//
// Copyright 2011-2012 Ettus Research LLC
//
// 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 3 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.
//
// You should have received a copy of the GNU General Public License
// along with this program.  If not, see <http://www.gnu.org/licenses/>.
//

#ifndef INCLUDED_LIBUHD_TRANSPORT_SUPER_RECV_PACKET_HANDLER_HPP
#define INCLUDED_LIBUHD_TRANSPORT_SUPER_RECV_PACKET_HANDLER_HPP

#include <uhd/config.hpp>
#include <uhd/exception.hpp>
#include <uhd/convert.hpp>
#include <uhd/stream.hpp>
#include <uhd/utils/msg.hpp>
#include <uhd/utils/tasks.hpp>
#include <uhd/utils/atomic.hpp>
#include <uhd/utils/byteswap.hpp>
#include <uhd/types/metadata.hpp>
#include <uhd/transport/vrt_if_packet.hpp>
#include <uhd/transport/zero_copy.hpp>
#include <boost/dynamic_bitset.hpp>
#include <boost/foreach.hpp>
#include <boost/function.hpp>
#include <boost/format.hpp>
#include <boost/bind.hpp>
#include <boost/make_shared.hpp>
#include <boost/thread/barrier.hpp>
#include <iostream>
#include <vector>

namespace uhd{ namespace transport{ namespace sph{

UHD_INLINE boost::uint32_t get_context_code(
    const boost::uint32_t *vrt_hdr, const vrt::if_packet_info_t &if_packet_info
){
    //extract the context word (we dont know the endianness so mirror the bytes)
    boost::uint32_t word0 = vrt_hdr[if_packet_info.num_header_words32] |
              uhd::byteswap(vrt_hdr[if_packet_info.num_header_words32]);
    return word0 & 0xff;
}

typedef boost::function<void(void)> handle_overflow_type;
static inline void handle_overflow_nop(void){}

/***********************************************************************
 * Super receive packet handler
 *
 * A receive packet handler represents a group of channels.
 * The channel group shares a common sample rate.
 * All channels are received in unison in recv().
 **********************************************************************/
class recv_packet_handler{
public:
    typedef boost::function<managed_recv_buffer::sptr(double)> get_buff_type;
    typedef void(*vrt_unpacker_type)(const boost::uint32_t *, vrt::if_packet_info_t &);
    //typedef boost::function<void(const boost::uint32_t *, vrt::if_packet_info_t &)> vrt_unpacker_type;

    /*!
     * Make a new packet handler for receive
     * \param size the number of transport channels
     */
    recv_packet_handler(const size_t size = 1):
        _queue_error_for_next_call(false),
        _buffers_infos_index(0)
    {
        this->resize(size);
        set_alignment_failure_threshold(1000);
    }

    ~recv_packet_handler(void){
        _task_handlers.clear();
    }

    //! Resize the number of transport channels
    void resize(const size_t size){
        if (this->size() == size) return;
        _task_handlers.clear();
        _props.resize(size);
        //re-initialize all buffers infos by re-creating the vector
        _buffers_infos = std::vector<buffers_info_type>(4, buffers_info_type(size));
        _task_barrier_entry.resize(size);
        _task_barrier_exit.resize(size);
        _task_handlers.resize(size);
        for (size_t i = 1/*skip 0*/; i < size; i++){
            _task_handlers[i] = task::make(boost::bind(&recv_packet_handler::converter_thread_task, this, i));
        };
    }

    //! Get the channel width of this handler
    size_t size(void) const{
        return _props.size();
    }

    //! Setup the vrt unpacker function and offset
    void set_vrt_unpacker(const vrt_unpacker_type &vrt_unpacker, const size_t header_offset_words32 = 0){
        _vrt_unpacker = vrt_unpacker;
        _header_offset_words32 = header_offset_words32;
    }

    /*!
     * Set the threshold for alignment failure.
     * How many packets throw out before giving up?
     * \param threshold number of packets per channel
     */
    void set_alignment_failure_threshold(const size_t threshold){
        _alignment_faulure_threshold = threshold*this->size();
    }

