       sw->vendor_id, sw->device_id, sw->revision,
+	       sw->thunderbolt_version);
+	tb_dbg(tb, "  Max Port Number: %d\n", sw->max_port_number);
+	tb_dbg(tb, "  Config:\n");
+	tb_dbg(tb,
 		"   Upstream Port Number: %d Depth: %d Route String: %#llx Enabled: %d, PlugEventsDelay: %dms\n",
-		sw->upstream_port_number, sw->depth,
-		(((u64) sw->route_hi) << 32) | sw->route_lo,
-		sw->enabled, sw->plug_events_delay);
-	tb_info(tb,
-		"   unknown1: %#x unknown4: %#x\n",
-		sw->__unknown1, sw->__unknown4);
+	       sw->upstream_port_number, sw->depth,
+	       (((u64) sw->route_hi) << 32) | sw->route_lo,
+	       sw->enabled, sw->plug_events_delay);
+	tb_dbg(tb, "   unknown1: %#x unknown4: %#x\n",
+	       sw->__unknown1, sw->__unknown4);
 }
 
 /**
@@ -634,7 +633,7 @@ int tb_switch_reset(struct tb *tb, u64 route)
 		header.route_lo = route,
 		header.enabled = true,
 	};
-	tb_info(tb, "resetting switch at %llx\n", route);
+	tb_dbg(tb, "resetting switch at %llx\n", route);
 	res.err = tb_cfg_write(tb->ctl, ((u32 *) &header) + 2, route,
 			0, 2, 2, 2);
 	if (res.err)
@@ -1139,7 +1138,7 @@ struct tb_switch *tb_switch_alloc(struct tb *tb, struct device *parent,
 	if (tb_cfg_read(tb->ctl, &sw->config, route, 0, TB_CFG_SWITCH, 0, 5))
 		goto err_free_sw_ports;
 
-	tb_info(tb, "current switch config:\n");
+	tb_dbg(tb, "current switch config:\n");
 	tb_dump_switch(tb, &sw->config);
 
 	/* configure switch */
@@ -1246,9 +1245,8 @@ int tb_switch_configure(struct tb_switch *sw)
 	int ret;
 
 	route = tb_route(sw);
-	tb_info(tb,
-		"initializing Switch at %#llx (depth: %d, up port: %d)\n",
-		route, tb_route_length(route), sw->config.upstream_port_number);
+	tb_dbg(tb, "initializing Switch at %#llx (depth: %d, up port: %d)\n",
+	       route, tb_route_length(route), sw->config.upstream_port_number);
 
 	if (sw->config.vendor_id != PCI_VENDOR_ID_INTEL)
 		tb_sw_warn(sw, "unknown switch vendor id %#x\n",
@@ -1386,13 +1384,13 @@ int tb_switch_add(struct tb_switch *sw)
 			tb_sw_warn(sw, "tb_eeprom_read_rom failed\n");
 			return ret;
 		}
-		tb_sw_info(sw, "uid: %#llx\n", sw->uid);
+		tb_sw_dbg(sw, "uid: %#llx\n", sw->uid);
 
 		tb_switch_set_uuid(sw);
 
 		for (i = 0; i <= sw->config.max_port_number; i++) {
 			if (sw->ports[i].disabled) {
-				tb_port_info(&sw->ports[i], "disabled by eeprom\n");
+				tb_port_dbg(&sw->ports[i], "disabled by eeprom\n");
 				continue;
 			}
 			ret = tb_init_port(&sw->ports[i]);
@@ -1405,6 +1403,14 @@ int tb_switch_add(struct tb_switch *sw)
 	if (ret)
 		return ret;
 
+	if (tb_route(sw)) {
+		dev_info(&sw->dev, "new device found, vendor=%#x device=%#x\n",
+			 sw->vendor, sw->device);
+		if (sw->vendor_name && sw->device_name)
+			dev_info(&sw->dev, "%s %s\n", sw->vendor_name,
+				 sw->device_name);
+	}
+
 	ret = tb_switch_nvm_add(sw);
 	if (ret) {
 		device_del(&sw->dev);
@@ -1456,6 +1462,9 @@ void tb_switch_remove(struct tb_switch *sw)
 		tb_plug_events_active(sw, false);
 
 	tb_switch_nvm_remove(sw);
+
+	if (tb_route(sw))
+		dev_info(&sw->dev, "device disconnected\n");
 	device_unregister(&sw->dev);
 }
 
