Could you try latest git master beignet? We have some major
performance improvement for some cases.
Not sure whether it is the case for JohnTheRipper, but worth to give a try.
Thanks,
Zhigang Gong.
On Sat, Oct 25, 2014 at 2:46 PM, Oleksii Shevchuk
wrote:
> Zhigang Gong writes:
>
>> This should
Zhigang Gong writes:
> This should be an application bug, according to OpenCL 1.2 spec:
>
> CL_PROFILING_INFO_NOT_AVAILABLE if the CL_QUEUE_PROFILING_ENABLE
> flag is not set for the command-queue, if the execution status of
> the command identified
> by event is not CL_COMPLETE or if event
As you are using 3.16.x kernel, the most possible solution is in the
README.md's known issue section:
Almost all unit tests fail on Linux kernel 3.15/3.16. There is a known
issue in some versions of linux kernel which enable register whitelist
feature but miss some necessary registers which are re
This should be an application bug, according to OpenCL 1.2 spec:
CL_PROFILING_INFO_NOT_AVAILABLE if the CL_QUEUE_PROFILING_ENABLE
flag is not set for the command-queue, if the execution status of
the command identified
by event is not CL_COMPLETE or if event is a user event object.
To make
Thanks for your help.
I will check.
Yan Wang
> Hi,
> The 2 bugs are:
> 1. https://jira01.devtools.intel.com/browse/VIZ-4490
> VIZ-4490 [HSW/IVB/BYT-M] After update opencv, 1 opencv case
> (opencv_test_imgproc/OCL_Imgproc/HoughLines)fail
>
> 2. https://jira01.devtools.intel.com/browse/VIZ-4489
Hi list!
I'm trying to use beignet (67650189c145a65addffdcd4d8ff452709bd4149)
with johntheripper (ceb8aa1899817965afd26ffb651932e475a7bbc7 at
https://github.com/magnumripper/JohnTheRipper) on i7/HD4000.
Test suite reports that tests are passed:
.
compiler_abs_diff_int()Instruction (#42) src
Jérôme,
Thanks for sharing the very interesting performance data.
We will look into the large image issue.
Thanks,
Zhigang Gong.
___
Beignet mailing list
Beignet@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/beignet
This assertion is just to make sure we will not get a NULL pointer
for a normal buffer. The OpenCL spec doesn't give a very specific
statement about a NULL buffer object. But it does allow to pass
a NULL to a buffer object. Thus some one may implement the following
kernel:
__kernel foo( global uin
Considering the following CFG, our previous liveness analysis will only
back traverse from the exit point BB, thus the BB 6 and 7 will not be handled.
4---
| |
| |
5 --|---> 10 ---> ret
| |
| |
6<--
|
|<--
7 |
| |
---
Although the CFG l
This patchset LGTM, thanks.
> -Original Message-
> From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of
> Zhenyu Wang
> Sent: Thursday, October 23, 2014 15:19
> To: beignet@lists.freedesktop.org
> Subject: [Beignet] [PATCH v2 1/5] Make use of write enable flag for mem
On 2014.10.22 08:49:53 +, Guo, Yejun wrote:
> three comments in line, thanks.
>
Thanks for review this.
>gpgpu->aux_offset.surface_heap_offset = size_aux;
>size_aux += sizeof(surface_heap_t);
>
> @@ -784,7 +782,10 @@ intel_gpgpu_state_init(intel_gpgpu_t *gpgpu,
>gpgpu->aux_offs
On Sun, 2014-10-19 at 22:15 +0800, Boxiang Sun wrote:
> Did you set the 'MESA_SOURCE_FOUND' to '1' and the path of mesa
> source
> directory in Beignet cmake file?
>
> I'm pretty sure that the current master could not compiles with Mesa
> 10.3.x
> if without some modifications.