    //! Set the rate of ticks per second
    void set_tick_rate(const double rate){
        _tick_rate = rate;
    }

    //! Set the rate of samples per second
    void set_samp_rate(const double rate){
        _samp_rate = rate;
    }

    /*!
     * Set the function to get a managed buffer.
     * \param xport_chan which transport channel
     * \param get_buff the getter function
     */
    void set_xport_chan_get_buff(const size_t xport_chan, const get_buff_type &get_buff, const bool flush = false){
        if (flush){
            while (get_buff(0.0));
        }
        _props.at(xport_chan).get_buff = get_buff;
    }

    //! Set the conversion routine for all channels
    void set_converter(const uhd::convert::id_type &id){
        _num_outputs = id.num_outputs;
        _converter = uhd::convert::get_converter(id)();
        this->set_scale_factor(1/32767.); //update after setting converter
        _bytes_per_otw_item = uhd::convert::get_bytes_per_item(id.input_format);
        _bytes_per_cpu_item = uhd::convert::get_bytes_per_item(id.output_format);
    }

    //! Set the transport channel's overflow handler
    void set_overflow_handler(const size_t xport_chan, const handle_overflow_type &handle_overflow){
        _props.at(xport_chan).handle_overflow = handle_overflow;
    }

    //! Set the scale factor used in float conversion
    void set_scale_factor(const double scale_factor){
        _converter->set_scalar(scale_factor);
    }

    /*******************************************************************
     * Receive:
     * The entry point for the fast-path receive calls.
     * Dispatch into combinations of single packet receive calls.
     ******************************************************************/
    UHD_INLINE size_t recv(
        const uhd::rx_streamer::buffs_type &buffs,
        const size_t nsamps_per_buff,
        uhd::rx_metadata_t &metadata,
        const double timeout,
        const bool one_packet
    ){
        //handle metadata queued from a previous receive
        if (_queue_error_for_next_call){
            _queue_error_for_next_call = false;
            metadata = _queue_metadata;
            //We want to allow a full buffer recv to be cut short by a timeout,
            //but do not want to generate an inline timeout message packet.
            if (_queue_metadata.error_code != rx_metadata_t::ERROR_CODE_TIMEOUT) return 0;
        }

        size_t accum_num_samps = recv_one_packet(
            buffs, nsamps_per_buff, metadata, timeout
        );

        if (one_packet) return accum_num_samps;

        //first recv had an error code set, return immediately
        if (metadata.error_code != rx_metadata_t::ERROR_CODE_NONE) return accum_num_samps;

        //loop until buffer is filled or error code
        while(accum_num_samps < nsamps_per_buff){
            size_t num_samps = recv_one_packet(
                buffs, nsamps_per_buff - accum_num_samps, _queue_metadata,
                timeout, accum_num_samps*_bytes_per_cpu_item
            );

            //metadata had an error code set, store for next call and return
            if (_queue_metadata.error_code != rx_metadata_t::ERROR_CODE_NONE){
                _queue_error_for_next_call = true;
                break;
            }
            accum_num_samps += num_samps;
        }
        return accum_num_samps;
    }

private:

    vrt_unpacker_type _vrt_unpacker;
    size_t _header_offset_words32;
    double _tick_rate, _samp_rate;
    bool _queue_error_for_next_call;
    size_t _alignment_faulure_threshold;
    rx_metadata_t _queue_metadata;
    struct xport_chan_props_type{
        xport_chan_props_type(void):
            packet_count(0),
            handle_overflow(&handle_overflow_nop)
        {}
        get_buff_type get_buff;
        size_t packet_count;
        handle_overflow_type handle_overflow;
    };
    std::vector<xport_chan_props_type> _props;
    size_t _num_outputs;
    size_t _bytes_per_otw_item; //used in conversion
    size_t _bytes_per_cpu_item; //used in conversion
    uhd::convert::converter::sptr _converter; //used in conversion

    //! information stored for a received buffer
    struct per_buffer_info_type{
        managed_recv_buffer::sptr buff;
        const boost::uint32_t *vrt_hdr;
        vrt::if_packet_info_t ifpi;
        time_spec_t time;
        const char *copy_buff;
    };

