gpio: rcar: Fix runtime PM imbalance on error

[linux/fpc-iii.git] / drivers / greybus / operation.c

// SPDX-License-Identifier: GPL-2.0
/*
 * Greybus operations
 *
 * Copyright 2014-2015 Google Inc.
 * Copyright 2014-2015 Linaro Ltd.
 */

#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/module.h>
#include <linux/sched.h>
#include <linux/wait.h>
#include <linux/workqueue.h>
#include <linux/greybus.h>

#include "greybus_trace.h"

static struct kmem_cache *gb_operation_cache;
static struct kmem_cache *gb_message_cache;

/* Workqueue to handle Greybus operation completions. */
static struct workqueue_struct *gb_operation_completion_wq;

/* Wait queue for synchronous cancellations. */
static DECLARE_WAIT_QUEUE_HEAD(gb_operation_cancellation_queue);

/*
 * Protects updates to operation->errno.
 */
static DEFINE_SPINLOCK(gb_operations_lock);

static int gb_operation_response_send(struct gb_operation *operation,
                                      int errno);

/*
 * Increment operation active count and add to connection list unless the
 * connection is going away.
 *
 * Caller holds operation reference.
 */
static int gb_operation_get_active(struct gb_operation *operation)
{
        struct gb_connection *connection = operation->connection;
        unsigned long flags;

        spin_lock_irqsave(&connection->lock, flags);
        switch (connection->state) {
        case GB_CONNECTION_STATE_ENABLED:
                break;
        case GB_CONNECTION_STATE_ENABLED_TX:
                if (gb_operation_is_incoming(operation))
                        goto err_unlock;
                break;
        case GB_CONNECTION_STATE_DISCONNECTING:
                if (!gb_operation_is_core(operation))
                        goto err_unlock;
                break;
        default:
                goto err_unlock;
        }

        if (operation->active++ == 0)
                list_add_tail(&operation->links, &connection->operations);

        trace_gb_operation_get_active(operation);

        spin_unlock_irqrestore(&connection->lock, flags);

        return 0;

err_unlock:
        spin_unlock_irqrestore(&connection->lock, flags);

        return -ENOTCONN;
}

/* Caller holds operation reference. */
static void gb_operation_put_active(struct gb_operation *operation)
{
        struct gb_connection *connection = operation->connection;
        unsigned long flags;

        spin_lock_irqsave(&connection->lock, flags);

        trace_gb_operation_put_active(operation);

        if (--operation->active == 0) {
                list_del(&operation->links);
                if (atomic_read(&operation->waiters))
                        wake_up(&gb_operation_cancellation_queue);
        }
        spin_unlock_irqrestore(&connection->lock, flags);
}

static bool gb_operation_is_active(struct gb_operation *operation)
{
        struct gb_connection *connection = operation->connection;
        unsigned long flags;
        bool ret;

        spin_lock_irqsave(&connection->lock, flags);
        ret = operation->active;
        spin_unlock_irqrestore(&connection->lock, flags);

        return ret;
}

/*
 * Set an operation's result.
 *
 * Initially an outgoing operation's errno value is -EBADR.
 * If no error occurs before sending the request message the only
 * valid value operation->errno can be set to is -EINPROGRESS,
 * indicating the request has been (or rather is about to be) sent.
 * At that point nobody should be looking at the result until the
 * response arrives.
 *
 * The first time the result gets set after the request has been
 * sent, that result "sticks."  That is, if two concurrent threads
 * race to set the result, the first one wins.  The return value
 * tells the caller whether its result was recorded; if not the
 * caller has nothing more to do.
 *
 * The result value -EILSEQ is reserved to signal an implementation
 * error; if it's ever observed, the code performing the request has
 * done something fundamentally wrong.  It is an error to try to set
 * the result to -EBADR, and attempts to do so result in a warning,
 * and -EILSEQ is used instead.  Similarly, the only valid result
 * value to set for an operation in initial state is -EINPROGRESS.
 * Attempts to do otherwise will also record a (successful) -EILSEQ
 * operation result.
 */
static bool gb_operation_result_set(struct gb_operation *operation, int result)
{
        unsigned long flags;
        int prev;

        if (result == -EINPROGRESS) {
                /*
                 * -EINPROGRESS is used to indicate the request is
                 * in flight.  It should be the first result value
                 * set after the initial -EBADR.  Issue a warning
                 * and record an implementation error if it's
                 * set at any other time.
                 */
                spin_lock_irqsave(&gb_operations_lock, flags);
                prev = operation->errno;
                if (prev == -EBADR)
                        operation->errno = result;
                else
                        operation->errno = -EILSEQ;
                spin_unlock_irqrestore(&gb_operations_lock, flags);
                WARN_ON(prev != -EBADR);