@@ -1483,7 +1492,7 @@ void tb_sw_set_unplugged(struct tb_switch *sw)
 int tb_switch_resume(struct tb_switch *sw)
 {
 	int i, err;
-	tb_sw_info(sw, "resuming switch\n");
+	tb_sw_dbg(sw, "resuming switch\n");
 
 	/*
 	 * Check for UID of the connected switches except for root
diff --git a/drivers/thunderbolt/tb.c b/drivers/thunderbolt/tb.c
index 1424581fd9af..30e02c716f6c 100644
--- a/drivers/thunderbolt/tb.c
+++ b/drivers/thunderbolt/tb.c
@@ -404,10 +404,10 @@ static int tb_suspend_noirq(struct tb *tb)
 {
 	struct tb_cm *tcm = tb_priv(tb);
 
-	tb_info(tb, "suspending...\n");
+	tb_dbg(tb, "suspending...\n");
 	tb_switch_suspend(tb->root_switch);
 	tcm->hotplug_active = false; /* signal tb_handle_hotplug to quit */
-	tb_info(tb, "suspend finished\n");
+	tb_dbg(tb, "suspend finished\n");
 
 	return 0;
 }
@@ -417,7 +417,7 @@ static int tb_resume_noirq(struct tb *tb)
 	struct tb_cm *tcm = tb_priv(tb);
 	struct tb_pci_tunnel *tunnel, *n;
 
-	tb_info(tb, "resuming...\n");
+	tb_dbg(tb, "resuming...\n");
 
 	/* remove any pci devices the firmware might have setup */
 	tb_switch_reset(tb, 0);
@@ -432,12 +432,12 @@ static int tb_resume_noirq(struct tb *tb)
 		 * the pcie links need some time to get going.
 		 * 100ms works for me...
 		 */
-		tb_info(tb, "tunnels restarted, sleeping for 100ms\n");
+		tb_dbg(tb, "tunnels restarted, sleeping for 100ms\n");
 		msleep(100);
 	}
 	 /* Allow tb_handle_hotplug to progress events */
 	tcm->hotplug_active = true;
-	tb_info(tb, "resume finished\n");
+	tb_dbg(tb, "resume finished\n");
 
 	return 0;
 }
diff --git a/drivers/thunderbolt/tb.h b/drivers/thunderbolt/tb.h
index 5067d69d0501..52584c4003e3 100644
--- a/drivers/thunderbolt/tb.h
+++ b/drivers/thunderbolt/tb.h
@@ -1,8 +1,9 @@
 /* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Thunderbolt Cactus Ridge driver - bus logic (NHI independent)
+ * Thunderbolt driver - bus logic (NHI independent)
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #ifndef TB_H_
@@ -327,7 +328,7 @@ static inline int tb_port_write(struct tb_port *port, const void *buffer,
 #define tb_WARN(tb, fmt, arg...) dev_WARN(&(tb)->nhi->pdev->dev, fmt, ## arg)
 #define tb_warn(tb, fmt, arg...) dev_warn(&(tb)->nhi->pdev->dev, fmt, ## arg)
 #define tb_info(tb, fmt, arg...) dev_info(&(tb)->nhi->pdev->dev, fmt, ## arg)
-
+#define tb_dbg(tb, fmt, arg...) dev_dbg(&(tb)->nhi->pdev->dev, fmt, ## arg)
 
 #define __TB_SW_PRINT(level, sw, fmt, arg...)           \
 	do {                                            \
@@ -338,7 +339,7 @@ static inline int tb_port_write(struct tb_port *port, const void *buffer,
 #define tb_sw_WARN(sw, fmt, arg...) __TB_SW_PRINT(tb_WARN, sw, fmt, ##arg)
 #define tb_sw_warn(sw, fmt, arg...) __TB_SW_PRINT(tb_warn, sw, fmt, ##arg)
 #define tb_sw_info(sw, fmt, arg...) __TB_SW_PRINT(tb_info, sw, fmt, ##arg)
-
+#define tb_sw_dbg(sw, fmt, arg...) __TB_SW_PRINT(tb_dbg, sw, fmt, ##arg)
 
