aboutsummaryrefslogtreecommitdiffstats
path: root/host/lib
Commit message (Collapse)AuthorAgeFilesLines
* Merge branch 'maint'Martin Braun2016-04-222-10/+18
|\ | | | | | | | | | | | | Conflicts: fpga-src host/cmake/Modules/UHDVersion.cmake host/lib/usrp/b200/b200_impl.cpp
| * b200: Fixed setting of max rateMartin Braun2016-04-222-10/+18
| | | | | | | | | | | | | | DSP rates are now being tracked as in whether or not they've been set. We can disregard unset DSPs for the automatic clock rate calculation. Reviewed-By: Derek Kozel <derek.kozel@ettus.com>
* | Merge branch 'maint'Martin Braun2016-04-201-0/+2
|\|
| * e3xx: spi: Fix issue introduced in 1b149f56Moritz Fischer2016-04-191-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Kernels (3.15+) introduce the possibility to do DUAL and QUAD spi operations via spidev. Prior to this commit nothing was setting the {tx,rx}_nbits members of the struct spi_ioc_transfer. from include/uapi/linux/spi/spidev.h struct spi_ioc_transfer { __u64 tx_buf; __u64 rx_buf; __u32 len; __u32 speed_hz; __u16 delay_usecs; __u8 bits_per_word; __u8 cs_change; __u8 tx_nbits; __u8 rx_nbits; __u16 pad; }; This turns into an issue on more recent kernels, where it turns all transactions into QUAD transactions, while the controller actually doesn't support that mode of operation. Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
* | vita core: Fix the hard reset issue by flushing the flow control windowPaul David2016-04-181-0/+13
| |
* | transport optimize: Integrated the transport offloading into the X3XX codebasePaul David2016-04-185-54/+227
| |
* | transport optimize: Added a thread transport offload to share the workloadPaul David2016-04-182-0/+159
| |
* | Added option for writes to specify a SPI speedDerek Kozel2016-04-082-7/+22
| |
* | Merge branch 'maint'Martin Braun2016-04-071-2/+2
|\| | | | | | | | | Conflicts: host/lib/usrp/cores/gpio_core_200.cpp
| * gpio core: Fixed a mismatch in the address assignment for RX and full duplexPaul David2016-04-061-2/+2
| |
* | E310: Radio reset fixmichael-west2016-03-291-1/+1
| | | | | | | | | | | | - Updated images - E310: Bumped compat number to 16 for radio reset fix - Update fpga-src submodule pointer
* | Merge branch 'maint'Martin Braun2016-03-291-1/+1
|\|
| * ubx: Changed member declaration to satisfy debug builds on WindowsMartin Braun2016-03-291-1/+1
| |
* | Merge branch 'maint'Martin Braun2016-03-251-2/+46
|\|
| * Unroll the loops in the NEON float to/from integer converters.Philip Balister2016-03-231-2/+46
| | | | | | | | | | | | | | | | Unrolling the RX loop showed marked improvement with perf. The TX path was only slightly better. Checked signal correctness with shinysdr to verify received signal and tx_waveforms into a spectrum analyzer for TX. Signed-off-by: Philip Balister <philip@opensdr.com>
| * Added missing stdint.h includeNicholas Corgan2016-03-221-1/+2
| |
| * transport: Also send flow control ACKs on sequence errorsMartin Braun2016-03-161-0/+6
| |
* | x300: DSP updates to support heterodyne dboardsAshish Chaudhari2016-03-212-2/+7
| |
* | dboard_iface: Added method to configure front connection and settingsAshish Chaudhari2016-03-217-0/+36
| |
* | usrp3: Added RX DSP support for heterodyne samplingAshish Chaudhari2016-03-215-17/+46
| | | | | | | | | | | | | | | | | | - rx_dsp_core: set_mux takes in a fe_connection obj to determine IQ mapping and sampling mode parameters - rx_dsp_core: Support a non-zero IF for downconversion The current strategy applies a software DSP freq offset which eats into the range of the CORDIC which is OK because heterodyne assumes real-mode sampling
* | usrp: Added fe_connection type and unit testAshish Chaudhari2016-03-212-0/+68
| | | | | | | | | | - Wraps a sampling mode and IF frequency - Built-in parser to deduce swap,invert,mode bits from string connection
* | experts: Print values of data nodes when debuggingAshish Chaudhari2016-03-212-2/+32
| |
* | experts: Added additional assignment operator to data writerAshish Chaudhari2016-03-211-1/+6
| |
* | Updated code to work for new minimum dependenciesNicholas Corgan2016-03-213-24/+5
| | | | | | | | | | * Removed code referencing now-unsupported versions of Boost * Added <stdint.h> includes where needed
* | transport: Also send flow control ACKs on sequence errorsMartin Braun2016-03-211-0/+6
| |
* | log: remove MSVC 2010 Boost bug workaroundNicholas Corgan2016-03-111-13/+1
| |
* | Merge branch 'maint'Martin Braun2016-03-115-7/+10
|\|
| * Fixed minor warnings:Nicholas Corgan2016-03-043-5/+5
| | | | | | | | | | | | * nirio_driver_iface_win: labeled unused variable for MinGW builds * b200_impl: fixed unreferenced variable warning * n200_image_loader: fixed signed vs. unsigned comparison
| * x300: Added power cycle message to uhd_image_loaderMartin Braun2016-03-031-0/+2
| |
| * e300: Fixed a memory leak in udev codeMarcus Müller2016-03-031-2/+3
| | | | | | | | | | When querying temp mboard sensor on e300, sysfs attributes are read through udev.