                return true;
        }

        /*
         * The first result value set after a request has been sent
         * will be the final result of the operation.  Subsequent
         * attempts to set the result are ignored.
         *
         * Note that -EBADR is a reserved "initial state" result
         * value.  Attempts to set this value result in a warning,
         * and the result code is set to -EILSEQ instead.
         */
        if (WARN_ON(result == -EBADR))
                result = -EILSEQ; /* Nobody should be setting -EBADR */

        spin_lock_irqsave(&gb_operations_lock, flags);
        prev = operation->errno;
        if (prev == -EINPROGRESS)
                operation->errno = result;      /* First and final result */
        spin_unlock_irqrestore(&gb_operations_lock, flags);

        return prev == -EINPROGRESS;
}

int gb_operation_result(struct gb_operation *operation)
{
        int result = operation->errno;

        WARN_ON(result == -EBADR);
        WARN_ON(result == -EINPROGRESS);

        return result;
}
EXPORT_SYMBOL_GPL(gb_operation_result);

/*
 * Looks up an outgoing operation on a connection and returns a refcounted
 * pointer if found, or NULL otherwise.
 */
static struct gb_operation *
gb_operation_find_outgoing(struct gb_connection *connection, u16 operation_id)
{
        struct gb_operation *operation;
        unsigned long flags;
        bool found = false;

        spin_lock_irqsave(&connection->lock, flags);
        list_for_each_entry(operation, &connection->operations, links)
                if (operation->id == operation_id &&
                    !gb_operation_is_incoming(operation)) {
                        gb_operation_get(operation);
                        found = true;
                        break;
                }
        spin_unlock_irqrestore(&connection->lock, flags);

        return found ? operation : NULL;
}

static int gb_message_send(struct gb_message *message, gfp_t gfp)
{
        struct gb_connection *connection = message->operation->connection;

        trace_gb_message_send(message);
        return connection->hd->driver->message_send(connection->hd,
                                        connection->hd_cport_id,
                                        message,
                                        gfp);
}

/*
 * Cancel a message we have passed to the host device layer to be sent.
 */
static void gb_message_cancel(struct gb_message *message)
{
        struct gb_host_device *hd = message->operation->connection->hd;

        hd->driver->message_cancel(message);
}

static void gb_operation_request_handle(struct gb_operation *operation)
{
        struct gb_connection *connection = operation->connection;
        int status;
        int ret;

        if (connection->handler) {
                status = connection->handler(operation);
        } else {
                dev_err(&connection->hd->dev,
                        "%s: unexpected incoming request of type 0x%02x\n",
                        connection->name, operation->type);

                status = -EPROTONOSUPPORT;
        }

        ret = gb_operation_response_send(operation, status);
        if (ret) {
                dev_err(&connection->hd->dev,
                        "%s: failed to send response %d for type 0x%02x: %d\n",
                        connection->name, status, operation->type, ret);
                return;
        }
}

/*
 * Process operation work.
 *
 * For incoming requests, call the protocol request handler. The operation
 * result should be -EINPROGRESS at this point.
 *
 * For outgoing requests, the operation result value should have
 * been set before queueing this.  The operation callback function
 * allows the original requester to know the request has completed
 * and its result is available.
 */
static void gb_operation_work(struct work_struct *work)
{
        struct gb_operation *operation;
        int ret;

        operation = container_of(work, struct gb_operation, work);

        if (gb_operation_is_incoming(operation)) {
                gb_operation_request_handle(operation);
        } else {
                ret = del_timer_sync(&operation->timer);
                if (!ret) {
                        /* Cancel request message if scheduled by timeout. */
                        if (gb_operation_result(operation) == -ETIMEDOUT)
                                gb_message_cancel(operation->request);
                }

                operation->callback(operation);
        }

        gb_operation_put_active(operation);
        gb_operation_put(operation);
}

static void gb_operation_timeout(struct timer_list *t)
{
        struct gb_operation *operation = from_timer(operation, t, timer);

        if (gb_operation_result_set(operation, -ETIMEDOUT)) {
                /*
                 * A stuck request message will be cancelled from the
                 * workqueue.
                 */
                queue_work(gb_operation_completion_wq, &operation->work);
        }
}

static void gb_operation_message_init(struct gb_host_device *hd,
                                      struct gb_message *message,
                                      u16 operation_id,
                                      size_t payload_size, u8 type)
{
        struct gb_operation_msg_hdr *header;

        header = message->buffer;

        message->header = header;
        message->payload = payload_size ? header + 1 : NULL;
        message->payload_size = payload_size;

