firmware: Add basic support for TI System Control Interface (TI-SCI) protocol

Texas Instrument's System Control Interface (TI-SCI) Message Protocol
is used in Texas Instrument's System on Chip (SoC) such as those
in keystone family K2G SoC to communicate between various compute
processors with a central system controller entity.

TI-SCI message protocol provides support for management of various
hardware entities within the SoC. Add support driver to allow
communication with system controller entity within the SoC using the
mailbox client.

We introduce the basic registration and query capability for the
driver protocol as part of this change. Subsequent patches add in
functionality specific to the TI-SCI features.

Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Tero Kristo <t-kristo@ti.com>

1 files changed, 790 insertions, 0 deletions

diff --git a/drivers/firmware/ti_sci.c b/drivers/firmware/ti_sci.c
new file mode 100644
index 000000000000..5e99d7c18276
--- /dev/null
+++ b/drivers/firmware/ti_sci.c
@@ -0,0 +1,790 @@
+/*
+ * Texas Instruments System Control Interface Protocol Driver
+ *
+ * Copyright (C) 2015-2016 Texas Instruments Incorporated - http://www.ti.com/
+ *	Nishanth Menon
+ *
+ * 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 "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#define pr_fmt(fmt) "%s: " fmt, __func__
+
+#include <linux/bitmap.h>
+#include <linux/debugfs.h>
+#include <linux/export.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/mailbox_client.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/semaphore.h>
+#include <linux/slab.h>
+#include <linux/soc/ti/ti-msgmgr.h>
+#include <linux/soc/ti/ti_sci_protocol.h>
+
+#include "ti_sci.h"
+
+/* List of all TI SCI devices active in system */
+static LIST_HEAD(ti_sci_list);
+/* Protection for the entire list */
+static DEFINE_MUTEX(ti_sci_list_mutex);
+
+/**
+ * struct ti_sci_xfer - Structure representing a message flow
+ * @tx_message:	Transmit message
+ * @rx_len:	Receive message length
+ * @xfer_buf:	Preallocated buffer to store receive message
+ *		Since we work with request-ACK protocol, we can
+ *		reuse the same buffer for the rx path as we
+ *		use for the tx path.
+ * @done:	completion event
+ */
+struct ti_sci_xfer {
+	struct ti_msgmgr_message tx_message;
+	u8 rx_len;
+	u8 *xfer_buf;
+	struct completion done;
+};
+
+/**
+ * struct ti_sci_xfers_info - Structure to manage transfer information
+ * @sem_xfer_count:	Counting Semaphore for managing max simultaneous
+ *			Messages.
+ * @xfer_block:		Preallocated Message array
+ * @xfer_alloc_table:	Bitmap table for allocated messages.
+ *			Index of this bitmap table is also used for message
+ *			sequence identifier.
+ * @xfer_lock:		Protection for message allocation
+ */
+struct ti_sci_xfers_info {
+	struct semaphore sem_xfer_count;
+	struct ti_sci_xfer *xfer_block;
+	unsigned long *xfer_alloc_table;
+	/* protect transfer allocation */
+	spinlock_t xfer_lock;
+};
+
+/**
+ * struct ti_sci_desc - Description of SoC integration
+ * @host_id:		Host identifier representing the compute entity
+ * @max_rx_timeout_ms:	Timeout for communication with SoC (in Milliseconds)
+ * @max_msgs: Maximum number of messages that can be pending
+ *		  simultaneously in the system
+ * @max_msg_size: Maximum size of data per message that can be handled.
+ */
+struct ti_sci_desc {
+	u8 host_id;
+	int max_rx_timeout_ms;
+	int max_msgs;
+	int max_msg_size;
+};
+
+/**
+ * struct ti_sci_info - Structure representing a TI SCI instance
+ * @dev:	Device pointer
+ * @desc:	SoC description for this instance
+ * @d:		Debugfs file entry
+ * @debug_region: Memory region where the debug message are available
+ * @debug_region_size: Debug region size
+ * @debug_buffer: Buffer allocated to copy debug messages.
+ * @handle:	Instance of TI SCI handle to send to clients.
