Merged in branch master

19 files changed, 637 insertions, 330 deletions

diff --git a/host/lib/convert/convert_unpack_sc12.cpp b/host/lib/convert/convert_unpack_sc12.cpp
index 6583eb21f..a2aec2ae5 100644
--- a/host/lib/convert/convert_unpack_sc12.cpp
+++ b/host/lib/convert/convert_unpack_sc12.cpp
@@ -32,6 +32,17 @@ struct item32_sc12_3x
     item32_t line2;
 };
 
+/*
+ * convert_sc12_item32_3_to_star_4 takes in 3 lines with 32 bit each
+ * and converts them 4 samples of type 'std::complex<type>'.
+ * The structure of the 3 lines is as follows:
+ *  _ _ _ _ _ _ _ _
+ * |_ _ _1_ _ _|_ _|
+ * |_2_ _ _|_ _ _3_|
+ * |_ _|_ _ _4_ _ _|
+ *
+ * The numbers mark the position of one complex sample.
+ */
 template <typename type, tohost32_type tohost>
 void convert_sc12_item32_3_to_star_4
 (
@@ -84,17 +95,48 @@ struct convert_sc12_item32_1_to_star_1 : public converter
         _scalar = scalar/unpack_growth;
     }
 
+    /*
+     * This converter takes in 24 bits complex samples, 12 bits I and 12 bits Q, and converts them to type 'std::complex<type>'.
+     * 'type' is usually 'float'.
+     * For the converter to work correctly the used managed_buffer which holds all samples of one packet has to be 32 bits aligned.
+     * We assume 32 bits to be one line. This said the converter must be aware where it is supposed to start within 3 lines.
+     *
+     */
     void operator()(const input_type &inputs, const output_type &outputs, const size_t nsamps)
     {
-        const item32_sc12_3x *input = reinterpret_cast<const item32_sc12_3x *>(size_t(inputs[0]) & ~0x3);
+        /*
+         * Looking at the line structure above we can identify 4 cases.
+         * Each corresponds to the start of a different sample within a 3 line block.
+         * head_samps derives the number of samples left within one block.
+         * Then the number of bytes the converter has to rewind are calculated.
+         */
+        const size_t head_samps = size_t(inputs[0]) & 0x3;
+        size_t rewind = 0;
+        switch(head_samps)
+        {
+            case 0: break;
+            case 1: rewind = 9; break;
+            case 2: rewind = 6; break;
+            case 3: rewind = 3; break;
+        }
+
+        /*
+         * The pointer *input now points to the head of a 3 line block.
+         */
+        const item32_sc12_3x *input = reinterpret_cast<const item32_sc12_3x *>(size_t(inputs[0]) - rewind);
         std::complex<type> *output = reinterpret_cast<std::complex<type> *>(outputs[0]);
 
         //helper variables
         std::complex<type> dummy0, dummy1, dummy2;
         size_t i = 0, o = 0;
 
-        //handle the head case
-        const size_t head_samps = size_t(inputs[0]) & 0x3;
+        /*
+         * handle the head case
+         * head_samps holds the number of samples left in a block.
+         * The 3 line converter is called for the whole block and already processed samples are dumped.
+         * We don't run into the risk of a SIGSEGV because input will always point to valid memory within a managed_buffer.
+         * Furthermore the bytes in a buffer remain unchanged after they have been copied into it.
+         */
         switch (head_samps)
         {
         case 0: break; //no head
@@ -111,7 +153,18 @@ struct convert_sc12_item32_1_to_star_1 : public converter
             i++; o += 4;
         }
 
-        //handle the tail case
+        /*
+         * handle the tail case
+         * The converter can be called with any number of samples to be converted.
+         * This can end up in only a part of a block to be converted in one call.
+         * We never have to worry about SIGSEGVs here as long as we end in the middle of a managed_buffer.
+         * If we are at the end of managed_buffer there are 2 precautions to prevent SIGSEGVs.
+         * Firstly only a read operation is performed.
+         * Secondly managed_buffers allocate a fixed size memory which is always larger than the actually used size.
+         * e.g. The current sample maximum is 2000 samples in a packet over USB.
+         * With sc12 samples a packet consists of 6000kb but managed_buffers allocate 16kb each.
+         * Thus we don't run into problems here either.
+         */
         const size_t tail_samps = nsamps - o;
         switch (tail_samps)
         {
diff --git a/host/lib/transport/libusb1_base.cpp b/host/lib/transport/libusb1_base.cpp
index 0ef53db0a..8bd0f4354 100644
--- a/host/lib/transport/libusb1_base.cpp
+++ b/host/lib/transport/libusb1_base.cpp
@@ -19,10 +19,12 @@
 #include <uhd/exception.hpp>
 #include <uhd/utils/msg.hpp>
 #include <uhd/utils/log.hpp>
+#include <uhd/utils/tasks.hpp>
 #include <uhd/types/dict.hpp>
 #include <boost/weak_ptr.hpp>
 #include <boost/thread/mutex.hpp>
 #include <boost/foreach.hpp>
+#include <boost/bind.hpp>
 #include <cstdlib>
 #include <iostream>
 
@@ -37,9 +39,11 @@ public:
     libusb_session_impl(void){
         UHD_ASSERT_THROW(libusb_init(&_context) == 0);
         libusb_set_debug(_context, debug_level);
+        task_handler = task::make(boost::bind(&libusb_session_impl::libusb_event_handler_task, this, _context));
     }
 
     ~libusb_session_impl(void){
+        task_handler.reset();
         libusb_exit(_context);
     }
 
@@ -49,6 +53,21 @@ public:
 
 private:
     libusb_context *_context;
+    task::sptr task_handler;
+
+    /*
+     * Task to handle libusb events.  There should only be one thread per libusb_context handling events.
+     * Using more than one thread can result in excessive CPU usage in kernel space (presumably from locking/waiting).
+     * The libusb documentation says it is safe, which it is, but it neglects to state the cost in CPU usage.
+     * Just don't do it!
+     */
+    UHD_INLINE void libusb_event_handler_task(libusb_context *context)
+    {
+        timeval tv;
+        tv.tv_sec = 0;
+        tv.tv_usec = 100000;
+        libusb_handle_events_timeout(context, &tv);
+    }
 };
 
 libusb::session::sptr libusb::session::get_global_session(void){
@@ -274,6 +293,10 @@ public:
         return libusb::device_descriptor::make(this->get_device())->get().idProduct;
     }
 
+    bool firmware_loaded() {
+        return (get_manufacturer() == "Ettus Research LLC");
+    }
+
 private:
     libusb::device::sptr _dev; //always keep a reference to device
 };
diff --git a/host/lib/transport/libusb1_zero_copy.cpp b/host/lib/transport/libusb1_zero_copy.cpp
index 197e257da..2d18e1623 100644
--- a/host/lib/transport/libusb1_zero_copy.cpp
+++ b/host/lib/transport/libusb1_zero_copy.cpp
@@ -73,6 +73,15 @@ struct lut_result_t
     int completed;
     libusb_transfer_status status;
     int actual_length;
+    boost::mutex mut;
+    boost::condition_variable usb_transfer_complete;
+};
+
+// Created to be used as an argument to boost::condition_variable::timed_wait() function
+struct lut_result_completed {
+    const lut_result_t& _result;
+    lut_result_completed(const lut_result_t& result):_result(result) {}
+    bool operator()() const {return (_result.completed ? true : false);}
 };
 
 /*!
@@ -84,48 +93,11 @@ struct lut_result_t
 static void LIBUSB_CALL libusb_async_cb(libusb_transfer *lut)
 {
     lut_result_t *r = (lut_result_t *)lut->user_data;
-    r->completed = 1;
+    boost::lock_guard<boost::mutex> lock(r->mut);
     r->status = lut->status;
     r->actual_length = lut->actual_length;
-}
-
-/*!
- * Wait for a managed buffer to become complete.
- *
- * This routine processes async events until the transaction completes.
- * We must call the libusb handle events in a loop because the handler
- * may complete managed buffers other than the one we are waiting on.
- *
- * We cannot determine if handle events timed out or processed an event.
- * Therefore, the timeout condition is handled by using boost system time.
- *
- * \param ctx the libusb context structure
- * \param timeout the wait timeout in seconds
- * \param completed a reference to the completed flag
- * \return true for completion, false for timeout
- */
-UHD_INLINE bool wait_for_completion(libusb_context *ctx, const double timeout, int &completed)
-{
-    //already completed by a previous call?
-    if (completed) return true;
-
-    //perform a non-blocking event handle
-    timeval tv;
-    tv.tv_sec = 0;
-    tv.tv_usec = 0;
-    libusb_handle_events_timeout_completed(ctx, &tv, &completed);
-    if (completed) return true;
-
-    //finish the rest with a timeout loop
-    const boost::system_time timeout_time = boost::get_system_time() + boost::posix_time::microseconds(long(timeout*1000000));
-    while (not completed and (boost::get_system_time() < timeout_time)){
-        timeval tv;
-        tv.tv_sec = 0;
-        tv.tv_usec = 10000; /*10ms*/
-        libusb_handle_events_timeout_completed(ctx, &tv, &completed);
-    }
-
-    return completed;
+    r->completed = 1;
+    r->usb_transfer_complete.notify_one();  // wake up thread waiting in wait_for_completion() member function below
 }
 
 /***********************************************************************
@@ -154,7 +126,7 @@ public:
     template <typename buffer_type>
     UHD_INLINE typename buffer_type::sptr get_new(const double timeout)
     {
-        if (wait_for_completion(_ctx, timeout, result.completed))
+        if (wait_for_completion(timeout))
         {
             if (result.status != LIBUSB_TRANSFER_COMPLETED) throw uhd::runtime_error(str(boost::format(
                 "usb %s transfer status: %d") % _name % int(result.status)));
@@ -164,9 +136,31 @@ public:
         return typename buffer_type::sptr();
     }
 