    //!information stored for a set of aligned buffers
    struct buffers_info_type : std::vector<per_buffer_info_type> {
        buffers_info_type(const size_t size):
            std::vector<per_buffer_info_type>(size),
            indexes_todo(size, true),
            alignment_time_valid(false),
            data_bytes_to_copy(0),
            fragment_offset_in_samps(0)
        {/* NOP */}
        boost::dynamic_bitset<> indexes_todo; //used in alignment logic
        time_spec_t alignment_time; //used in alignment logic
        bool alignment_time_valid; //used in alignment logic
        size_t data_bytes_to_copy; //keeps track of state
        size_t fragment_offset_in_samps; //keeps track of state
        rx_metadata_t metadata; //packet description
    };

    //! a circular queue of buffer infos
    std::vector<buffers_info_type> _buffers_infos;
    size_t _buffers_infos_index;
    buffers_info_type &get_curr_buffer_info(void){return _buffers_infos[_buffers_infos_index];}
    buffers_info_type &get_prev_buffer_info(void){return _buffers_infos[(_buffers_infos_index + 3)%4];}
    buffers_info_type &get_next_buffer_info(void){return _buffers_infos[(_buffers_infos_index + 1)%4];}
    void increment_buffer_info(void){_buffers_infos_index = (_buffers_infos_index + 1)%4;}

    //! possible return options for the packet receiver
    enum packet_type{
        PACKET_IF_DATA,
        PACKET_TIMESTAMP_ERROR,
        PACKET_INLINE_MESSAGE,
        PACKET_TIMEOUT_ERROR,
        PACKET_SEQUENCE_ERROR
    };

    /*******************************************************************
     * Get and process a single packet from the transport:
     * Receive a single packet at the given index.
     * Extract all the relevant info and store.
     * Check the info to determine the return code.
     ******************************************************************/
    UHD_INLINE packet_type get_and_process_single_packet(
        const size_t index,
        buffers_info_type &prev_buffer_info,
        buffers_info_type &curr_buffer_info,
        double timeout
    ){
        //get a single packet from the transport layer
        managed_recv_buffer::sptr &buff = curr_buffer_info[index].buff;
        buff = _props[index].get_buff(timeout);
        if (buff.get() == NULL) return PACKET_TIMEOUT_ERROR;

        //bounds check before extract
        size_t num_packet_words32 = buff->size()/sizeof(boost::uint32_t);
        if (num_packet_words32 <= _header_offset_words32){
            throw std::runtime_error("recv buffer smaller than vrt packet offset");
        }

        //extract packet info
        per_buffer_info_type &info = curr_buffer_info[index];
        info.ifpi.num_packet_words32 = num_packet_words32 - _header_offset_words32;
        info.vrt_hdr = buff->cast<const boost::uint32_t *>() + _header_offset_words32;
        _vrt_unpacker(info.vrt_hdr, info.ifpi);
        info.time = time_spec_t::from_ticks(info.ifpi.tsf, _tick_rate); //assumes has_tsf is true
        info.copy_buff = reinterpret_cast<const char *>(info.vrt_hdr + info.ifpi.num_header_words32);

        //--------------------------------------------------------------
        //-- Determine return conditions:
        //-- The order of these checks is HOLY.
        //--------------------------------------------------------------

        //1) check for inline IF message packets
        if (info.ifpi.packet_type != vrt::if_packet_info_t::PACKET_TYPE_DATA){
            return PACKET_INLINE_MESSAGE;
        }

        //2) check for sequence errors
        #ifndef SRPH_DONT_CHECK_SEQUENCE
        const size_t expected_packet_count = _props[index].packet_count;
        _props[index].packet_count = (info.ifpi.packet_count + 1)%16;
        if (expected_packet_count != info.ifpi.packet_count){
            return PACKET_SEQUENCE_ERROR;
        }
        #endif