+ * @cl:		Mailbox Client
+ * @chan_tx:	Transmit mailbox channel
+ * @chan_rx:	Receive mailbox channel
+ * @minfo:	Message info
+ * @node:	list head
+ * @users:	Number of users of this instance
+ */
+struct ti_sci_info {
+	struct device *dev;
+	const struct ti_sci_desc *desc;
+	struct dentry *d;
+	void __iomem *debug_region;
+	char *debug_buffer;
+	size_t debug_region_size;
+	struct ti_sci_handle handle;
+	struct mbox_client cl;
+	struct mbox_chan *chan_tx;
+	struct mbox_chan *chan_rx;
+	struct ti_sci_xfers_info minfo;
+	struct list_head node;
+	/* protected by ti_sci_list_mutex */
+	int users;
+};
+
+#define cl_to_ti_sci_info(c)	container_of(c, struct ti_sci_info, cl)
+#define handle_to_ti_sci_info(h) container_of(h, struct ti_sci_info, handle)
+
+#ifdef CONFIG_DEBUG_FS
+
+/**
+ * ti_sci_debug_show() - Helper to dump the debug log
+ * @s:	sequence file pointer
+ * @unused:	unused.
+ *
+ * Return: 0
+ */
+static int ti_sci_debug_show(struct seq_file *s, void *unused)
+{
+	struct ti_sci_info *info = s->private;
+
+	memcpy_fromio(info->debug_buffer, info->debug_region,
+		      info->debug_region_size);
+	/*
+	 * We don't trust firmware to leave NULL terminated last byte (hence
+	 * we have allocated 1 extra 0 byte). Since we cannot guarantee any
+	 * specific data format for debug messages, We just present the data
+	 * in the buffer as is - we expect the messages to be self explanatory.
+	 */
+	seq_puts(s, info->debug_buffer);
+	return 0;
+}
+
+/**
+ * ti_sci_debug_open() - debug file open
+ * @inode:	inode pointer
+ * @file:	file pointer
+ *
+ * Return: result of single_open
+ */
+static int ti_sci_debug_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, ti_sci_debug_show, inode->i_private);
+}
+
+/* log file operations */
+static const struct file_operations ti_sci_debug_fops = {
+	.open = ti_sci_debug_open,
+	.read = seq_read,
+	.llseek = seq_lseek,
+	.release = single_release,
+};
+
+/**
+ * ti_sci_debugfs_create() - Create log debug file
+ * @pdev:	platform device pointer
+ * @info:	Pointer to SCI entity information
+ *
+ * Return: 0 if all went fine, else corresponding error.
+ */
+static int ti_sci_debugfs_create(struct platform_device *pdev,
+				 struct ti_sci_info *info)
+{
+	struct device *dev = &pdev->dev;
+	struct resource *res;
+	char debug_name[50] = "ti_sci_debug@";
+
+	/* Debug region is optional */
+	res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
+					   "debug_messages");
+	info->debug_region = devm_ioremap_resource(dev, res);
+	if (IS_ERR(info->debug_region))
+		return 0;
+	info->debug_region_size = resource_size(res);
+
+	info->debug_buffer = devm_kcalloc(dev, info->debug_region_size + 1,
+					  sizeof(char), GFP_KERNEL);
+	if (!info->debug_buffer)
+		return -ENOMEM;
+	/* Setup NULL termination */
+	info->debug_buffer[info->debug_region_size] = 0;
+
+	info->d = debugfs_create_file(strncat(debug_name, dev_name(dev),
+					      sizeof(debug_name)),
+				      0444, NULL, info, &ti_sci_debug_fops);
+	if (IS_ERR(info->d))
+		return PTR_ERR(info->d);
+
+	dev_dbg(dev, "Debug region => %p, size = %zu bytes, resource: %pr\n",
+		info->debug_region, info->debug_region_size, res);
+	return 0;
+}
+
+/**
+ * ti_sci_debugfs_destroy() - clean up log debug file
+ * @pdev:	platform device pointer
+ * @info:	Pointer to SCI entity information
+ */
+static void ti_sci_debugfs_destroy(struct platform_device *pdev,
+				   struct ti_sci_info *info)
+{
+	if (IS_ERR(info->debug_region))
+		return;
+
+	debugfs_remove(info->d);
+}
+#else /* CONFIG_DEBUG_FS */
+static inline int ti_sci_debugfs_create(struct platform_device *dev,
+					struct ti_sci_info *info)
+{
+	return 0;
+}
+
+static inline void ti_sci_debugfs_destroy(struct platform_device *dev,
+					  struct ti_sci_info *info)
+{
+}
+#endif /* CONFIG_DEBUG_FS */
+
+/**
+ * ti_sci_dump_header_dbg() - Helper to dump a message header.
