Re: [Beignet] [PATCH] GBE: clean llvm module's clone and release.

LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Thursday, June 22, 2017 14:04 To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] GBE: clean llvm module's clone and

Re: [Beignet] [PATCH] Backend: Fix double free of the cloned_module

OK, the chaos in llvmToGen is the root cause of this bug, refine that function may solve the problem. -Original Message- From: Yang, Rong R Sent: Thursday, June 22, 2017 09:38 To: Pan, Xiuli <xiuli@intel.com>; beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@

Re: [Beignet] [PATCH 2/3] Runtime: Add new API enums for cl_intel_required_subgroup_size extension

Add with the next version macro. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Pan, Xiuli Sent: Tuesday, June 13, 2017 17:06 To: Yang, Rong R <rong.r.y...@intel.com>; beignet@lists.freedesktop.org Subject: Re: [Beignet] [PATCH 2/3] R

Re: [Beignet] [PATCH 2/3] Runtime: Add new API enums for cl_intel_required_subgroup_size extension

Now we have no clang support this extension, I think we can add that in next LLVM support. -Original Message- From: Yang, Rong R Sent: Tuesday, June 13, 2017 16:41 To: Pan, Xiuli <xiuli@intel.com>; beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@intel.com>

Re: [Beignet] [PATCH 1/3] Backend: Add intel_reqd_sub_group_size support

The spec has required the subgroup size to be 8 or 16, and I think we may need to fail the build in some other place. -Original Message- From: Yang, Rong R Sent: Tuesday, June 13, 2017 16:44 To: Pan, Xiuli <xiuli@intel.com>; beignet@lists.freedesktop.org Cc: Pan, Xiuli

Re: [Beignet] [PATCH] Fix libva examples build error

Ping. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Xiuli Pan Sent: Friday, May 5, 2017 16:23 To: beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@intel.com> Subject: [Beignet] [PATCH] Fix libva examples build error From: Pan

Re: [Beignet] [PATCH] do constant folding for kernel struct args

Rander seems to have a similar optimization about imm value at sel ir. If your case here need the optimization done in GEN IR level then rander's patch may no longer be needed. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Guo, Yejun Sent:

Re: [Beignet] [PATCH] backend: refine hypot function

LGTM. Only test about correctness performance need recheck. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of rander.wang Sent: Thursday, May 18, 2017 16:18 To: beig...@freedesktop.org Cc: Wang, Rander Subject: [Beignet]

Re: [Beignet] [PATCH] backend: refine exp function with float input

LGTM. Thanks -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of rander.wang Sent: Wednesday, June 7, 2017 15:47 To: beig...@freedesktop.org Cc: Wang, Rander Subject: [Beignet] [PATCH] backend: refine exp function with

Re: [Beignet] [PATCH] backend: refine exp function with float input

I only check the correctness, there is no problem It seems mix using tab and space in the function __gen_ocl_internal_simple_exp and there some intend problem. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of rander.wang Sent: Wednesday, June

Re: [Beignet] [PATCH] backend: add sqrt-div pattern to instruction select

In the utest case, it change sqrtsqrt div ==> rqrt div rqrt div rqrt which case will run faster? -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of rander.wang Sent: Friday, May 19, 2017 16:19 To:

Re: [Beignet] [PATCH] utests: added for optimization negtiveAdd

I found the test case name does not follow the style, maybe you can replace up case with _ Thanks -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Pan, Xiuli Sent: Monday, May 22, 2017 11:47 To: Wang, Rander <rander.w...@intel.com>

Re: [Beignet] [PATCH] utests: added for optimization negtiveAdd

LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of rander.wang Sent: Friday, May 19, 2017 16:42 To: beig...@freedesktop.org Cc: Wang, Rander Subject: [Beignet] [PATCH] utests: added for optimization negtiveAdd