        /*
         * The type supplied for incoming message buffers will be
         * GB_REQUEST_TYPE_INVALID. Such buffers will be overwritten by
         * arriving data so there's no need to initialize the message header.
         */
        if (type != GB_REQUEST_TYPE_INVALID) {
                u16 message_size = (u16)(sizeof(*header) + payload_size);

                /*
                 * For a request, the operation id gets filled in
                 * when the message is sent.  For a response, it
                 * will be copied from the request by the caller.
                 *
                 * The result field in a request message must be
                 * zero.  It will be set just prior to sending for
                 * a response.
                 */
                header->size = cpu_to_le16(message_size);
                header->operation_id = 0;
                header->type = type;
                header->result = 0;
        }
}

/*
 * Allocate a message to be used for an operation request or response.
 * Both types of message contain a common header.  The request message
 * for an outgoing operation is outbound, as is the response message
 * for an incoming operation.  The message header for an outbound
 * message is partially initialized here.
 *
 * The headers for inbound messages don't need to be initialized;
 * they'll be filled in by arriving data.
 *
 * Our message buffers have the following layout:
 *      message header  \_ these combined are
 *      message payload /  the message size
 */
static struct gb_message *
gb_operation_message_alloc(struct gb_host_device *hd, u8 type,
                           size_t payload_size, gfp_t gfp_flags)
{
        struct gb_message *message;
        struct gb_operation_msg_hdr *header;
        size_t message_size = payload_size + sizeof(*header);

        if (message_size > hd->buffer_size_max) {
                dev_warn(&hd->dev, "requested message size too big (%zu > %zu)\n",
                         message_size, hd->buffer_size_max);
                return NULL;
        }

        /* Allocate the message structure and buffer. */
        message = kmem_cache_zalloc(gb_message_cache, gfp_flags);
        if (!message)
                return NULL;

        message->buffer = kzalloc(message_size, gfp_flags);
        if (!message->buffer)
                goto err_free_message;

        /* Initialize the message.  Operation id is filled in later. */
        gb_operation_message_init(hd, message, 0, payload_size, type);

        return message;

err_free_message:
        kmem_cache_free(gb_message_cache, message);

        return NULL;
}

static void gb_operation_message_free(struct gb_message *message)
{
        kfree(message->buffer);
        kmem_cache_free(gb_message_cache, message);
}

/*
 * Map an enum gb_operation_status value (which is represented in a
 * message as a single byte) to an appropriate Linux negative errno.
 */
static int gb_operation_status_map(u8 status)
{
        switch (status) {
        case GB_OP_SUCCESS:
                return 0;
        case GB_OP_INTERRUPTED:
                return -EINTR;
        case GB_OP_TIMEOUT:
                return -ETIMEDOUT;
        case GB_OP_NO_MEMORY:
                return -ENOMEM;
        case GB_OP_PROTOCOL_BAD:
                return -EPROTONOSUPPORT;
        case GB_OP_OVERFLOW:
                return -EMSGSIZE;
        case GB_OP_INVALID:
                return -EINVAL;
        case GB_OP_RETRY:
                return -EAGAIN;
        case GB_OP_NONEXISTENT:
                return -ENODEV;
        case GB_OP_MALFUNCTION:
                return -EILSEQ;
        case GB_OP_UNKNOWN_ERROR:
        default:
                return -EIO;
        }
}

/*
 * Map a Linux errno value (from operation->errno) into the value
 * that should represent it in a response message status sent
 * over the wire.  Returns an enum gb_operation_status value (which
 * is represented in a message as a single byte).
 */
static u8 gb_operation_errno_map(int errno)
{
        switch (errno) {
        case 0:
                return GB_OP_SUCCESS;
        case -EINTR:
                return GB_OP_INTERRUPTED;
        case -ETIMEDOUT:
                return GB_OP_TIMEOUT;
        case -ENOMEM:
                return GB_OP_NO_MEMORY;
        case -EPROTONOSUPPORT:
                return GB_OP_PROTOCOL_BAD;
        case -EMSGSIZE:
                return GB_OP_OVERFLOW;  /* Could be underflow too */
        case -EINVAL:
                return GB_OP_INVALID;
        case -EAGAIN:
                return GB_OP_RETRY;
        case -EILSEQ:
                return GB_OP_MALFUNCTION;
        case -ENODEV:
                return GB_OP_NONEXISTENT;
        case -EIO:
        default:
                return GB_OP_UNKNOWN_ERROR;
        }
}

bool gb_operation_response_alloc(struct gb_operation *operation,
                                 size_t response_size, gfp_t gfp)
{
        struct gb_host_device *hd = operation->connection->hd;
        struct gb_operation_msg_hdr *request_header;
        struct gb_message *response;
        u8 type;

        type = operation->type | GB_MESSAGE_TYPE_RESPONSE;
        response = gb_operation_message_alloc(hd, type, response_size, gfp);
        if (!response)
                return false;
        response->operation = operation;

