March 2017 Archives by author

• Messages sorted by: [ thread ] [ subject ] [ date ]
• More info on this list...

Starting: Thu Mar 2 03:33:31 UTC 2017
Ending: Fri Mar 31 06:37:02 UTC 2017
Messages: 199

• [Beignet] Limit get_program_global_data() calls to OpenCL 2.0   Jan Beich
• [Beignet] [PATCH] Properly check return value from __cxa_demangle   Jan Beich
• [Beignet] [PATCHv2] Properly check return value from __cxa_demangle   Jan Beich
• [Beignet] [PATCH newRT] Wrap all memory allocate functions.   He Junyan
• [Beignet] [PATCH] Docs: Fix spelling and grammar   Rebecca N. Palmer
• [Beignet] [PATCH] Docs: Fix broken link   Rebecca N. Palmer
• [Beignet] [PATCH newRT] Add cl_gen_device_common.h file.   Rebecca N. Palmer
• [Beignet] [PATCH V2] CMAKE: Refine builtin kernel bin generator   Xiuli Pan
• [Beignet] [PATCH 1/2] implement extension cl_intel_media_block_io WRITE related function   Pan, Xiuli
• [Beignet] [PATCH 1/4] Backend: Add missing Unaligned OWord Block Read disasm   Xiuli Pan
• [Beignet] [PATCH 2/4] Backend: Store the spill register information   Xiuli Pan
• [Beignet] [PATCH 1/3] add extension cl_intel_media_block_io READ related function   Pan, Xiuli
• [Beignet] [PATCH 2/3] cl_intel_media_block_io READ related unit tests.   Pan, Xiuli
• [Beignet] [PATCH 3/3] fix build error log not output issue.   Pan, Xiuli
• [Beignet] [PATCH 2/2] cl_intel_media_block_io WRITE related unit tests.   Pan, Xiuli
• [Beignet] [PATCH] Backend: Fix flag and subflag seting for src 3 instruction   Xiuli Pan
• [Beignet] [PATCH v2 5/5] cl_intel_media_block_io WRITE related unit tests.   Pan, Xiuli
• [Beignet] [PATCH] Backend: Fix physical register usage in selection   Xiuli Pan
• [Beignet] [PATCH 1/7] Backend: Fix flag and subflag seting for src 3 instruction   Xiuli Pan
• [Beignet] [PATCH 2/7] Backend: Remove old llvm support code.   Xiuli Pan
• [Beignet] [PATCH 3/7] Backend: Fix an include file error problem   Xiuli Pan
• [Beignet] [PATCH 4/7] Backend: Refine GEP lowering code   Xiuli Pan
• [Beignet] [PATCH 5/7] Backend: Refine LLVM version check macro   Xiuli Pan
• [Beignet] [PATCH 6/7] Backend: Refine FCmp one and une   Xiuli Pan
• [Beignet] [PATCH 7/7] Backend: Add LLVM40 support   Xiuli Pan
• [Beignet] [PATCH 7/7] Backend: LLVM40 support   Xiuli Pan
• [Beignet] [PATCH 5/7] Backend: Refine LLVM version check macro   Pan, Xiuli
• [Beignet] [PATCH] Properly check return value from __cxa_demangle   Pan, Xiuli
• [Beignet] [PATCH] CMAKE: Add standalone benchmark build   Xiuli Pan
• [Beignet] [PATCH 1/2] Backend: Store the spill register information   Xiuli Pan
• [Beignet] [PATCH 2/2] Backend: Add hole reuse in reg alloction   Xiuli Pan
• [Beignet] [PATCH] Backend: add double support to convert_uchar|char|short|ushort|int|uint|long|ulong_sat(double x) HW support Double to int16, int32 from IVB, others done by software. Double to int64 is supported by BWD+, now skip it and refine it later   Simon Richter
• [Beignet] [PATCH] intel: Check that we can reserve the zero-offset   Yang Rong
• [Beignet] [PATCH] GBE: set memcpy and memset functions's linkage to LinkOnceAnyLinkage at last call.   Yang Rong
• [Beignet] [PATCH] Backend: for BDW and after, According to BSpec no need to split CMP when src is DW DF   Song, Ruiling
• [Beignet] [PATCH 2/2] Backend: refine the geometry function   Song, Ruiling
• [Beignet] [PATCH 2/2] Backend: refine the geometry function   Song, Ruiling
• [Beignet] [PATCH 7/7] Backend: for BDW and after, According to BSpec no need to split CMP when src is DW DF   Song, Ruiling