+ * @dev:	Device pointer corresponding to the SCI entity
+ * @hdr:	pointer to header.
+ */
+static inline void ti_sci_dump_header_dbg(struct device *dev,
+					  struct ti_sci_msg_hdr *hdr)
+{
+	dev_dbg(dev, "MSGHDR:type=0x%04x host=0x%02x seq=0x%02x flags=0x%08x\n",
+		hdr->type, hdr->host, hdr->seq, hdr->flags);
+}
+
+/**
+ * ti_sci_rx_callback() - mailbox client callback for receive messages
+ * @cl:	client pointer
+ * @m:	mailbox message
+ *
+ * Processes one received message to appropriate transfer information and
+ * signals completion of the transfer.
+ *
+ * NOTE: This function will be invoked in IRQ context, hence should be
+ * as optimal as possible.
+ */
+static void ti_sci_rx_callback(struct mbox_client *cl, void *m)
+{
+	struct ti_sci_info *info = cl_to_ti_sci_info(cl);
+	struct device *dev = info->dev;
+	struct ti_sci_xfers_info *minfo = &info->minfo;
+	struct ti_msgmgr_message *mbox_msg = m;
+	struct ti_sci_msg_hdr *hdr = (struct ti_sci_msg_hdr *)mbox_msg->buf;
+	struct ti_sci_xfer *xfer;
+	u8 xfer_id;
+
+	xfer_id = hdr->seq;
+
+	/*
+	 * Are we even expecting this?
+	 * NOTE: barriers were implicit in locks used for modifying the bitmap
+	 */
+	if (!test_bit(xfer_id, minfo->xfer_alloc_table)) {
+		dev_err(dev, "Message for %d is not expected!\n", xfer_id);
+		return;
+	}
+
+	xfer = &minfo->xfer_block[xfer_id];
+
+	/* Is the message of valid length? */
+	if (mbox_msg->len > info->desc->max_msg_size) {
+		dev_err(dev, "Unable to handle %d xfer(max %d)\n",
+			mbox_msg->len, info->desc->max_msg_size);
+		ti_sci_dump_header_dbg(dev, hdr);
+		return;
+	}
+	if (mbox_msg->len < xfer->rx_len) {
+		dev_err(dev, "Recv xfer %d < expected %d length\n",
+			mbox_msg->len, xfer->rx_len);
+		ti_sci_dump_header_dbg(dev, hdr);
+		return;
+	}
+
+	ti_sci_dump_header_dbg(dev, hdr);
+	/* Take a copy to the rx buffer.. */
+	memcpy(xfer->xfer_buf, mbox_msg->buf, xfer->rx_len);
+	complete(&xfer->done);
+}
+
+/**
+ * ti_sci_get_one_xfer() - Allocate one message
+ * @info:	Pointer to SCI entity information
+ * @msg_type:	Message type
+ * @msg_flags:	Flag to set for the message
+ * @tx_message_size: transmit message size
+ * @rx_message_size: receive message size
+ *
+ * Helper function which is used by various command functions that are
+ * exposed to clients of this driver for allocating a message traffic event.
+ *
+ * This function can sleep depending on pending requests already in the system
+ * for the SCI entity. Further, this also holds a spinlock to maintain integrity
+ * of internal data structures.
+ *
+ * Return: 0 if all went fine, else corresponding error.
+ */
+static struct ti_sci_xfer *ti_sci_get_one_xfer(struct ti_sci_info *info,
+					       u16 msg_type, u32 msg_flags,
+					       size_t tx_message_size,
+					       size_t rx_message_size)
+{
+	struct ti_sci_xfers_info *minfo = &info->minfo;
+	struct ti_sci_xfer *xfer;
+	struct ti_sci_msg_hdr *hdr;
+	unsigned long flags;
+	unsigned long bit_pos;
+	u8 xfer_id;
+	int ret;
+	int timeout;
+
+	/* Ensure we have sane transfer sizes */
+	if (rx_message_size > info->desc->max_msg_size ||
+	    tx_message_size > info->desc->max_msg_size ||
+	    rx_message_size < sizeof(*hdr) || tx_message_size < sizeof(*hdr))
+		return ERR_PTR(-ERANGE);
+
+	/*
+	 * Ensure we have only controlled number of pending messages.