Re: [Beignet] [PATCH 1/2] Backend: Add optimization for negtive modifier

LGTM. Thanks -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of rander.wang Sent: Wednesday, May 17, 2017 16:21 To: beig...@freedesktop.org Cc: Wang, Rander Subject: [Beignet] [PATCH 1/2] Backend: Add optimization for

Re: [Beignet] [PATCH] backend: add sqrt-div pattern to instruction select

I discuss with Yang Rong, and found you can do more merge for sqrt r1, r2; load r4, 2.0; ===> rqrt r5, r2 div r3, r4, r1; mul r3, r5, 2.0 To get less loadi. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of

Re: [Beignet] [PATCH 1/2] backend: fix tgamma error after restructure

LGTM. Thanks. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of rander.wang Sent: Friday, May 19, 2017 11:02 To: beig...@freedesktop.org Cc: Wang, Rander Subject: [Beignet] [PATCH 1/2] backend: fix tgamma error after

Re: [Beignet] [PATCH] Backend: Fix llvm40 assert about literal structs

Ping for review. If llvm is debug version will cause assert for device enqueue cases. -Original Message- From: Pan, Xiuli Sent: Tuesday, April 25, 2017 13:27 To: beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@intel.com> Subject: [PATCH] Backend: Fix llvm40 assert about l

Re: [Beignet] [PATCH 3/3] GLK: add geminilake runtime support.

LGTM. Thanks -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Monday, May 15, 2017 16:33 To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH 3/3] GLK: add geminilake runtime

Re: [Beignet] issue with cl

Hi Michal, Could you report a bug on https://bugs.freedesktop.org/ and provide the full log. Thanks Xiuli -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Michal B Sent: Sunday, April 9, 2017 11:41 PM To: beignet@lists.freedesktop.org

Re: [Beignet] [PATCH 7/7] Backend: Add LLVM40 support

Ping for review. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Xiuli Pan Sent: Friday, March 17, 2017 2:16 PM To: beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@intel.com> Subject: [Beignet] [PATCH 7/7] Backend: Add LLVM40 s

Re: [Beignet] [PATCH] Properly check return value from __cxa_demangle

From the source code of the __cxa_demangle or libcxxrt: https://github.com/pathscale/libcxxrt/blob/c94e7e03165e26a085cd1f7b887c5aec15f387a9/src/typeinfo.cc#L95 It seems in this case the buf and n are all NULL, thus the status won’t be changed when the function returned. I have tried it works with

Re: [Beignet] [PATCH 5/7] Backend: Refine LLVM version check macro

2:23 PM To: Pan, Xiuli <xiuli@intel.com>; beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@intel.com> Subject: RE: [Beignet] [PATCH 5/7] Backend: Refine LLVM version check macro > -#if LLVM_VERSION_MAJOR == 3 && LLVM_VERSION_MINOR >= 9 > +#if LLVM_VERSION_MAJ

Re: [Beignet] [PATCH v2 5/5] cl_intel_media_block_io WRITE related unit tests.

LGTM, thanks. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of xionghu@intel.com Sent: Wednesday, March 8, 2017 11:01 PM To: beignet@lists.freedesktop.org Cc: Luo, Xionghu Subject: [Beignet] [PATCH v2 5/5]

Re: [Beignet] [PATCH 2/2] cl_intel_media_block_io WRITE related unit tests.

1. white space in compiler_subgroup_image_block_write.cl on line if (out[vsz * image_row_pitch + i] != expected[WG_GLOBAL_SIZE_X * vsz + i]) 2. one printf in kernel compiler_subgroup_media_block_write_ui2 Others LGTM. -Original Message- From: Beignet

Re: [Beignet] [PATCH 3/3] fix build error log not output issue.

I think llvm::errs() << options; should be llvm::errs() << err; And there seems some more this kind of bugs in program.cpp. Such as programCompileFromSource, no need for output options. For there is no reason to output some options here. -Original Message- From: Beignet

Re: [Beignet] [PATCH 2/3] cl_intel_media_block_io READ related unit tests.

