8204d20644
Describe HIP env vars
...
[ROCm/hip commit: 2dc7e9ccfa
2016-03-19 03:09:57 -05:00
662cd0bb30
Fix copy and sync bugs. Remove extra sync in default stream.
...
- NULL stream was waiting for itself to be empty before each command.
- Force "blocking" streams to wait for NULL to empty. This was missing
before.
- async copy was disabling itself via trueAsync=false for common cases.
Refactor:
- rename _null_stream to _default_stream.
- move some null sync function to defaultSync, move to dev member func.
[ROCm/hip commit: 6472c8b915
2016-03-19 02:44:26 -05:00
b6de799f47
Add beastperiteration and onesize for testing.
...
onesize allows running tests at one specific size.
[ROCm/hip commit: 7ff5b16d2a
2016-03-19 02:43:04 -05:00
79d0d442d3
Improve formatting - line up cols
...
[ROCm/hip commit: 85fce5f21e
2016-03-18 23:43:04 -05:00
5996a5c0a1
Print Pinned or Unpinned in result summary
...
[ROCm/hip commit: c2102847a4
2016-03-18 21:28:29 -05:00
8a47efb22a
Supported --aliged mode. Add results check for H2D and D2H.
...
[ROCm/hip commit: 618556eaf9
2016-03-18 03:09:52 -05:00
0134651419
Refactor copy code.
...
-Move staging buffer locks inside the staging buffer code.
-Remove dedicated per-device completion_signal + per-device lock -
instead allocated signal from the per-stream pool. This elimintes
the lock and allows more concurrency.
-remove switch HIP_DISABLE_BIDIR_MEMCPY
[ROCm/hip commit: 0af4d3623f
2016-03-18 03:02:00 -05:00
3320975a80
Refactor staging buffer and sync copies.
...
- refactor staging buffer to operate on hsa* data structures not
hc::accelerator.
- use hsa_memory_allocate to allocate staging buffers rather than
am_alloc.
- Refactor device reset with single member function. Don't reallocate
staging buffers on reset.
- Properly track dependencies based on command type. Add new deps for
H2D and D2D rather than overloading H2D.
[ROCm/hip commit: 7d500599fa
2016-03-17 20:09:10 -05:00
fc27c61c58
Refactor to isolate staging buffer code.
...
[ROCm/hip commit: e7586adb33
2016-03-17 00:20:56 -05:00
0ae7bc7e14
Start separaration of staging_buffer.cpp code.
...
Still #include staging_buffer.cpp into hip_hcc.cpp.
Directed tests compile hip_hcc to static library and use the library.
[ROCm/hip commit: 28ee7aff71
2016-03-16 22:26:49 -05:00
910017a20f
Add aligned alloc
...
[ROCm/hip commit: c02bccc153
2016-03-16 21:55:57 -05:00
0b617788a8
Checkpoint code cleanup.
...
- Refactor ihipStream in prep for thread-safe implementation.
- Do some work on PinInPlace implementation.
[ROCm/hip commit: ec50cbb3aa
2016-03-16 21:16:29 -05:00
bef8dc7bcc
print device config info
...
[ROCm/hip commit: 70c5f5e3f5
2016-03-14 23:02:49 -05:00
0abf5db89e
Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging
...
Conflicts:
src/hip_hcc.cpp
tests/src/CMakeLists.txt
[ROCm/hip commit: e1617b9604
2016-03-14 15:01:26 -05:00
5951d581d9
enable DB, comments
...
[ROCm/hip commit: 1a27e5134e
2016-03-14 14:40:41 -05:00
10ca8ed91d
Add Bidir copy test and help.
...
[ROCm/hip commit: 5606bee076
2016-03-14 14:39:23 -05:00
3427873d7c
refactor, add support for speccing xfers in bytes
...
[ROCm/hip commit: ac6ed35ba0
2016-03-13 09:41:06 -05:00
e08baf41d7
Remove ROCR_V2, assume=1
...
[ROCm/hip commit: b0dca6bd0c
2016-03-12 04:08:13 -06:00
30fbe45694
Remove USE_AM_TRACKER (assume 1)
...
