aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* usrp: fix rx dsp core scaling factor off by 2Josh Blum2012-09-111-1/+1
| | | | The scale factor is a 18 bit number, this should be 1 << 17
* xcvr2450: disable transmit PAs when receivingJosh Blum2012-08-011-2/+2
|
* uhd: bumped patch version number to 3.4.3release_003_004_003Josh Blum2012-07-301-1/+1
|
* e100: set vita header offset for previous FPGA changesetJosh Blum2012-07-161-1/+4
|
* Merge branch 'fpga_maint' into maintJosh Blum2012-07-163-43/+6
|\
| * e100: offset gpmc to fifo writes by 2 transfersJosh Blum2012-07-152-5/+5
| | | | | | | | This effectivly works around bus initial transaction issues.
| * e100: reverted commit registering in gpmcJosh Blum2012-07-151-38/+1
| | | | | | | | There is a subtle bus issue that the last changset did not address.
* | n2xx: fix for usrp_n2xx_net_burner addr decode failureJosh Blum2012-07-121-1/+1
| | | | | | | | | | | | | | | | In the windows network address extractor: The except: continue line could get use stuck in an infinite loop. This fix sets addr to None so the code below it does not execute, and the next node in the chain is tested as expected.
* | uhd: docstring typo fix version.hppJosh Blum2012-07-041-1/+1
| |
* | cmake: got rid of remaining references of UHD_RELEASE_MODE being treated as ↵Nicholas Corgan2012-07-021-0/+3
| | | | | | | | a boolean
* | usrp1: revert calculation for DAC freq outside of 1st NyquistJosh Blum2012-06-281-36/+2
| | | | | | | | This patch does not work. A possibile solution will be added to master.
* | xcvr2450: fix to disable LO offset for TX sideJosh Blum2012-06-181-1/+1
| | | | | | | | | | | | | | | | | | | | XCVR2450 has a common LO for RX and TX. The use LO offset should be the same for both sides. When different, a tune for one side will override the other side, because the DSP will not know the compensation factor after the other side is tuned. Setting both use_lo_offset to false is how this was pre 3.4 release.
* | n2x0: fix for net_burner, continue on errorJosh Blum2012-06-081-1/+1
| |
* | usrp1: shutoff DAC digital w/ TX state machineJosh Blum2012-05-301-1/+6
| |
* | examples: fix tx bursts so we only set EOB at the endJosh Blum2012-05-301-9/+5
| |
* | uhd: bumped patch version number to 3.4.2release_003_004_002Josh Blum2012-05-221-1/+1
| |
* | Merge branch 'fpga_maint' into maintJosh Blum2012-05-221-1/+38
|\|
| * Added registers for gpmc-to-fifo interface to address sequence errors for ↵Al Fayez2012-05-221-1/+38
| | | | | | | | E100/E110
* | b100: fix for 61.44 MHz clock mod docsJosh Blum2012-05-211-6/+6
| |
* | usrp1: removed print from codec_ctrl last commitJosh Blum2012-05-171-1/+1
| |
* | Fixing USRP1 aliasing logicNicholas Corgan2012-05-171-2/+36
| |
* | docs: removed unsupported gps_gpgsa from docs as wellJosh Blum2012-05-161-1/+1
| |
* | Removed GPGSA from GPS driver since Firefly doesn't actually support it.Nick Foster2012-05-161-5/+1
| |
* | b100: fix use of boost cstdint hereJosh Blum2012-05-161-1/+1
| |
* | Throwing here causes difficulty initializing new USRPsJason Abele2012-05-141-1/+1
| |
* | usrp1: ensure frontend specs are init'd to somethingJosh Blum2012-05-111-0/+2
| | | | | | | | | | This helps the case of 4x DDC no DUC for example, that way at least something empty is set to the property.
* | Merge branch 'fpga_maint' into maintJosh Blum2012-05-102-7/+9
|\|
| * e100: bump compat minor for xclock reader fixJosh Blum2012-05-101-1/+1
| |
| * fpga: xclock fix for edge case conditionJosh Blum2012-05-081-6/+8
| |
* | n2xx: fix possible attribute error in netburner win_get_interfacesJosh Blum2012-05-091-3/+3
| | | | | | | | | | | | It is possible on some machines (at least one in this case), for adNode to not have the attribute next. Unexplained... This may be a fix, and if not, should be harmless.