• [Beignet] [PATCH v2] Provide more possible candidate of load/store as possible.   Song, Ruiling
• [Beignet] [PATCH 4/4] Backend:add double support for some relation function   Song, Ruiling
• [Beignet] [PATCH 5/5] Backend: add double support to bitselect   Song, Ruiling
• [Beignet] [PATCH v2] Provide more possible candidate of load/store as possible.   Song, Ruiling
• [Beignet] [PATCH] intel: Check that we can reserve the zero-offset   Song, Ruiling
• [Beignet] [PATCH 5/7] Backend: Refine LLVM version check macro   Song, Ruiling
• [Beignet] [PATCH] utests: add utest to double version of log, log2, log10   Song, Ruiling
• [Beignet] [PATCH] utests: add utest to double version of log, log2, log10   Song, Ruiling
• [Beignet] [PATCH] backend: add double support to fmax   Song, Ruiling
• [Beignet] [PATCH] utests: add utest to double version of log, log2, log10   Wang, Rander
• [Beignet] [PATCH] backend: add double support to fmax   Wang, Rander
• [Beignet] [PATCH 7/7] Backend: for BDW and after, According to BSpec no need to split CMP when src is DW DF   Yang, Rong R
• [Beignet] [PATCH 2/2] Backend: refine the geometry function   Yang, Rong R
• [Beignet] [PATCH V2] CMAKE: Refine builtin kernel bin generator   Yang, Rong R
• [Beignet] [PATCH 1/4] Backend: Add missing Unaligned OWord Block Read disasm   Yang, Rong R
• [Beignet] [PATCHv2] Properly check return value from __cxa_demangle   Yang, Rong R
• [Beignet] [Patch V2 2/3] fix regression on pre-BDW platform.   Yang, Rong R
• [Beignet] Limit get_program_global_data() calls to OpenCL 2.0   Yang, Rong R
• [Beignet] [PATCH newRT] Wrap all memory allocate functions.   Yang, Rong R
• [Beignet] [PATCH newRT] Rename intel dir to gen.   junyan.he at inbox.com
• [Beignet] [PATCH V3 newRT] Add GenProgram::toBinaryFormat to generate ELF format binary.   junyan.he at inbox.com
• [Beignet] [PATCH V4 newRT] Add GenProgram::toBinaryFormat to generate ELF format binary.   junyan.he at inbox.com
• [Beignet] [PATCH V2 newRT] Add compiler API functions.   junyan.he at inbox.com
• [Beignet] [PATCH] Add device API struct to define device behavior   junyan.he at inbox.com
• [Beignet] [PATCH] Add libelf check in CMakeList   junyan.he at inbox.com
• [Beignet] [PATCH] Add Gen's specific program struct.   junyan.he at inbox.com
• [Beignet] [PATCH] Add Gen's kernel specific struct.   junyan.he at inbox.com
• [Beignet] [PATCH] Move intel_gpgpu.c's functions to new file.   junyan.he at inbox.com
• [Beignet] [PATCH] Move compiler load/unload logic to gen specific file.   junyan.he at inbox.com
• [Beignet] [PATCH] Refine command queue's enqueue ndrang.   junyan.he at inbox.com
• [Beignet] [PATCH newRT] Wrap all memory allocate functions.   junyan.he at inbox.com
• [Beignet] [PATCH newRT] Fix some resource leak point in utests.   junyan.he at inbox.com
• [Beignet] [PATCH newRT] Move intel's batch buffer souce code to gen dir.   junyan.he at inbox.com
• [Beignet] [PATCH newRT] Move intel_structs.h to gen dir and format its style.   junyan.he at inbox.com
• [Beignet] [PATCH newRT] Move intel_driver.c to gen dir and format its style.   junyan.he at inbox.com
• [Beignet] [PATCH 1/6 newRT] Refine intel_driver file.   junyan.he at inbox.com
• [Beignet] [PATCH 2/6 newRT] Move X11 files to gen dir.   junyan.he at inbox.com
• [Beignet] [PATCH 3/6 newRT] Fix two bugs in gen kernel.   junyan.he at inbox.com
• [Beignet] [PATCH 4/6 newRT] Add cl_device_id_gen file in gen dir.   junyan.he at inbox.com
• [Beignet] [PATCH 5/6 newRT] Move the enqueue Copy/Fill to gen dir.   junyan.he at inbox.com
