beignet - Beignet OpenCL Library for Intel Ivy Bridge and newer GPUs (mirrored from https://gitlab.freedesktop.org/beignet/beignet)

Age	Commit message (Expand)	Author	Files	Lines
2017-09-21	backend: use simd-1 for scalar dst in indirectMov.	Song, Ruiling	1	-14/+24
2017-09-21	GBE: remove static context to fix Segmentation fault.	Yang Rong	4	-33/+39
2017-09-21	GBE: enable llvm5.0 support.	Yang Rong	6	-33/+87
2017-09-21	libocl: enable llvm5.0 support.	Yang Rong	3	-32/+59
2017-09-21	libocl: Consider only bottom ilogb(2m-1)+1 bits	Jan Vesely	1	-30/+30
2017-09-21	libocl: Add shuffle and shuffle2 builtins for half type	Jan Vesely	2	-0/+4
2017-07-27	GBE: fix a errMsg uninitialized build warning.	Yang, Rong R	1	-3/+3
2017-07-27	backend: Fix a bug in load-store optimization.	Song, Ruiling	1	-25/+46
2017-07-20	backend: refine global immediate optimization	rander	1	-4/+0
2017-07-20	Fix GCC6 build bug	Pan Xiuli	1	-0/+1
2017-07-18	backend: refine load store optimization	Song, Ruiling	1	-37/+88
2017-07-12	Implement extension cl_intel_device_side_avc_motion_estimation.	Chuanbo Weng	24	-33/+2113
2017-07-06	backend: improve add zero pattern	rander	1	-5/+5
2017-07-06	backend: refine fdiv to rcp at some cases	rander	1	-0/+28
2017-07-04	backend: refine math log function	rander	1	-40/+10
2017-07-04	backend: refine pow function	rander	1	-146/+148
2017-06-23	backend: refine load/store merging algorithm	rander	1	-9/+78
2017-06-23	backend: add global immediate optimization	rander	1	-25/+342
2017-06-23	GBE: clean llvm module's clone and release.	Yang, Rong R	7	-57/+69
2017-06-16	backend: refine the local copy propagation.	rander.wang	1	-0/+34
2017-06-16	Runtime: Add new API enums for cl_intel_required_subgroup_size extension	Pan Xiuli	1	-0/+4
2017-06-16	Backend: Add intel_reqd_sub_group_size support	Pan Xiuli	3	-13/+45
2017-06-16	do constant folding for kernel struct args	Guo Yejun	6	-0/+213
2017-06-09	Backend: Add optimization for negative modifier	rander	1	-4/+28
2017-06-09	backend: add sqrt-div pattern to instruction select	rander	1	-0/+69
2017-06-09	keep GEN IR as SSA style	Guo Yejun	1	-3/+5
2017-06-09	backend: refine exp function with float input	rander	1	-2/+56
2017-06-09	backend: refine hypot function	rander	1	-14/+60
2017-05-23	refresh DAG when an arg has both direct and indirect read	Guo Yejun	1	-1/+16
2017-05-23	Backend: Add sel ir output for MATH function	Pan Xiuli	1	-0/+42
2017-05-23	backend: fix tgamma error after restructure	rander	1	-25/+31
2017-05-18	Backend: Fix performance regression with sampler refine fro LLVM40	Pan Xiuli	2	-9/+41
2017-05-18	Backend: Fix llvm40 assert about literal structs	Pan Xiuli	1	-1/+2
2017-05-17	backend: refine asin function	rander.wang	1	-21/+7
2017-05-17	backend: refine atan	rander.wang	1	-53/+58
2017-05-17	backend: refine acos	rander.wang	1	-4/+9
2017-05-17	backend: refine sincos	rander.wang	2	-13/+277
2017-05-17	backend: refine tan function	rander.wang	1	-16/+45
2017-05-17	backend: refine cos function	rander.wang	1	-26/+25
2017-05-17	backend: refine sin function	rander.wang	1	-20/+22
2017-05-17	backend: refine the argue reduce	rander.wang	1	-24/+14
2017-05-17	backend: refine pow function	rander.wang	1	-134/+141
2017-05-17	backend: refine the code structure of math	rander.wang	9	-7538/+4073
2017-05-15	GLK: add geminilake backend support.	Yang Rong	5	-2/+47
2017-05-15	GBE: set memcpy and memset functions's linkage to LinkOnceAnyLinkage at last ...	Yang, Rong R	3	-7/+14
2017-05-04	backend: refine normalize function to pass cft	rander.wang	1	-16/+33
2017-05-04	backend: refine powr to pass cft	rander	1	-77/+25
2017-05-04	backend: refine modf to pass the cft	rander	2	-8/+84
2017-05-04	backend: refine min\|max mag	rander	1	-20/+6
2017-05-04	backend: refine remquo to pass cft	rander	2	-65/+396