summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2017-05-15GLK: add Geminilake pciids.Yang Rong1-1/+8
Signed-off-by: Yang Rong <rong.r.yang@intel.com> Reviewed-by: Pan Xiuli <xiuli.pan@intel.com>
2017-05-15GBE: set memcpy and memset functions's linkage to LinkOnceAnyLinkage at last ↵Yang, Rong R3-7/+14
call. LLVM IR pass will produce memcpy and memset, if set LinkOnceAnyLinkage, memcpy and memset will be delete before and cause fail. Signed-off-by: Yang Rong <rong.r.yang@intel.com> Reviewed-by: Pan Xiuli <xiuli.pan@intel.com>
2017-05-04utest: modify compiler_if_else to not rely on compiler behaviour.Ruiling Song2-12/+5
the test case modify src as well as dst. and it introduce cross workitem memory dependency in dst[id] = src[id+1]; The compiler may order 'then' and 'else' block not as written. If compiler order the else block first. src[3+1] will be modified in else part. And the utest will get wrong result. If user want to get the old behaviour, It should use two if-then and order them as required. Signed-off-by: Ruiling Song <ruiling.song@intel.com> Reviewed-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine normalize function to pass cftrander.wang1-16/+33
remove the call of length to (1)make faster (2) to pass param with 0x1.xxxp1023 Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine powr to pass cftrander1-77/+25
refine corner case according to spec Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine modf to pass the cftrander2-8/+84
Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine min|max magrander1-20/+6
do it according to spec to make cft pass Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine remquo to pass cftrander2-65/+396
do the thing according to spec Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine fma to pass cftrander1-1/+201
port from openlibm and refine it to round to even Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine rootn to pass cftrander1-30/+48
do it according to spec, and refine some corner case Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine atan2 to pass cftrander1-37/+29
refine the corner case like inf and +-0 according to spec Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine powr to pass cftrander1-13/+43
add path to all the corner case Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine fract to pass the cftrander2-4/+60
do it as spec and add path to corner case Signed-off-by: rander.wang <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-05-04backend: refine sin|cos|tan pi to pass cftrander1-40/+137
reduce parameter to small range [0, 0.5] Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of geometry functionrander2-0/+125
Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: refine batch file for doublerander4-39/+11
refined batch file to generate vector2,3,4,16 of math function with double typed parameters Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of modfrander4-0/+32
need to be refined to pass cft Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add missed functionrander1-0/+1
Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of common functionrander1-0/+30
Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of rsqrtrander2-0/+6
do it by 1/sqrt. need to be refined for performance Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of fmin fmaxrander2-0/+32
do it according to spec Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of exp2 exp10rander2-0/+12
convert them to pow. need to be refined for performance Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of sincosrander4-0/+28
convert it to sin + cos. need to be refined for performance Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of remquorander4-0/+199
cp from fdlibm, need to be refined to pass cft Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of fmarander2-0/+6
simulated by mad, need to be refined to pass cft Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of tgammarander2-0/+163
modify from lgamma, need refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of hypotrander2-0/+66
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of rootnrander2-0/+42
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of ldexprander2-0/+10
Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of powrrander2-0/+30
convert powr to pow and pass the cft. for powr is the subset of pow, some optimizations can be done later Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of pownrander2-0/+19
convert pown to pow, pass the cft. And also there are some optimizations can be done, do it later Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of powrander2-0/+260
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of remainderrander2-0/+43
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of lgamma_rrander4-0/+664
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of lgammarander2-0/+163
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of erfcrander2-0/+124
cp from fdlibm and refine it Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of erfrander2-0/+115
cp from fdlibm and refine it Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of cbrtrander2-0/+57
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of tanpirander2-0/+6
it is a simple implementation and need to be refined to pass cft Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of sinpirander2-0/+50
it is a simple implementation and need to be refined to pass cft Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of cospirander2-0/+6
it is a simple implementation and need to be refined to pass cft Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of tanhrander2-0/+35
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of tanrander2-0/+128
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of sinhrander2-0/+41
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of sinrander2-0/+29
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of coshrander2-0/+221
Also add double version of exp & expm1, for cosh depends on them cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of cosrander2-0/+409
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of atanhrander2-0/+27
cp from fdlibm and pass the cft after refined Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of atanpi atan2pirander2-0/+12
it is a simple implementation and need to be refined to pass cft Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
2017-04-17backend: add double version of atan2rander2-0/+86
cp from fdlibm, and need to be refined to pass cft Signed-off-by: rander <rander.wang@intel.com> Tested-by: Yang Rong <rong.r.yang@intel.com>
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-24 06:16:31 +0000