• [Beignet] [PATCH 6/6 newRT] Add cl_context_gen file.   junyan.he at inbox.com
• [Beignet] [PATCH newRT] Add cl_gen_device_common.h file.   junyan.he at inbox.com
• [Beignet] [PATCH newRT] Move pci id for gen to gen dir.   junyan.he at inbox.com
• [Beignet] [PATCH 1/3] add extension cl_intel_media_block_io READ related function   xionghu.luo at intel.com
• [Beignet] [PATCH 2/3] cl_intel_media_block_io READ related unit tests.   xionghu.luo at intel.com
• [Beignet] [PATCH 3/3] fix build error log not output issue.   xionghu.luo at intel.com
• [Beignet] [PATCH 1/2] implement extension cl_intel_media_block_io WRITE related function   xionghu.luo at intel.com
• [Beignet] [PATCH 2/2] cl_intel_media_block_io WRITE related unit tests.   xionghu.luo at intel.com
• [Beignet] [PATCH v2 1/5] add extension cl_intel_media_block_io READ related function   xionghu.luo at intel.com
• [Beignet] [PATCH v2 2/5] cl_intel_media_block_io READ related unit tests.   xionghu.luo at intel.com
• [Beignet] [PATCH v2 3/5] fix build error log not output issue.   xionghu.luo at intel.com
• [Beignet] [PATCH v2 4/5] implement extension cl_intel_media_block_io WRITE related function   xionghu.luo at intel.com
• [Beignet] [PATCH v2 5/5] cl_intel_media_block_io WRITE related unit tests.   xionghu.luo at intel.com
• [Beignet] [Patch V2 1/2] add extension intel_planar_yuv.   xionghu.luo at intel.com
• [Beignet] [Patch V2 2/2] add utest for extension intel_planar_yuv.   xionghu.luo at intel.com
• [Beignet] [PATCH 1/3] clang 3.9 will report error when use read_write in OpenCL-1.2.   xionghu.luo at intel.com
• [Beignet] [PATCH 2/3] fix regression on pre-BDW platform.   xionghu.luo at intel.com
• [Beignet] [PATCH 3/3] llvm3.9 will assert if ouput is empty string.   xionghu.luo at intel.com
• [Beignet] [Patch V2 2/3] fix regression on pre-BDW platform.   xionghu.luo at intel.com
• [Beignet] [PATCH] Provide more possible candidate of load/store as possible.   yan.wang at linux.intel.com
• [Beignet] [PATCH v2] Provide more possible candidate of load/store as possible.   yan.wang at linux.intel.com
• [Beignet] [PATCH] Backend: for BDW and after, According to BSpec no need to split CMP when src is DW DF   rander
• [Beignet] [PATCH 2/2] Backend: refine the geometry function   rander
• [Beignet] [PATCH 1/2] Backend: for BDW and after, According to BSpec no need to split CMP when src is DW DF   rander
• [Beignet] [PATCH 2/2] Backend: refine the geometry function   rander
• [Beignet] [PATCH 4/4] Backend:add double support for some relation function   rander
• [Beignet] [PATCH 5/5] Backend: add double support to bitselect   rander
• [Beignet] [PATCH 7/7] Backend: for BDW and after, According to BSpec no need to split CMP when src is DW DF   rander
• [Beignet] [PATCH 8/8] Backend: add double support for shuffle   rander
• [Beignet] [PATCH 9/9] Backend: add double support to prefetch. Actually it does nothing   rander
• [Beignet] [PATCH 10/10] Backend: add double support to max, min, min, step Also add other function in headfile to compile successfully   rander
• [Beignet] [PATCH 9/9] Backend: add double support to prefetch. Actually it does nothing   rander
• [Beignet] [PATCH] Backend: add double support to convert_uchar|char|short|ushort|int|uint|long|ulong_sat(double x) HW support Double to int16, int32 from IVB, others done by software. Double to int64 is supported by BWD+, now skip it and refine it later   rander
• [Beignet] [PATCH] Backend: add double support to convert_u|long_rtn(double)   rander
• [Beignet] [PATCH] Backend: add double support to convert_u|int_rtn(double x)   rander