LGTM, except the extension should be in first patch. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of xionghu@intel.com Sent: Friday, March 3, 2017 12:08 AM To: beignet@lists.freedesktop.org Cc: Luo, Xionghu Subject:

Re: [Beignet] [PATCH 1/3] add extension cl_intel_media_block_io READ related function

1. Missing a #define in libocl header 2. Extension should be added in this patch otherwise in utest patch 3. In IR MediaBlockReadInstruction the width and height no longer need default value Others LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On

Re: [Beignet] [PATCH] MAD compact instrcution could not support "absolute" attribute.

LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of yan.w...@linux.intel.com Sent: Thursday, February 23, 2017 4:20 PM To: beignet@lists.freedesktop.org Cc: Yan Wang Subject: [Beignet] [PATCH] MAD compact

Re: [Beignet] [PATCH] Runtime: add a warning when load gen binary fail.

LGTM! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Friday, February 10, 2017 2:38 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] Runtime: add a warning when load gen

Re: [Beignet] [PATCH] CMAKE: if user disable ICD support, must pass cmake option explicitly.

LGTM! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Wednesday, February 8, 2017 12:53 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] CMAKE: if user disable ICD

Re: [Beignet] [PATCH] Only provide device IDs if the type matches

Hi Giuseppe, You point about the device type is right, but the git commit you found is not right the function is just copy form somewhere else. And the this patch is not follow the OpenCL spec. As the Beignet does not have a CPU device, we should just return CL_INVALID_DEVICE_TYPE. I will

Re: [Beignet] [PATCH v2] Enable OpenCL 2.0 only where supported

...@gmail.com] Sent: Monday, January 23, 2017 5:21 PM To: Pan, Xiuli <xiuli@intel.com>; Rebecca N. Palmer <rebecca_pal...@zoho.com>; beignet@lists.freedesktop.org; ignate...@redhat.com Subject: Re: [Beignet] [PATCH v2] Enable OpenCL 2.0 only where supported Yes indeed, I think R

Re: [Beignet] [PATCH] FindLLVM: Accept LLVM 3.9

We have no test with llvm 3.4 now, but as we have no big changes I think llvm 3.4 can still work. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Simon Richter Sent: Monday, January 23, 2017 7:12 PM To: beignet@lists.freedesktop.org Subject:

Re: [Beignet] [PATCH v2] Enable OpenCL 2.0 only where supported

I am writing these changes now and will try to start test soon. And I will spilt the patch into runtime change and utest change. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Pan, Xiuli Sent: Monday, January 23, 2017 10:45 AM To: Rebecca N

Re: [Beignet] Graphical Glitches and Local Memory Performance Issue

Hi, Could you make this as a bug report to Beignet on https://bugs.freedesktop.org/ It will help us to track the status and you can also provide some more information and reproduce codes. As for the Local buffer type, it seems to be our query API's bug. Our implementation does not use the

Re: [Beignet] [PATCH v2] Enable OpenCL 2.0 only where supported

I think if we want to enable OpenCL 2.0 as default we must check if there is a LLVM 3.9 first, or there will have errors to ask people to update llvm. We may turn the OCL 20 off when no llvm39 is found. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On

Re: [Beignet] [PATCH] FindLLVM: Accept LLVM 3.9

LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Rebecca N. Palmer Sent: Monday, January 23, 2017 1:07 AM To: beignet@lists.freedesktop.org Subject: [Beignet] [PATCH] FindLLVM: Accept LLVM 3.9 Signed-off-by: Rebecca N. Palmer

Re: [Beignet] [PATCH] Fix typo

LGTM, thanks. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Rebecca N. Palmer Sent: Monday, January 23, 2017 1:09 AM To: beignet@lists.freedesktop.org Subject: [Beignet] [PATCH] Fix typo Signed-off-by: Rebecca N. Palmer

Re: [Beignet] Building beignet with OpenCL 2.0 support in distros

Hi all, Thank you for pay close attention to our latest release. We are very sorry that we are not very thoughtful about the package making things. We did not have a performance test for OCL2.0 so we did not set it as a default option and we did not think about how to run a binary version on

Re: [Beignet] [PATCH] GBE: fix llvm3.5 version build error.