+    // This is public because it is accessed from the libusb_zero_copy_single constructor
     lut_result_t result;
 
+    /*!
+     * Wait for a managed buffer to become complete.
+     *
+     * \param timeout the wait timeout in seconds.  A negative value will wait forever.
+     * \return true for completion, false for timeout
+     */
+    UHD_INLINE bool wait_for_completion(const double timeout)
+    {
+        boost::unique_lock<boost::mutex> lock(result.mut);
+        if (!result.completed) {
+            if (timeout < 0.0) {
+                result.usb_transfer_complete.wait(lock);
+            } else {
+                const boost::system_time timeout_time = boost::get_system_time() + boost::posix_time::microseconds(long(timeout*1000000));
+                result.usb_transfer_complete.timed_wait(lock, timeout_time, lut_result_completed(result));
+            }
+        }
+        return result.completed;
+    }
+
 private:
+
     boost::function<void(libusb_zero_copy_mb *)> _release_cb;
     const bool _is_recv;
     const std::string _name;
@@ -252,8 +246,6 @@ public:
 
     ~libusb_zero_copy_single(void)
     {
-        libusb_context *ctx = libusb::session::get_global_session()->get_context();
-
         //cancel all transfers
         BOOST_FOREACH(libusb_transfer *lut, _all_luts)
         {
@@ -261,8 +253,10 @@ public:
         }
 
         //process all transfers until timeout occurs
-        int completed = 0;
-        wait_for_completion(ctx, 0.01, completed);
+        BOOST_FOREACH(libusb_zero_copy_mb *mb, _enqueued)
+        {
+            mb->wait_for_completion(0.01);
+        }
 
         //free all transfers
         BOOST_FOREACH(libusb_transfer *lut, _all_luts)
@@ -276,19 +270,18 @@ public:
     {
         typename buffer_type::sptr buff;
         libusb_zero_copy_mb *front = NULL;
+        boost::mutex::scoped_lock lock(_mutex);
+        if (_enqueued.empty())
         {
-            boost::mutex::scoped_lock l(_mutex);
-            if (_enqueued.empty())
-            {
-                _cond.timed_wait(l, boost::posix_time::microseconds(long(timeout*1e6)));
-            }
-            if (_enqueued.empty()) return buff;
-            front = _enqueued.front();
+            _cond.timed_wait(lock, boost::posix_time::microseconds(long(timeout*1e6)));
         }
+        if (_enqueued.empty()) return buff;
+        front = _enqueued.front();
 
+        lock.unlock();
         buff = front->get_new<buffer_type>(timeout);
+        lock.lock();
 
-        boost::mutex::scoped_lock l(_mutex);
         if (buff) _enqueued.pop_front();
         this->submit_what_we_can();
         return buff;
diff --git a/host/lib/usrp/b100/b100_impl.cpp b/host/lib/usrp/b100/b100_impl.cpp
index 19df3d6af..305ba42a7 100644
--- a/host/lib/usrp/b100/b100_impl.cpp
+++ b/host/lib/usrp/b100/b100_impl.cpp
@@ -310,12 +310,14 @@ b100_impl::b100_impl(const device_addr_t &device_addr){
     _tree->create<std::string>(rx_codec_path / "name").set("ad9522");
     _tree->create<meta_range_t>(rx_codec_path / "gains/pga/range").set(b100_codec_ctrl::rx_pga_gain_range);
     _tree->create<double>(rx_codec_path / "gains/pga/value")
-        .coerce(boost::bind(&b100_impl::update_rx_codec_gain, this, _1));
+        .coerce(boost::bind(&b100_impl::update_rx_codec_gain, this, _1))
+        .set(0.0);
     _tree->create<std::string>(tx_codec_path / "name").set("ad9522");
     _tree->create<meta_range_t>(tx_codec_path / "gains/pga/range").set(b100_codec_ctrl::tx_pga_gain_range);
     _tree->create<double>(tx_codec_path / "gains/pga/value")
         .subscribe(boost::bind(&b100_codec_ctrl::set_tx_pga_gain, _codec_ctrl, _1))
-        .publish(boost::bind(&b100_codec_ctrl::get_tx_pga_gain, _codec_ctrl));
+        .publish(boost::bind(&b100_codec_ctrl::get_tx_pga_gain, _codec_ctrl))
+        .set(0.0);
 
     ////////////////////////////////////////////////////////////////////
     // and do the misc mboard sensors
diff --git a/host/lib/usrp/b200/b200_iface.cpp b/host/lib/usrp/b200/b200_iface.cpp
index 457b380b6..e0809998e 100644
--- a/host/lib/usrp/b200/b200_iface.cpp
+++ b/host/lib/usrp/b200/b200_iface.cpp
@@ -69,6 +69,11 @@ const static boost::uint8_t FX3_STATE_RUNNING = 0x04;
 const static boost::uint8_t FX3_STATE_UNCONFIGURED = 0x05;
 const static boost::uint8_t FX3_STATE_ERROR = 0x06;
 
+const static int VREQ_MAX_SIZE_USB2 = 64;
+const static int VREQ_MAX_SIZE_USB3 = 512;
+const static int VREQ_DEFAULT_SIZE  = VREQ_MAX_SIZE_USB2;
+const static int VREQ_MAX_SIZE      = VREQ_MAX_SIZE_USB3;
+
 typedef boost::uint32_t hash_type;
 
 
@@ -243,10 +248,12 @@ public:
         size_t num_bytes
     ){
         byte_vector_t recv_bytes(num_bytes);
-        fx3_control_read(B200_VREQ_EEPROM_READ,
+        int bytes_read = fx3_control_read(B200_VREQ_EEPROM_READ,
                          0, offset | (boost::uint16_t(addr) << 8),
                          (unsigned char*) &recv_bytes[0],
                          num_bytes);
+        if (bytes_read != num_bytes)
+            throw uhd::io_error("Failed to read data from EEPROM.");
         return recv_bytes;
     }
 
@@ -492,10 +499,24 @@ public:
         hash_type hash = generate_hash(filename);
         hash_type loaded_hash; usrp_get_fpga_hash(loaded_hash);
         if (hash == loaded_hash) return 0;
-
-        unsigned char out_buff[64];
-        memset(out_buff, 0x00, sizeof(out_buff));
-        fx3_control_write(B200_VREQ_FPGA_CONFIG, 0, 0, out_buff, 1, 1000);
+        
+        // Establish default largest possible control request transfer size based on operating USB speed
+        int transfer_size = VREQ_DEFAULT_SIZE;
+        int current_usb_speed = get_usb_speed();
+        if (current_usb_speed == 3)
+            transfer_size = VREQ_MAX_SIZE_USB3;
+        else if (current_usb_speed != 2)
+            throw uhd::io_error("load_fpga: get_usb_speed returned invalid USB speed (not 2 or 3).");
+        
+        UHD_ASSERT_THROW(transfer_size <= VREQ_MAX_SIZE);
+        
+        unsigned char out_buff[VREQ_MAX_SIZE];
+        
+        // Request loopback read, which will indicate the firmware's current control request buffer size
+        int nread = fx3_control_read(B200_VREQ_LOOP, 0, 0, out_buff, sizeof(out_buff), 1000);
+        if (nread <= 0)
+            throw uhd::io_error("load_fpga: unable to complete firmware loopback request.");
+        transfer_size = std::min(transfer_size, nread); // Select the smaller value
 
         size_t file_size = 0;
         {
@@ -510,6 +531,9 @@ public:
             throw uhd::io_error("load_fpga: cannot open FPGA input file.");
         }
 
+        memset(out_buff, 0x00, sizeof(out_buff));
+        fx3_control_write(B200_VREQ_FPGA_CONFIG, 0, 0, out_buff, 1, 1000);
+
         wait_count = 0;
         do {
             fx3_state = get_fx3_status();
@@ -543,14 +567,18 @@ public:
 
         size_t bytes_sent = 0;
         while(!file.eof()) {
-            file.read((char *) out_buff, sizeof(out_buff));
+            file.read((char *) out_buff, transfer_size);
             const std::streamsize n = file.gcount();
             if(n == 0) continue;
 
             boost::uint16_t transfer_count = boost::uint16_t(n);
 
             /* Send the data to the device. */
-            fx3_control_write(B200_VREQ_FPGA_DATA, 0, 0, out_buff, transfer_count, 5000);
+            int nwritten = fx3_control_write(B200_VREQ_FPGA_DATA, 0, 0, out_buff, transfer_count, 5000);
+            if (nwritten <= 0)
+                throw uhd::io_error("load_fpga: cannot write bitstream to FX3.");
+            else if (nwritten != transfer_count)
+                throw uhd::io_error("load_fpga: short write while transferring bitstream to FX3.");
 
             if (load_img_msg)
             {
diff --git a/host/lib/usrp/b200/b200_iface.hpp b/host/lib/usrp/b200/b200_iface.hpp
index 1247d1f86..5b6eeede4 100644
--- a/host/lib/usrp/b200/b200_iface.hpp
+++ b/host/lib/usrp/b200/b200_iface.hpp
@@ -25,7 +25,17 @@
 #include <boost/utility.hpp>
 #include "ad9361_ctrl.hpp"
 
