diff options
Diffstat (limited to 'host/docs/transport.rst')
| -rw-r--r-- | host/docs/transport.rst | 134 |
1 files changed, 134 insertions, 0 deletions
diff --git a/host/docs/transport.rst b/host/docs/transport.rst new file mode 100644 index 000000000..b601cd8ff --- /dev/null +++ b/host/docs/transport.rst @@ -0,0 +1,134 @@ +======================================================================== +UHD - Transport Application Notes +======================================================================== + +.. contents:: Table of Contents + +------------------------------------------------------------------------ +Introduction +------------------------------------------------------------------------ +A transport is the layer between the packet interface and a device IO interface. +The advanced user can pass optional parameters +into the underlying transport layer through the device address. +These optional parameters control how the transport object allocates memory, +resizes kernel buffers, spawns threads, etc. +When not spcified, the transport layer will use values for these parameters +that are known to perform well on a variety of systems. +The transport parameters are defined below for the various transports in the UHD: + +------------------------------------------------------------------------ +UDP transport (sockets) +------------------------------------------------------------------------ +The UDP transport is implemented with user-space sockets: + +* **UNIX:** standard Berkeley sockets API using send()/recv() +* **Windows:** Windows Sockets API (WSA) using overlapped IO + +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +Transport parameters +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +The following parameters can be used to alter the transport's default behavior: + +* **recv_frame_size:** The size of a single receive buffer in bytes +* **num_recv_frames:** The number of receive buffers to allocate +* **send_frame_size:** The size of a single send buffer in bytes +* **num_send_frames:** The number of send buffers to allocate + +**Note1:** +num_recv_frames does not affect performance (all platforms). + +**Note2:** +num_send_frames does not affect performance (UNIX only). + +**Note3:** +recv_frame_size and send_frame_size can be used to +increase or decrease the maximum number of samples per packet. +The frame sizes default to an MTU of 1472 bytes per IP/UDP packet, +and may be increased if permitted by your network hardware. + +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +Flow control parameters +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +The host-based flow control expects periodic update packets from the device. +These update packets inform the host of the last packet consumed by the device, +which allows the host to determine throttling conditions for the transmission of packets. +The following mechanisms affect the transmission of periodic update packets: + +* **ups_per_fifo:** The number of update packets for each FIFO's worth of bytes sent into the device +* **ups_per_sec:** The number of update packets per second (defaults to 20 updates per second) + +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +Resize socket buffers +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +It may be useful increase the size of the socket buffers to +move the burden of buffering samples into the kernel, or to +buffer incoming samples faster than they can be processed. +However, if your application cannot process samples fast enough, +no amount of buffering can save you. +The following parameters can be used to alter socket's buffer sizes: + +* **recv_buff_size:** The desired size of the receive buffer in bytes +* **send_buff_size:** The desired size of the send buffer in bytes + +**Note:** Large send buffers tend to decrease transmit performance. + +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +Latency Optimization +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +Latency is a measurement of the time it takes a sample to travel between the host and device. +Most computer hardware and software is bandwidth optimized which may negatively affect latency. +If your application has strict latency requirements, please consider the following notes: + +**Note1:** +The time taken by the device to populate a packet is proportional to the sample rate. +Therefore, to improve receive latency, configure the transport for a smaller frame size. + +**Note2:** +For overall latency improvements, +look for "Interrupt Coalescing" settings for your OS and ethernet chipset. +It seems the Intel ethernet chipsets offer fine-grained control in Linux. +Also, consult: + +* http://publib.boulder.ibm.com/infocenter/pseries/v5r3/index.jsp?topic=/com.ibm.aix.prftungd/doc/prftungd/interrupt_coal.htm + +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +Linux specific notes +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +On linux, the maximum buffer sizes are capped by the sysctl values +**net.core.rmem_max** and **net.core.wmem_max**. +To change the maximum values, run the following commands: +:: + + sudo sysctl -w net.core.rmem_max=<new value> + sudo sysctl -w net.core.wmem_max=<new value> + +Set the values permanently by editing */etc/sysctl.conf* + +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +Windows specific notes +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +On Windows, it is important to change the default UDP behavior such that +1500 byte packets still travel through the fast path of the sockets stack. +FastSendDatagramThreshold registry key to change documented here: + +* http://www.microsoft.com/windows/windowsmedia/howto/articles/optimize_web.aspx#appendix_e + +------------------------------------------------------------------------ +USB transport (libusb) +------------------------------------------------------------------------ +The USB transport is implemented with libusb. +Libusb provides an asynchronous API for USB bulk transfers. +The transport implementation allocates a number of buffers +and submits asynchronous requests through libusb. +Event handler threads run in the background to process these requests. + +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +Transport parameters +^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ +The following parameters can be used to alter the transport's default behavior: + +* **recv_frame_size:** The size of a single receive transfers in bytes +* **num_recv_frames:** The number of simultaneous receive transfers +* **send_frame_size:** The size of a single send transfers in bytes +* **num_send_frames:** The number of simultaneous send transfers +* **concurrency_hint:** The number of threads to run the event handler |