LGTM, thanks. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Thursday, January 19, 2017 6:15 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] GBE: fix llvm3.5 version

Re: [Beignet] [PATCH] Android.mk: update Android.mk for android build.

LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Thursday, January 19, 2017 5:39 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] Android.mk: update Android.mk for

Re: [Beignet] [PATCH] Add some pointer access check.

LGTM. Thanks. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Wednesday, January 11, 2017 6:16 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] Add some pointer access

Re: [Beignet] [PATCH] GBE: fix long work group fail.

The patch LGTM, thanks. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Friday, January 6, 2017 10:17 AM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] GBE: fix long work

Re: [Beignet] [PATCH] utest: fix i386 system long ctz fail.

LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Tuesday, January 3, 2017 6:14 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] utest: fix i386 system long ctz fail.

Re: [Beignet] [Patch V2] GBE: fix a src/dst register reuse bug.

LGTM, it will solve the overlap register problem. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Friday, December 30, 2016 12:39 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet]

Re: [Beignet] [PATCH V2] add sends support for oword/media block write

LGTM! Thanks -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Guo, Yejun Sent: Wednesday, December 28, 2016 3:18 PM To: beignet@lists.freedesktop.org Cc: Guo, Yejun Subject: [Beignet] [PATCH V2] add sends support for

Re: [Beignet] [PATCH V2] output more detail of GEN IR for workgroup op

LGTM! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Guo, Yejun Sent: Wednesday, December 21, 2016 6:33 PM To: beignet@lists.freedesktop.org Cc: Guo, Yejun Subject: [Beignet] [PATCH V2] output more detail of GEN IR for

Re: [Beignet] [PATCH] add sends support for oword/media block write

It seems you missed the gen7_encoder part for the media block write, gen7 has a different data port and the virtual function for MBWRITE should also be changed for Gen7Encoder. Others LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of

Re: [Beignet] Proposed patch: add CMake option for toggling OCL ICD Loader compatibility

It seems the mail list could not see this mail's attachment, could you try with git send-email to send the patch directly? Thanks Xiuli -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Marek Szuba Sent: Thursday, December 1, 2016 8:24 PM To:

Re: [Beignet] [PATCH V4] Add profiling feature based on new event implementation.

It seems still have bugs. Here are some logs I got, the gen timestamps is print in the function cl_event_update_timestamp_gen, and the final result is print last as timestamp. gen timestamp[0] is d88bddb30 gen timestamp[1] is d88bde2b0 run for 8 times gen timestamp[2] is d88bddae0 // It is

Re: [Beignet] [PATCH] GBE: fix a src/dst register reuse bug.

(2608) mad(8) g5<1>:F g4.1<0,1,0>:F g127.7<0,1,0>:F g47:F { align16 WE_normal 2Q }; From: Ruiling Song [mailto:ruiling.son...@gmail.com] Sent: Wednesday, December 21, 2016 10:59 AM To: Pan, Xiuli <xiuli@intel.com> Cc: Song, Ruiling <ruiling.s...@int

Re: [Beignet] [PATCH] GBE: fix a src/dst register reuse bug.

I got the point. In some case, such as block read gemm. We may get 16 mad with g4, thus there may be some other mad using g4.2 or g4.3 later. If we reuse the g4, the g4.2 or g4.3 may be overwrite. In this case may be g4.1 is just a scalar value, and g4.2 or g4.* can be some other scalar value.

Re: [Beignet] [PATCH V3] Add profiling feature based on new event implementation.

When test with gemm case, it will get random CL_INVALID_VALUE from the CL_EVENT_INVALID_TIMESTAMP. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of junyan...@inbox.com Sent: Friday, December 16, 2016 10:29 PM To: beignet@lists.freedesktop.org

Re: [Beignet] [PATCH 10/19] OCL20: enable -cl-std=CL2.0.

