Re: [Beignet] [PATCH] Runtime: Add SKL device id for new SKL device

2016-04-07 Thread Yang, Rong R
LGTM, pushed, thanks. > -Original Message- > From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of > Xiuli Pan > Sent: Friday, April 1, 2016 5:15 > To: beignet@lists.freedesktop.org > Cc: in...@terre-adelie.org; Pan, Xiuli > Subject: [Beignet] [PATCH] Runtime: Add SKL

Re: [Beignet] [PATCH] Increase size for compile log output

2016-04-07 Thread Yang, Rong R
LGTM, pushed, thanks. > -Original Message- > From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of > Giuseppe Bilotta > Sent: Friday, March 25, 2016 19:15 > To: beignet@lists.freedesktop.org > Cc: Giuseppe Bilotta > Subject: [Beignet] [PATCH] Increase size for compile

Re: [Beignet] [PATCH] libocl: Refine return type of workitem built-in functions

2016-04-07 Thread Yang, Rong R
LGTM, pushed, merge your patch "[Beignet] [PATCH] Utest: Add utest load spir for spir64" to one commit to avoid utest assert. > -Original Message- > From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of > Xiuli Pan > Sent: Tuesday, March 29, 2016 16:12 > To: beignet@li

Re: [Beignet] [PATCH] utest: do not check MV near image border

2016-04-07 Thread Yang, Rong R
Pushed, thanks. > -Original Message- > From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of > yan.w...@linux.intel.com > Sent: Thursday, March 17, 2016 17:33 > To: Guo, Yejun > Cc: Guo, Yejun ; beignet@lists.freedesktop.org > Subject: Re: [Beignet] [PATCH] utest: do n

Re: [Beignet] [PATCH] Prepend std namespace to isnan and isinf calls.

2016-04-07 Thread Yang, Rong R
The format have some error, fix it by manual. Pushed, thanks for your contribution. From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Pan, Xiuli Sent: Wednesday, April 6, 2016 12:35 To: Pan, Xiuli ; Sean Lynch Cc: beignet@lists.freedesktop.org Subject: Re: [Beignet] [PAT

[Beignet] [PATCH] Backend: Fix workgroup broadcast on QWORD, 1D works

2016-04-07 Thread Grigore Lupescu
From: Grigore Lupescu Signed-off-by: Grigore Lupescu --- backend/src/backend/gen_insn_selection.cpp | 62 +- 1 file changed, 43 insertions(+), 19 deletions(-) diff --git a/backend/src/backend/gen_insn_selection.cpp b/backend/src/backend/gen_insn_selection.cpp index

Re: [Beignet] GROMACS on beignet

2016-04-07 Thread Szilárd Páll
Hi, On Wed, Apr 6, 2016 at 5:53 AM, Pan, Xiuli wrote: > Hi Szilard, > > The enqueuer seems to be our runtime implementation limitation, we are now > refining our runtime. Thanks for the confirmation! Note that we have CPU-side cycle-counter timers with statistics printed at the end of the log

Re: [Beignet] [PATCH V2] GBE: add ocl 2.0 work_group_barrier support.

2016-04-07 Thread Pan, Xiuli
LGTM! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Ruiling Song Sent: Tuesday, April 5, 2016 3:20 PM To: beignet@lists.freedesktop.org Cc: Song, Ruiling Subject: [Beignet] [PATCH V2] GBE: add ocl 2.0 work_group_barrier support. to do an im

[Beignet] [PATCH 5/5] GBE: Refine program scope variable logic.

2016-04-07 Thread Ruiling Song
The program scope variable may appear in Module's global list as unordered. So I choose to split the program scope logic into two passes. The first pass will just create these constants. The second pass to initialize the data. Signed-off-by: Ruiling Song --- backend/src/ir/constant.cpp

[Beignet] [PATCH 2/5] runtime: prog->global_data may get 64bit address

2016-04-07 Thread Ruiling Song
Signed-off-by: Ruiling Song --- src/cl_program.c | 1 + 1 file changed, 1 insertion(+) diff --git a/src/cl_program.c b/src/cl_program.c index ebe227d..281525e 100644 --- a/src/cl_program.c +++ b/src/cl_program.c @@ -210,6 +210,7 @@ static cl_int get_program_global_data(cl_program prog) { pr

[Beignet] [PATCH 4/5] GBE: handle ConstantExpr in program-scope variable handling.

2016-04-07 Thread Ruiling Song
although we have eliminate ConstantExpr in llvm instructions, but in program scope variable, we still meet ConstantExpr. So, we handle it here. also enhance the test case to hit it. Signed-off-by: Ruiling Song --- backend/src/llvm/llvm_gen_backend.cpp | 45 ++- ke

[Beignet] [PATCH 3/5] GBE: imm64 should not be in src1 per hardware spec.

2016-04-07 Thread Ruiling Song
Signed-off-by: Ruiling Song --- backend/src/backend/gen8_encoder.cpp | 5 +++-- backend/src/backend/gen_insn_selection.cpp | 12 2 files changed, 11 insertions(+), 6 deletions(-) diff --git a/backend/src/backend/gen8_encoder.cpp b/backend/src/backend/gen8_encoder.cpp index a

[Beignet] [PATCH 1/5] GBE: Fix a typo in RelocTable

2016-04-07 Thread Ruiling Song
Signed-off-by: Ruiling Song --- backend/src/ir/unit.hpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/backend/src/ir/unit.hpp b/backend/src/ir/unit.hpp index b8df145..ce603a0 100644 --- a/backend/src/ir/unit.hpp +++ b/backend/src/ir/unit.hpp @@ -58,7 +58,7 @@ namespace ir {