-class b200_iface: boost::noncopyable, public virtual uhd::i2c_iface,
+const static boost::uint16_t B200_VENDOR_ID  = 0x2500;
+const static boost::uint16_t B200_PRODUCT_ID = 0x0020;
+const static boost::uint16_t FX3_VID = 0x04b4;
+const static boost::uint16_t FX3_DEFAULT_PID = 0x00f3;
+const static boost::uint16_t FX3_REENUM_PID = 0x00f0;
+
+static const std::string     B200_FW_FILE_NAME = "usrp_b200_fw.hex";
+static const std::string     B200_FPGA_FILE_NAME = "usrp_b200_fpga.bin";
+static const std::string     B210_FPGA_FILE_NAME = "usrp_b210_fpga.bin";
+
+class UHD_API b200_iface: boost::noncopyable, public virtual uhd::i2c_iface,
                   public ad9361_ctrl_iface_type {
 public:
     typedef boost::shared_ptr<b200_iface> sptr;
@@ -64,6 +74,10 @@ public:
     //! send SPI through the FX3
     virtual void transact_spi( unsigned char *tx_data, size_t num_tx_bits, \
             unsigned char *rx_data, size_t num_rx_bits) = 0;
+
+    virtual void write_eeprom(boost::uint16_t addr, boost::uint16_t offset, const uhd::byte_vector_t &bytes) = 0;
+
+    virtual uhd::byte_vector_t read_eeprom(boost::uint16_t addr, boost::uint16_t offset, size_t num_bytes) = 0;
 };
 
 
diff --git a/host/lib/usrp/b200/b200_impl.cpp b/host/lib/usrp/b200/b200_impl.cpp
index 0da388b93..132b1198d 100644
--- a/host/lib/usrp/b200/b200_impl.cpp
+++ b/host/lib/usrp/b200/b200_impl.cpp
@@ -37,10 +37,6 @@ using namespace uhd;
 using namespace uhd::usrp;
 using namespace uhd::transport;
 
-const boost::uint16_t B200_VENDOR_ID  = 0x2500;
-const boost::uint16_t B200_PRODUCT_ID = 0x0020;
-const boost::uint16_t INIT_PRODUCT_ID = 0x00f0;
-
 static const boost::posix_time::milliseconds REENUMERATION_TIMEOUT_MS(3000);
 
 //! mapping of frontend to radio perif index
@@ -99,7 +95,7 @@ static device_addrs_t b200_find(const device_addr_t &hint)
         catch(const uhd::exception &){continue;} //ignore claimed
 
         //check if fw was already loaded
-        if (handle->get_manufacturer() != "Ettus Research LLC")
+        if (!(handle->firmware_loaded()))
         {
             b200_iface::make(control)->load_firmware(b200_fw_image);
         }
@@ -160,8 +156,15 @@ b200_impl::b200_impl(const device_addr_t &device_addr)
     const fs_path mb_path = "/mboards/0";
 
     //try to match the given device address with something on the USB bus
+    uint16_t vid = B200_VENDOR_ID;
+    uint16_t pid = B200_PRODUCT_ID;
+    if (device_addr.has_key("vid"))
+            sscanf(device_addr.get("vid").c_str(), "%x", &vid);
+    if (device_addr.has_key("pid"))
+            sscanf(device_addr.get("pid").c_str(), "%x", &pid);
+
     std::vector<usb_device_handle::sptr> device_list =
-        usb_device_handle::get_device_list(B200_VENDOR_ID, B200_PRODUCT_ID);
+        usb_device_handle::get_device_list(vid, pid);
 
     //locate the matching handle in the device list
     usb_device_handle::sptr handle;
@@ -252,7 +255,7 @@ b200_impl::b200_impl(const device_addr_t &device_addr)
     ////////////////////////////////////////////////////////////////////
     _async_task_data.reset(new AsyncTaskData());
     _async_task_data->async_md.reset(new async_md_type(1000/*messages deep*/));
-    _async_task = uhd::task::make(boost::bind(&b200_impl::handle_async_task, this, _ctrl_transport, _async_task_data));
+    _async_task = uhd::msg_task::make(boost::bind(&b200_impl::handle_async_task, this, _ctrl_transport, _async_task_data));
 
     ////////////////////////////////////////////////////////////////////
     // Local control endpoint
@@ -474,7 +477,7 @@ b200_impl::b200_impl(const device_addr_t &device_addr)
 
 b200_impl::~b200_impl(void)
 {
-    UHD_SAFE_CALL
+	UHD_SAFE_CALL
     (
         _async_task.reset();
     )
@@ -612,7 +615,7 @@ void b200_impl::setup_radio(const size_t dspno)
 /***********************************************************************
  * loopback tests
  **********************************************************************/
- 
+
 void b200_impl::register_loopback_self_test(wb_iface::sptr iface)
 {
     bool test_fail = false;
diff --git a/host/lib/usrp/b200/b200_impl.hpp b/host/lib/usrp/b200/b200_impl.hpp
index eced4a539..362c45347 100644
--- a/host/lib/usrp/b200/b200_impl.hpp
+++ b/host/lib/usrp/b200/b200_impl.hpp
@@ -44,10 +44,6 @@
 #include <uhd/transport/bounded_buffer.hpp>
 #include <boost/weak_ptr.hpp>
 #include "recv_packet_demuxer_3000.hpp"
-
-static const std::string     B200_FW_FILE_NAME = "usrp_b200_fw.hex";
-static const std::string     B200_FPGA_FILE_NAME = "usrp_b200_fpga.bin";
-static const std::string     B210_FPGA_FILE_NAME = "usrp_b210_fpga.bin";
 static const boost::uint8_t  B200_FW_COMPAT_NUM_MAJOR = 0x03;
 static const boost::uint8_t  B200_FW_COMPAT_NUM_MINOR = 0x00;
 static const boost::uint16_t B200_FPGA_COMPAT_NUM = 0x02;
@@ -120,7 +116,7 @@ struct b200_impl : public uhd::device
     boost::weak_ptr<uhd::tx_streamer> _tx_streamer;
 
     //async ctrl + msgs
-    uhd::task::sptr _async_task;
+    uhd::msg_task::sptr _async_task;
     typedef uhd::transport::bounded_buffer<uhd::async_metadata_t> async_md_type;
     struct AsyncTaskData
     {
@@ -130,7 +126,7 @@ struct b200_impl : public uhd::device
         b200_uart::sptr gpsdo_uart;
     };
     boost::shared_ptr<AsyncTaskData> _async_task_data;
-    void handle_async_task(uhd::transport::zero_copy_if::sptr, boost::shared_ptr<AsyncTaskData>);
+    boost::optional<uhd::msg_task::msg_type_t> handle_async_task(uhd::transport::zero_copy_if::sptr, boost::shared_ptr<AsyncTaskData>);
 
     void register_loopback_self_test(uhd::wb_iface::sptr iface);
     void codec_loopback_self_test(uhd::wb_iface::sptr iface);
diff --git a/host/lib/usrp/b200/b200_io_impl.cpp b/host/lib/usrp/b200/b200_io_impl.cpp
index d643ef855..4fe90bd4a 100644
--- a/host/lib/usrp/b200/b200_io_impl.cpp
+++ b/host/lib/usrp/b200/b200_io_impl.cpp
@@ -139,27 +139,44 @@ bool b200_impl::recv_async_msg(
     return _async_task_data->async_md->pop_with_timed_wait(async_metadata, timeout);
 }
 
-void b200_impl::handle_async_task(
+/*
+ * This method is constantly called in a msg_task loop.
+ * Incoming messages are dispatched in to the hosts radio_ctrl_cores.
+ * The radio_ctrl_core queues are accessed via a weak_ptr to them, stored in AsyncTaskData.
+ * During shutdown the radio_ctrl_core dtor's are called.
+ * An empty peek32(0) is sent out to flush pending async messages.
+ * The response to those messages can't be delivered to the ctrl_core queues anymore
+ * because the shared pointer corresponding to the weak_ptrs is no longer valid.
+ * Those stranded messages are put into a dump_queue implemented in msg_task.
+ * A radio_ctrl_core can search for missing messages there.
+ */
+boost::optional<uhd::msg_task::msg_type_t> b200_impl::handle_async_task(
     uhd::transport::zero_copy_if::sptr xport,
     boost::shared_ptr<AsyncTaskData> data
 )
 {
     managed_recv_buffer::sptr buff = xport->get_recv_buff();
-    if (not buff or buff->size() < 8) return;
+    if (not buff or buff->size() < 8)
+        return NULL;
+
     const boost::uint32_t sid = uhd::wtohx(buff->cast<const boost::uint32_t *>()[1]);
-    switch (sid)
-    {
+    switch (sid) {
 
     //if the packet is a control response
     case B200_RESP0_MSG_SID:
     case B200_RESP1_MSG_SID:
     case B200_LOCAL_RESP_SID:
     {
-        radio_ctrl_core_3000::sptr ctrl;
+    	radio_ctrl_core_3000::sptr ctrl;
         if (sid == B200_RESP0_MSG_SID) ctrl = data->radio_ctrl[0].lock();
         if (sid == B200_RESP1_MSG_SID) ctrl = data->radio_ctrl[1].lock();
         if (sid == B200_LOCAL_RESP_SID) ctrl = data->local_ctrl.lock();
-        if (ctrl) ctrl->push_response(buff->cast<const boost::uint32_t *>());
+        if (ctrl){
+        	ctrl->push_response(buff->cast<const boost::uint32_t *>());
+        }
+        else{
+            return std::make_pair(sid, uhd::msg_task::buff_to_vector(buff->cast<boost::uint8_t *>(), buff->size() ) );
+        }
         break;
     }
 
@@ -204,6 +221,7 @@ void b200_impl::handle_async_task(
     default:
         UHD_MSG(error) << "Got a ctrl packet with unknown SID " << sid << std::endl;
     }
+    return NULL;
 }
 