        //3) check for out of order timestamps
        if (info.ifpi.has_tsf and prev_buffer_info[index].time > info.time){
            return PACKET_TIMESTAMP_ERROR;
        }

        //4) otherwise the packet is normal!
        return PACKET_IF_DATA;
    }

    /*******************************************************************
     * Alignment check:
     * Check the received packet for alignment and mark accordingly.
     ******************************************************************/
    UHD_INLINE void alignment_check(
        const size_t index, buffers_info_type &info
    ){
        //if alignment time was not valid or if the sequence id is newer:
        //  use this index's time as the alignment time
        //  reset the indexes list and remove this index
        if (not info.alignment_time_valid or info[index].time > info.alignment_time){
            info.alignment_time_valid = true;
            info.alignment_time = info[index].time;
            info.indexes_todo.set();
            info.indexes_todo.reset(index);
            info.data_bytes_to_copy = info[index].ifpi.num_payload_bytes;
        }

        //if the sequence id matches:
        //  remove this index from the list and continue
        else if (info[index].time == info.alignment_time){
            info.indexes_todo.reset(index);
        }

        //if the sequence id is older:
        //  continue with the same index to try again
        //else if (info[index].time < info.alignment_time)...
    }

    /*******************************************************************
     * Get aligned buffers:
     * Iterate through each index and try to accumulate aligned buffers.
     * Handle all of the edge cases like inline messages and errors.
     * The logic will throw out older packets until it finds a match.
     ******************************************************************/
    UHD_INLINE void get_aligned_buffs(double timeout){

        increment_buffer_info(); //increment to next buffer
        buffers_info_type &prev_info = get_prev_buffer_info();
        buffers_info_type &curr_info = get_curr_buffer_info();
        buffers_info_type &next_info = get_next_buffer_info();

        //Loop until we get a message of an aligned set of buffers:
        // - Receive a single packet and extract its info.
        // - Handle the packet type yielded by the receive.
        // - Check the timestamps for alignment conditions.
        size_t iterations = 0;
        while (curr_info.indexes_todo.any()){

            //get the index to process for this iteration
            const size_t index = curr_info.indexes_todo.find_first();
            packet_type packet;

            //receive a single packet from the transport
            try{
                packet = get_and_process_single_packet(
                    index, prev_info, curr_info, timeout
                );
            }

            //handle the case when the get packet throws
            catch(const std::exception &e){
                UHD_MSG(error) << boost::format(
                    "The receive packet handler caught an exception.\n%s"
                ) % e.what() << std::endl;
                std::swap(curr_info, next_info); //save progress from curr -> next
                curr_info.metadata.has_time_spec = false;
                curr_info.metadata.time_spec = time_spec_t(0.0);
                curr_info.metadata.more_fragments = false;
                curr_info.metadata.fragment_offset = 0;
                curr_info.metadata.start_of_burst = false;
                curr_info.metadata.end_of_burst = false;
                curr_info.metadata.error_code = rx_metadata_t::ERROR_CODE_BAD_PACKET;
                return;
            }

            switch(packet){
            case PACKET_IF_DATA:
                alignment_check(index, curr_info);
                break;

            case PACKET_TIMESTAMP_ERROR:
                //If the user changes the device time while streaming or without flushing,
                //we can receive a packet that comes before the previous packet in time.
                //This could cause the alignment logic to discard future received packets.
                //Therefore, when this occurs, we reset the info to restart from scratch.
                if (curr_info.alignment_time_valid and curr_info.alignment_time != curr_info[index].time){
                    curr_info.alignment_time_valid = false;
                }
                alignment_check(index, curr_info);
                break;

            case PACKET_INLINE_MESSAGE:
                std::swap(curr_info, next_info); //save progress from curr -> next
                curr_info.metadata.has_time_spec = next_info[index].ifpi.has_tsf;
                curr_info.metadata.time_spec = next_info[index].time;
                curr_info.metadata.more_fragments = false;
                curr_info.metadata.fragment_offset = 0;
                curr_info.metadata.start_of_burst = false;
                curr_info.metadata.end_of_burst = false;
                curr_info.metadata.error_code = rx_metadata_t::error_code_t(get_context_code(next_info[index].vrt_hdr, next_info[index].ifpi));
                if (curr_info.metadata.error_code == rx_metadata_t::ERROR_CODE_OVERFLOW){
                    _props[index].handle_overflow();
                    UHD_MSG(fastpath) << "O";
                }
                return;