* | Merge branch 'maint'Martin Braun2016-02-291-4/+6
|\|
| * Enable multiple programs to use USB USRPs on WindowsDerek Kozel2016-02-291-4/+6
| | | | | | | | | | | | | | Window's WinUSB driver doesn't support multiple processes accessing a single USB device and libusb_open returns LIBUSB_ACCESS_ERROR when trying to access an already claimed USRP. One device access did not catch this exception and caused UHD to error during USRP discovery.
* | dboards: Added APIs to get RX and TX frontend namesAshish Chaudhari2016-02-261-1/+21
| |
* | experts: Multiple minor fixes/enhancementsAshish Chaudhari2016-02-263-10/+31
| | | | | | | | | | | | | | | | - Fixed segfault issue for graph modification errors - Demangled node type names and added quotes to "dot" labels to allow fully qualified C++ names - add_prop_node functions initialize the property objects in the tree in addition to data nodes - Passed in resolve mutex to desired data node correctly - Added == and != operators to data accessors
* | dboards: Added ability to register a per-dboard container classAshish Chaudhari2016-02-263-40/+101
| | | | | | | | | | | | | | | | | | | | - The typical dboard classes are actually "subdev" classes i.e. there is one instance per dboard subdev (front-end). This makes it hard to implement shared functionality between multiple front-ends. - This changes adds the ability to create a container class which is created per group of subdevs and each subdev gets a pointer to the container class for cross linkage.
* | dboard: Made dboard class dtors virtualAshish Chaudhari2016-02-2612-39/+43
| | | | | | | | | | - Interface and all base classes have virtual destructors - Wrapped UBX dtor in UHD_SAFE_CALL
* | Merge branch 'maint'Ashish Chaudhari2016-02-1810-105/+346
|\| | | | | | | | | | | Conflicts: host/lib/usrp/cores/gpio_core_200.cpp host/lib/usrp/dboard/db_ubx.cpp
| * UBX: Phase synchronizationmichael-west2016-02-187-99/+340
| | | | | | | | | | | | | | - Disabled MAX2871 VCO auto selection for phase sync - Added checks for new phase sync constraints recently published by Maxim - Added dboard_clock_rate option for X300 - Adjusted timing of SYNC signal relative to dboard referenc clock
| * Corrected the UHD behavior in the event of a USB disconnectDaulPavid2016-02-122-5/+5
| |
| * Rectified the usage of UHD_IMAGES_DIR_WINREG_KEY in search paths.vkakade2016-02-111-1/+1
| |
* | x300: Added capability to write_spi to BOTH channels in dboard_ifaceAshish Chaudhari2016-02-181-6/+8
| |
* | db_manager: Cleaned up iface pointer in subtree after errorAshish Chaudhari2016-02-181-0/+1
| |
* | x300,e300: Fixed IQ swapping issue in RX frontendAshish Chaudhari2016-02-172-10/+4
| | | | | | | | - DB connection mapping is implemented in DDC chain. FE corrections module does not swap
* | usrp: Refactored dboard_iface for all productsAshish Chaudhari2016-02-1611-264/+411
| | | | | | | | | | | | | | - Made dboard_iface an interface! Removed PIMPL - Added unit "BOTH" to dboard API and expanded GPIO API width to 32 - Removed gpio_debug_mux. No product ever used that - Refactored gpio_atr cores to work with new dboard_iface
* | dboard: Added restricted dboard registration capabilityAshish Chaudhari2016-02-1611-24/+42
| | | | | | | | | | | | - Moved dboard iface initialization to dboard_manager - Added a restricted register function. Restricted dboards don't expose their control iface in the property tree
* | b200: Removed superfluous function declarationMartin Braun2016-02-151-1/+0
| |
* | adf435x: Refactored ADF435X control codeAshish Chaudhari2016-02-1213-678/+524
| | | | | | | | | | | | - Removed adf435x_common and replaced with a real encapsulated interface - Looks similar to the MAX287X code - Updated all DB classes to use the new common code
* | uhd: Added expert dependency tracking infrastructureAshish Chaudhari2016-02-127-0/+1551
| | | | | | | | | | | | | | - Code location uhd/lib/experts - expert_nodes.hpp contains all node definitions: data and worker - expert_container.hpp contains the memory manager and resolver for expert nodes - expert_factory contains the initializer and modifier for expert_container
* | prop_tree: Added advanced coercion capability to propertyAshish Chaudhari2016-02-111-1/+1
| | | | | | | | | | | | - Added auto and manual coerce modes - Added set_coerced API for manual coercion - Added detailed doxy comments describing behavior of property class
* | prop_tree: Multiple API enhancements to uhd::propertyAshish Chaudhari2016-02-1128-403/+403
| | | | | | | | | | | | | | | | | | - Added desired and coerced values and accessors to property - Added support to register desired subscribers - set APIs don't reallocate storage for a property value - Renamed callback method registration APIs - Registering 2 coercers or publishers for a property will throw - Registering a coercer and a publisher for the same property will throw