 #define __TB_PORT_PRINT(level, _port, fmt, arg...)                      \
 	do {                                                            \
@@ -352,6 +353,8 @@ static inline int tb_port_write(struct tb_port *port, const void *buffer,
 	__TB_PORT_PRINT(tb_warn, port, fmt, ##arg)
 #define tb_port_info(port, fmt, arg...) \
 	__TB_PORT_PRINT(tb_info, port, fmt, ##arg)
+#define tb_port_dbg(port, fmt, arg...) \
+	__TB_PORT_PRINT(tb_dbg, port, fmt, ##arg)
 
 struct tb *icm_probe(struct tb_nhi *nhi);
 struct tb *tb_probe(struct tb_nhi *nhi);
diff --git a/drivers/thunderbolt/tb_msgs.h b/drivers/thunderbolt/tb_msgs.h
index 2487e162c885..02c84aa3d018 100644
--- a/drivers/thunderbolt/tb_msgs.h
+++ b/drivers/thunderbolt/tb_msgs.h
@@ -1,12 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Thunderbolt control channel messages
  *
  * Copyright (C) 2014 Andreas Noever <andreas.noever@gmail.com>
  * Copyright (C) 2017, Intel Corporation
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
  */
 
 #ifndef _TB_MSGS
diff --git a/drivers/thunderbolt/tb_regs.h b/drivers/thunderbolt/tb_regs.h
index 693b0353c3fe..6f1ff04ee195 100644
--- a/drivers/thunderbolt/tb_regs.h
+++ b/drivers/thunderbolt/tb_regs.h
@@ -1,12 +1,13 @@
 /* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Thunderbolt Cactus Ridge driver - Port/Switch config area registers
+ * Thunderbolt driver - Port/Switch config area registers
  *
  * Every thunderbolt device consists (logically) of a switch with multiple
  * ports. Every port contains up to four config regions (HOPS, PORT, SWITCH,
  * COUNTERS) which are used to configure the device.
  *
  * Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
+ * Copyright (C) 2018, Intel Corporation
  */
 
 #ifndef _TB_REGS
diff --git a/drivers/thunderbolt/xdomain.c b/drivers/thunderbolt/xdomain.c
index db8bece63327..e27dd8beb94b 100644
--- a/drivers/thunderbolt/xdomain.c
+++ b/drivers/thunderbolt/xdomain.c
@@ -1,13 +1,10 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Thunderbolt XDomain discovery protocol support
  *
  * Copyright (C) 2017, Intel Corporation
  * Authors: Michael Jamet <michael.jamet@intel.com>
  *          Mika Westerberg <mika.westerberg@linux.intel.com>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
  */
 
 #include <linux/device.h>
diff --git a/drivers/uio/uio.c b/drivers/uio/uio.c
index 70a7981b94b3..85644669fbe7 100644
--- a/drivers/uio/uio.c
+++ b/drivers/uio/uio.c
@@ -274,6 +274,8 @@ static struct class uio_class = {
 	.dev_groups = uio_groups,
 };
 
+static bool uio_class_registered;
+
 /*
  * device functions
  */
@@ -668,7 +670,7 @@ static vm_fault_t uio_vma_fault(struct vm_fault *vmf)
 	struct page *page;
 	unsigned long offset;
 	void *addr;
-	int ret = 0;
+	vm_fault_t ret = 0;
 	int mi;
 
 	mutex_lock(&idev->info_lock);
@@ -736,7 +738,8 @@ static int uio_mmap_physical(struct vm_area_struct *vma)
 		return -EINVAL;
 
 	vma->vm_ops = &uio_physical_vm_ops;
-	vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+	if (idev->info->mem[mi].memtype == UIO_MEM_PHYS)
+		vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
 
 	/*
 	 * We cannot use the vm_iomap_memory() helper here,
@@ -793,18 +796,19 @@ static int uio_mmap(struct file *filep, struct vm_area_struct *vma)
 	}
 
 	switch (idev->info->mem[mi].memtype) {
-		case UIO_MEM_PHYS:
-			ret = uio_mmap_physical(vma);
-			break;
-		case UIO_MEM_LOGICAL:
-		case UIO_MEM_VIRTUAL:
-			ret = uio_mmap_logical(vma);
-			break;
-		default:
-			ret = -EINVAL;
+	case UIO_MEM_IOVA:
+	case UIO_MEM_PHYS:
+		ret = uio_mmap_physical(vma);
+		break;
+	case UIO_MEM_LOGICAL:
+	case UIO_MEM_VIRTUAL:
+		ret = uio_mmap_logical(vma);
+		break;
+	default:
+		ret = -EINVAL;
 	}
 
