looks good to me, thanks.
-Original Message-
From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Xiuli
Pan
Sent: Wednesday, July 27, 2016 10:11 AM
To: beignet@lists.freedesktop.org
Cc: Pan, Xiuli
Subject: [Beignet] [PATCH V3] Utset: Add check for workgroup tests
From: Pan Xiuli
Workgroup is also an OpenCL 2.0 feature, add check for these cases.
V2: Move check before kernel build
V3: Remove uselees check
Signed-off-by: Pan Xiuli
---
utests/compiler_workgroup_broadcast.cpp | 12 ++
I've submited a new patchset [1] to address the issues related to lgamma and
lgamma_r.
The problem should now be fixed - all tests defined in utest and conformance
are passing on BDW.
If all tests pass on other configurations as well I think we could merge this
patchset [1] to master.
In
From: Grigore Lupescu
Affected functions:
__gen_ocl_internal_log
__gen_ocl_internal_log10
__gen_ocl_internal_log2
__gen_ocl_internal_log_valid
__kernel_sinf
__kernel_cosf
__gen_ocl_internal_cbrt
__gen_ocl_internal_asinh
__gen_ocl_internal_atan
__gen_ocl_asin_util
tan
log1p
lgamma_r
lgamma
From: Grigore Lupescu
Affected functions:
log
log2
log10
exp
exp2
Signed-off-by: Grigore Lupescu
---
backend/src/libocl/tmpl/ocl_math.tmpl.cl | 18 +-
1 file changed, 17 insertions(+), 1 deletion(-)
diff --git a/backend/src/libocl/tmpl/ocl_math.tmpl.cl
From: Grigore Lupescu
Use lower grade polynomials for approximations, keep conformance passing.
LOG Use polynomial grade 4 (was 7)
LOG2Use polynomial grade 4 (was 7)
SIN Use polynomial grade 4 (was 6)
COS Use polynomial grade 3 (was 6)
TANFUse polynomial grade 7 (was 12)
From: Pan Xiuli
Workgroup is also an OpenCL 2.0 feature, add check for these cases.
V2: Move check before kernel build
Signed-off-by: Pan Xiuli
---
utests/compiler_workgroup_broadcast.cpp | 12 ++
utests/compiler_workgroup_reduce.cpp
From: Pan Xiuli
Workgroup is also an OpenCL 2.0 feature, add check for these cases.
Signed-off-by: Pan Xiuli
---
utests/compiler_workgroup_broadcast.cpp | 2 ++
utests/compiler_workgroup_reduce.cpp | 2 ++
LGTM, thanks.
-Original Message-
From: Beignet [mailto:beignet-boun...@lists.freedesktop.org] On Behalf Of Yang
Rong
Sent: Tuesday, July 26, 2016 4:50 PM
To: beignet@lists.freedesktop.org
Cc: Yang, Rong R
Subject: [Beignet] [PATCH] Runtime: fix a userptr bug.
Userptr also require size
Userptr also require size cache alignment, otherwise, the remained
memory may be allocated in CPU side, when gpu flush the last cacheline
to memory, will override the value changed by CPU.
Signed-off-by: Yang Rong
---
src/cl_mem.c | 4 +++-
1 file changed, 3
the return type of CL_PROGRAM_BINARY_SIZES query is unsigned char*[],
and param_value_size must be >= size of the return type, see spec 1.2
section 5.6.7 (P151)
Signed-off-by: Guo Yejun
---
utests/get_cl_info.cpp | 10 ++
1 file changed, 6 insertions(+), 4
11 matches
Mail list logo