summaryrefslogtreecommitdiffstats
path: root/vec_test.h
Commit message (Collapse)AuthorAgeFilesLines
* Reformat source code with clang-formatErik Schnetter2015-10-161-1442/+1248
| | | | | | | | | Note: If you have an existing checkout with modifications, then you should: (1) save your current state on a branch (2) run clang-format on this branch (3) compare this branch to a fresh checkout of the clang-formatted master Since both your branch and master have been formatted with clang-format, this will lead to a very small diff, avoiding spurious changes due to formatting differences.
* Don’t define << or >> for VML_NO_IOSTREAMErik Schnetter2013-11-261-0/+2
|
* If VML_NO_IOSTREAM is set, then don’t provide any iostream operationsErik Schnetter2013-11-251-0/+5
| | | | This reduces the size of the library.
* Provide lrintErik Schnetter2013-11-251-0/+7
|
* Introduce mad(), as low-accuracy version of fma()Erik Schnetter2013-09-101-0/+12
|
* Use nicer type names in vec_testErik Schnetter2013-09-021-150/+164
|
* Provide clean set of libm elemental functions in namespace vml_stdErik Schnetter2013-09-021-2/+2
| | | | Different C++ standard versions and different compilers provide different sets of elemental functions in std. Thus, we create a clean set in a new namespace vml_std.
* Implement more integer functions: abs bitifthen clz isignbit max min ↵Erik Schnetter2013-09-011-15/+78
| | | | | | popcount rotate Rename integer signbit to isignbit to avoid name conflicts.
* Make frexp take a pointer argument instead of a referenceErik Schnetter2013-08-211-2/+2
|
* Implement ifthen with boolvec argumentsErik Schnetter2013-07-071-0/+20
|
* Simplify all() and any() implementationsErik Schnetter2013-07-051-4/+4
|
* Remove superfluous inline declarationsErik Schnetter2013-07-041-1/+1
|
* Simplify prod and sumErik Schnetter2013-07-041-4/+4
|
* Implement maxval and minval reduction operationsErik Schnetter2013-07-041-0/+24
|
* Whitespace change in asm intrinsicsErik Schnetter2013-07-021-6/+6
|
* Implement frexpErik Schnetter2013-06-281-0/+8
|
* Correct case in commentErik Schnetter2013-06-191-1/+1
|
* Support Intel compilerErik Schnetter2013-06-171-1/+7
|
* Correct generic barrier on ARMErik Schnetter2013-06-121-1/+1
|
* Implement barrier for ARMErik Schnetter2013-06-121-0/+2
|
* Declare some functions without auto keywordErik Schnetter2013-06-121-2/+2
|
* Implement signbit function for integer vectorsErik Schnetter2013-06-121-0/+13
|
* Implement nextafterErik Schnetter2013-06-091-0/+11
|
* Remove commented-out duplicate function declarationsErik Schnetter2013-06-071-10/+8
|
* Implement atan2Erik Schnetter2013-06-061-0/+8
|
* Correct operator~Erik Schnetter2013-06-051-1/+1
|
* Continue implementing QPX supportErik Schnetter2013-06-051-7/+17
|
* Add integer comparison operators for SSE2 float and pseudovec classesErik Schnetter2013-06-031-0/+24
|
* Suggest PPC barrier implementation near PPC64 barrier implementationErik Schnetter2013-06-021-1/+1
|
* Use different macro to detect PPC architectureErik Schnetter2013-06-021-1/+1
|
* Output error when barrier instruction is needed, but is not knownErik Schnetter2013-06-021-0/+2
|
* Define barrier for all PPC architectures, not just when Altivec is presentErik Schnetter2013-06-021-1/+1
|
* Update alignment checks: use new "alignment" entryErik Schnetter2013-04-231-6/+6
|
* Reduce alignment requirement for vec_pseudo and vec_testErik Schnetter2013-04-231-6/+6
|
* Introduce field "alignment" describing the necessary alignment for vector typesErik Schnetter2013-04-231-3/+6
|
* Describe aligned via attribute((aligned)) instead of alignasErik Schnetter2013-04-231-3/+7
|
* Don't use +X asm constraint with gcc; gcc crashes with itErik Schnetter2013-04-201-0/+5
|
* Use architecture-independent +X asm constraint instead of +xErik Schnetter2013-04-191-1/+1
|
* Don't use typeinfo in vec_testErik Schnetter2013-03-221-9/+4
|
* Fixes to make it build with Clang++ 3.2 using libstdc++ from Debian 6.0.Pekka Jääskeläinen2013-03-211-1/+4
|
* Add rint(), correct round()Erik Schnetter2013-02-191-3/+19
|
* Add cbrt, hypot, trunc; rename scalbn to ldexpErik Schnetter2013-02-161-19/+41
|
* Clean up vector classesErik Schnetter2013-02-141-0/+1205
realvec<T,N>: machine-specific, optimised implementations realpseudovec<T,N>: scalarised implementation based on libm (for benchmark comparisons) realtestvec<T,N>: scalarised implementation without libm, using only vecmathlib's algorithms (for correctness testing)
OpenPOWER on IntegriCloud