        /*
         * Size and type get initialized when the message is
         * allocated.  The errno will be set before sending.  All
         * that's left is the operation id, which we copy from the
         * request message header (as-is, in little-endian order).
         */
        request_header = operation->request->header;
        response->header->operation_id = request_header->operation_id;
        operation->response = response;

        return true;
}
EXPORT_SYMBOL_GPL(gb_operation_response_alloc);

/*
 * Create a Greybus operation to be sent over the given connection.
 * The request buffer will be big enough for a payload of the given
 * size.
 *
 * For outgoing requests, the request message's header will be
 * initialized with the type of the request and the message size.
 * Outgoing operations must also specify the response buffer size,
 * which must be sufficient to hold all expected response data.  The
 * response message header will eventually be overwritten, so there's
 * no need to initialize it here.
 *
 * Request messages for incoming operations can arrive in interrupt
 * context, so they must be allocated with GFP_ATOMIC.  In this case
 * the request buffer will be immediately overwritten, so there is
 * no need to initialize the message header.  Responsibility for
 * allocating a response buffer lies with the incoming request
 * handler for a protocol.  So we don't allocate that here.
 *
 * Returns a pointer to the new operation or a null pointer if an
 * error occurs.
 */
static struct gb_operation *
gb_operation_create_common(struct gb_connection *connection, u8 type,
                           size_t request_size, size_t response_size,
                           unsigned long op_flags, gfp_t gfp_flags)
{
        struct gb_host_device *hd = connection->hd;
        struct gb_operation *operation;

        operation = kmem_cache_zalloc(gb_operation_cache, gfp_flags);
        if (!operation)
                return NULL;
        operation->connection = connection;

        operation->request = gb_operation_message_alloc(hd, type, request_size,
                                                        gfp_flags);
        if (!operation->request)
                goto err_cache;
        operation->request->operation = operation;

        /* Allocate the response buffer for outgoing operations */
        if (!(op_flags & GB_OPERATION_FLAG_INCOMING)) {
                if (!gb_operation_response_alloc(operation, response_size,
                                                 gfp_flags)) {
                        goto err_request;
                }

                timer_setup(&operation->timer, gb_operation_timeout, 0);
        }

        operation->flags = op_flags;
        operation->type = type;
        operation->errno = -EBADR;  /* Initial value--means "never set" */

        INIT_WORK(&operation->work, gb_operation_work);
        init_completion(&operation->completion);
        kref_init(&operation->kref);
        atomic_set(&operation->waiters, 0);

        return operation;

err_request:
        gb_operation_message_free(operation->request);
err_cache:
        kmem_cache_free(gb_operation_cache, operation);

        return NULL;
}

/*
 * Create a new operation associated with the given connection.  The
 * request and response sizes provided are the number of bytes
 * required to hold the request/response payload only.  Both of
 * these are allowed to be 0.  Note that 0x00 is reserved as an
 * invalid operation type for all protocols, and this is enforced
 * here.
 */
struct gb_operation *
gb_operation_create_flags(struct gb_connection *connection,
                          u8 type, size_t request_size,
                          size_t response_size, unsigned long flags,
                          gfp_t gfp)
{
        struct gb_operation *operation;

        if (WARN_ON_ONCE(type == GB_REQUEST_TYPE_INVALID))
                return NULL;
        if (WARN_ON_ONCE(type & GB_MESSAGE_TYPE_RESPONSE))
                type &= ~GB_MESSAGE_TYPE_RESPONSE;

        if (WARN_ON_ONCE(flags & ~GB_OPERATION_FLAG_USER_MASK))
                flags &= GB_OPERATION_FLAG_USER_MASK;

        operation = gb_operation_create_common(connection, type,
                                               request_size, response_size,
                                               flags, gfp);
        if (operation)
                trace_gb_operation_create(operation);

        return operation;
}
EXPORT_SYMBOL_GPL(gb_operation_create_flags);

struct gb_operation *
gb_operation_create_core(struct gb_connection *connection,
                         u8 type, size_t request_size,
                         size_t response_size, unsigned long flags,
                         gfp_t gfp)
{
        struct gb_operation *operation;

        flags |= GB_OPERATION_FLAG_CORE;

        operation = gb_operation_create_common(connection, type,
                                               request_size, response_size,
                                               flags, gfp);
        if (operation)
                trace_gb_operation_create_core(operation);

        return operation;
}

/* Do not export this function. */

size_t gb_operation_get_payload_size_max(struct gb_connection *connection)
{
        struct gb_host_device *hd = connection->hd;

        return hd->buffer_size_max - sizeof(struct gb_operation_msg_hdr);
}
EXPORT_SYMBOL_GPL(gb_operation_get_payload_size_max);

static struct gb_operation *
gb_operation_create_incoming(struct gb_connection *connection, u16 id,
                             u8 type, void *data, size_t size)
{
        struct gb_operation *operation;
        size_t request_size;
        unsigned long flags = GB_OPERATION_FLAG_INCOMING;