>
> Regards,
> S
From: Junyan He
Signed-off-by: Junyan He
---
backend/src/backend/gen/gen_mesa_disasm.c | 1110 +++--
backend/src/backend/gen_defs.hpp |2 +
2 files changed, 582 insertions(+), 530 deletions(-)
diff --git a/backend/src/backend/gen/gen_mesa_disasm.c
b/backen
On Sat, Oct 18, 2014 at 04:26:02PM +0200, Martin Hauke wrote:
> Hi Zhigang,
>
> On 18.10.2014 04:36, you wrote:
> >> building git master on OpenSUSE 13.1 actually fails since this OpenSUSE
> >> version ships only with libdrm2-2.4.46-3.2.2
> > As you may know, we are working on enable BDW. And the
three comments in line, thanks.
-Original Message-
From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of
Zhenyu Wang
Sent: Wednesday, October 22, 2014 4:11 PM
To: beignet@lists.freedesktop.org
Subject: [Beignet] [PATCH] Fix AUX buffer for really page aligned
Apply ALI
From: Junyan He
For BDW, the vertical align is 4 at least.
This cause the slice pitch twice as big as
the Gen7 for 1D buffer array.
Because the buffer tiling alignment may change
for different GENs, we move it from run time to
intel driver.
V2:
Fix all the bugs about 1d and 2d image array.
A
Should check hstride when unpack.
Signed-off-by: Yang Rong
---
backend/src/backend/gen_encoder.cpp | 34 --
1 file changed, 28 insertions(+), 6 deletions(-)
diff --git a/backend/src/backend/gen_encoder.cpp
b/backend/src/backend/gen_encoder.cpp
index 847ab7b..c11
In function cl_get_kernel_max_wg_sz, hsw's thread count may large than 64,
add a max limit.
Signed-off-by: Yang Rong
---
src/cl_device_id.c | 10 +++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/src/cl_device_id.c b/src/cl_device_id.c
index a0d0db6..7944ca4 100644
--- a/s
This patch LGTM, will push latter, thanks.
On Fri, Oct 10, 2014 at 03:01:25PM +0800, Ruiling Song wrote:
> we should use simd1, instead of simd8/simd16.
>
> Signed-off-by: Ruiling Song
> ---
> backend/src/backend/gen_context.cpp | 20 ++--
> backend/src/backend/gen_context.hpp
This patch itself LGTM. But I will move it after other patches.
Thus once this patch get applied, it can accept Gen8 device and
work as expected.
Thanks.
On Mon, Sep 29, 2014 at 01:37:44PM +0800, Yang Rong wrote:
> From: Junyan He
>
> Signed-off-by: Junyan He
> ---
> src/cl_command_queue.c
Yes, I am also plan to change GenContext to a pure virtual class, and I think
It is better to do this when optimize the long operations in BDW.
-Original Message-
From: He Junyan [mailto:junyan...@inbox.com]
Sent: Thursday, October 9, 2014 13:09
To: Yang, Rong R
Cc: beignet@lists.freedes
This patchset LGTM
On 一, 2014-09-29 at 13:37 +0800, Yang Rong wrote:
> From: Junyan He
>
> Must call cl_bind_buf instead of intel_gpgpu_bind_buf directly in intel_gpgpu.
>
> Signed-off-by: Junyan He
> ---
> src/intel/intel_gpgpu.c | 36 +++-
> 1 file changed, 2
Hello,
I'm trying to run beignet from Wolfram Mathematica 10.0.1, I have
beignet 0.9.2, Fedora 20 x86_64, kernel 3.16.3, i5-3230M.
First, there is an issue which is probably to be blamed on Mathematica
(you need to manually load /usr/lib64/beignet/libcl.so with
LibraryLoad), I suppose that Math
Hi,
The 2 bugs are:
1. https://jira01.devtools.intel.com/browse/VIZ-4490
VIZ-4490 [HSW/IVB/BYT-M] After update opencv, 1 opencv case
(opencv_test_imgproc/OCL_Imgproc/HoughLines)fail
2. https://jira01.devtools.intel.com/browse/VIZ-4489
VIZ-4489 [IVB/HSW/BYT-M] After update opencv, some openc
Could you give me one URL?