 /***********************************************************************
@@ -231,7 +249,7 @@ rx_streamer::sptr b200_impl::get_rx_stream(const uhd::stream_args_t &args_)
         //calculate packet size
         static const size_t hdr_size = 0
             + vrt::max_if_hdr_words32*sizeof(boost::uint32_t)
-            + sizeof(vrt::if_packet_info_t().tlr) //forced to have trailer
+            //+ sizeof(vrt::if_packet_info_t().tlr) //forced to have trailer
             - sizeof(vrt::if_packet_info_t().cid) //no class id ever used
             - sizeof(vrt::if_packet_info_t().tsi) //no int time ever used
         ;
diff --git a/host/lib/usrp/cores/radio_ctrl_core_3000.cpp b/host/lib/usrp/cores/radio_ctrl_core_3000.cpp
index 5298fd213..0d6e1c665 100644
--- a/host/lib/usrp/cores/radio_ctrl_core_3000.cpp
+++ b/host/lib/usrp/cores/radio_ctrl_core_3000.cpp
@@ -35,35 +35,27 @@ using namespace uhd::transport;
 
 static const double ACK_TIMEOUT = 2.0; //supposed to be worst case practical timeout
 static const double MASSIVE_TIMEOUT = 10.0; //for when we wait on a timed command
-static const size_t SR_READBACK  = 32;
+static const size_t SR_READBACK = 32;
 
-class radio_ctrl_core_3000_impl : public radio_ctrl_core_3000
+class radio_ctrl_core_3000_impl: public radio_ctrl_core_3000
 {
 public:
 
-    radio_ctrl_core_3000_impl(
-        const bool big_endian,
-        uhd::transport::zero_copy_if::sptr ctrl_xport,
-        uhd::transport::zero_copy_if::sptr resp_xport,
-        const boost::uint32_t sid,
-        const std::string &name
-    ):
-        _link_type(vrt::if_packet_info_t::LINK_TYPE_CHDR),
-        _packet_type(vrt::if_packet_info_t::PACKET_TYPE_CONTEXT),
-        _bige(big_endian),
-        _ctrl_xport(ctrl_xport),
-        _resp_xport(resp_xport),
-        _sid(sid),
-        _name(name),
-        _seq_out(0),
-        _timeout(ACK_TIMEOUT),
-        _resp_queue(128/*max response msgs*/),
-        _resp_queue_size(_resp_xport? _resp_xport->get_num_recv_frames() : 3)
+    radio_ctrl_core_3000_impl(const bool big_endian,
+            uhd::transport::zero_copy_if::sptr ctrl_xport,
+            uhd::transport::zero_copy_if::sptr resp_xport,
+            const boost::uint32_t sid, const std::string &name) :
+            _link_type(vrt::if_packet_info_t::LINK_TYPE_CHDR), _packet_type(
+                    vrt::if_packet_info_t::PACKET_TYPE_CONTEXT), _bige(
+                    big_endian), _ctrl_xport(ctrl_xport), _resp_xport(
+                    resp_xport), _sid(sid), _name(name), _seq_out(0), _timeout(
+                    ACK_TIMEOUT), _resp_queue(128/*max response msgs*/), _resp_queue_size(
+                    _resp_xport ? _resp_xport->get_num_recv_frames() : 3)
     {
-        UHD_LOG << "radio_ctrl_core_3000_impl() " << _name << std::endl;
+        UHD_LOG<< "radio_ctrl_core_3000_impl() " << _name << std::endl;
         if (resp_xport)
         {
-            while (resp_xport->get_recv_buff(0.0)){} //flush
+            while (resp_xport->get_recv_buff(0.0)) {} //flush
         }
         this->set_time(uhd::time_spec_t(0.0));
         this->set_tick_rate(1.0); //something possible but bogus
@@ -74,8 +66,8 @@ public:
         UHD_LOG << "~radio_ctrl_core_3000_impl() " << _name << std::endl;
         _timeout = ACK_TIMEOUT; //reset timeout to something small
         UHD_SAFE_CALL(
-            this->peek32(0); //dummy peek with the purpose of ack'ing all packets
-            _async_task.reset(); //now its ok to release the task
+            this->peek32(0);//dummy peek with the purpose of ack'ing all packets
+            _async_task.reset();//now its ok to release the task
         )
     }
 
@@ -95,7 +87,6 @@ public:
     {
         boost::mutex::scoped_lock lock(_mutex);
         UHD_LOGV(always) << _name << std::hex << " addr 0x" << addr << std::dec << std::endl;
-
         this->send_pkt(SR_READBACK, addr/8);
         this->wait_for_ack(false);
 
@@ -136,6 +127,11 @@ public:
     }
 
 private:
+    // This is the buffer type for messages in radio control core.
+    struct resp_buff_type
+    {
+        boost::uint32_t data[8];
+    };
 
     /*******************************************************************
      * Primary control and interaction private methods
@@ -143,7 +139,7 @@ private:
     UHD_INLINE void send_pkt(const boost::uint32_t addr, const boost::uint32_t data = 0)
     {
         managed_send_buffer::sptr buff = _ctrl_xport->get_send_buff(0.0);
-        if (not buff){
+        if (not buff) {
             throw uhd::runtime_error("fifo ctrl timed out getting a send buffer");
         }
         boost::uint32_t *pkt = buff->cast<boost::uint32_t *>();
@@ -173,12 +169,11 @@ private:
         pkt[packet_info.num_header_words32+0] = (_bige)? uhd::htonx(addr) : uhd::htowx(addr);
         pkt[packet_info.num_header_words32+1] = (_bige)? uhd::htonx(data) : uhd::htowx(data);
         //UHD_MSG(status) << boost::format("0x%08x, 0x%08x\n") % addr % data;
-
         //send the buffer over the interface
         _outstanding_seqs.push(_seq_out);
         buff->commit(sizeof(boost::uint32_t)*(packet_info.num_packet_words32));
 
-        _seq_out++; //inc seq for next call
+        _seq_out++;//inc seq for next call
     }
 
     UHD_INLINE boost::uint64_t wait_for_ack(const bool readback)
@@ -186,7 +181,6 @@ private:
         while (readback or (_outstanding_seqs.size() >= _resp_queue_size))
         {
             UHD_LOGV(always) << _name << " wait_for_ack: " << "readback = " << readback << " outstanding_seqs.size() " << _outstanding_seqs.size() << std::endl;
-
             //get seq to ack from outstanding packets list
             UHD_ASSERT_THROW(not _outstanding_seqs.empty());
             const size_t seq_to_ack = _outstanding_seqs.front();
@@ -218,7 +212,27 @@ private:
             //get buffer from response endpoint - or die in timeout
             else
             {
-                UHD_ASSERT_THROW(_resp_queue.pop_with_timed_wait(resp_buff, _timeout));
+                /*
+                 * Couldn't get message with haste.
+                 * Now check both possible queues for messages.
+                 * Messages should come in on _resp_queue,
+                 * but could end up in dump_queue.
+                 * If we don't get a message --> Die in timeout.
+                 */
+                double accum_timeout = 0.0;
+                const double short_timeout = 0.005; // == 5ms
+                while(not (_resp_queue.pop_with_haste(resp_buff)
+                        || check_dump_queue(resp_buff)
+                        || _resp_queue.pop_with_timed_wait(resp_buff, short_timeout)
+                        )){
+                    /*
+                     * If a message couldn't be received within a given timeout
+                     * --> throw AssertionError!
+                     */
+                    accum_timeout += short_timeout;
+                    UHD_ASSERT_THROW(accum_timeout < _timeout);
+                }
+
                 pkt = resp_buff.data;
                 packet_info.num_packet_words32 = sizeof(resp_buff)/sizeof(boost::uint32_t);
             }
@@ -262,9 +276,33 @@ private:
                 return ((hi << 32) | lo);
             }
         }
+
         return 0;
     }
 
+    /*
+     * If ctrl_core waits for a message that didn't arrive it can search for it in the dump queue.
+     * This actually happens during shutdown.
+     * handle_async_task can't access radio_ctrl_cores queue anymore thus it returns the corresponding message.
+     * msg_task class implements a dump_queue to store such messages.
+     * With check_dump_queue we can check if a message we are waiting for got stranded there.
+     * If a message got stuck we get it here and push it onto our own message_queue.
+     */
+    bool check_dump_queue(resp_buff_type b) {
+        boost::uint32_t recv_sid = (((_sid)<<16)|((_sid)>>16));
+        uhd::msg_task::msg_payload_t msg;
+        do{
+            msg = _async_task->get_msg_from_dump_queue(recv_sid);
+        }
+        while(msg.size() < 8 && msg.size() != 0);
+
+        if(msg.size() >= 8) {
+            memcpy(b.data, &msg.front(), 8);
+            return true;
+        }
+        return false;
+    }
+
     void push_response(const boost::uint32_t *buff)
     {
         resp_buff_type resp_buff;
@@ -272,7 +310,7 @@ private:
         _resp_queue.push_with_haste(resp_buff);
     }
 
-    void hold_task(boost::shared_ptr<void> task)
+    void hold_task(uhd::msg_task::sptr task)
     {
         _async_task = task;
     }
@@ -282,7 +320,7 @@ private:
     const bool _bige;
     const uhd::transport::zero_copy_if::sptr _ctrl_xport;
     const uhd::transport::zero_copy_if::sptr _resp_xport;
-    boost::shared_ptr<void> _async_task;
+    uhd::msg_task::sptr _async_task;
     const boost::uint32_t _sid;
     const std::string _name;
     boost::mutex _mutex;
@@ -292,22 +330,15 @@ private:
     double _tick_rate;
     double _timeout;
     std::queue<size_t> _outstanding_seqs;
-    struct resp_buff_type
-    {
-        boost::uint32_t data[8];
-    };
     bounded_buffer<resp_buff_type> _resp_queue;
     const size_t _resp_queue_size;
 };
 