-out:
+ out:
 	mutex_unlock(&idev->info_lock);
 	return ret;
 }
@@ -876,6 +880,9 @@ static int init_uio_class(void)
 		printk(KERN_ERR "class_register failed for uio\n");
 		goto err_class_register;
 	}
+
+	uio_class_registered = true;
+
 	return 0;
 
 err_class_register:
@@ -886,6 +893,7 @@ exit:
 
 static void release_uio_class(void)
 {
+	uio_class_registered = false;
 	class_unregister(&uio_class);
 	uio_major_cleanup();
 }
@@ -912,6 +920,9 @@ int __uio_register_device(struct module *owner,
 	struct uio_device *idev;
 	int ret = 0;
 
+	if (!uio_class_registered)
+		return -EPROBE_DEFER;
+
 	if (!parent || !info || !info->name || !info->version)
 		return -EINVAL;
 
diff --git a/drivers/uio/uio_dmem_genirq.c b/drivers/uio/uio_dmem_genirq.c
index e1134a4d97f3..003badaef5f3 100644
--- a/drivers/uio/uio_dmem_genirq.c
+++ b/drivers/uio/uio_dmem_genirq.c
@@ -163,7 +163,8 @@ static int uio_dmem_genirq_probe(struct platform_device *pdev)
 			dev_err(&pdev->dev, "unable to kmalloc\n");
 			goto bad2;
 		}
-		uioinfo->name = pdev->dev.of_node->name;
+		uioinfo->name = devm_kasprintf(&pdev->dev, GFP_KERNEL, "%pOFn",
+					       pdev->dev.of_node);
 		uioinfo->version = "devicetree";
 
 		/* Multiple IRQs are not supported */
diff --git a/drivers/uio/uio_fsl_elbc_gpcm.c b/drivers/uio/uio_fsl_elbc_gpcm.c
index bbc17effae5e..9cc37fe07d35 100644
--- a/drivers/uio/uio_fsl_elbc_gpcm.c
+++ b/drivers/uio/uio_fsl_elbc_gpcm.c
@@ -382,8 +382,7 @@ static int uio_fsl_elbc_gpcm_probe(struct platform_device *pdev)
 	}
 
 	/* set all UIO data */
-	if (node->name)
-		info->mem[0].name = kstrdup(node->name, GFP_KERNEL);
+	info->mem[0].name = kasprintf(GFP_KERNEL, "%pOFn", node);
 	info->mem[0].addr = res.start;
 	info->mem[0].size = resource_size(&res);
 	info->mem[0].memtype = UIO_MEM_PHYS;
diff --git a/drivers/uio/uio_hv_generic.c b/drivers/uio/uio_hv_generic.c
index e401be8321ab..c2493d011225 100644
--- a/drivers/uio/uio_hv_generic.c
+++ b/drivers/uio/uio_hv_generic.c
@@ -17,7 +17,6 @@
  * # echo -n "ed963694-e847-4b2a-85af-bc9cfc11d6f3" \
  *    > /sys/bus/vmbus/drivers/uio_hv_generic/bind
  */
-#define DEBUG 1
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
 #include <linux/device.h>
@@ -33,13 +32,13 @@
 
 #include "../hv/hyperv_vmbus.h"
 
-#define DRIVER_VERSION	"0.02.0"
+#define DRIVER_VERSION	"0.02.1"
 #define DRIVER_AUTHOR	"Stephen Hemminger <sthemmin at microsoft.com>"
 #define DRIVER_DESC	"Generic UIO driver for VMBus devices"
 
 #define HV_RING_SIZE	 512	/* pages */
-#define SEND_BUFFER_SIZE (15 * 1024 * 1024)
-#define RECV_BUFFER_SIZE (15 * 1024 * 1024)
+#define SEND_BUFFER_SIZE (16 * 1024 * 1024)
+#define RECV_BUFFER_SIZE (31 * 1024 * 1024)
 