I will review it asap. Thanks Xiuli -Original Message- From: Song, Ruiling Sent: Monday, December 19, 2016 2:47 PM To: Yang, Rong R <rong.r.y...@intel.com>; beignet@lists.freedesktop.org; Pan, Xiuli <xiuli@intel.com> Cc: Yang, Rong R <rong.r.y...@intel.com> Sub

Re: [Beignet] [PATCH V2 5/5] enable sends for typed write

V2 LGTM. Thanks. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Guo, Yejun Sent: Thursday, December 15, 2016 1:43 PM To: beignet@lists.freedesktop.org Cc: Guo, Yejun Subject: [Beignet] [PATCH V2 5/5] enable sends for

Re: [Beignet] [PATCH] Backend: Refine block read/write instruction selection

. -Original Message- From: Guo, Yejun Sent: Wednesday, December 14, 2016 3:43 PM To: Pan, Xiuli <xiuli@intel.com>; beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@intel.com> Subject: RE: [Beignet] [PATCH] Backend: Refine block read/write instruction selection two comments

Re: [Beignet] [PATCH] Backend: Fix GenRegister::offset sub reg offset

Ping for review. -Original Message- From: Pan, Xiuli Sent: Monday, November 7, 2016 4:06 PM To: beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@intel.com> Subject: [PATCH] Backend: Fix GenRegister::offset sub reg offset From: Pan Xiuli <xiuli@intel.com> We us

Re: [Beignet] [PATCH] save host_ptr when create sub buffer from CL_MEM_ALLOC_HOST_PTR

LGMT, the test code works fine. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Guo, Yejun Sent: Thursday, December 1, 2016 1:42 PM To: beignet@lists.freedesktop.org Cc: Guo, Yejun Subject: [Beignet] [PATCH] save

Re: [Beignet] fatal error: 'ocl.h' file not found

Hi Shengquan, You should also copy the include for CLANG usage and keep the clang the some version. Thanks Xiuli From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yuan, Shengquan Sent: Tuesday, November 8, 2016 1:32 PM To: beignet@lists.freedesktop.org Subject:

Re: [Beignet] [PATCH] GBE: Fix a logical insn with flag bug.

LGTM! It is a flag conflict condition. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Monday, October 31, 2016 6:44 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] GBE:

Re: [Beignet] Problems compiling the OCL20 branch

: Thursday, October 27, 2016 1:15 AM To: Pan, Xiuli <xiuli@intel.com>; beignet@lists.freedesktop.org Subject: Re: [Beignet] Problems compiling the OCL20 branch Hi Xiuli Thanks for the patches. I managed to compile the OCL20 branch and use the shared virtual memory features. I have looked

Re: [Beignet] Problems compiling the OCL20 branch

Hi Jesper, The OCL20 branch can only build with LLVM3.8+ for some OpenCL 2.0 features need frontend support, and for LLVM3.8 we need some patch for clang. And for LLVM3.9 there are some change we have not merged into OCL20 branch. So if you would try the OCL20 branch, I suggest you build a

Re: [Beignet] [PATCH] Add new BXT and KBL pciids to GetGenID.sh.

LGTM. Thanks -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Monday, September 12, 2016 6:25 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] Add new BXT and KBL pciids to

Re: [Beignet] [PATCH] GBE: fix half immediate negate assert.

LGTM! Thanks! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of rong.r.y...@intel.com Sent: Sunday, September 18, 2016 10:50 AM To: beignet@lists.freedesktop.org Cc: Luo, Xionghu ; Yang, Rong R

Re: [Beignet] [PATCH] Runtime: avoid clang warning "warning: expression result unused".

