Re: [Beignet] Gen 8 support

Yes, it is supported. You can have a try. Please take a look at Beignet Wiki page: http://www.freedesktop.org/wiki/Software/Beignet/ If you still have problem, you can send email to this mail list. Thanks! Ruiling From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Andrés

[Beignet] [Printf][PATCH 08/11] Implement ASM generation of printf.

Contributor: Junyan He Signed-off-by: Yan Wang --- backend/src/backend/gen8_context.cpp | 36 +++ backend/src/backend/gen8_context.hpp | 1 + backend/src/backend/gen_context.cpp | 70

[Beignet] [Printf][PATCH 06/11] Implement emision of printf instruction.

Contributor: Junyan He Signed-off-by: Yan Wang --- backend/src/ir/context.hpp| 5 ++ backend/src/llvm/llvm_gen_backend.cpp | 89 --- 2 files changed, 78 insertions(+), 16 deletions(-) diff --git

[Beignet] [Printf][PATCH 10/11] Output printf result.

Contributor: Junyan He Signed-off-by: Yan Wang --- backend/src/ir/printf.cpp | 122 +- backend/src/ir/printf.hpp | 2 +- 2 files changed, 112 insertions(+), 12 deletions(-) diff --git

[Beignet] [Printf][PATCH 04/11] Add the implementation of printf ir instruction.

Contributor: Junyan He Signed-off-by: Yan Wang --- backend/src/ir/function.hpp| 8 ++ backend/src/ir/instruction.cpp | 57 +- backend/src/ir/instruction.hpp | 13 ++

Re: [Beignet] Most of utest cases run failed

All errores are CL_INVALID_BUFFER_SIZE, seems device's global_mem_size or/and max_mem_alloc_size too small. These values are calculated from drm_intel_get_aperture_sizes and sysinfo, can you provide these functions return value in file src/cl_device_id.c and src/intel/intel_driver.c. Thanks. >

[Beignet] [Printf][PATCH 01/11] Change printf data structure and remove old code.

Contributor: Junyan He Signed-off-by: Yan Wang --- backend/src/backend/program.cpp | 14 -- backend/src/backend/program.hpp | 10 +- backend/src/gbe_bin_interpreter.cpp | 2 - backend/src/ir/printf.cpp |

Re: [Beignet] beignet git with llvm 3.7.1 + haswell (gen 7.5) failing

Yes, I agree with you. But at least as I know haswell hardware does not support double well. Hi Junyan, Can we handle it graciously? Do you have any idea? Thanks! Ruiling From: Paulo Dias [mailto:paulo.miguel.d...@gmail.com] Sent: Wednesday, January 20, 2016 8:37 PM To: Song, Ruiling

[Beignet] [PATCH 10/15] Backend: Add state register into schedule consideration.

From: Junyan He Because the workgroup OP has forwarding msg and wait functions, it needs all the threads to sync with each other. It has very similar behavior as BARRIER, so we add it into schedule consideration accordingly. Signed-off-by: Junyan He

[Beignet] [PATCH 03/15] Backend: Add sr0 reg helper function.

From: Junyan He sr0 is used to specify the state reigster where we can get the state of each EU thread. Signed-off-by: Junyan He Reviewed-by: Yang Rong --- backend/src/backend/gen75_context.cpp | 8 +---

[Beignet] [PATCH 04/15] Backend: Add tidMapSLM and wgBroadcastSLM to each function.

From: Junyan He We will use SLM to store the value to broadcast and the map between real hw thread and logical workgroup thread. These two values give the offset in the SLM. Signed-off-by: Junyan He Reviewed-by: Yang Rong

[Beignet] [PATCH 05/15] Backend: Add threadid as a curbe register.

From: Junyan He We need to build a map between logical workgroup thread and real hw thread. This curbe will hold the logical workgroup thread ID when it is executed on some HW thread. Signed-off-by: Junyan He Reviewed-by: Yang Rong

[Beignet] [PATCH 06/15] libocl: Refine the workgroup functions, add signed info.

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- backend/src/libocl/src/ocl_work_group.cl | 114 +++ 1 file changed, 57 insertions(+), 57 deletions(-) diff --git

[Beignet] [PATCH 09/15] Backend: Add WORKGROUP_OP instruction selection.