            case PACKET_TIMEOUT_ERROR:
                std::swap(curr_info, next_info); //save progress from curr -> next
                curr_info.metadata.has_time_spec = false;
                curr_info.metadata.time_spec = time_spec_t(0.0);
                curr_info.metadata.more_fragments = false;
                curr_info.metadata.fragment_offset = 0;
                curr_info.metadata.start_of_burst = false;
                curr_info.metadata.end_of_burst = false;
                curr_info.metadata.error_code = rx_metadata_t::ERROR_CODE_TIMEOUT;
                return;

            case PACKET_SEQUENCE_ERROR:
                alignment_check(index, curr_info);
                std::swap(curr_info, next_info); //save progress from curr -> next
                curr_info.metadata.has_time_spec = prev_info.metadata.has_time_spec;
                curr_info.metadata.time_spec = prev_info.metadata.time_spec + time_spec_t::from_ticks(
                    prev_info[index].ifpi.num_payload_words32*sizeof(boost::uint32_t)/_bytes_per_otw_item, _samp_rate);
                curr_info.metadata.more_fragments = false;
                curr_info.metadata.fragment_offset = 0;
                curr_info.metadata.start_of_burst = false;
                curr_info.metadata.end_of_burst = false;
                curr_info.metadata.error_code = rx_metadata_t::ERROR_CODE_OVERFLOW;
                UHD_MSG(fastpath) << "O";
                return;

            }

            //too many iterations: detect alignment failure
            if (iterations++ > _alignment_faulure_threshold){
                UHD_MSG(error) << boost::format(
                    "The receive packet handler failed to time-align packets.\n"
                    "%u received packets were processed by the handler.\n"
                    "However, a timestamp match could not be determined.\n"
                ) % iterations << std::endl;
                std::swap(curr_info, next_info); //save progress from curr -> next
                curr_info.metadata.has_time_spec = false;
                curr_info.metadata.time_spec = time_spec_t(0.0);
                curr_info.metadata.more_fragments = false;
                curr_info.metadata.fragment_offset = 0;
                curr_info.metadata.start_of_burst = false;
                curr_info.metadata.end_of_burst = false;
                curr_info.metadata.error_code = rx_metadata_t::ERROR_CODE_ALIGNMENT;
                return;
            }

        }

        //set the metadata from the buffer information at index zero
        curr_info.metadata.has_time_spec = curr_info[0].ifpi.has_tsf;
        curr_info.metadata.time_spec = curr_info[0].time;
        curr_info.metadata.more_fragments = false;
        curr_info.metadata.fragment_offset = 0;
        curr_info.metadata.start_of_burst = curr_info[0].ifpi.sob;
        curr_info.metadata.end_of_burst = curr_info[0].ifpi.eob;
        curr_info.metadata.error_code = rx_metadata_t::ERROR_CODE_NONE;

    }

    /*******************************************************************
     * Receive a single packet:
     * Handles fragmentation, messages, errors, and copy-conversion.
     * When no fragments are available, call the get aligned buffers.
     * Then copy-convert available data into the user's IO buffers.
     ******************************************************************/
    UHD_INLINE size_t recv_one_packet(
        const uhd::rx_streamer::buffs_type &buffs,
        const size_t nsamps_per_buff,
        uhd::rx_metadata_t &metadata,
        const double timeout,
        const size_t buffer_offset_bytes = 0
    ){
        //get the next buffer if the current one has expired
        if (get_curr_buffer_info().data_bytes_to_copy == 0){

            //reset current buffer info members for reuse
            get_curr_buffer_info().fragment_offset_in_samps = 0;
            get_curr_buffer_info().alignment_time_valid = false;
            get_curr_buffer_info().indexes_todo.set();

