aboutsummaryrefslogtreecommitdiffstats
path: root/host/lib/convert
Commit message (Collapse)AuthorAgeFilesLines
* sc12: fixed byte alignment issueJohannes Demel2013-11-191-4/+57
|
* convert: added sc12 and fc32 convertersJosh Blum2013-08-284-1/+413
|
* convert: sc8 swap for vita complianceJosh Blum2013-08-288-55/+54
|
* orc: added std cmake finder script for orcJosh Blum2013-03-291-5/+4
|
* convert: add sc8 host to/from sc8 wire formatJosh Blum2012-06-122-0/+27
|
* convert: applied sc8 signed convert fixJosh Blum2012-05-101-8/+12
|
* convert: some platform specific fixesJosh Blum2012-05-094-8/+15
| | | | | | | neon: missing header windows: shuffle needs to be deemed "constant expression", fixed with a template parameter.
* convert: squashed converter and sse2 workJosh Blum2012-05-0913-497/+714
|
* Merge branch 'maint'Josh Blum2012-05-091-16/+24
|\
| * 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.
* | convert: added prio param to get converterJosh Blum2012-04-271-20/+23
|/ | | | We can now test generic conversion implementations against SIMD (for example)
* 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
* uhd: fix sc16 to sc8 conversion tableJosh Blum2012-02-291-12/+12
| | | | | | | 1) this was registered as the sc8 to sc16 converter, probably messed that up as well 2) the cast to index was wrong, now unit test passes
* Merge branch 'next'Josh Blum2012-02-1710-14/+508
|\
| * uhd: fixed sse2 conversion fc32 to sc8_item32_beJosh Blum2012-02-091-2/+2
| |
| * uhd: fixed orc conversion fc32 to sc8_item32_beJosh Blum2012-02-091-1/+1
| |
| * uhd: added sse2 conversions for fc32 to sc8Josh Blum2012-02-082-0/+151
| |
| * uhd: added sse2 conversions for fc64 to sc8Josh Blum2012-02-084-9/+166
| |
| * Add Orc functions to convert to sc8. bswap version is a bit of a hack.Nick Foster2012-02-082-0/+28
| |
| * uhd: added sc8 conversion testsJosh Blum2012-02-073-3/+72
| |
| * dsp rework: tx trailer, scaling work (peak)Josh Blum2012-01-313-13/+72
| |
| * uhd: implement convert_sc8to_sc16 table w/ scalarJosh Blum2012-01-311-0/+30
| |
* | usrp1: big endian compile fix, conversion should cast to unsignedJosh Blum2012-02-121-2/+2
|/ | | | The htonx only takes unsigned integers, cast the int16 to uint16.
* convert: added generic conversion for sc8 wire -> sc8 hostJosh Blum2011-11-152-2/+23
|
* convert: move priorities to implementation, different for armJosh Blum2011-11-152-5/+17
|
* convert: msvc warning fixes for sc8 table genJosh Blum2011-11-131-1/+3
|
* uhd: fixed sc8 table conversion, and simplified shiftsJosh Blum2011-11-131-19/+15
|
* convert: added table conversion routines for sc8Josh Blum2011-11-121-17/+95
|
* convert: simplify table conversion with templatesJosh Blum2011-11-121-52/+25
|
* convert: added table conversion routine for sc16 to floatsJosh Blum2011-11-123-2/+141
|
* convert: made conversion functions into classes so they can keep stateJosh Blum2011-11-123-25/+15
|
* uhd: convert should use register_bytes_per_itemJosh Blum2011-11-111-11/+13
|
* uhd: dont pass 0 sample buffs to converter (avoid segfaults)Josh Blum2011-11-071-1/+0
|
* uhd: added converter for item32 as cpu typeJosh Blum2011-11-031-0/+28
|
* uhd: added trailer parsing for occupancyJosh Blum2011-11-031-0/+1
|
* uhd: fixed 8sc item32 converter on head/tail casesJosh Blum2011-11-031-14/+10
|
* uhd: renamed convert markup to formatJosh Blum2011-11-036-47/+42
| | | | | | | | removed convert args added simd level got orc and neon updated
* usrp2: work on alternative OTW formatsJosh Blum2011-11-034-26/+96
|
* usrp1: various tweaks related to streamingJosh Blum2011-11-031-19/+0
|
* usrp1: type conversions and 8-bit workJosh Blum2011-11-032-25/+89
|
* usrp1: multi channel receive workingJosh Blum2011-11-031-1/+1
|
* uhd: restored super packet handler functionalityJosh Blum2011-11-031-0/+10
|
* convert: reworked convert to use new identification standardJosh Blum2011-11-037-265/+114
|
* uhd: fill in other fields of pkg-config fileJosh Blum2011-09-011-8/+8
|
* uhd: if found, require a liborc version orc-0.4 > 0.4.11Josh Blum2011-08-041-1/+1
|
* e100: perfer orc if available and dont build neon intrinsic supportJosh Blum2011-07-221-41/+48
|
* uhd: created SSE2 conversion routines for fc64Josh Blum2011-07-173-7/+221
|
* uhd: pulled misc changes from other branches into masterJosh Blum2011-07-141-20/+20
|
* uhd: fixed sse2 conversion bounds checkJosh Blum2011-06-271-4/+4
|
* uhd: use aligned loads and stores for floats in sse2 conversion routinesJosh Blum2011-06-151-78/+126
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-27 10:08:59 +0000