• [Beignet] [PATCH] Backend: add double support to convert_uchar|short_rtn(double x)   rander
• [Beignet] [PATCH] Backend: add double support to convert_float_rtn(double x)   rander
• [Beignet] [PATCH] Backend: add double support to convert_u|long_rte(double x)   rander
• [Beignet] [PATCH] Backend: add double support to convert_u|char|u|short|u|int_rte(double x)   rander
• [Beignet] [PATCH] Backend: add double support to convert_u|char|short|int|long_rtz(double x) rtz can be done with rtn with usigned type. for signed type, rtn with abs(x), then add the sign effect   rander
• [Beignet] [PATCH] utests: add utests for relation functions of double type   rander
• [Beignet] [PATCH] utest: add utest to common function of double   rander
• [Beignet] [PATCH] utests: add utest to convert_double2int8   rander
• [Beignet] [PATCH] utests: add utest to double convert int16   rander
• [Beignet] [PATCH] utests: add utest to double convert int32   rander
• [Beignet] [PATCH] utests: add utest to convert double to int64   rander
• [Beignet] [PATCH] utests: add utest to convert_double2int8   rander
• [Beignet] [PATCH] utests: add utest to double convert int16   rander
• [Beignet] [PATCH] utests: add utest to double convert int32   rander
• [Beignet] [PATCH] utests: add utest to double convert int32   rander
• [Beignet] [PATCH] utests: add utest to convert double to int64   rander
• [Beignet] [PATCH] utests: add utest to int8 convert to double. the algorithm is very simple, for convert_double_rte|z|p|n(int8 x) the input from -128 ~ 127 or 0 ~ 255 should get the same result   rander
• [Beignet] [PATCH] utests: add utest to convert_double_rte|z|p|n(int 16) it is very simple, for double can cover the range of short just check every short value with the double value, they should be the same   rander
• [Beignet] [PATCH] Backend: add double support to convert_u|long_rtp(double x)   rander
• [Beignet] [PATCH] backend: add double support to convert_u|char|short|int|long_sat_rte|z|n|p(double x) Algorithm: do the operation as rte|z|n|p without sat when the data range is in. And if outof range, just clamp to the max|min.   rander
• [Beignet] [PATCH] backend: add double support to convert_u|char|short|int_rtp(double x) first convert double to u|long, then convert to smaller type And converting double directly to smaller type does not save any instructions   rander
• [Beignet] [PATCH] backend: add double support to convert_double_rte|n|z|p(double x) just call convert_double(double x). actually just a mov   rander
• [Beignet] [PATCH] backend: add convert_double_R(float x)   rander
• [Beignet] [PATCH] utests: add utest to convert_float_roundmode(double x)   rander
• [Beignet] [PATCH] backend: refine convert_float_rtn(double x)   rander
• [Beignet] [PATCH] backend: add convert_float_rtz(double)   rander
• [Beignet] [PATCH] backend: add convert_float_rtp(double)   rander
• [Beignet] [PATCH] backend: add convert_float_rte(double)   rander
• [Beignet] [PATCH] utests: add utest to convert_double_rd(ulong)   rander
• [Beignet] [PATCH] backend: add convert_double_rtp(ulong)   rander
• [Beignet] [PATCH] backend: add convert_double_rtz(ulong)   rander
• [Beignet] [PATCH] backend: add convert_double_rtn(ulong)   rander
• [Beignet] [PATCH] backend: add convert_double_rte(ulong)   rander
• [Beignet] [PATCH] backend: add convert_double(long)   rander
• [Beignet] [PATCH] backend: add convert_double_rtn(long)   rander
• [Beignet] [PATCH] backend: add convert_double_rtz(long)   rander
• [Beignet] [PATCH] backend: add convert_double_rte(long)   rander