            //perform receive with alignment logic
            get_aligned_buffs(timeout);
        }

        buffers_info_type &info = get_curr_buffer_info();
        metadata = info.metadata;

        //interpolate the time spec (useful when this is a fragment)
        metadata.time_spec += time_spec_t::from_ticks(info.fragment_offset_in_samps, _samp_rate);

        //extract the number of samples available to copy
        const size_t nsamps_available = info.data_bytes_to_copy/_bytes_per_otw_item;
        const size_t nsamps_to_copy = std::min(nsamps_per_buff*_num_outputs, nsamps_available);
        const size_t bytes_to_copy = nsamps_to_copy*_bytes_per_otw_item;
        const size_t nsamps_to_copy_per_io_buff = nsamps_to_copy/_num_outputs;

        //setup the data to share with converter threads
        _convert_nsamps = nsamps_to_copy_per_io_buff;
        _convert_buffs = &buffs;
        _convert_buffer_offset_bytes = buffer_offset_bytes;
        _convert_bytes_to_copy = bytes_to_copy;

        //perform N channels of conversion
        converter_thread_task(0);

        //update the copy buffer's availability
        info.data_bytes_to_copy -= bytes_to_copy;

        //setup the fragment flags and offset
        metadata.more_fragments = info.data_bytes_to_copy != 0;
        metadata.fragment_offset = info.fragment_offset_in_samps;
        info.fragment_offset_in_samps += nsamps_to_copy; //set for next call

        return nsamps_to_copy_per_io_buff;
    }

    /*******************************************************************
     * Perform one thread's work of the conversion task.
     * The entry and exit use a dual synchronization barrier,
     * to wait for data to become ready and block until completion.
     ******************************************************************/
    UHD_INLINE void converter_thread_task(const size_t index)
    {
        _task_barrier_entry.wait();

        //shortcut references to local data structures
        buffers_info_type &buff_info = get_curr_buffer_info();
        per_buffer_info_type &info = buff_info[index];
        const rx_streamer::buffs_type &buffs = *_convert_buffs;

        //fill IO buffs with pointers into the output buffer
        void *io_buffs[4/*max interleave*/];
        for (size_t i = 0; i < _num_outputs; i++){
            char *b = reinterpret_cast<char *>(buffs[index*_num_outputs + i]);
            io_buffs[i] = b + _convert_buffer_offset_bytes;
        }
        const ref_vector<void *> out_buffs(io_buffs, _num_outputs);

        //perform the conversion operation
        _converter->conv(info.copy_buff, out_buffs, _convert_nsamps);

        //advance the pointer for the source buffer
        info.copy_buff += _convert_bytes_to_copy;

        //release the buffer if fully consumed
        if (buff_info.data_bytes_to_copy == _convert_bytes_to_copy){
            info.buff.reset(); //effectively a release
        }

        _task_barrier_exit.wait();
    }

    //! Shared variables for the worker threads
    reusable_barrier _task_barrier_entry, _task_barrier_exit;
    std::vector<task::sptr> _task_handlers;
    size_t _convert_nsamps;
    const rx_streamer::buffs_type *_convert_buffs;
    size_t _convert_buffer_offset_bytes;
    size_t _convert_bytes_to_copy;

};

class recv_packet_streamer : public recv_packet_handler, public rx_streamer{
public:
    recv_packet_streamer(const size_t max_num_samps){
        _max_num_samps = max_num_samps;
    }

    size_t get_num_channels(void) const{
        return this->size();
    }

    size_t get_max_num_samps(void) const{
        return _max_num_samps;
    }

    size_t recv(
        const rx_streamer::buffs_type &buffs,
        const size_t nsamps_per_buff,
        uhd::rx_metadata_t &metadata,
        const double timeout,
        const bool one_packet
    ){
        return recv_packet_handler::recv(buffs, nsamps_per_buff, metadata, timeout, one_packet);
    }

private:
    size_t _max_num_samps;
};

}}} //namespace

#endif /* INCLUDED_LIBUHD_TRANSPORT_SUPER_RECV_PACKET_HANDLER_HPP */