        /* Caller has made sure we at least have a message header. */
        request_size = size - sizeof(struct gb_operation_msg_hdr);

        if (!id)
                flags |= GB_OPERATION_FLAG_UNIDIRECTIONAL;

        operation = gb_operation_create_common(connection, type,
                                               request_size,
                                               GB_REQUEST_TYPE_INVALID,
                                               flags, GFP_ATOMIC);
        if (!operation)
                return NULL;

        operation->id = id;
        memcpy(operation->request->header, data, size);
        trace_gb_operation_create_incoming(operation);

        return operation;
}

/*
 * Get an additional reference on an operation.
 */
void gb_operation_get(struct gb_operation *operation)
{
        kref_get(&operation->kref);
}
EXPORT_SYMBOL_GPL(gb_operation_get);

/*
 * Destroy a previously created operation.
 */
static void _gb_operation_destroy(struct kref *kref)
{
        struct gb_operation *operation;

        operation = container_of(kref, struct gb_operation, kref);

        trace_gb_operation_destroy(operation);

        if (operation->response)
                gb_operation_message_free(operation->response);
        gb_operation_message_free(operation->request);

        kmem_cache_free(gb_operation_cache, operation);
}

/*
 * Drop a reference on an operation, and destroy it when the last
 * one is gone.
 */
void gb_operation_put(struct gb_operation *operation)
{
        if (WARN_ON(!operation))
                return;

        kref_put(&operation->kref, _gb_operation_destroy);
}
EXPORT_SYMBOL_GPL(gb_operation_put);

/* Tell the requester we're done */
static void gb_operation_sync_callback(struct gb_operation *operation)
{
        complete(&operation->completion);
}

/**
 * gb_operation_request_send() - send an operation request message
 * @operation:  the operation to initiate
 * @callback:   the operation completion callback
 * @timeout:    operation timeout in milliseconds, or zero for no timeout
 * @gfp:        the memory flags to use for any allocations
 *
 * The caller has filled in any payload so the request message is ready to go.
 * The callback function supplied will be called when the response message has
 * arrived, a unidirectional request has been sent, or the operation is
 * cancelled, indicating that the operation is complete. The callback function
 * can fetch the result of the operation using gb_operation_result() if
 * desired.
 *
 * Return: 0 if the request was successfully queued in the host-driver queues,
 * or a negative errno.
 */
int gb_operation_request_send(struct gb_operation *operation,
                              gb_operation_callback callback,
                              unsigned int timeout,
                              gfp_t gfp)
{
        struct gb_connection *connection = operation->connection;
        struct gb_operation_msg_hdr *header;
        unsigned int cycle;
        int ret;

        if (gb_connection_is_offloaded(connection))
                return -EBUSY;

        if (!callback)
                return -EINVAL;

        /*
         * Record the callback function, which is executed in
         * non-atomic (workqueue) context when the final result
         * of an operation has been set.
         */
        operation->callback = callback;

        /*
         * Assign the operation's id, and store it in the request header.
         * Zero is a reserved operation id for unidirectional operations.
         */
        if (gb_operation_is_unidirectional(operation)) {
                operation->id = 0;
        } else {
                cycle = (unsigned int)atomic_inc_return(&connection->op_cycle);
                operation->id = (u16)(cycle % U16_MAX + 1);
        }

        header = operation->request->header;
        header->operation_id = cpu_to_le16(operation->id);

        gb_operation_result_set(operation, -EINPROGRESS);

        /*
         * Get an extra reference on the operation. It'll be dropped when the
         * operation completes.
         */
        gb_operation_get(operation);
        ret = gb_operation_get_active(operation);
        if (ret)
                goto err_put;

        ret = gb_message_send(operation->request, gfp);
        if (ret)
                goto err_put_active;

        if (timeout) {
                operation->timer.expires = jiffies + msecs_to_jiffies(timeout);
                add_timer(&operation->timer);
        }

        return 0;

err_put_active:
        gb_operation_put_active(operation);
err_put:
        gb_operation_put(operation);

        return ret;
}
EXPORT_SYMBOL_GPL(gb_operation_request_send);

