8111fd3b8b
Remove device mapping from shareWithAll memory
...
When shareWithAll memory (e.g., host memory) is allocated, set appId
in hc::AmPointerInfo to -1 to indicate that this memory is not mapped
to any device. Peer checking in ihipStream_t::canSeeMemory is not
necessary if memory is shared with all devices. Thus, it is skipped.
Note that earlier host memory is always mapped to device 0 and HIP
always performs peer checking for all kinds of hipMemcpy. Since the
peer checking process requires context locking, hipMemcpy from/to host
memory always grabs device 0's context lock. Therefore, if there is
another thread holding the context lock of device 0 (e.g.,
hipDeviceSynchronize on device 0), hipMemcpy will have to wait for the
lock until it can actually perform memcpy. This can significantly
deteriorate execution performance.
Signed-off-by: Sarunya Pumma <sarunya.pumma@amd.com >
2018-07-28 23:15:16 -07:00
4e3a75c019
Merge pull request #607 from emankov/hipBLAS
...
[HIPIFY][BLAS] cuBLAS support revise
2018-07-28 21:11:03 +03:00
92eb165875
[HIPIFY][BLAS] cuBLAS support revise
...
+ doc for CUBLAS API supported by HIP
+ 3 tests: 0 & 1 based indexing, sgemm matmul
2018-07-28 21:00:44 +03:00
1a09b9d4d4
Merge pull request #596 from mangupta/patch5-fixhiplaunchparm
...
[tests] Fixed hipLaunchParm test on nvcc path
2018-07-26 06:17:50 +05:30
90a192661c
Merge pull request #595 from mangupta/patch4-fixhipStdComplex
...
[tests] Fixed build & disabled run of hipStdComplex on nvcc path
2018-07-26 06:17:04 +05:30
1d7fb3cc85
Merge pull request #592 from mangupta/patch1-initialize-cudapath
...
[hipcc] Initialize CUDA_PATH before printing
2018-07-26 06:16:31 +05:30
f37d054c8a
Merge pull request #590 from ROCm-Developer-Tools/fix-ballot-anyall-return
...
Fix return EXIT_FAILURE if ballot or anyall fails
2018-07-26 06:16:23 +05:30
4c41d62435
[tests] Fixed hipLaunchParm test on nvcc path
...
- Uses c++11 features. Added it to nvcc options
- Arguments for some kernels exceeded 4096 bytes which is the limit
imposed by nvcc. Reduced BLOCK_DIM_SIZE to 512 to handle this
- Fixed compilation issues on nvcc path
Change-Id: I14f6b28afcb7c6b24a085fd707b2104e2ed64627
2018-07-25 15:41:04 +05:30
9d8dd3a4e6
[tests] Fixed build & disabled run of hipStdComplex on nvcc path
...
Change-Id: I885e96902faa065f6559183f33b6a7aa1f37e393
2018-07-25 15:39:45 +05:30
7615257ec3
[hipcc] Initialize CUDA_PATH before printing
...
Change-Id: Ifd448b628fa4a7055f4a439ed2c817c5f11f9bef
2018-07-25 15:29:57 +05:30
73ef531208
Merge pull request #589 from gargrahul/set_offset_gettexalignoffset
...
Set offset in hipGetTextureAlignmentOffset
2018-07-25 12:13:31 +05:30
2b73bccbef
Fix return EXIT_FAILURE if ballot or anyall fails
2018-07-24 20:39:48 +00:00
bc4cdf7e41
Null check before setting offset
2018-07-24 12:25:40 +05:30
867a4aa971
Set offset in hipGetTextureAlignmentOffset
2018-07-24 10:11:26 +05:30
8dc1e456a7
Merge pull request #582 from ROCm-Developer-Tools/set-vdi-path
...
Let hipcc use proper include and lib path for HIP/VDI
2018-07-24 07:32:54 +05:30
aa50cdc620
Merge pull request #580 from ROCm-Developer-Tools/document-fatbin
...
Document hip-clang fatbinary format and initialization function
2018-07-24 07:31:19 +05:30
762786e091
Fix typo
2018-07-24 07:30:05 +05:30
bd31e83d18
Fix warnings about unused command line option --hip-link
...
and set search path for libhip_hcc.so
2018-07-23 14:55:07 -04:00
edbfc6f5c5
Document hip-clang fatbinary format and initialization function
...
This is for documenting the change in hip-clang: https://reviews.llvm.org/D49083
2018-07-23 09:28:40 -04:00
c59aa75f25
Fix hipcc for -M for hip-clang
2018-07-23 09:26:42 -04:00
025afa1cb5
Let hipcc use proper include and lib path for HIP/VDI
...
Add support of environment variable HIP_VDI_HOME.
By setting environment variable HIP_VDI_HOME to the distribution directory of HIP/VDI, hipcc
will choose proper include and lib path for hip-clang.
2018-07-23 09:26:42 -04:00
d9547227cc
Merge pull request #583 from emankov/cuDNN
...