+	 * Ideally, we might just have to wait a single message, be
+	 * conservative and wait 5 times that..
+	 */
+	timeout = msecs_to_jiffies(info->desc->max_rx_timeout_ms) * 5;
+	ret = down_timeout(&minfo->sem_xfer_count, timeout);
+	if (ret < 0)
+		return ERR_PTR(ret);
+
+	/* Keep the locked section as small as possible */
+	spin_lock_irqsave(&minfo->xfer_lock, flags);
+	bit_pos = find_first_zero_bit(minfo->xfer_alloc_table,
+				      info->desc->max_msgs);
+	set_bit(bit_pos, minfo->xfer_alloc_table);
+	spin_unlock_irqrestore(&minfo->xfer_lock, flags);
+
+	/*
+	 * We already ensured in probe that we can have max messages that can
+	 * fit in  hdr.seq - NOTE: this improves access latencies
+	 * to predictable O(1) access, BUT, it opens us to risk if
+	 * remote misbehaves with corrupted message sequence responses.
+	 * If that happens, we are going to be messed up anyways..
+	 */
+	xfer_id = (u8)bit_pos;
+
+	xfer = &minfo->xfer_block[xfer_id];
+
+	hdr = (struct ti_sci_msg_hdr *)xfer->tx_message.buf;
+	xfer->tx_message.len = tx_message_size;
+	xfer->rx_len = (u8)rx_message_size;
+
+	reinit_completion(&xfer->done);
+
+	hdr->seq = xfer_id;
+	hdr->type = msg_type;
+	hdr->host = info->desc->host_id;
+	hdr->flags = msg_flags;
+
+	return xfer;
+}
+
+/**
+ * ti_sci_put_one_xfer() - Release a message
+ * @minfo:	transfer info pointer
+ * @xfer:	message that was reserved by ti_sci_get_one_xfer
+ *
+ * This holds a spinlock to maintain integrity of internal data structures.
+ */
+static void ti_sci_put_one_xfer(struct ti_sci_xfers_info *minfo,
+				struct ti_sci_xfer *xfer)
+{
+	unsigned long flags;
+	struct ti_sci_msg_hdr *hdr;
+	u8 xfer_id;
+
+	hdr = (struct ti_sci_msg_hdr *)xfer->tx_message.buf;
+	xfer_id = hdr->seq;
+
+	/*
+	 * Keep the locked section as small as possible
+	 * NOTE: we might escape with smp_mb and no lock here..
+	 * but just be conservative and symmetric.
+	 */
+	spin_lock_irqsave(&minfo->xfer_lock, flags);
+	clear_bit(xfer_id, minfo->xfer_alloc_table);
+	spin_unlock_irqrestore(&minfo->xfer_lock, flags);
+
+	/* Increment the count for the next user to get through */
+	up(&minfo->sem_xfer_count);
+}
+
+/**
+ * ti_sci_do_xfer() - Do one transfer
+ * @info:	Pointer to SCI entity information
+ * @xfer:	Transfer to initiate and wait for response
+ *
+ * Return: -ETIMEDOUT in case of no response, if transmit error,
+ *	   return corresponding error, else if all goes well,
+ *	   return 0.
+ */
+static inline int ti_sci_do_xfer(struct ti_sci_info *info,
+				 struct ti_sci_xfer *xfer)
+{
+	int ret;
+	int timeout;
+	struct device *dev = info->dev;
+
+	ret = mbox_send_message(info->chan_tx, &xfer->tx_message);
+	if (ret < 0)
+		return ret;
+
+	ret = 0;
+
+	/* And we wait for the response. */
+	timeout = msecs_to_jiffies(info->desc->max_rx_timeout_ms);
+	if (!wait_for_completion_timeout(&xfer->done, timeout)) {
+		dev_err(dev, "Mbox timedout in resp(caller: %pF)\n",
+			(void *)_RET_IP_);
+		ret = -ETIMEDOUT;
+	}
+	/*
+	 * NOTE: we might prefer not to need the mailbox ticker to manage the
+	 * transfer queueing since the protocol layer queues things by itself.