/*
 * Send a synchronous operation.  This function is expected to
 * block, returning only when the response has arrived, (or when an
 * error is detected.  The return value is the result of the
 * operation.
 */
int gb_operation_request_send_sync_timeout(struct gb_operation *operation,
                                           unsigned int timeout)
{
        int ret;

        ret = gb_operation_request_send(operation, gb_operation_sync_callback,
                                        timeout, GFP_KERNEL);
        if (ret)
                return ret;

        ret = wait_for_completion_interruptible(&operation->completion);
        if (ret < 0) {
                /* Cancel the operation if interrupted */
                gb_operation_cancel(operation, -ECANCELED);
        }

        return gb_operation_result(operation);
}
EXPORT_SYMBOL_GPL(gb_operation_request_send_sync_timeout);

/*
 * Send a response for an incoming operation request.  A non-zero
 * errno indicates a failed operation.
 *
 * If there is any response payload, the incoming request handler is
 * responsible for allocating the response message.  Otherwise the
 * it can simply supply the result errno; this function will
 * allocate the response message if necessary.
 */
static int gb_operation_response_send(struct gb_operation *operation,
                                      int errno)
{
        struct gb_connection *connection = operation->connection;
        int ret;

        if (!operation->response &&
            !gb_operation_is_unidirectional(operation)) {
                if (!gb_operation_response_alloc(operation, 0, GFP_KERNEL))
                        return -ENOMEM;
        }

        /* Record the result */
        if (!gb_operation_result_set(operation, errno)) {
                dev_err(&connection->hd->dev, "request result already set\n");
                return -EIO;    /* Shouldn't happen */
        }

        /* Sender of request does not care about response. */
        if (gb_operation_is_unidirectional(operation))
                return 0;

        /* Reference will be dropped when message has been sent. */
        gb_operation_get(operation);
        ret = gb_operation_get_active(operation);
        if (ret)
                goto err_put;

        /* Fill in the response header and send it */
        operation->response->header->result = gb_operation_errno_map(errno);

        ret = gb_message_send(operation->response, GFP_KERNEL);
        if (ret)
                goto err_put_active;

        return 0;

err_put_active:
        gb_operation_put_active(operation);
err_put:
        gb_operation_put(operation);

        return ret;
}

/*
 * This function is called when a message send request has completed.
 */
void greybus_message_sent(struct gb_host_device *hd,
                          struct gb_message *message, int status)
{
        struct gb_operation *operation = message->operation;
        struct gb_connection *connection = operation->connection;

        /*
         * If the message was a response, we just need to drop our
         * reference to the operation.  If an error occurred, report
         * it.
         *
         * For requests, if there's no error and the operation in not
         * unidirectional, there's nothing more to do until the response
         * arrives. If an error occurred attempting to send it, or if the
         * operation is unidrectional, record the result of the operation and
         * schedule its completion.
         */
        if (message == operation->response) {
                if (status) {
                        dev_err(&connection->hd->dev,
                                "%s: error sending response 0x%02x: %d\n",
                                connection->name, operation->type, status);
                }

                gb_operation_put_active(operation);
                gb_operation_put(operation);
        } else if (status || gb_operation_is_unidirectional(operation)) {
                if (gb_operation_result_set(operation, status)) {
                        queue_work(gb_operation_completion_wq,
                                   &operation->work);
                }
        }
}
EXPORT_SYMBOL_GPL(greybus_message_sent);

/*
 * We've received data on a connection, and it doesn't look like a
 * response, so we assume it's a request.
 *
 * This is called in interrupt context, so just copy the incoming
 * data into the request buffer and handle the rest via workqueue.
 */
static void gb_connection_recv_request(struct gb_connection *connection,
                                const struct gb_operation_msg_hdr *header,
                                void *data, size_t size)
{
        struct gb_operation *operation;
        u16 operation_id;
        u8 type;
        int ret;

        operation_id = le16_to_cpu(header->operation_id);
        type = header->type;

        operation = gb_operation_create_incoming(connection, operation_id,
                                                 type, data, size);
        if (!operation) {
                dev_err(&connection->hd->dev,
                        "%s: can't create incoming operation\n",
                        connection->name);
                return;
        }

        ret = gb_operation_get_active(operation);
        if (ret) {
                gb_operation_put(operation);
                return;
        }
        trace_gb_message_recv_request(operation->request);

        /*
         * The initial reference to the operation will be dropped when the
         * request handler returns.
         */
        if (gb_operation_result_set(operation, -EINPROGRESS))
                queue_work(connection->wq, &operation->work);
}