[HIPIFY][DNN] cuDNN support revise
2018-07-23 14:13:54 +03:00
b1ca77147b
Merge pull request #568 from Srinivasuluch/sprint2_9tests
...
Adding a few more struct scenarions and a ResultValidation()
2018-07-23 14:44:06 +05:30
896a614f51
Merge pull request #558 from Jorghi12/patch-8
...
Fix the FindHIP.cmake script.
2018-07-23 11:48:42 +05:30
8bc73424d1
Fix hipHostRegister mem leak and copy kind in hipMemcpyAsync ( #587 )
2018-07-23 11:04:14 +05:30
9e2d33c866
Fix typo in FindHIP.cmake
2018-07-23 10:39:45 +05:30
1a2c16decb
Merge pull request #559 from rpathani/master
...
Adding functor unit tests
2018-07-23 10:15:16 +05:30
06db076b10
Merge pull request #584 from ROCm-Developer-Tools/fix-vector-explicit-cast
...
Fix explicit cast required
2018-07-23 10:13:05 +05:30
b6397143a6
Merge pull request #581 from ROCm-Developer-Tools/fix-warning-math-decl
...
[HIPClang] Switch ordering on math functions
2018-07-23 10:12:01 +05:30
23b4e46ca4
Merge pull request #554 from gargrahul/cleanup_hipmemset3d_test
...
Cleaned up hipMemset3D test
2018-07-23 10:11:49 +05:30
4ddb81424f
Fix explicit cast required
...
This is required for PyTorch which runs into an issue with narrowing types. Requires an explicit cast.
2018-07-20 16:34:15 -04:00
7ef05de0d5
[HIPIFY][DNN] cuDNN support revise
...
+ add doc for CUDDN API supported by HIP
+ cuDNN softMax test
2018-07-20 21:00:24 +03:00
ffefae77d7
[HIPClang] Switch ordering on math functions
...
Header math_functions.h should be included after including __clang_cuda_math_forward_declares.h to avoid warning: attribute declaration must precede definition.
2018-07-20 13:25:04 -04:00
b37c56efed
Update hipLaunchParm.cpp
...
Added struct bit field test, test number#21
2018-07-20 12:38:06 +05:30
ac3d5af867
Merge pull request #578 from emankov/testing
...
[HIPIFY][test] Add static and dynamic shared memory tests
2018-07-19 20:48:25 +03:00
bf9832d957
Merge branch 'master' of https://github.com/rpathani/HIP
...
# Conflicts:
# tests/src/kernel/hipLaunchParmFunctor.cpp
2018-07-19 12:11:23 +05:30
8434f2ed6b
Merge branch 'master' of https://github.com/rpathani/HIP
...
# Conflicts:
# tests/src/kernel/hipLaunchParmFunctor.cpp
2018-07-19 12:02:42 +05:30
2b6a736318
Update FindHIP.cmake
...
Using Bool expression to handle empty case.
2018-07-18 23:02:11 -04:00
12202dae90
Merge pull request #543 from ROCm-Developer-Tools/add-std-complex
...
Support std::complex for hip-clang
2018-07-19 07:14:45 +05:30
70bd255d71
[HIPIFY][test] Add static and dynamic shared memory tests
2018-07-18 18:33:44 +03:00
39374f4208
Fix test hipStdComplex.cpp to emit expected output for passing
2018-07-18 07:21:24 -04:00
4c707232b9
Update hipLaunchParm.cpp
...
submitting changes as per Sam review/suggestions
2018-07-18 12:41:51 +05:30
80438035c2
Support std::complex for hip-clang
2018-07-18 00:08:04 -04:00
ebc1c248cf
Merge pull request #576 from ROCm-Developer-Tools/fix_missing_vector_bits
...
Constrain variadic constructor to ranks > 1 and add missing `__device__`s
2018-07-18 09:34:25 +05:30
4948fa0cce
Merge pull request #575 from ROCm-Developer-Tools/fix_functional_gridlaunch_nonsense
...
It is unclear what I was thinking when authoring the original code...
2018-07-18 09:34:16 +05:30
05ecc35825
And a few more oversights.
2018-07-17 19:33:13 +01:00
a7f3101a7f
Merge branch 'fix_missing_vector_bits' of https://github.com/ROCm-Developer-Tools/HIP into fix_missing_vector_bits
2018-07-17 17:29:52 +01:00
487bb6f2ef
Update hip_vector_types.h
2018-07-17 17:28:48 +01:00
fe9859b79f
Not yet.
2018-07-17 17:24:48 +01:00
73d0091328
Not yet.
2018-07-17 17:21:10 +01:00
Sarunya Pumma
Evgeny Mankov
Evgeny Mankov
Maneesh Gupta
Aaron Enye Shi
Rahul Garg
Yaxun Sam Liu
saleelk
Srinivasuluch
rohit pathania
Jorghi12
Alex Voicu