 /*
  * List of resources to be mapped to user space
@@ -56,6 +55,7 @@ enum hv_uio_map {
 struct hv_uio_private_data {
 	struct uio_info info;
 	struct hv_device *device;
+	atomic_t refcnt;
 
 	void	*recv_buf;
 	u32	recv_gpadl;
@@ -129,13 +129,12 @@ static int hv_uio_ring_mmap(struct file *filp, struct kobject *kobj,
 {
 	struct vmbus_channel *channel
 		= container_of(kobj, struct vmbus_channel, kobj);
-	struct hv_device *dev = channel->primary_channel->device_obj;
-	u16 q_idx = channel->offermsg.offer.sub_channel_index;
+	void *ring_buffer = page_address(channel->ringbuffer_page);
 
-	dev_dbg(&dev->device, "mmap channel %u pages %#lx at %#lx\n",
-		q_idx, vma_pages(vma), vma->vm_pgoff);
+	if (channel->state != CHANNEL_OPENED_STATE)
+		return -ENODEV;
 
-	return vm_iomap_memory(vma, virt_to_phys(channel->ringbuffer_pages),
+	return vm_iomap_memory(vma, virt_to_phys(ring_buffer),
 			       channel->ringbuffer_pagecount << PAGE_SHIFT);
 }
 
@@ -176,58 +175,104 @@ hv_uio_new_channel(struct vmbus_channel *new_sc)
 	}
 }
 
+/* free the reserved buffers for send and receive */
 static void
 hv_uio_cleanup(struct hv_device *dev, struct hv_uio_private_data *pdata)
 {
-	if (pdata->send_gpadl)
+	if (pdata->send_gpadl) {
 		vmbus_teardown_gpadl(dev->channel, pdata->send_gpadl);
-	vfree(pdata->send_buf);
+		pdata->send_gpadl = 0;
+		vfree(pdata->send_buf);
+	}
 
-	if (pdata->recv_gpadl)
+	if (pdata->recv_gpadl) {
 		vmbus_teardown_gpadl(dev->channel, pdata->recv_gpadl);
-	vfree(pdata->recv_buf);
+		pdata->recv_gpadl = 0;
+		vfree(pdata->recv_buf);
+	}
+}
+
+/* VMBus primary channel is opened on first use */
+static int
+hv_uio_open(struct uio_info *info, struct inode *inode)
+{
+	struct hv_uio_private_data *pdata
+		= container_of(info, struct hv_uio_private_data, info);
+	struct hv_device *dev = pdata->device;
+	int ret;
+
+	if (atomic_inc_return(&pdata->refcnt) != 1)
+		return 0;
+
+	ret = vmbus_connect_ring(dev->channel,
+				 hv_uio_channel_cb, dev->channel);
+
+	if (ret == 0)
+		dev->channel->inbound.ring_buffer->interrupt_mask = 1;
+	else
+		atomic_dec(&pdata->refcnt);
+
+	return ret;
+}
+
+/* VMBus primary channel is closed on last close */
+static int
+hv_uio_release(struct uio_info *info, struct inode *inode)
+{
+	struct hv_uio_private_data *pdata
+		= container_of(info, struct hv_uio_private_data, info);
+	struct hv_device *dev = pdata->device;
+	int ret = 0;
+
+	if (atomic_dec_and_test(&pdata->refcnt))
+		ret = vmbus_disconnect_ring(dev->channel);
+
+	return ret;
 }
 
 static int
 hv_uio_probe(struct hv_device *dev,
 	     const struct hv_vmbus_device_id *dev_id)
 {
+	struct vmbus_channel *channel = dev->channel;
 	struct hv_uio_private_data *pdata;
+	void *ring_buffer;
 	int ret;
 
+	/* Communicating with host has to be via shared memory not hypercall */
+	if (!channel->offermsg.monitor_allocated) {
+		dev_err(&dev->device, "vmbus channel requires hypercall\n");
+		return -ENOTSUPP;
+	}
+
 	pdata = kzalloc(sizeof(*pdata), GFP_KERNEL);
 	if (!pdata)
 		return -ENOMEM;
 
-	ret = vmbus_open(dev->channel, HV_RING_SIZE * PAGE_SIZE,
-			 HV_RING_SIZE * PAGE_SIZE, NULL, 0,
-			 hv_uio_channel_cb, dev->channel);
+	ret = vmbus_alloc_ring(channel, HV_RING_SIZE * PAGE_SIZE,
+			       HV_RING_SIZE * PAGE_SIZE);
 	if (ret)
 		goto fail;
 