-
-radio_ctrl_core_3000::sptr radio_ctrl_core_3000::make(
-    const bool big_endian,
-    zero_copy_if::sptr ctrl_xport,
-    zero_copy_if::sptr resp_xport,
-    const boost::uint32_t sid,
-    const std::string &name
-)
+radio_ctrl_core_3000::sptr radio_ctrl_core_3000::make(const bool big_endian,
+        zero_copy_if::sptr ctrl_xport, zero_copy_if::sptr resp_xport,
+        const boost::uint32_t sid, const std::string &name)
 {
-    return sptr(new radio_ctrl_core_3000_impl(big_endian, ctrl_xport, resp_xport, sid, name));
+    return sptr(
+            new radio_ctrl_core_3000_impl(big_endian, ctrl_xport, resp_xport,
+                    sid, name));
 }
diff --git a/host/lib/usrp/cores/radio_ctrl_core_3000.hpp b/host/lib/usrp/cores/radio_ctrl_core_3000.hpp
index a49ca2a4b..51a307c10 100644
--- a/host/lib/usrp/cores/radio_ctrl_core_3000.hpp
+++ b/host/lib/usrp/cores/radio_ctrl_core_3000.hpp
@@ -18,11 +18,12 @@
 #ifndef INCLUDED_LIBUHD_USRP_RADIO_CTRL_3000_HPP
 #define INCLUDED_LIBUHD_USRP_RADIO_CTRL_3000_HPP
 
+#include <uhd/utils/msg_task.hpp>
 #include <uhd/types/time_spec.hpp>
 #include <uhd/transport/zero_copy.hpp>
+#include <uhd/types/wb_iface.hpp>
 #include <boost/shared_ptr.hpp>
 #include <boost/utility.hpp>
-#include <uhd/types/wb_iface.hpp>
 #include <string>
 
 /*!
@@ -43,7 +44,7 @@ public:
     );
 
     //! Hold a ref to a task thats feeding push response
-    virtual void hold_task(boost::shared_ptr<void> task) = 0;
+    virtual void hold_task(uhd::msg_task::sptr task) = 0;
 
     //! Push a response externall (resp_xport is NULL)
     virtual void push_response(const boost::uint32_t *buff) = 0;
diff --git a/host/lib/usrp/dboard/db_dbsrx2.cpp b/host/lib/usrp/dboard/db_dbsrx2.cpp
index 013f3178a..8a8f61a69 100644
--- a/host/lib/usrp/dboard/db_dbsrx2.cpp
+++ b/host/lib/usrp/dboard/db_dbsrx2.cpp
@@ -358,12 +358,12 @@ double dbsrx2::set_gain(double gain, const std::string &name){
  * Bandwidth Handling
  **********************************************************************/
 double dbsrx2::set_bandwidth(double bandwidth){
-    //convert complex bandpass to lowpass bandwidth
-    bandwidth = bandwidth/2.0;
-
     //clip the input
     bandwidth = dbsrx2_bandwidth_range.clip(bandwidth);
 
+    //convert complex bandpass to lowpass bandwidth
+    bandwidth = bandwidth/2.0;
+
     _max2112_write_regs.lp = int((bandwidth/1e6 - 4)/0.29 + 12);
     _bandwidth = double(4 + (_max2112_write_regs.lp - 12) * 0.29)*1e6;
 
diff --git a/host/lib/usrp/dboard/db_sbx_common.cpp b/host/lib/usrp/dboard/db_sbx_common.cpp
index 9db29e65a..5b713c6d7 100644
--- a/host/lib/usrp/dboard/db_sbx_common.cpp
+++ b/host/lib/usrp/dboard/db_sbx_common.cpp
@@ -21,6 +21,137 @@ using namespace uhd;
 using namespace uhd::usrp;
 using namespace boost::assign;
 
+/***********************************************************************
+ * ADF 4350/4351 Tuning Utility
+ **********************************************************************/
+sbx_xcvr::sbx_versionx::adf435x_tuning_settings sbx_xcvr::sbx_versionx::_tune_adf435x_synth(
+    double target_freq,
+    double ref_freq,
+    const adf435x_tuning_constraints& constraints,
+    double& actual_freq)
+{
+    //Default invalid value for actual_freq
+    actual_freq = 0;
+
+    double pfd_freq = 0;
+    boost::uint16_t R = 0, BS = 0, N = 0, FRAC = 0, MOD = 0;
+    boost::uint16_t RFdiv = static_cast<boost::uint16_t>(constraints.rf_divider_range.start());
+    bool D = false, T = false;
+
+    //Reference doubler for 50% duty cycle
+    //If ref_freq < 12.5MHz enable the reference doubler
+    D = (ref_freq <= constraints.ref_doubler_threshold);
+
+    static const double MIN_VCO_FREQ = 2.2e9;
+    static const double MAX_VCO_FREQ = 4.4e9;
+
+    //increase RF divider until acceptable VCO frequency
+    double vco_freq = target_freq;
+    while (vco_freq < MIN_VCO_FREQ && RFdiv < static_cast<boost::uint16_t>(constraints.rf_divider_range.stop())) {
+        vco_freq *= 2;
+        RFdiv *= 2;
+    }
+
+    /*
+     * The goal here is to loop though possible R dividers,
+     * band select clock dividers, N (int) dividers, and FRAC
+     * (frac) dividers.
+     *
+     * Calculate the N and F dividers for each set of values.
+     * The loop exits when it meets all of the constraints.
+     * The resulting loop values are loaded into the registers.
+     *
+     * from pg.21
+     *
+     * f_pfd = f_ref*(1+D)/(R*(1+T))
+     * f_vco = (N + (FRAC/MOD))*f_pfd
+     *    N = f_vco/f_pfd - FRAC/MOD = f_vco*((R*(T+1))/(f_ref*(1+D))) - FRAC/MOD
+     * f_rf = f_vco/RFdiv)
+     * f_actual = f_rf/2
+     */
+    for(R = 1; R <= 1023; R+=1){
+        //PFD input frequency = f_ref/R ... ignoring Reference doubler/divide-by-2 (D & T)
+        pfd_freq = ref_freq*(D?2:1)/(R*(T?2:1));
+
+        //keep the PFD frequency at or below 25MHz (Loop Filter Bandwidth)
+        if (pfd_freq > constraints.pfd_freq_max) continue;
+
+        //ignore fractional part of tuning
+        //N is computed from target_freq and not vco_freq because the feedback
+        //mode is set to FEEDBACK_SELECT_DIVIDED
+        N = boost::uint16_t(std::floor(target_freq/pfd_freq));
+
+        //keep N > minimum int divider requirement
+        if (N < static_cast<boost::uint16_t>(constraints.int_range.start())) continue;
+
+        for(BS=1; BS <= 255; BS+=1){
+            //keep the band select frequency at or below band_sel_freq_max
+            //constraint on band select clock
+            if (pfd_freq/BS > constraints.band_sel_freq_max) continue;
+            goto done_loop;
+        }
+    } done_loop:
+
+    //Fractional-N calculation
+    MOD = 4095; //max fractional accuracy
+    //N is computed from target_freq and not vco_freq because the feedback
+    //mode is set to FEEDBACK_SELECT_DIVIDED
+    FRAC = static_cast<boost::uint16_t>((target_freq/pfd_freq - N)*MOD);
+    if (constraints.force_frac0) {
+        if (FRAC > (MOD / 2)) { //Round integer such that actual freq is closest to target
+            N++;
+        }
+        FRAC = 0;
+    }
+
+    //Reference divide-by-2 for 50% duty cycle
+    // if R even, move one divide by 2 to to regs.reference_divide_by_2
+    if(R % 2 == 0) {
+        T = true;
+        R /= 2;
+    }
+
+    //Typical phase resync time documented in data sheet pg.24
+    static const double PHASE_RESYNC_TIME = 400e-6;
+
+    //actual frequency calculation
+    actual_freq = double((N + (double(FRAC)/double(MOD)))*ref_freq*(D?2:1)/(R*(T?2:1)));
+
+    //load the settings
+    adf435x_tuning_settings settings;
+    settings.frac_12_bit = FRAC;
+    settings.int_16_bit = N;
+    settings.mod_12_bit = MOD;
+    settings.clock_divider_12_bit = std::max<boost::uint16_t>(1, std::ceil(PHASE_RESYNC_TIME*pfd_freq/MOD));
+    settings.r_counter_10_bit = R;
+    settings.r_divide_by_2_en = T;
+    settings.r_doubler_en = D;
+    settings.band_select_clock_div = BS;
+    settings.rf_divider = RFdiv;
+    settings.feedback_after_divider = true;
+
+    UHD_LOGV(often)
+        << boost::format("ADF 435X Frequencies (MHz): REQUESTED=%0.9f, ACTUAL=%0.9f"
+        ) % (target_freq/1e6) % (actual_freq/1e6) << std::endl
+        << boost::format("ADF 435X Intermediates (MHz): VCO=%0.2f, PFD=%0.2f, BAND=%0.2f, REF=%0.2f"
+        ) % (vco_freq/1e6) % (pfd_freq/1e6) % (pfd_freq/BS/1e6) % (ref_freq/1e6) << std::endl
+        << boost::format("ADF 435X Settings: R=%d, BS=%d, N=%d, FRAC=%d, MOD=%d, T=%d, D=%d, RFdiv=%d"
+        ) % R % BS % N % FRAC % MOD % T % D % RFdiv << std::endl;
+
+    UHD_ASSERT_THROW((settings.frac_12_bit          & ((boost::uint16_t)~0xFFF)) == 0);
+    UHD_ASSERT_THROW((settings.mod_12_bit           & ((boost::uint16_t)~0xFFF)) == 0);
+    UHD_ASSERT_THROW((settings.clock_divider_12_bit & ((boost::uint16_t)~0xFFF)) == 0);
+    UHD_ASSERT_THROW((settings.r_counter_10_bit     & ((boost::uint16_t)~0x3FF)) == 0);
+
+    UHD_ASSERT_THROW(vco_freq >= MIN_VCO_FREQ and vco_freq <= MAX_VCO_FREQ);
+    UHD_ASSERT_THROW(settings.rf_divider >= static_cast<boost::uint16_t>(constraints.rf_divider_range.start()));
+    UHD_ASSERT_THROW(settings.rf_divider <= static_cast<boost::uint16_t>(constraints.rf_divider_range.stop()));
+    UHD_ASSERT_THROW(settings.int_16_bit >= static_cast<boost::uint16_t>(constraints.int_range.start()));
+    UHD_ASSERT_THROW(settings.int_16_bit <= static_cast<boost::uint16_t>(constraints.int_range.stop()));
+
+    return settings;
+}
+
 