Thanks.
Yan Wang
> All of these three failures are already tracked in JIRA.
> If you have access to JIRA, you can check them easily.
>
> Thanks,
> Zhigang Gong.
>
> On Fri, Oct 24, 2014 at 10:33 PM, wrote:
>> Sure. I could try Yang Rong's patch.
>> BTW, I also meet th
All of these three failures are already tracked in JIRA.
If you have access to JIRA, you can check them easily.
Thanks,
Zhigang Gong.
On Fri, Oct 24, 2014 at 10:33 PM, wrote:
> Sure. I could try Yang Rong's patch.
> BTW, I also meet the following failed cases. Could you please confirm
> them? T
Sure. I could try Yang Rong's patch.
BTW, I also meet the following failed cases. Could you please confirm
them? Thanks.
[ FAILED ] 3 tests, listed below:
[ FAILED ] OCL_ImgProc/Canny.Accuracy/8, where GetParam() =
(Channels(3), AppertureSize(3), L2gradient(false), UseRoi(false))
[ FAILED ]
Hi Zhigang,
On Fri, Oct 24, 2014 at 12:13 PM, Zhigang Gong
wrote:
> Hi,
>
> Luxmark (both 2.0/2.1) works fine on my IVB machine. The back trace
> you provided below doesn't indicate it's a beignet related problem.
> It hadn't enter beignet domain and just crashed in luxmark internal.
I'm testing
Per OpenCL spec 1.2:
CL_DEVICE_IMAGE_MAX_BUFFER_SIZE should be size_t type rather
than cl_ulong.
This bug will cause problems on i386 platform.
Signed-off-by: Zhigang Gong
---
src/cl_device_id.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/cl_device_id.h b/src/cl_devi
Hi,
Luxmark (both 2.0/2.1) works fine on my IVB machine. The back trace
you provided below doesn't indicate it's a beignet related problem.
It hadn't enter beignet domain and just crashed in luxmark internal.
On Fri, Oct 24, 2014 at 12:04:29PM +0300, Vasily Khoruzhick wrote:
> Hi Zhigang,
>
> Lu
I have BYT box, an IVB machine and a HSW notebook. All of them haven't
this issue.
Yang rong is working on a race condition patch. This issue may be
related. You may
try again once he send out the patch.
On Fri, Oct 24, 2014 at 5:46 PM, wrote:
> Hi, Zhigang,
> May your platform also get failed
Hi, Zhigang,
May your platform also get failed cases when run
OCL_ImageProc/Filter2D.Mat* although no crash, because I have only
Baytail T platfrom. I am not sure the other platform has the same issue.
If yes, I could try to investigate the reason. It may have issue of
filter2D OpenCL implement
Hi Zhigang,
Luxmark crashes with following backtrace:
Program received signal SIGSEGV, Segmentation fault.
0x004cd8b0 in slg::PathOCLRenderEngine::StopLockLess() ()
(gdb) bt
#0 0x004cd8b0 in slg::PathOCLRenderEngine::StopLockLess() ()
#1 0x00482236 in slg::RenderEngine::
I just checked again, and found both 3.17 and 3.17.1 should work
fine on IVB with beignet. I just tested beignet on IVB with kernel
3.17.1, all unit tests passed successfully. For IVB user, no need
to wait for 3.18.
Don't know which application you were testing on your IVB machine.
If it could be
LGTM, pushed, thanks.
On Sat, Oct 25, 2014 at 03:10:02AM +0800, Meng Mengmeng wrote:
> ---
> src/CMakeLists.txt | 5 ++---
> src/git_sha1.sh| 4 ++--
> 2 files changed, 4 insertions(+), 5 deletions(-)
>
> diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
> index 0d22589..9e65856 100644
>
35 matches
Mail list logo