* | convert: fixed sc8 convert issue w/ undefined behaviourJosh Blum2012-05-081-16/+24
| | | | | | | | | | | | | | Its important to use a signed cast when converting float to int. Then assign that signed int to an unsigned type of the same width. Its undefined behaviour when converting a negative float to an unsigned int.
* | usrp2: fix ad9510 register map typosJosh Blum2012-04-241-13/+13
| | | | | | | | | | | | These register addresses should be in hex. Fortunately, they are not set in the code, so the typo did not break anything in UHD.
* | uhd: fix missing API export for tasks.hppJosh Blum2012-04-231-2/+2
| |
* | usrp2: removed errors='ignore' for burners for py26Josh Blum2012-04-222-2/+2
| |
* | master: Fixing comments that say SPI is LSB.Ben Hilburn2012-04-181-2/+2
| |
* | n2x0: card burner fixes for python3.0 and IPv6Josh Blum2012-04-181-3/+5
| | | | | | | | | | | | has_key is not in python3.0, use key in dict put try catch around interface enumeration, might be an ipv6 issue
* | usb: do not release recv buffer in wrapperJosh Blum2012-04-171-3/+2
| | | | | | | | | | | | | | The actual recv buffer will be automatically released when dereferenced. By releasing this buffer early we allowed for a race condition: Subsequent wrapper buffers that shared the same actual buffer could get their memory filled by new recvd packets from the USB layer.
* | recv: only inspect tsf on timestamp error checkJosh Blum2012-04-171-1/+1
| |
* | convert: give SIMD conversions prio over tableJosh Blum2012-04-161-2/+2
| | | | | | | | | | | | | | | | given performance benchmarks involving the converter, when the driver is also doing other things sc8 conversions will still involve tables since no SIMD implementations exist yet
* | cmake: UHDVersion.cmake gets rid of git commit number if 'release' is ↵Nicholas Corgan2012-04-131-1/+5
| | | | | | | | specified in UHD_RELEASE_MODE
* | utils: card/net burner string decodeJosh Blum2012-04-122-2/+2
| |
* | uhd: bumped version num to 3.4.1 for releaserelease_003_004_001Josh Blum2012-04-101-1/+1
| |
* | Merge branch 'fpga_maint' into maintJosh Blum2012-04-095-6/+6
|\|
| * vita: moved clear register to overlap with nchan registerJosh Blum2012-04-095-6/+6
| | | | | | | | | | | | | | | | | | | | | | | | This fixes the bug where setting the format clears the vita RX. This is only an issue when the noclear option is set by UHD, because the format register is always so, so it always clears. Note: noclear is there to support the backwards compat API (pre streamer). Now, numchans and clear overlap. This is ok because in the host code, clear and numchans are always used together. All timing meets on N2xx and USRP2.
* | dsp: clear register now overlaps with numchan register.Josh Blum2012-04-091-4/+2
| | | | | | | | This fixes the bug of unwanted clearing when setting format.
* | usrp: fix set_clock_config typo for external and mimo clock refJosh Blum2012-04-061-2/+2
| | | | | | | | | | | | This is a typo in the set_clock_config implementation. However, the enum values are the same, so this would not cause a bug. Fixed although set_clock_config is a deprecated interface.
* | Merge branch 'fpga_maint' into maintJosh Blum2012-04-022-5/+5
|\|
| * b100: fix slave fifo data xfer exit conditionJosh Blum2012-04-012-5/+5
| | | | | | | | | | | | | | When exiting the read/write data state, when the transfer count maxes out/peaks, the fifo read/write signals were getting this condition the cycle after with the state change.
* | examples: fix bsd compilation for network relay exampleJosh Blum2012-04-011-3/+3
| | | | | | | | | | | | | | Fixes compilation on platforms that dont have one of the OS defines by providing #else case for rx_dsp_buff_size constant. This also makes OSX the special case and assumes other OS can handle the large RX socket buffer size.
* | usrp1: stop threads in deconstructorJosh Blum2012-03-293-4/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Its important to stop the threads before we let the other smart point objects naturally deconstruct to avoid thread-based race conditions. The attempt to deconstruct the tree and soft time ctrl had a bug because the tree had references in subtrees within the dboard manager class. Rather than continue to fix this method and deconstruct the tree to free up soft time ctrl, it seems simpler to just stop the thread in soft time ctrl, and then let it naturally deconstruct later by ref count.
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-20 12:27:44 +0000