LGTM. Thanks -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Wednesday, August 31, 2016 7:28 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] Runtime: avoid clang warning

Re: [Beignet] Weird clinfo ouput @ FreeBSD-11.0/Intel GPU

Hi Igor, It seems something wrong with the libdrm, what the libdrm you are using? Thanks Xiuli From: Igor Pavlov [mailto:igor.arabesc.pav...@gmail.com] Sent: Wednesday, October 12, 2016 6:13 PM To: Pan, Xiuli <xiuli@intel.com> Cc: beignet@lists.freedesktop.org Subject: Re: [Beignet]

Re: [Beignet] Weird clinfo ouput @ FreeBSD-11.0/Intel GPU

libdrm Thanks Xiuli From: Igor Pavlov [mailto:igor.arabesc.pav...@gmail.com] Sent: Wednesday, October 12, 2016 8:00 AM To: Pan, Xiuli <xiuli@intel.com> Cc: beignet@lists.freedesktop.org Subject: Re: [Beignet] Weird clinfo ouput @ FreeBSD-11.0/Intel GPU It's a headless system, there i

Re: [Beignet] Weird clinfo ouput @ FreeBSD-11.0/Intel GPU

Hi Igor, Could you provide the PCI ID of your device by lspci –v. It seems we could not get the device, it may because that you build with X but running without a screen open, or build without X but running by non-root user. Thanks Xiuli From: Beignet

Re: [Beignet] [PATCH] Add clGetKernelSubGroupInfoKHR to _cl_icd_dispatch table

Hi, The api clGetKernelSubGroupInfoKHR for OpenCL 1.2 is added for the extension cl_intel_subgroups. https://www.khronos.org/registry/cl/extensions/intel/cl_intel_subgroups.txt And I add it in https://cgit.freedesktop.org/beignet/commit/?id=aa3ee67c825fbef5f68b31345c060c981bf35ad3 Now I see

Re: [Beignet] Status of LLVM 3.9 support?

This patch is still under review, and it will soon be pushed as it is reviewed. Thanks Xiuli -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Rebecca N. Palmer Sent: Monday, September 26, 2016 6:38 AM To: beignet@lists.freedesktop.org Subject:

Re: [Beignet] [PATCH] add another broxton pciid 0x5A85

I think the pciid patch can go first and then the stack size one, for the stack size patch will influence not only BXT but also CHV. -Original Message- From: Guo, Yejun Sent: Monday, September 12, 2016 10:39 AM To: Pan, Xiuli <xiuli@intel.com>; beignet@lists.freedesktop.org S

Re: [Beignet] [PATCH] add another broxton pciid 0x5A85

I think this patch can be spilt into two patch. One is add this pciid. and the other is for the stack size bug. Others LGTM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Guo Yejun Sent: Saturday, September 10, 2016 8:49 AM To:

Re: [Beignet] [PATCH] GBE: fix uninitialized build warning.

LGTM! Thanks! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Tuesday, September 6, 2016 12:09 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] GBE: fix uninitialized

Re: [Beignet] [PATCH 2/6] Utest: Fix sub group broadcast for simd8

The spec only defined the index should be a sub group local id. In simd 8 mode, it should be less than 8. The fix here is change the index to make the test fit for both simd 16 and simd 8. -Original Message- From: Yang, Rong R Sent: Monday, September 5, 2016 1:35 PM To: Pan, Xiuli

Re: [Beignet] [PATCH] libocl: add vload/vstore for half type

Hi Jan, Thanks for your patch, but I have already sent a patch set to deal with missing half type builtin functions with an utest case. You can find them here: https://lists.freedesktop.org/archives/beignet/2016-July/007783.html Thanks Xiuli -Original Message- From: Beignet

Re: [Beignet] [PATCH] use different pointer alignment for different implementation

LGTM! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Guo Yejun Sent: Tuesday, August 2, 2016 4:16 AM To: beignet@lists.freedesktop.org Cc: Guo, Yejun Subject: [Beignet] [PATCH] use different pointer alignment for

Re: [Beignet] [PATCH 2/2] GBE: Improve spill policy by considering use count.

-Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Ruiling Song Sent: Tuesday, June 21, 2016 9:03 AM To: beignet@lists.freedesktop.org Cc: Song, Ruiling Subject: [Beignet] [PATCH 2/2] GBE: Improve spill policy by

Re: [Beignet] [PATCH] GBE: fix a memset typo.

LGTM! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Monday, June 20, 2016 6:03 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject: [Beignet] [PATCH] GBE: fix a memset typo. Signed-off-by:

Re: [Beignet] cl_intel_subgroups support

...@lists.freedesktop.org] On Behalf Of Pan, Xiuli Sent: Tuesday, May 17, 2016 1:12 PM To: Andrew Lavin <aj.la...@gmail.com>; Beignet <beignet@lists.freedesktop.org> Subject: Re: [Beignet] cl_intel_subgroups support Hi Andrew, We do not have this extension right now, but I am working on th

Re: [Beignet] [PATCH 2/8] OCL20: add device enqueue builtins.

This patch set LGTM. Some comments and debug code need to fix. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Friday, May 20, 2016 3:46 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject:

Re: [Beignet] [PATCH 4/8] OCL20: handle device enqueue helper functions in the backend.

The comments may need change here. All seems copied from SLM. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang Rong Sent: Friday, May 20, 2016 3:46 PM To: beignet@lists.freedesktop.org Cc: Yang, Rong R Subject:

Re: [Beignet] cl_intel_subgroups support

Hi Andrew, We do not have this extension right now, but I am working on this extension now, please keep an eye on this mail list for patches. Thanks Xiuli From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Andrew Lavin Sent: Tuesday, May 17, 2016 12:00 PM To: Beignet

Re: [Beignet] [PATCH 1/5] Backend: Fix printfs mem leak

The PrintSlot in printfs is still copy by value, this patch only changes pointer to printfs. -Original Message- From: Luo, Xionghu Sent: Friday, May 6, 2016 9:53 AM To: Pan, Xiuli <xiuli@intel.com>; beignet@lists.freedesktop.org Cc: Pan, Xiuli <xiuli@intel.com>

Re: [Beignet] [PATCH] Remove unncessary assertion in printf processing.

LGTM! Now conformance can pass. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of yan.w...@linux.intel.com Sent: Tuesday, May 3, 2016 2:41 PM To: beignet@lists.freedesktop.org Cc: Yan Wang Subject: [Beignet] [PATCH]

Re: [Beignet] [PATCH] runtime: handle PROGRAM_BUILD_GLOBAL_VARIABLE_TOTAL_SIZE

Ping to push! -Original Message- From: Pan, Xiuli Sent: Wednesday, April 6, 2016 3:15 PM To: Ruiling Song <ruiling.s...@intel.com>; beignet@lists.freedesktop.org Cc: Song, Ruiling <ruiling.s...@intel.com> Subject: RE: [Beignet] [PATCH] run

Re: [Beignet] [PATCH] GBE: Fix destination grf register type for cmp instruction.

LGTM! Thanks! -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Ruiling Song Sent: Friday, April 22, 2016 2:27 PM To: beignet@lists.freedesktop.org Cc: Song, Ruiling Subject: [Beignet] [PATCH] GBE: Fix destination grf

Re: [Beignet] GROMACS on beignet

-boun...@lists.freedesktop.org] On Behalf Of Szilárd Páll Sent: Thursday, April 7, 2016 8:57 PM To: Pan, Xiuli <xiuli@intel.com> Cc: beignet@lists.freedesktop.org Subject: Re: [Beignet] GROMACS on beignet Hi, On Wed, Apr 6, 2016 at 5:53 AM, Pan, Xiuli <xiuli@intel.com> wrote:

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

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

Re: [Beignet] [PATCH] runtime: handle PROGRAM_BUILD_GLOBAL_VARIABLE_TOTAL_SIZE

LGTM! This is more considerate than my version. -Original Message- From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Ruiling Song Sent: Wednesday, April 6, 2016 2:29 PM To: beignet@lists.freedesktop.org Cc: Song, Ruiling Subject: [Beignet]