+	 * Unfortunately, we have to kick the mailbox framework after we have
+	 * received our message.
+	 */
+	mbox_client_txdone(info->chan_tx, ret);
+
+	return ret;
+}
+
+/**
+ * ti_sci_cmd_get_revision() - command to get the revision of the SCI entity
+ * @info:	Pointer to SCI entity information
+ *
+ * Updates the SCI information in the internal data structure.
+ *
+ * Return: 0 if all went fine, else return appropriate error.
+ */
+static int ti_sci_cmd_get_revision(struct ti_sci_info *info)
+{
+	struct device *dev = info->dev;
+	struct ti_sci_handle *handle = &info->handle;
+	struct ti_sci_version_info *ver = &handle->version;
+	struct ti_sci_msg_resp_version *rev_info;
+	struct ti_sci_xfer *xfer;
+	int ret;
+
+	/* No need to setup flags since it is expected to respond */
+	xfer = ti_sci_get_one_xfer(info, TI_SCI_MSG_VERSION,
+				   0x0, sizeof(struct ti_sci_msg_hdr),
+				   sizeof(*rev_info));
+	if (IS_ERR(xfer)) {
+		ret = PTR_ERR(xfer);
+		dev_err(dev, "Message alloc failed(%d)\n", ret);
+		return ret;
+	}
+
+	rev_info = (struct ti_sci_msg_resp_version *)xfer->xfer_buf;
+
+	ret = ti_sci_do_xfer(info, xfer);
+	if (ret) {
+		dev_err(dev, "Mbox send fail %d\n", ret);
+		goto fail;
+	}
+
+	ver->abi_major = rev_info->abi_major;
+	ver->abi_minor = rev_info->abi_minor;
+	ver->firmware_revision = rev_info->firmware_revision;
+	strncpy(ver->firmware_description, rev_info->firmware_description,
+		sizeof(ver->firmware_description));
+
+fail:
+	ti_sci_put_one_xfer(&info->minfo, xfer);
+	return ret;
+}
+
+/**
+ * ti_sci_get_handle() - Get the TI SCI handle for a device
+ * @dev:	Pointer to device for which we want SCI handle
+ *
+ * NOTE: The function does not track individual clients of the framework
+ * and is expected to be maintained by caller of TI SCI protocol library.
+ * ti_sci_put_handle must be balanced with successful ti_sci_get_handle
+ * Return: pointer to handle if successful, else:
+ * -EPROBE_DEFER if the instance is not ready
+ * -ENODEV if the required node handler is missing
+ * -EINVAL if invalid conditions are encountered.
+ */
+const struct ti_sci_handle *ti_sci_get_handle(struct device *dev)
+{
+	struct device_node *ti_sci_np;
+	struct list_head *p;
+	struct ti_sci_handle *handle = NULL;
+	struct ti_sci_info *info;
+
+	if (!dev) {
+		pr_err("I need a device pointer\n");
+		return ERR_PTR(-EINVAL);
+	}
+	ti_sci_np = of_get_parent(dev->of_node);
+	if (!ti_sci_np) {
+		dev_err(dev, "No OF information\n");
+		return ERR_PTR(-EINVAL);
+	}
+
+	mutex_lock(&ti_sci_list_mutex);
+	list_for_each(p, &ti_sci_list) {
+		info = list_entry(p, struct ti_sci_info, node);
+		if (ti_sci_np == info->dev->of_node) {
+			handle = &info->handle;
+			info->users++;
+			break;
+		}
+	}
+	mutex_unlock(&ti_sci_list_mutex);
+	of_node_put(ti_sci_np);
+
+	if (!handle)
+		return ERR_PTR(-EPROBE_DEFER);
+
+	return handle;
+}
+EXPORT_SYMBOL_GPL(ti_sci_get_handle);
+
+/**
+ * ti_sci_put_handle() - Release the handle acquired by ti_sci_get_handle
+ * @handle:	Handle acquired by ti_sci_get_handle
+ *
+ * NOTE: The function does not track individual clients of the framework
+ * and is expected to be maintained by caller of TI SCI protocol library.