/*
 * We've received data that appears to be an operation response
 * message.  Look up the operation, and record that we've received
 * its response.
 *
 * This is called in interrupt context, so just copy the incoming
 * data into the response buffer and handle the rest via workqueue.
 */
static void gb_connection_recv_response(struct gb_connection *connection,
                                const struct gb_operation_msg_hdr *header,
                                void *data, size_t size)
{
        struct gb_operation *operation;
        struct gb_message *message;
        size_t message_size;
        u16 operation_id;
        int errno;

        operation_id = le16_to_cpu(header->operation_id);

        if (!operation_id) {
                dev_err_ratelimited(&connection->hd->dev,
                                    "%s: invalid response id 0 received\n",
                                    connection->name);
                return;
        }

        operation = gb_operation_find_outgoing(connection, operation_id);
        if (!operation) {
                dev_err_ratelimited(&connection->hd->dev,
                                    "%s: unexpected response id 0x%04x received\n",
                                    connection->name, operation_id);
                return;
        }

        errno = gb_operation_status_map(header->result);
        message = operation->response;
        message_size = sizeof(*header) + message->payload_size;
        if (!errno && size > message_size) {
                dev_err_ratelimited(&connection->hd->dev,
                                    "%s: malformed response 0x%02x received (%zu > %zu)\n",
                                    connection->name, header->type,
                                    size, message_size);
                errno = -EMSGSIZE;
        } else if (!errno && size < message_size) {
                if (gb_operation_short_response_allowed(operation)) {
                        message->payload_size = size - sizeof(*header);
                } else {
                        dev_err_ratelimited(&connection->hd->dev,
                                            "%s: short response 0x%02x received (%zu < %zu)\n",
                                            connection->name, header->type,
                                            size, message_size);
                        errno = -EMSGSIZE;
                }
        }

        /* We must ignore the payload if a bad status is returned */
        if (errno)
                size = sizeof(*header);

        /* The rest will be handled in work queue context */
        if (gb_operation_result_set(operation, errno)) {
                memcpy(message->buffer, data, size);

                trace_gb_message_recv_response(message);

                queue_work(gb_operation_completion_wq, &operation->work);
        }

        gb_operation_put(operation);
}

/*
 * Handle data arriving on a connection.  As soon as we return the
 * supplied data buffer will be reused (so unless we do something
 * with, it's effectively dropped).
 */
void gb_connection_recv(struct gb_connection *connection,
                        void *data, size_t size)
{
        struct gb_operation_msg_hdr header;
        struct device *dev = &connection->hd->dev;
        size_t msg_size;

        if (connection->state == GB_CONNECTION_STATE_DISABLED ||
            gb_connection_is_offloaded(connection)) {
                dev_warn_ratelimited(dev, "%s: dropping %zu received bytes\n",
                                     connection->name, size);
                return;
        }

        if (size < sizeof(header)) {
                dev_err_ratelimited(dev, "%s: short message received\n",
                                    connection->name);
                return;
        }

        /* Use memcpy as data may be unaligned */
        memcpy(&header, data, sizeof(header));
        msg_size = le16_to_cpu(header.size);
        if (size < msg_size) {
                dev_err_ratelimited(dev,
                                    "%s: incomplete message 0x%04x of type 0x%02x received (%zu < %zu)\n",
                                    connection->name,
                                    le16_to_cpu(header.operation_id),
                                    header.type, size, msg_size);
                return;         /* XXX Should still complete operation */
        }

        if (header.type & GB_MESSAGE_TYPE_RESPONSE) {
                gb_connection_recv_response(connection, &header, data,
                                            msg_size);
        } else {
                gb_connection_recv_request(connection, &header, data,
                                           msg_size);
        }
}

/*
 * Cancel an outgoing operation synchronously, and record the given error to
 * indicate why.
 */
void gb_operation_cancel(struct gb_operation *operation, int errno)
{
        if (WARN_ON(gb_operation_is_incoming(operation)))
                return;

        if (gb_operation_result_set(operation, errno)) {
                gb_message_cancel(operation->request);
                queue_work(gb_operation_completion_wq, &operation->work);
        }
        trace_gb_message_cancel_outgoing(operation->request);

        atomic_inc(&operation->waiters);
        wait_event(gb_operation_cancellation_queue,
                   !gb_operation_is_active(operation));
        atomic_dec(&operation->waiters);
}
EXPORT_SYMBOL_GPL(gb_operation_cancel);

/*
 * Cancel an incoming operation synchronously. Called during connection tear
 * down.
 */
void gb_operation_cancel_incoming(struct gb_operation *operation, int errno)
{
        if (WARN_ON(!gb_operation_is_incoming(operation)))
                return;

        if (!gb_operation_is_unidirectional(operation)) {
                /*
                 * Make sure the request handler has submitted the response
                 * before cancelling it.
                 */
                flush_work(&operation->work);
                if (!gb_operation_result_set(operation, errno))
                        gb_message_cancel(operation->response);
        }
        trace_gb_message_cancel_incoming(operation->response);

        atomic_inc(&operation->waiters);
        wait_event(gb_operation_cancellation_queue,
                   !gb_operation_is_active(operation));
        atomic_dec(&operation->waiters);
}

