summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2012-02-13core: fixed alignment call for char_to_short.Tom Rondeau
2012-02-13volk: adding an examples directory with scripts to benchmark and compare ↵Tom Rondeau
volk-optimized GR blocks.
2012-02-13core: using volk for conjugate block and added QA code for it.Tom Rondeau
2012-02-13volk: adding complex conjugate kernel.Tom Rondeau
2012-02-13volk: fixed complex multiply and conjugate kernel to use xor for conjugation.Tom Rondeau
2012-02-13core: added Volk-optimized gr_multiply_conjugate_cc at one block with QA code.Tom Rondeau
2012-02-13volk: added a 32fc multiply conjugate kernel.Tom Rondeau
2012-02-13core: added volk version of add_ff.Tom Rondeau
2012-02-13volk: added unaligned version of adding 2 vectors.Tom Rondeau
2012-02-13core: new multiply_const_ff and multiply_ff blocks done using Volk.Tom Rondeau
2012-02-13volk: new unaligned versions of float multipliers.Tom Rondeau
2012-02-13sched: some added protections and checks for the alignment states.Tom Rondeau
2012-02-13core: moving multiply_const_ff from gengen to general to take advantage of volk.Tom Rondeau
Also adds SSE and AVX and unaligned Volk versions for this.
2012-02-02core: more fixes when moving files from gengen to general.Tom Rondeau
2012-02-02volk: adding new functions to Makefile.Tom Rondeau
2012-02-02volk: adding new Volk functions added in this branch to the volk_profile ↵Tom Rondeau
tool and installing it into $prefix/bin.
2012-02-02core: moving multiply_cc and multiply_const_cc out of gengen and into ↵Tom Rondeau
general so they can make use of volk calls. QA code now explicitly tests the cc versions of these blocks.
2012-02-02volk: improving performance of multiply_const and multiply two streams.Tom Rondeau
2012-02-02sched: better working alignment handling.Tom Rondeau
Works with max_noutput_items and set_output_multiple.
2012-02-02volk: adding unaligned versions of complex multiply a constant and complex ↵Tom Rondeau
multiply 2 streams.
2012-01-30core: redo fft_filter (complex and float) with Volk.Tom Rondeau
No need for sse implementation now but keeping code for reference.
2012-01-30core: fixing up complex_to_xxx for using Volk where appropriate. Speed ↵Tom Rondeau
benchmark were used to decide which implementation to use.
2012-01-30core: complex_to_xxx (float, real, imag, arg) to volk.Tom Rondeau
2012-01-30volk: adding complex to imag kernel.Tom Rondeau
2012-01-29core: minor edits.Tom Rondeau
2012-01-29core: switched complex to mag and mag_squared to use Volk functions.Tom Rondeau
2012-01-29volk: added volk magnitiude squared functions (aligned/unaligned) for ↵Tom Rondeau
complex numbers.
2012-01-29volk: added unaligned volk function for magnitude of a complex number.Tom Rondeau
2012-01-28volk: fix lower bound of int conversion.Tom Rondeau
2012-01-28core: update Makefile.am for new type converters.Tom Rondeau
2012-01-28core: more type converters in Volk; adding char_to_short converter.Tom Rondeau
2012-01-28core: more conversion work to Volk for type converters.Tom Rondeau
2012-01-28core: QA codes for float_to_X and short_to_X (note: float_to_int needs work).Tom Rondeau
2012-01-28core: cleanup short_to_float.Tom Rondeau
2012-01-28core: short_to_X: adding short_to_char block using Volk; made Volk version ↵Tom Rondeau
of short_to_float.
2012-01-28core: minor tweaks to float_to_XTom Rondeau
2012-01-28volk: cleaned up float_to_X to be Volk-only and added vlen and scale to ↵Tom Rondeau
constructor to set the vector length and a free scaling with the data.
2012-01-26QA: type converter QA codes added and updated for Volk. Some small ↵Tom Rondeau
differences between Volk and non-Volk for rounding issues are made here.
2012-01-26core: moved float_to_X type converters over to use Volk calls.Tom Rondeau
2012-01-26volk: float_to_int and float_to_char updated to clip instead of wrap around. ↵Tom Rondeau
The float to int clips at smaller than 2^32 because of the limits of the float representation.
2012-01-26volk: fix a warning.Tom Rondeau
2012-01-26volk: float_to_short now clips the values instead of wrapping around.Tom Rondeau
2012-01-23sched: better comments. Handling of noutput_items adjustment done better and ↵Tom Rondeau
documented.
2012-01-23build: look for local volk headers instead of installed.Tom Rondeau
2012-01-22sched: forgot to turn debugging off.Tom Rondeau
2012-01-22core: test case of using Volk to convert from float to short.Tom Rondeau
2012-01-22core: link to Volk.Tom Rondeau
2012-01-22sched: first attempt at working with aligned data sets. A block has an ↵Tom Rondeau
indicator on whether or not the buffers are aligned; they can use this to determine which Volk function to use or if to use Volk at all.
2012-01-07build: update revision to 3.5.1Johnathan Corgan
2012-01-07build: update canonical config filesJohnathan Corgan