-	/* Communicating with host has to be via shared memory not hypercall */
-	if (!dev->channel->offermsg.monitor_allocated) {
-		dev_err(&dev->device, "vmbus channel requires hypercall\n");
-		ret = -ENOTSUPP;
-		goto fail_close;
-	}
-
-	dev->channel->inbound.ring_buffer->interrupt_mask = 1;
-	set_channel_read_mode(dev->channel, HV_CALL_ISR);
+	set_channel_read_mode(channel, HV_CALL_ISR);
 
 	/* Fill general uio info */
 	pdata->info.name = "uio_hv_generic";
 	pdata->info.version = DRIVER_VERSION;
 	pdata->info.irqcontrol = hv_uio_irqcontrol;
+	pdata->info.open = hv_uio_open;
+	pdata->info.release = hv_uio_release;
 	pdata->info.irq = UIO_IRQ_CUSTOM;
+	atomic_set(&pdata->refcnt, 0);
 
 	/* mem resources */
 	pdata->info.mem[TXRX_RING_MAP].name = "txrx_rings";
+	ring_buffer = page_address(channel->ringbuffer_page);
 	pdata->info.mem[TXRX_RING_MAP].addr
-		= (uintptr_t)dev->channel->ringbuffer_pages;
+		= (uintptr_t)virt_to_phys(ring_buffer);
 	pdata->info.mem[TXRX_RING_MAP].size
-		= dev->channel->ringbuffer_pagecount << PAGE_SHIFT;
-	pdata->info.mem[TXRX_RING_MAP].memtype = UIO_MEM_LOGICAL;
+		= channel->ringbuffer_pagecount << PAGE_SHIFT;
+	pdata->info.mem[TXRX_RING_MAP].memtype = UIO_MEM_IOVA;
 
 	pdata->info.mem[INT_PAGE_MAP].name = "int_page";
 	pdata->info.mem[INT_PAGE_MAP].addr
@@ -247,7 +292,7 @@ hv_uio_probe(struct hv_device *dev,
 		goto fail_close;
 	}
 
-	ret = vmbus_establish_gpadl(dev->channel, pdata->recv_buf,
+	ret = vmbus_establish_gpadl(channel, pdata->recv_buf,
 				    RECV_BUFFER_SIZE, &pdata->recv_gpadl);
 	if (ret)
 		goto fail_close;
@@ -261,14 +306,13 @@ hv_uio_probe(struct hv_device *dev,
 	pdata->info.mem[RECV_BUF_MAP].size = RECV_BUFFER_SIZE;
 	pdata->info.mem[RECV_BUF_MAP].memtype = UIO_MEM_VIRTUAL;
 
-
 	pdata->send_buf = vzalloc(SEND_BUFFER_SIZE);
 	if (pdata->send_buf == NULL) {
 		ret = -ENOMEM;
 		goto fail_close;
 	}
 
-	ret = vmbus_establish_gpadl(dev->channel, pdata->send_buf,
+	ret = vmbus_establish_gpadl(channel, pdata->send_buf,
 				    SEND_BUFFER_SIZE, &pdata->send_gpadl);
 	if (ret)
 		goto fail_close;
@@ -290,10 +334,10 @@ hv_uio_probe(struct hv_device *dev,
 		goto fail_close;
 	}
 
-	vmbus_set_chn_rescind_callback(dev->channel, hv_uio_rescind);
-	vmbus_set_sc_create_callback(dev->channel, hv_uio_new_channel);
+	vmbus_set_chn_rescind_callback(channel, hv_uio_rescind);
+	vmbus_set_sc_create_callback(channel, hv_uio_new_channel);
 
-	ret = sysfs_create_bin_file(&dev->channel->kobj, &ring_buffer_bin_attr);
+	ret = sysfs_create_bin_file(&channel->kobj, &ring_buffer_bin_attr);
 	if (ret)
 		dev_notice(&dev->device,
 			   "sysfs create ring bin file failed; %d\n", ret);
@@ -304,7 +348,6 @@ hv_uio_probe(struct hv_device *dev,
 
 fail_close:
 	hv_uio_cleanup(dev, pdata);
-	vmbus_close(dev->channel);
 fail:
 	kfree(pdata);
 