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- backend/src/backend/gen_context.cpp| 3 ++ backend/src/backend/gen_context.hpp| 1 +

[Beignet] [PATCH 08/15] Add forward message function for gen encoder.

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- backend/src/backend/gen_encoder.cpp | 12 backend/src/backend/gen_encoder.hpp | 2 ++ 2 files changed, 14 insertions(+) diff --git

[Beignet] [PATCH 12/15] Runtime: Add the threadid calculation for curbe.

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- src/cl_command_queue_gen7.c | 12 +++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/src/cl_command_queue_gen7.c

[Beignet] [PATCH 14/15] Backend: Add reduce add to gen_context.

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- backend/src/backend/gen_context.cpp | 24 ++-- 1 file changed, 22 insertions(+), 2 deletions(-) diff --git

[Beignet] [PATCH 11/15] Backend: Implement reduce min and max in gen_context

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- backend/src/backend/gen_context.cpp | 277 1 file changed, 277 insertions(+) diff --git

[Beignet] [PATCH 15/15] Utests: Add test cases for reduce add.

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- kernels/compiler_workgroup_reduce.cl | 12 +++ utests/compiler_workgroup_reduce.cpp | 69 2 files changed,

[Beignet] [PATCH 07/15] Backend: Establishing the thread/TID-EUID map.

From: Junyan He We need to use forward message to send data and sync threads within the same work group. The HW lack the feature to get the TID and EUID of other threads. So we need to establish a map for this usage. Signed-off-by: Junyan He

[Beignet] [PATCH 01/15] Backend: Add a auxiliary function to convert GenReg to uniform.

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- backend/src/backend/gen_register.hpp | 9 + 1 file changed, 9 insertions(+) diff --git a/backend/src/backend/gen_register.hpp

[Beignet] [PATCH 13/15] Utests: Add test cases for workgroup reduce max/min.

From: Junyan He Signed-off-by: Junyan He Reviewed-by: Yang Rong --- kernels/compiler_workgroup_reduce.cl | 28 ++ utests/CMakeLists.txt| 1 + utests/compiler_workgroup_reduce.cpp | 174

[Beignet] [PATCH] Backend:LLVM 3.8 release support

Add more support for relaese 3.8 Signed-off-by: Pan Xiuli --- backend/src/backend/gen_program.cpp | 2 +- backend/src/backend/program.cpp | 2 +- backend/src/llvm/ExpandConstantExpr.cpp | 2 +- backend/src/llvm/llvm_to_gen.cpp| 4

[Beignet] Gen 8 support

Hi All, I'll develop a board based on the Intel x5-Z8300. In this board must be run an application based on OpenCL, this its possible right now? Currently I'm testing over on "Tronsmart Ara x5" running Debian Jessie, Kernel 3.16.0-4-amd64 Does Beignet supports the Gen 8 GPU? Best regards, --

[Beignet] Gen 8 support

Hi All, I'll develop a board based on the Intel x5-Z8300. In this board must be run an application based on OpenCL, this its possible right now? Currently I'm testing over on "Tronsmart Ara x5" running Debian Jessie, kernel 3.16.0-4-amd64 Does Beignet supports the Gen 8 GPU? Best regards, --

[Beignet] [PATCH V2] Fix the bug of crash when we pass -I path with spaces.

From: Junyan He We failed to handle -I "/XX X/YY YY/" like path passed from the build option. We need to consider the spaces here and pass it correctly to Clang. Signed-off-by: Junyan He --- backend/src/backend/program.cpp | 51

Re: [Beignet] [PATCH V2] Fix the bug of crash when we pass -I path with spaces.

V2 just fix some typo, please just ignore the previous one. This patch should be merget to master and release1.1 On Wed, Jan 20, 2016 at 05:57:20PM +0800, junyan...@inbox.com wrote: > Date: Wed, 20 Jan 2016 17:57:20 +0800 > From: junyan...@inbox.com > To: beignet@lists.freedesktop.org > Subject:

Re: [Beignet] beignet git with llvm 3.7.1 + haswell (gen 7.5) failing

it does, but beignet should fail graciously with an error message then, not segfault. and it used to work even with haswell. groo@hydra:~/devel/opencl/tools-master$ ./cl-demo 1 10 Choose platform: [0] Intel [1] Mesa Enter choice: 0 Choose device: [0] Intel(R) HD Graphics Haswell Ultrabook GT2