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

Ping for pushed for mater and OCL2.0 branch. Face same problem in Ubuntu 16.04. From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Pan, Xiuli Sent: Tuesday, March 29, 2016 4:08 PM To: Sean Lynch <se...@literati.org> Cc: beignet@lists.freedesktop.org Subject: Re: [B

Re: [Beignet] GROMACS on beignet

...@gmail.com] Sent: Wednesday, April 6, 2016 7:35 AM To: Pan, Xiuli <xiuli@intel.com> Cc: beignet@lists.freedesktop.org Subject: Re: [Beignet] GROMACS on beignet Hi Xiuli, On Tue, Apr 5, 2016 at 5:09 AM, Pan, Xiuli <xiuli@intel.com> wrote: > Hi Szilárd, > > > > 1.

Re: [Beignet] GROMACS on beignet

, April 1, 2016 6:46 PM To: Pan, Xiuli <xiuli@intel.com> Cc: beignet@lists.freedesktop.org Subject: Re: [Beignet] GROMACS on beignet Hi Xiuli, Apologies if I were not clear enough with my questions. 1. By saying that there is only 64K local memory, I assume you meant t

Re: [Beignet] clCreateFromGLTexture working?

Hi Sean, It seems we used to support the GL extension, and there are some problem with the wiki I think. Thanks Xiuli From: Song, Ruiling Sent: Wednesday, March 30, 2016 10:29 AM To: Sean Lynch <se...@literati.org>; Pan, Xiuli <xiuli@intel.com> Cc: beignet@lists.freedesktop

Re: [Beignet] clCreateFromGLTexture working?

that your final target is to run on Android, and we have tried to build on Android and you can find the patchset here https://lists.freedesktop.org/archives/beignet/2015-November/006638.html Thanks Xiuli From: Sean Lynch [mailto:se...@literati.org] Sent: Tuesday, March 29, 2016 9:41 PM To: Pan

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

out what is happening! Thanks! From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Sean Lynch Sent: Tuesday, March 29, 2016 2:20 PM To: Pan, Xiuli <xiuli@intel.com> Cc: beignet@lists.freedesktop.org Subject: [Beignet] [PATCH] Prepend std namespace to isnan and isinf

Re: [Beignet] clCreateFromGLTexture working?

Hi Sean, It seems we do not support API clCreateFromGLTexture in beignet, you can get what api we support from the file src/cl_khr_icd.c. If PyOpenCL can run into this maybe there is something wrong with the device extension check either for them or for beignet. I will recheck about this and

Re: [Beignet] patch to get beignet to compile on gcc 3.5.0

using ‘git format-patch -s’ to this mail list so that we can easily merge it into the master branch. Thanks Xiuli From: Sean Lynch [mailto:se...@literati.org] Sent: Tuesday, March 29, 2016 11:44 AM To: Pan, Xiuli <xiuli@intel.com> Cc: beignet@lists.freedesktop.org Subject: Re: [Beignet]

Re: [Beignet] [PATCH] Backend: Clang now support static, fix now

Hi Ruiling, I will do some test. But I think I should try to find which clang version start to support static and may be use some macro to handle this. Thanks Xiuli -Original Message- From: Song, Ruiling Sent: Tuesday, March 29, 2016 11:52 AM To: Pan, Xiuli <xiuli@intel.

Re: [Beignet] patch to get beignet to compile on gcc 3.5.0

Hi Sean, I was quite confused with the gcc 3.5.0, I searched on gcc.gnu.org and found there is no version named 3.5.0. If you mean the newest version 5.3.0, I have built one release version and try to build beignet and there is no error. Could you post the change log or anything related to the

Re: [Beignet] Low latency io

implement that. Sorry for the wrong information. Thanks Xiuli From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Pan, Xiuli Sent: Monday, March 28, 2016 1:08 PM To: Svein Berge <sve...@pvv.org> Cc: beignet@lists.freedesktop.org Subject: Re: [Beignet] Low latency io Hi

  1   2   3   >