 /***********************************************************************
  * Register the SBX dboard (min freq, max freq, rx div2, tx div2)
@@ -362,4 +493,3 @@ void sbx_xcvr::flash_leds(void) {
     this->get_iface()->set_gpio_ddr(dboard_iface::UNIT_TX, (TXIO_MASK|TX_LED_IO));
     this->get_iface()->set_gpio_ddr(dboard_iface::UNIT_RX, (RXIO_MASK|RX_LED_IO));
 }
-
diff --git a/host/lib/usrp/dboard/db_sbx_common.hpp b/host/lib/usrp/dboard/db_sbx_common.hpp
index 4f3a2eeaa..e9bb2434c 100644
--- a/host/lib/usrp/dboard/db_sbx_common.hpp
+++ b/host/lib/usrp/dboard/db_sbx_common.hpp
@@ -181,6 +181,34 @@ protected:
         ~sbx_versionx(void) {}
 
         virtual double set_lo_freq(dboard_iface::unit_t unit, double target_freq) = 0;
+    protected:
+        struct adf435x_tuning_constraints {
+            bool            force_frac0;
+            double          ref_doubler_threshold;
+            double          pfd_freq_max;
+            double          band_sel_freq_max;
+            uhd::range_t    rf_divider_range;
+            uhd::range_t    int_range;
+        };
+
+        struct adf435x_tuning_settings {
+            boost::uint16_t frac_12_bit;
+            boost::uint16_t int_16_bit;
+            boost::uint16_t mod_12_bit;
+            boost::uint16_t r_counter_10_bit;
+            bool            r_doubler_en;
+            bool            r_divide_by_2_en;
+            boost::uint16_t clock_divider_12_bit;
+            boost::uint8_t  band_select_clock_div;
+            boost::uint16_t rf_divider;
+            bool            feedback_after_divider;
+        };
+
+        adf435x_tuning_settings _tune_adf435x_synth(
+            double target_freq,
+            double ref_freq,
+            const adf435x_tuning_constraints& constraints,
+            double& actual_freq);
     };
 
     /*!
diff --git a/host/lib/usrp/dboard/db_sbx_version3.cpp b/host/lib/usrp/dboard/db_sbx_version3.cpp
index 2765d530c..b0c9cd18f 100644
--- a/host/lib/usrp/dboard/db_sbx_version3.cpp
+++ b/host/lib/usrp/dboard/db_sbx_version3.cpp
@@ -63,85 +63,21 @@ double sbx_xcvr::sbx_version3::set_lo_freq(dboard_iface::unit_t unit, double tar
         (16, adf4350_regs_t::RF_DIVIDER_SELECT_DIV16)
     ;
 
-    double actual_freq, pfd_freq;
-    double ref_freq = self_base->get_iface()->get_clock_rate(unit);
-    int R=0, BS=0, N=0, FRAC=0, MOD=0;
-    int RFdiv = 1;
-    adf4350_regs_t::reference_divide_by_2_t T     = adf4350_regs_t::REFERENCE_DIVIDE_BY_2_DISABLED;
-    adf4350_regs_t::reference_doubler_t     D     = adf4350_regs_t::REFERENCE_DOUBLER_DISABLED;    
-
-    //Reference doubler for 50% duty cycle
-    // if ref_freq < 12.5MHz enable regs.reference_divide_by_2
-    if(ref_freq <= 12.5e6) D = adf4350_regs_t::REFERENCE_DOUBLER_ENABLED;
-
-    //increase RF divider until acceptable VCO frequency
-    double vco_freq = target_freq;
-    while (vco_freq < 2.2e9) {
-        vco_freq *= 2;
-        RFdiv *= 2;
-    }
-
     //use 8/9 prescaler for vco_freq > 3 GHz (pg.18 prescaler)
     adf4350_regs_t::prescaler_t prescaler = target_freq > 3e9 ? adf4350_regs_t::PRESCALER_8_9 : adf4350_regs_t::PRESCALER_4_5;
 
-    /*
-     * The goal here is to loop though possible R dividers,
-     * band select clock dividers, N (int) dividers, and FRAC 
-     * (frac) dividers.
-     *
-     * Calculate the N and F dividers for each set of values.
-     * The loop exits when it meets all of the constraints.
-     * The resulting loop values are loaded into the registers.
-     *
-     * from pg.21
-     *
-     * f_pfd = f_ref*(1+D)/(R*(1+T))
-     * f_vco = (N + (FRAC/MOD))*f_pfd
-     *    N = f_vco/f_pfd - FRAC/MOD = f_vco*((R*(T+1))/(f_ref*(1+D))) - FRAC/MOD
-     * f_rf = f_vco/RFdiv)
-     * f_actual = f_rf/2
-     */
-    for(R = 1; R <= 1023; R+=1){
-        //PFD input frequency = f_ref/R ... ignoring Reference doubler/divide-by-2 (D & T)
-        pfd_freq = ref_freq*(1+D)/(R*(1+T));
-
-        //keep the PFD frequency at or below 25MHz (Loop Filter Bandwidth)
-        if (pfd_freq > 25e6) continue;
-
-        //ignore fractional part of tuning
-        N = int(std::floor(target_freq/pfd_freq));
-
-        //keep N > minimum int divider requirement
-        if (N < prescaler_to_min_int_div[prescaler]) continue;
-
-        for(BS=1; BS <= 255; BS+=1){
-            //keep the band select frequency at or below 100KHz
-            //constraint on band select clock
-            if (pfd_freq/BS > 100e3) continue;
-            goto done_loop;
-        }
-    } done_loop:
-
-    //Fractional-N calculation
-    MOD = 4095; //max fractional accuracy
-    FRAC = int((target_freq/pfd_freq - N)*MOD);
-
-    //Reference divide-by-2 for 50% duty cycle
-    // if R even, move one divide by 2 to to regs.reference_divide_by_2
-    if(R % 2 == 0){
-        T = adf4350_regs_t::REFERENCE_DIVIDE_BY_2_ENABLED;
-        R /= 2;
-    }
-
-    //actual frequency calculation
-    actual_freq = double((N + (double(FRAC)/double(MOD)))*ref_freq*(1+int(D))/(R*(1+int(T))));
+    adf435x_tuning_constraints tuning_constraints;
+    tuning_constraints.force_frac0 = false;
+    tuning_constraints.band_sel_freq_max = 100e3;
+    tuning_constraints.ref_doubler_threshold = 12.5e6;
+    tuning_constraints.int_range = uhd::range_t(prescaler_to_min_int_div[prescaler], 4095);  //INT is a 12-bit field
+    tuning_constraints.pfd_freq_max = 25e6;
+    tuning_constraints.rf_divider_range = uhd::range_t(1, 16);
 
-    UHD_LOGV(often)
-        << boost::format("SBX Intermediates: ref=%0.2f, outdiv=%f, fbdiv=%f") % (ref_freq*(1+int(D))/(R*(1+int(T)))) % double(RFdiv*2) % double(N + double(FRAC)/double(MOD)) << std::endl
-        << boost::format("SBX tune: R=%d, BS=%d, N=%d, FRAC=%d, MOD=%d, T=%d, D=%d, RFdiv=%d"
-            ) % R % BS % N % FRAC % MOD % T % D % RFdiv << std::endl
-        << boost::format("SBX Frequencies (MHz): REQ=%0.2f, ACT=%0.2f, VCO=%0.2f, PFD=%0.2f, BAND=%0.2f"
-            ) % (target_freq/1e6) % (actual_freq/1e6) % (vco_freq/1e6) % (pfd_freq/1e6) % (pfd_freq/BS/1e6) << std::endl;
+    double actual_freq;
+    adf435x_tuning_settings tuning_settings = _tune_adf435x_synth(
+        target_freq, self_base->get_iface()->get_clock_rate(unit),
+        tuning_constraints, actual_freq);
 
     //load the register values
     adf4350_regs_t regs;
@@ -151,19 +87,25 @@ double sbx_xcvr::sbx_version3::set_lo_freq(dboard_iface::unit_t unit, double tar
     else
         regs.output_power = adf4350_regs_t::OUTPUT_POWER_5DBM;
 