+ * ti_sci_put_handle must be balanced with successful ti_sci_get_handle
+ *
+ * Return: 0 is successfully released
+ * if an error pointer was passed, it returns the error value back,
+ * if null was passed, it returns -EINVAL;
+ */
+int ti_sci_put_handle(const struct ti_sci_handle *handle)
+{
+	struct ti_sci_info *info;
+
+	if (IS_ERR(handle))
+		return PTR_ERR(handle);
+	if (!handle)
+		return -EINVAL;
+
+	info = handle_to_ti_sci_info(handle);
+	mutex_lock(&ti_sci_list_mutex);
+	if (!WARN_ON(!info->users))
+		info->users--;
+	mutex_unlock(&ti_sci_list_mutex);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ti_sci_put_handle);
+
+static void devm_ti_sci_release(struct device *dev, void *res)
+{
+	const struct ti_sci_handle **ptr = res;
+	const struct ti_sci_handle *handle = *ptr;
+	int ret;
+
+	ret = ti_sci_put_handle(handle);
+	if (ret)
+		dev_err(dev, "failed to put handle %d\n", ret);
+}
+
+/**
+ * devm_ti_sci_get_handle() - Managed get handle
+ * @dev:	device for which we want SCI handle for.
+ *
+ * NOTE: This releases the handle once the device resources are
+ * no longer needed. MUST NOT BE released with ti_sci_put_handle.
+ * The function does not track individual clients of the framework
+ * and is expected to be maintained by caller of TI SCI protocol library.
+ *
+ * Return: 0 if all went fine, else corresponding error.
+ */
+const struct ti_sci_handle *devm_ti_sci_get_handle(struct device *dev)
+{
+	const struct ti_sci_handle **ptr;
+	const struct ti_sci_handle *handle;
+
+	ptr = devres_alloc(devm_ti_sci_release, sizeof(*ptr), GFP_KERNEL);
+	if (!ptr)
+		return ERR_PTR(-ENOMEM);
+	handle = ti_sci_get_handle(dev);
+
+	if (!IS_ERR(handle)) {
+		*ptr = handle;
+		devres_add(dev, ptr);
+	} else {
+		devres_free(ptr);
+	}
+
+	return handle;
+}
+EXPORT_SYMBOL_GPL(devm_ti_sci_get_handle);
+
+/* Description for K2G */
+static const struct ti_sci_desc ti_sci_pmmc_k2g_desc = {
+	.host_id = 2,
+	/* Conservative duration */
+	.max_rx_timeout_ms = 1000,
+	/* Limited by MBOX_TX_QUEUE_LEN. K2G can handle upto 128 messages! */
+	.max_msgs = 20,
+	.max_msg_size = 64,
+};
+
+static const struct of_device_id ti_sci_of_match[] = {
+	{.compatible = "ti,k2g-sci", .data = &ti_sci_pmmc_k2g_desc},
+	{ /* Sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, ti_sci_of_match);
+
+static int ti_sci_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	const struct of_device_id *of_id;
+	const struct ti_sci_desc *desc;
+	struct ti_sci_xfer *xfer;
+	struct ti_sci_info *info = NULL;
+	struct ti_sci_xfers_info *minfo;
+	struct mbox_client *cl;
+	int ret = -EINVAL;
+	int i;
+
+	of_id = of_match_device(ti_sci_of_match, dev);
+	if (!of_id) {
+		dev_err(dev, "OF data missing\n");
+		return -EINVAL;
+	}
+	desc = of_id->data;
+
+	info = devm_kzalloc(dev, sizeof(*info), GFP_KERNEL);
+	if (!info)
+		return -ENOMEM;
+
+	info->dev = dev;
+	info->desc = desc;
+	INIT_LIST_HEAD(&info->node);
+	minfo = &info->minfo;
+
+	/*
+	 * Pre-allocate messages
+	 * NEVER allocate more than what we can indicate in hdr.seq
+	 * if we have data description bug, force a fix..