• [Beignet] [PATCH] utests: add utest to fxxx double typed math functions   rander
• [Beignet] [PATCH] backend: add double support to ceil   rander
• [Beignet] [PATCH] backend: add doube support to fabs   rander
• [Beignet] [PATCH] backend: add double support to floor   rander
• [Beignet] [PATCH] backend: add double support to fract   rander
• [Beignet] [PATCH] utests: add utest to double version of log, log2, log10   rander
• [Beignet] [PATCH] backend: add double support to log   rander
• [Beignet] [PATCH] backend: add double support to log2   rander
• [Beignet] [PATCH] backend: add double support to log10   rander
• [Beignet] [PATCH] backend: add double support to log1p   rander
• [Beignet] [PATCH] backend: add double support to logb   rander
• [Beignet] [PATCH] backend: add double support to ilogb   rander
• [Beignet] [PATCH] utests: add utest to double type of simple math functions   rander
• [Beignet] [PATCH] backend: add double support to mad   rander
• [Beignet] [PATCH] backend: add double support to fmax   rander
• [Beignet] [PATCH] backend: add double support to fmin   rander
• [Beignet] [PATCH] backend: add double support to copysign   rander
• [Beignet] [PATCH] backend: add double support to fdim   rander
• [Beignet] [PATCH] backend: add double support to frexp   rander
• [Beignet] [PATCH] backend: add double support to trunc   rander
• [Beignet] [PATCH] backend:add double support to round   rander
• [Beignet] [PATCH] backend: add double support to rint   rander
• [Beignet] [PATCH] backend: add double support to nan   rander
• [Beignet] [PATCH] backend: add double support to nextafter   rander
• [Beignet] [PATCH] backend: add double support to fmod   rander
• [Beignet] [PATCH] backend: refine code structure of ocl math function   rander
• [Beignet] [PATCH] backend: refine utests to check capability of double   rander
• [Beignet] [PATCH] utests: add utests to trigonometric of double version   rander
• [Beignet] [PATCH] backend: add double version of acos   rander
• [Beignet] [PATCH] backend: add double version of acospi   rander
• [Beignet] [PATCH] backend: add double version of acosh   rander
• [Beignet] [PATCH] backend: add double version of asin   rander
• [Beignet] [PATCH] backend: add double version asinpi   rander
• [Beignet] [PATCH] backend: add double version of asinh   rander
• [Beignet] [PATCH] backend: add double version of atan   rander
• [Beignet] [PATCH] backend: add double version of atan2   rander
• [Beignet] [PATCH] backend: add double version of atanpi atan2pi   rander
• [Beignet] [PATCH] backend: add double version of atanh   rander
• [Beignet] [PATCH] backend: add double version of cos   rander
• [Beignet] [PATCH] backend: add double version of cosh   rander
• [Beignet] [PATCH] backend: add double version of sin   rander
• [Beignet] [PATCH] backend: add double version of sinh   rander
• [Beignet] [PATCH] backend: add double version of tan   rander
• [Beignet] [PATCH] backend: add double version of tanh   rander
• [Beignet] [PATCH] backend: add double version of cospi   rander
• [Beignet] [PATCH] backend: add double version of sinpi   rander
• [Beignet] [PATCH] backend: add double version of tanpi   rander
• [Beignet] [PATCH v2] Provide more possible candidate of load/store as possible.   yan.wang
• [Beignet] [PATCH v2] Provide more possible candidate of load/store as possible.   yan.wang

Last message date: Fri Mar 31 06:37:02 UTC 2017
Archived on: Fri Mar 31 06:39:41 UTC 2017

• Messages sorted by: [ thread ] [ subject ] [ date ]
• More info on this list...