@@ -322,7 +365,8 @@ hv_uio_remove(struct hv_device *dev)
 	uio_unregister_device(&pdata->info);
 	hv_uio_cleanup(dev, pdata);
 	hv_set_drvdata(dev, NULL);
-	vmbus_close(dev->channel);
+
+	vmbus_free_ring(dev->channel);
 	kfree(pdata);
 	return 0;
 }
diff --git a/drivers/uio/uio_pdrv_genirq.c b/drivers/uio/uio_pdrv_genirq.c
index f598ecddc8a7..6c759934bff3 100644
--- a/drivers/uio/uio_pdrv_genirq.c
+++ b/drivers/uio/uio_pdrv_genirq.c
@@ -118,7 +118,8 @@ static int uio_pdrv_genirq_probe(struct platform_device *pdev)
 			dev_err(&pdev->dev, "unable to kmalloc\n");
 			return -ENOMEM;
 		}
-		uioinfo->name = pdev->dev.of_node->name;
+		uioinfo->name = devm_kasprintf(&pdev->dev, GFP_KERNEL, "%pOFn",
+					       pdev->dev.of_node);
 		uioinfo->version = "devicetree";
 		/* Multiple IRQs are not supported */
 	}
diff --git a/drivers/vme/vme.c b/drivers/vme/vme.c
index 92500f6bdad1..520a5f9c27de 100644
--- a/drivers/vme/vme.c
+++ b/drivers/vme/vme.c
@@ -1890,7 +1890,6 @@ static int __vme_register_driver_bus(struct vme_driver *drv,
 
 err_reg:
 	put_device(&vdev->dev);
-	kfree(vdev);
 err_devalloc:
 	list_for_each_entry_safe(vdev, tmp, &drv->devices, drv_list) {
 		list_del(&vdev->drv_list);
diff --git a/drivers/w1/masters/omap_hdq.c b/drivers/w1/masters/omap_hdq.c
index 83fc9aab34e8..3099052e1243 100644
--- a/drivers/w1/masters/omap_hdq.c
+++ b/drivers/w1/masters/omap_hdq.c
@@ -763,6 +763,8 @@ static int omap_hdq_remove(struct platform_device *pdev)
 	/* remove module dependency */
 	pm_runtime_disable(&pdev->dev);
 
+	w1_remove_master_device(&omap_w1_master);
+
 	return 0;
 }
 
diff --git a/drivers/w1/slaves/w1_ds2438.c b/drivers/w1/slaves/w1_ds2438.c
index bf641a191d07..7c4e33dbee4d 100644
--- a/drivers/w1/slaves/w1_ds2438.c
+++ b/drivers/w1/slaves/w1_ds2438.c
@@ -186,8 +186,8 @@ static int w1_ds2438_change_config_bit(struct w1_slave *sl, u8 mask, u8 value)
 	return -1;
 }
 
-static uint16_t w1_ds2438_get_voltage(struct w1_slave *sl,
-				      int adc_input, uint16_t *voltage)
+static int w1_ds2438_get_voltage(struct w1_slave *sl,
+				 int adc_input, uint16_t *voltage)
 {
 	unsigned int retries = W1_DS2438_RETRIES;
 	u8 w1_buf[DS2438_PAGE_SIZE + 1 /*for CRC*/];
@@ -235,6 +235,25 @@ post_unlock:
 	return ret;
 }
 
+static int w1_ds2438_get_current(struct w1_slave *sl, int16_t *voltage)
+{
+	u8 w1_buf[DS2438_PAGE_SIZE + 1 /*for CRC*/];
+	int ret;
+
+	mutex_lock(&sl->master->bus_mutex);
+
+	if (w1_ds2438_get_page(sl, 0, w1_buf) == 0) {
+		/* The voltage measured across current sense resistor RSENS. */
+		*voltage = (((int16_t) w1_buf[DS2438_CURRENT_MSB]) << 8) | ((int16_t) w1_buf[DS2438_CURRENT_LSB]);
+		ret = 0;
+	} else
+		ret = -1;
+
+	mutex_unlock(&sl->master->bus_mutex);
+
+	return ret;
+}
+
 static ssize_t iad_write(struct file *filp, struct kobject *kobj,
 			 struct bin_attribute *bin_attr, char *buf,
 			 loff_t off, size_t count)
@@ -257,6 +276,27 @@ static ssize_t iad_write(struct file *filp, struct kobject *kobj,
 	return ret;
 }
 