+	 */
+	if (WARN_ON(desc->max_msgs >=
+		    1 << 8 * sizeof(((struct ti_sci_msg_hdr *)0)->seq)))
+		return -EINVAL;
+
+	minfo->xfer_block = devm_kcalloc(dev,
+					 desc->max_msgs,
+					 sizeof(*minfo->xfer_block),
+					 GFP_KERNEL);
+	if (!minfo->xfer_block)
+		return -ENOMEM;
+
+	minfo->xfer_alloc_table = devm_kzalloc(dev,
+					       BITS_TO_LONGS(desc->max_msgs)
+					       * sizeof(unsigned long),
+					       GFP_KERNEL);
+	if (!minfo->xfer_alloc_table)
+		return -ENOMEM;
+	bitmap_zero(minfo->xfer_alloc_table, desc->max_msgs);
+
+	/* Pre-initialize the buffer pointer to pre-allocated buffers */
+	for (i = 0, xfer = minfo->xfer_block; i < desc->max_msgs; i++, xfer++) {
+		xfer->xfer_buf = devm_kcalloc(dev, 1, desc->max_msg_size,
+					      GFP_KERNEL);
+		if (!xfer->xfer_buf)
+			return -ENOMEM;
+
+		xfer->tx_message.buf = xfer->xfer_buf;
+		init_completion(&xfer->done);
+	}
+
+	ret = ti_sci_debugfs_create(pdev, info);
+	if (ret)
+		dev_warn(dev, "Failed to create debug file\n");
+
+	platform_set_drvdata(pdev, info);
+
+	cl = &info->cl;
+	cl->dev = dev;
+	cl->tx_block = false;
+	cl->rx_callback = ti_sci_rx_callback;
+	cl->knows_txdone = true;
+
+	spin_lock_init(&minfo->xfer_lock);
+	sema_init(&minfo->sem_xfer_count, desc->max_msgs);
+
+	info->chan_rx = mbox_request_channel_byname(cl, "rx");
+	if (IS_ERR(info->chan_rx)) {
+		ret = PTR_ERR(info->chan_rx);
+		goto out;
+	}
+
+	info->chan_tx = mbox_request_channel_byname(cl, "tx");
+	if (IS_ERR(info->chan_tx)) {
+		ret = PTR_ERR(info->chan_tx);
+		goto out;
+	}
+	ret = ti_sci_cmd_get_revision(info);
+	if (ret) {
+		dev_err(dev, "Unable to communicate with TISCI(%d)\n", ret);
+		goto out;
+	}
+
+	dev_info(dev, "ABI: %d.%d (firmware rev 0x%04x '%s')\n",
+		 info->handle.version.abi_major, info->handle.version.abi_minor,
+		 info->handle.version.firmware_revision,
+		 info->handle.version.firmware_description);
+
+	mutex_lock(&ti_sci_list_mutex);
+	list_add_tail(&info->node, &ti_sci_list);
+	mutex_unlock(&ti_sci_list_mutex);
+
+	return of_platform_populate(dev->of_node, NULL, NULL, dev);
+out:
+	if (!IS_ERR(info->chan_tx))
+		mbox_free_channel(info->chan_tx);
+	if (!IS_ERR(info->chan_rx))
+		mbox_free_channel(info->chan_rx);
+	debugfs_remove(info->d);
+	return ret;
+}
+
+static int ti_sci_remove(struct platform_device *pdev)
+{
+	struct ti_sci_info *info;
+	struct device *dev = &pdev->dev;
+	int ret = 0;
+
+	of_platform_depopulate(dev);
+
+	info = platform_get_drvdata(pdev);
+
+	mutex_lock(&ti_sci_list_mutex);
+	if (info->users)
+		ret = -EBUSY;
+	else
+		list_del(&info->node);
+	mutex_unlock(&ti_sci_list_mutex);
+
+	if (!ret) {
+		ti_sci_debugfs_destroy(pdev, info);
+
+		/* Safe to free channels since no more users */
+		mbox_free_channel(info->chan_tx);
+		mbox_free_channel(info->chan_rx);
+	}
+
+	return ret;
+}
+
+static struct platform_driver ti_sci_driver = {
+	.probe = ti_sci_probe,
+	.remove = ti_sci_remove,
+	.driver = {
+		   .name = "ti-sci",
+		   .of_match_table = of_match_ptr(ti_sci_of_match),
+	},
+};
+module_platform_driver(ti_sci_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("TI System Control Interface(SCI) driver");
+MODULE_AUTHOR("Nishanth Menon");
+MODULE_ALIAS("platform:ti-sci");