-    regs.frac_12_bit = FRAC;
-    regs.int_16_bit = N;
-    regs.mod_12_bit = MOD;
-    regs.clock_divider_12_bit = std::max(1, int(std::ceil(400e-6*pfd_freq/MOD)));
-    regs.feedback_select = adf4350_regs_t::FEEDBACK_SELECT_DIVIDED;
-    regs.clock_div_mode = adf4350_regs_t::CLOCK_DIV_MODE_RESYNC_ENABLE;
-    regs.prescaler = prescaler;
-    regs.r_counter_10_bit = R;
-    regs.reference_divide_by_2 = T;
-    regs.reference_doubler = D;
-    regs.band_select_clock_div = BS;
-    UHD_ASSERT_THROW(rfdivsel_to_enum.has_key(RFdiv));
-    regs.rf_divider_select = rfdivsel_to_enum[RFdiv];
+    regs.frac_12_bit            = tuning_settings.frac_12_bit;
+    regs.int_16_bit             = tuning_settings.int_16_bit;
+    regs.mod_12_bit             = tuning_settings.mod_12_bit;
+    regs.clock_divider_12_bit   = tuning_settings.clock_divider_12_bit;
+    regs.feedback_select        = tuning_settings.feedback_after_divider ?
+                                    adf4350_regs_t::FEEDBACK_SELECT_DIVIDED :
+                                    adf4350_regs_t::FEEDBACK_SELECT_FUNDAMENTAL;
+    regs.clock_div_mode         = adf4350_regs_t::CLOCK_DIV_MODE_RESYNC_ENABLE;
+    regs.prescaler              = prescaler;
+    regs.r_counter_10_bit       = tuning_settings.r_counter_10_bit;
+    regs.reference_divide_by_2  = tuning_settings.r_divide_by_2_en ?
+                                    adf4350_regs_t::REFERENCE_DIVIDE_BY_2_ENABLED :
+                                    adf4350_regs_t::REFERENCE_DIVIDE_BY_2_DISABLED;
+    regs.reference_doubler      = tuning_settings.r_doubler_en ?
+                                    adf4350_regs_t::REFERENCE_DOUBLER_ENABLED :
+                                    adf4350_regs_t::REFERENCE_DOUBLER_DISABLED;
+    regs.band_select_clock_div  = tuning_settings.band_select_clock_div;
+    UHD_ASSERT_THROW(rfdivsel_to_enum.has_key(tuning_settings.rf_divider));
+    regs.rf_divider_select      = rfdivsel_to_enum[tuning_settings.rf_divider];
 
     //reset the N and R counter
     regs.counter_reset = adf4350_regs_t::COUNTER_RESET_ENABLED;
diff --git a/host/lib/usrp/dboard/db_sbx_version4.cpp b/host/lib/usrp/dboard/db_sbx_version4.cpp
index 27fd68b05..8d95b0655 100644
--- a/host/lib/usrp/dboard/db_sbx_version4.cpp
+++ b/host/lib/usrp/dboard/db_sbx_version4.cpp
@@ -66,85 +66,21 @@ double sbx_xcvr::sbx_version4::set_lo_freq(dboard_iface::unit_t unit, double tar
         (64, adf4351_regs_t::RF_DIVIDER_SELECT_DIV64)
     ;
 
-    double actual_freq, pfd_freq;
-    double ref_freq = self_base->get_iface()->get_clock_rate(unit);
-    int R=0, BS=0, N=0, FRAC=0, MOD=0;
-    int RFdiv = 1;
-    adf4351_regs_t::reference_divide_by_2_t T     = adf4351_regs_t::REFERENCE_DIVIDE_BY_2_DISABLED;
-    adf4351_regs_t::reference_doubler_t     D     = adf4351_regs_t::REFERENCE_DOUBLER_DISABLED;    
-
-    //Reference doubler for 50% duty cycle
-    // if ref_freq < 12.5MHz enable regs.reference_divide_by_2
-    if(ref_freq <= 12.5e6) D = adf4351_regs_t::REFERENCE_DOUBLER_ENABLED;
-
-    //increase RF divider until acceptable VCO frequency
-    double vco_freq = target_freq;
-    while (vco_freq < 2.2e9) {
-        vco_freq *= 2;
-        RFdiv *= 2;
-    }
-
     //use 8/9 prescaler for vco_freq > 3 GHz (pg.18 prescaler)
-    adf4351_regs_t::prescaler_t prescaler = target_freq > 3e9 ? adf4351_regs_t::PRESCALER_8_9 : adf4351_regs_t::PRESCALER_4_5;
-
-    /*
-     * The goal here is to loop though possible R dividers,
-     * band select clock dividers, N (int) dividers, and FRAC 
-     * (frac) dividers.
-     *
-     * Calculate the N and F dividers for each set of values.
-     * The loop exits when it meets all of the constraints.
-     * The resulting loop values are loaded into the registers.
-     *
-     * from pg.21
-     *
-     * f_pfd = f_ref*(1+D)/(R*(1+T))
-     * f_vco = (N + (FRAC/MOD))*f_pfd
-     *    N = f_vco/f_pfd - FRAC/MOD = f_vco*((R*(T+1))/(f_ref*(1+D))) - FRAC/MOD
-     * f_rf = f_vco/RFdiv)
-     * f_actual = f_rf/2
-     */
-    for(R = 1; R <= 1023; R+=1){
-        //PFD input frequency = f_ref/R ... ignoring Reference doubler/divide-by-2 (D & T)
-        pfd_freq = ref_freq*(1+D)/(R*(1+T));
-
-        //keep the PFD frequency at or below 25MHz (Loop Filter Bandwidth)
-        if (pfd_freq > 25e6) continue;
-
-        //ignore fractional part of tuning
-        N = int(std::floor(vco_freq/pfd_freq));
-
-        //keep N > minimum int divider requirement
-        if (N < prescaler_to_min_int_div[prescaler]) continue;
-
-        for(BS=1; BS <= 255; BS+=1){
-            //keep the band select frequency at or below 100KHz
-            //constraint on band select clock
-            if (pfd_freq/BS > 100e3) continue;
-            goto done_loop;
-        }
-    } done_loop:
-
-    //Fractional-N calculation
-    MOD = 4095; //max fractional accuracy
-    FRAC = int((target_freq/pfd_freq - N)*MOD);
-
-    //Reference divide-by-2 for 50% duty cycle
-    // if R even, move one divide by 2 to to regs.reference_divide_by_2
-    if(R % 2 == 0){
-        T = adf4351_regs_t::REFERENCE_DIVIDE_BY_2_ENABLED;
-        R /= 2;
-    }
+    adf4351_regs_t::prescaler_t prescaler = target_freq > 3.6e9 ? adf4351_regs_t::PRESCALER_8_9 : adf4351_regs_t::PRESCALER_4_5;
 
-    //actual frequency calculation
-    actual_freq = double((N + (double(FRAC)/double(MOD)))*ref_freq*(1+int(D))/(R*(1+int(T))));
+    adf435x_tuning_constraints tuning_constraints;
+    tuning_constraints.force_frac0 = false;
+    tuning_constraints.band_sel_freq_max = 100e3;
+    tuning_constraints.ref_doubler_threshold = 12.5e6;
+    tuning_constraints.int_range = uhd::range_t(prescaler_to_min_int_div[prescaler], 4095);  //INT is a 12-bit field
+    tuning_constraints.pfd_freq_max = 25e6;
+    tuning_constraints.rf_divider_range = uhd::range_t(1, 64);
 
-    UHD_LOGV(often)
-        << boost::format("SBX Intermediates: ref=%0.2f, outdiv=%f, fbdiv=%f") % (ref_freq*(1+int(D))/(R*(1+int(T)))) % double(RFdiv*2) % double(N + double(FRAC)/double(MOD)) << std::endl
-        << boost::format("SBX tune: R=%d, BS=%d, N=%d, FRAC=%d, MOD=%d, T=%d, D=%d, RFdiv=%d"
-            ) % R % BS % N % FRAC % MOD % T % D % RFdiv << std::endl
-        << boost::format("SBX Frequencies (MHz): REQ=%0.2f, ACT=%0.2f, VCO=%0.2f, PFD=%0.2f, BAND=%0.2f"
-            ) % (target_freq/1e6) % (actual_freq/1e6) % (vco_freq/1e6) % (pfd_freq/1e6) % (pfd_freq/BS/1e6) << std::endl;
+    double actual_freq;
+    adf435x_tuning_settings tuning_settings = _tune_adf435x_synth(
+        target_freq, self_base->get_iface()->get_clock_rate(unit),
+        tuning_constraints, actual_freq);
 
     //load the register values
     adf4351_regs_t regs;
@@ -154,19 +90,25 @@ double sbx_xcvr::sbx_version4::set_lo_freq(dboard_iface::unit_t unit, double tar
     else
         regs.output_power = adf4351_regs_t::OUTPUT_POWER_5DBM;
 