+static ssize_t iad_read(struct file *filp, struct kobject *kobj,
+			struct bin_attribute *bin_attr, char *buf,
+			loff_t off, size_t count)
+{
+	struct w1_slave *sl = kobj_to_w1_slave(kobj);
+	int ret;
+	int16_t voltage;
+
+	if (off != 0)
+		return 0;
+	if (!buf)
+		return -EINVAL;
+
+	if (w1_ds2438_get_current(sl, &voltage) == 0) {
+		ret = snprintf(buf, count, "%i\n", voltage);
+	} else
+		ret = -EIO;
+
+	return ret;
+}
+
 static ssize_t page0_read(struct file *filp, struct kobject *kobj,
 			  struct bin_attribute *bin_attr, char *buf,
 			  loff_t off, size_t count)
@@ -272,9 +312,13 @@ static ssize_t page0_read(struct file *filp, struct kobject *kobj,
 
 	mutex_lock(&sl->master->bus_mutex);
 
+	/* Read no more than page0 size */
+	if (count > DS2438_PAGE_SIZE)
+		count = DS2438_PAGE_SIZE;
+
 	if (w1_ds2438_get_page(sl, 0, w1_buf) == 0) {
-		memcpy(buf, &w1_buf, DS2438_PAGE_SIZE);
-		ret = DS2438_PAGE_SIZE;
+		memcpy(buf, &w1_buf, count);
+		ret = count;
 	} else
 		ret = -EIO;
 
@@ -289,7 +333,6 @@ static ssize_t temperature_read(struct file *filp, struct kobject *kobj,
 {
 	struct w1_slave *sl = kobj_to_w1_slave(kobj);
 	int ret;
-	ssize_t c = PAGE_SIZE;
 	int16_t temp;
 
 	if (off != 0)
@@ -298,8 +341,7 @@ static ssize_t temperature_read(struct file *filp, struct kobject *kobj,
 		return -EINVAL;
 
 	if (w1_ds2438_get_temperature(sl, &temp) == 0) {
-		c -= snprintf(buf + PAGE_SIZE - c, c, "%d\n", temp);
-		ret = PAGE_SIZE - c;
+		ret = snprintf(buf, count, "%i\n", temp);
 	} else
 		ret = -EIO;
 
@@ -312,7 +354,6 @@ static ssize_t vad_read(struct file *filp, struct kobject *kobj,
 {
 	struct w1_slave *sl = kobj_to_w1_slave(kobj);
 	int ret;
-	ssize_t c = PAGE_SIZE;
 	uint16_t voltage;
 
 	if (off != 0)
@@ -321,8 +362,7 @@ static ssize_t vad_read(struct file *filp, struct kobject *kobj,
 		return -EINVAL;
 
 	if (w1_ds2438_get_voltage(sl, DS2438_ADC_INPUT_VAD, &voltage) == 0) {
-		c -= snprintf(buf + PAGE_SIZE - c, c, "%d\n", voltage);
-		ret = PAGE_SIZE - c;
+		ret = snprintf(buf, count, "%u\n", voltage);
 	} else
 		ret = -EIO;
 
@@ -335,7 +375,6 @@ static ssize_t vdd_read(struct file *filp, struct kobject *kobj,
 {
 	struct w1_slave *sl = kobj_to_w1_slave(kobj);
 	int ret;
-	ssize_t c = PAGE_SIZE;
 	uint16_t voltage;
 
 	if (off != 0)
@@ -344,15 +383,14 @@ static ssize_t vdd_read(struct file *filp, struct kobject *kobj,
 		return -EINVAL;
 
 	if (w1_ds2438_get_voltage(sl, DS2438_ADC_INPUT_VDD, &voltage) == 0) {
-		c -= snprintf(buf + PAGE_SIZE - c, c, "%d\n", voltage);
-		ret = PAGE_SIZE - c;
+		ret = snprintf(buf, count, "%u\n", voltage);
 	} else
 		ret = -EIO;
 
 	return ret;
 }
 
-static BIN_ATTR(iad, S_IRUGO | S_IWUSR | S_IWGRP, NULL, iad_write, 1);
+static BIN_ATTR(iad, S_IRUGO | S_IWUSR | S_IWGRP, iad_read, iad_write, 0);
 static BIN_ATTR_RO(page0, DS2438_PAGE_SIZE);
 static BIN_ATTR_RO(temperature, 0/* real length varies */);
 static BIN_ATTR_RO(vad, 0/* real length varies */);