/**
 * gb_operation_sync_timeout() - implement a "simple" synchronous operation
 * @connection: the Greybus connection to send this to
 * @type: the type of operation to send
 * @request: pointer to a memory buffer to copy the request from
 * @request_size: size of @request
 * @response: pointer to a memory buffer to copy the response to
 * @response_size: the size of @response.
 * @timeout: operation timeout in milliseconds
 *
 * This function implements a simple synchronous Greybus operation.  It sends
 * the provided operation request and waits (sleeps) until the corresponding
 * operation response message has been successfully received, or an error
 * occurs.  @request and @response are buffers to hold the request and response
 * data respectively, and if they are not NULL, their size must be specified in
 * @request_size and @response_size.
 *
 * If a response payload is to come back, and @response is not NULL,
 * @response_size number of bytes will be copied into @response if the operation
 * is successful.
 *
 * If there is an error, the response buffer is left alone.
 */
int gb_operation_sync_timeout(struct gb_connection *connection, int type,
                              void *request, int request_size,
                              void *response, int response_size,
                              unsigned int timeout)
{
        struct gb_operation *operation;
        int ret;

        if ((response_size && !response) ||
            (request_size && !request))
                return -EINVAL;

        operation = gb_operation_create(connection, type,
                                        request_size, response_size,
                                        GFP_KERNEL);
        if (!operation)
                return -ENOMEM;

        if (request_size)
                memcpy(operation->request->payload, request, request_size);

        ret = gb_operation_request_send_sync_timeout(operation, timeout);
        if (ret) {
                dev_err(&connection->hd->dev,
                        "%s: synchronous operation id 0x%04x of type 0x%02x failed: %d\n",
                        connection->name, operation->id, type, ret);
        } else {
                if (response_size) {
                        memcpy(response, operation->response->payload,
                               response_size);
                }
        }

        gb_operation_put(operation);

        return ret;
}
EXPORT_SYMBOL_GPL(gb_operation_sync_timeout);

/**
 * gb_operation_unidirectional_timeout() - initiate a unidirectional operation
 * @connection:         connection to use
 * @type:               type of operation to send
 * @request:            memory buffer to copy the request from
 * @request_size:       size of @request
 * @timeout:            send timeout in milliseconds
 *
 * Initiate a unidirectional operation by sending a request message and
 * waiting for it to be acknowledged as sent by the host device.
 *
 * Note that successful send of a unidirectional operation does not imply that
 * the request as actually reached the remote end of the connection.
 */
int gb_operation_unidirectional_timeout(struct gb_connection *connection,
                                        int type, void *request,
                                        int request_size,
                                        unsigned int timeout)
{
        struct gb_operation *operation;
        int ret;

        if (request_size && !request)
                return -EINVAL;

        operation = gb_operation_create_flags(connection, type,
                                              request_size, 0,
                                              GB_OPERATION_FLAG_UNIDIRECTIONAL,
                                              GFP_KERNEL);
        if (!operation)
                return -ENOMEM;

        if (request_size)
                memcpy(operation->request->payload, request, request_size);

        ret = gb_operation_request_send_sync_timeout(operation, timeout);
        if (ret) {
                dev_err(&connection->hd->dev,
                        "%s: unidirectional operation of type 0x%02x failed: %d\n",
                        connection->name, type, ret);
        }

        gb_operation_put(operation);

        return ret;
}
EXPORT_SYMBOL_GPL(gb_operation_unidirectional_timeout);

int __init gb_operation_init(void)
{
        gb_message_cache = kmem_cache_create("gb_message_cache",
                                             sizeof(struct gb_message), 0, 0,
                                             NULL);
        if (!gb_message_cache)
                return -ENOMEM;

        gb_operation_cache = kmem_cache_create("gb_operation_cache",
                                               sizeof(struct gb_operation), 0,
                                               0, NULL);
        if (!gb_operation_cache)
                goto err_destroy_message_cache;

        gb_operation_completion_wq = alloc_workqueue("greybus_completion",
                                                     0, 0);
        if (!gb_operation_completion_wq)
                goto err_destroy_operation_cache;

        return 0;

err_destroy_operation_cache:
        kmem_cache_destroy(gb_operation_cache);
        gb_operation_cache = NULL;
err_destroy_message_cache:
        kmem_cache_destroy(gb_message_cache);
        gb_message_cache = NULL;

        return -ENOMEM;
}

void gb_operation_exit(void)
{
        destroy_workqueue(gb_operation_completion_wq);
        gb_operation_completion_wq = NULL;
        kmem_cache_destroy(gb_operation_cache);
        gb_operation_cache = NULL;
        kmem_cache_destroy(gb_message_cache);
        gb_message_cache = NULL;
}