-    regs.frac_12_bit = FRAC;
-    regs.int_16_bit = N;
-    regs.mod_12_bit = MOD;
-    regs.clock_divider_12_bit = std::max(1, int(std::ceil(400e-6*pfd_freq/MOD)));
-    regs.feedback_select = adf4351_regs_t::FEEDBACK_SELECT_DIVIDED;
-    regs.clock_div_mode = adf4351_regs_t::CLOCK_DIV_MODE_RESYNC_ENABLE;
-    regs.prescaler = prescaler;
-    regs.r_counter_10_bit = R;
-    regs.reference_divide_by_2 = T;
-    regs.reference_doubler = D;
-    regs.band_select_clock_div = BS;
-    UHD_ASSERT_THROW(rfdivsel_to_enum.has_key(RFdiv));
-    regs.rf_divider_select = rfdivsel_to_enum[RFdiv];
+    regs.frac_12_bit            = tuning_settings.frac_12_bit;
+    regs.int_16_bit             = tuning_settings.int_16_bit;
+    regs.mod_12_bit             = tuning_settings.mod_12_bit;
+    regs.clock_divider_12_bit   = tuning_settings.clock_divider_12_bit;
+    regs.feedback_select        = tuning_settings.feedback_after_divider ?
+                                    adf4351_regs_t::FEEDBACK_SELECT_DIVIDED :
+                                    adf4351_regs_t::FEEDBACK_SELECT_FUNDAMENTAL;
+    regs.clock_div_mode         = adf4351_regs_t::CLOCK_DIV_MODE_RESYNC_ENABLE;
+    regs.prescaler              = prescaler;
+    regs.r_counter_10_bit       = tuning_settings.r_counter_10_bit;
+    regs.reference_divide_by_2  = tuning_settings.r_divide_by_2_en ?
+                                    adf4351_regs_t::REFERENCE_DIVIDE_BY_2_ENABLED :
+                                    adf4351_regs_t::REFERENCE_DIVIDE_BY_2_DISABLED;
+    regs.reference_doubler      = tuning_settings.r_doubler_en ?
+                                    adf4351_regs_t::REFERENCE_DOUBLER_ENABLED :
+                                    adf4351_regs_t::REFERENCE_DOUBLER_DISABLED;
+    regs.band_select_clock_div  = tuning_settings.band_select_clock_div;
+    UHD_ASSERT_THROW(rfdivsel_to_enum.has_key(tuning_settings.rf_divider));
+    regs.rf_divider_select      = rfdivsel_to_enum[tuning_settings.rf_divider];
 
     //reset the N and R counter
     regs.counter_reset = adf4351_regs_t::COUNTER_RESET_ENABLED;
diff --git a/host/lib/usrp/usrp1/usrp1_impl.cpp b/host/lib/usrp/usrp1/usrp1_impl.cpp
index 253ac1d6f..625926f36 100644
--- a/host/lib/usrp/usrp1/usrp1_impl.cpp
+++ b/host/lib/usrp/usrp1/usrp1_impl.cpp
@@ -258,12 +258,14 @@ usrp1_impl::usrp1_impl(const device_addr_t &device_addr){
         _tree->create<std::string>(rx_codec_path / "name").set("ad9522");
         _tree->create<meta_range_t>(rx_codec_path / "gains/pga/range").set(usrp1_codec_ctrl::rx_pga_gain_range);
         _tree->create<double>(rx_codec_path / "gains/pga/value")
-            .coerce(boost::bind(&usrp1_impl::update_rx_codec_gain, this, db, _1));
+            .coerce(boost::bind(&usrp1_impl::update_rx_codec_gain, this, db, _1))
+            .set(0.0);
         _tree->create<std::string>(tx_codec_path / "name").set("ad9522");
         _tree->create<meta_range_t>(tx_codec_path / "gains/pga/range").set(usrp1_codec_ctrl::tx_pga_gain_range);
         _tree->create<double>(tx_codec_path / "gains/pga/value")
             .subscribe(boost::bind(&usrp1_codec_ctrl::set_tx_pga_gain, _dbc[db].codec, _1))
-            .publish(boost::bind(&usrp1_codec_ctrl::get_tx_pga_gain, _dbc[db].codec));
+            .publish(boost::bind(&usrp1_codec_ctrl::get_tx_pga_gain, _dbc[db].codec))
+            .set(0.0);
     }
 
     ////////////////////////////////////////////////////////////////////
@@ -407,7 +409,7 @@ usrp1_impl::usrp1_impl(const device_addr_t &device_addr){
         _tree->access<subdev_spec_t>(mb_path / "rx_subdev_spec").set(_rx_subdev_spec);
     if (_tree->list(mb_path / "tx_dsps").size() > 0)
         _tree->access<subdev_spec_t>(mb_path / "tx_subdev_spec").set(_tx_subdev_spec);
- 
+
 }
 
 usrp1_impl::~usrp1_impl(void){
diff --git a/host/lib/utils/paths.cpp b/host/lib/utils/paths.cpp
index f9d8b613c..3e2bea1c6 100644
--- a/host/lib/utils/paths.cpp
+++ b/host/lib/utils/paths.cpp
@@ -34,8 +34,19 @@
 namespace fs = boost::filesystem;
 
 /***********************************************************************
+ * Get a list of paths for an environment variable
+ **********************************************************************/
+static std::string get_env_var(const std::string &var_name, const std::string &def_val = ""){
+    const char *var_value_ptr = std::getenv(var_name.c_str());
+    return (var_value_ptr == NULL)? def_val : var_value_ptr;
+}
+
+static std::vector<fs::path> get_env_paths(const std::string &var_name){
+
+/***********************************************************************
  * Determine the paths separator
  **********************************************************************/
+
 #ifdef UHD_PLATFORM_WIN32
     static const std::string env_path_sep = ";";
 #else
@@ -46,16 +57,6 @@ namespace fs = boost::filesystem;
     boost::tokenizer<boost::char_separator<char> > \
     (inp, boost::char_separator<char>(env_path_sep.c_str()))
 
-/***********************************************************************
- * Get a list of paths for an environment variable
- **********************************************************************/
-static std::string get_env_var(const std::string &var_name, const std::string &def_val = ""){
-    const char *var_value_ptr = std::getenv(var_name.c_str());
-    return (var_value_ptr == NULL)? def_val : var_value_ptr;
-}
-
-static std::vector<fs::path> get_env_paths(const std::string &var_name){
-
     std::string var_value = get_env_var(var_name);
 
     //convert to filesystem path, filter blank paths
@@ -85,6 +86,7 @@ std::vector<fs::path> get_image_paths(void){
 std::vector<fs::path> get_module_paths(void){
     std::vector<fs::path> paths = get_env_paths("UHD_MODULE_PATH");
     paths.push_back(fs::path(uhd::get_pkg_path()) / UHD_LIB_DIR / "uhd" / "modules");
+    paths.push_back(fs::path(uhd::get_pkg_path()) / "share" / "uhd" / "modules");
     return paths;
 }
 
diff --git a/host/lib/utils/tasks.cpp b/host/lib/utils/tasks.cpp
index 1f735de06..08c32a5fb 100644
--- a/host/lib/utils/tasks.cpp
+++ b/host/lib/utils/tasks.cpp
@@ -16,11 +16,13 @@
 //
 
 #include <uhd/utils/tasks.hpp>
+#include <uhd/utils/msg_task.hpp>
 #include <uhd/utils/msg.hpp>
 #include <boost/thread/thread.hpp>
 #include <boost/thread/barrier.hpp>
 #include <exception>
 #include <iostream>
+#include <vector>
 
 using namespace uhd;
 
@@ -80,3 +82,100 @@ private:
 task::sptr task::make(const task_fcn_type &task_fcn){
     return task::sptr(new task_impl(task_fcn));
 }
+
+/*
+ * During shutdown pointers to queues for radio_ctrl_core might not be available anymore.
+ * msg_task_impl provides a dump_queue for such messages.
+ * ctrl_cores can check this queue for stranded messages.
+ */
+
+class msg_task_impl : public msg_task{
+public:
+
+    msg_task_impl(const task_fcn_type &task_fcn):
+        _spawn_barrier(2)
+    {
+        _thread_group.create_thread(boost::bind(&msg_task_impl::task_loop, this, task_fcn));
+        _spawn_barrier.wait();
+    }
+
+    ~msg_task_impl(void){
+        _running = false;
+        _thread_group.interrupt_all();
+        _thread_group.join_all();
+    }
+
+    /*
+     * Returns the first message for the given SID.
+     * This way a radio_ctrl_core doesn't have to die in timeout but can check for stranded messages here.
+     * This might happen during shutdown when dtors are called.
+     * See also: comments in b200_io_impl->handle_async_task
+     */
+    msg_payload_t get_msg_from_dump_queue(boost::uint32_t sid)
+    {
+        boost::mutex::scoped_lock lock(_mutex);
+        msg_payload_t b;
+        for (size_t i = 0; i < _dump_queue.size(); i++) {
+            if (sid == _dump_queue[i].first) {
+                b = _dump_queue[i].second;
+                _dump_queue.erase(_dump_queue.begin() + i);
+                break;
+            }
+        }
+        return b;
+    }
+
+private:
+
+    void task_loop(const task_fcn_type &task_fcn){
+        _running = true;
+        _spawn_barrier.wait();
+
+        try{
+            while (_running){
+            	boost::optional<msg_type_t> buff = task_fcn();
+            	if(buff != boost::none){
+            	    /*
+            	     * If a message gets stranded it is returned by task_fcn and then pushed to the dump_queue.
+            	     * This way ctrl_cores can check dump_queue for missing messages.
+            	     */
+            	    boost::mutex::scoped_lock lock(_mutex);
+            	    _dump_queue.push_back(buff.get() );
+            	}
+            }
+        }
+        catch(const boost::thread_interrupted &){
+            //this is an ok way to exit the task loop
+        }
+        catch(const std::exception &e){
+            do_error_msg(e.what());
+        }
+        catch(...){
+            //FIXME
+            //Unfortunately, this is also an ok way to end a task,
+            //because on some systems boost throws uncatchables.
+        }
+    }
+
+    void do_error_msg(const std::string &msg){
+        UHD_MSG(error)
+            << "An unexpected exception was caught in a task loop." << std::endl
+            << "The task loop will now exit, things may not work." << std::endl
+            << msg << std::endl
+        ;
+    }
+
+    boost::mutex _mutex;
+    boost::thread_group _thread_group;
+    boost::barrier _spawn_barrier;
+    bool _running;
+
+    /*
+     * This queue holds stranded messages until a radio_ctrl_core grabs them via 'get_msg_from_dump_queue'.
+     */
+    std::vector <msg_type_t> _dump_queue;
+};
+
+msg_task::sptr msg_task::make(const task_fcn_type &task_fcn){
+    return msg_task::sptr(new msg_task_impl(task_fcn));
+}