[ROCm/hip commit: a5a2c7f33c
2016-03-12 04:04:57 -06:00
4900ebb39f
Improve error reporting.
...
use throw with error class.
fix bug when memcpyDefault resolved to D2D copy.
[ROCm/hip commit: 250739666d
2016-03-12 04:02:04 -06:00
db1ce3ba84
Added hipHostRegister for hip with tests and added copyright
...
[ROCm/hip commit: 102f173396
2016-03-08 12:57:22 -06:00
963c0f6076
Added a random test for memcpyAsync
...
[ROCm/hip commit: cbb42c6b6a
2016-03-08 03:40:56 -06:00
ab68dc4240
v2 added support for unpinned async memcpy
...
[ROCm/hip commit: 18c0e69f51
2016-03-08 03:19:50 -06:00
bf3d3a9745
Revert "Fixed passing unpinned memory to async memcpy"
...
This reverts commit c272c5380845965aaa2d
2016-03-08 03:17:27 -06:00
7f8a1a3e17
Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging
...
[ROCm/hip commit: 81824926be
2016-03-08 03:15:33 -06:00
c272c53808
Fixed passing unpinned memory to async memcpy
...
[ROCm/hip commit: 216af71480
2016-03-08 03:14:49 -06:00
6d77b45f88
Refactor memccpy tests.
...
- Add "hipMemcpy_simple" - some simple smoke tests.
- Run memcpy 3 times w/ different switches for kinds, sizes,
multi-thread.
[ROCm/hip commit: cc7517b57e
2016-03-07 17:15:48 -06:00
c05f4abd71
Added hipHostRegister flags
...
[ROCm/hip commit: d9429dd4ec
2016-03-07 10:52:40 -06:00
02760925a9
Added hipHostRegister feature for CUDA backend and its tests
...
[ROCm/hip commit: 4ed0b1cb1a
2016-03-07 03:42:50 -06:00
ee6a02b4fa
added copyrights for newly added tests
...
[ROCm/hip commit: de85c80eb0
2016-03-07 02:01:25 -06:00
02925c2393
Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging
...
[ROCm/hip commit: 3799fd55e2
2016-03-07 01:41:10 -06:00
e33e27232e
Increased size of arrays for single stream tests
...
[ROCm/hip commit: faaee7e9bf
2016-03-07 01:40:31 -06:00
82116d905d
Enhance HIP trace debug functions.
...
- Control with HIP_DB=mask (env var). See src/hip_hcc.cpp for mask
values:
#define DB_API 0 /* 0x01 - shortcut to enable HIP_TRACE_API on single switch */
#define DB_SYNC 1 /* 0x02 - trace synchronization pieces */
#define DB_MEM 2 /* 0x04 - trace memory allocation / deallocation */
#define DB_COPY1 3 /* 0x08 - trace memory copy commands. . */
#define DB_SIGNAL 4 /* 0x10 - trace signal pool commands */
- Combine with HIP_TRACE to see debug with API trace.
- Use colors to distinguish different flows of debug.
- Add define COMPILE_DB_TRACE to allow removing all debug at compile-time
[ROCm/hip commit: aa03e1264c
2016-03-06 23:50:52 -06:00
2de8a1a4a6
Fix typo in nvcc_detail/hip_runtime_api.h
...
[ROCm/hip commit: 39d5a2c079
2016-03-07 09:40:15 +05:30
91dbc3114d
added feature for hipHostGetFlags for CUDA and HIP
...
[ROCm/hip commit: 75952029d6
2016-03-06 12:17:30 -06:00
5346c57411
added flags support to be used later
...
[ROCm/hip commit: 1f3695b224
2016-03-06 08:40:33 -06:00
f1b8758919
corrected hipDeviceGetProperties to hipGetDeviceProperties - not docs
...
[ROCm/hip commit: d3ba2b9782
2016-03-06 08:31:04 -06:00
3c91a6d0a7
Added hipHostAlloc with hipHostAllocMapped flag
...
[ROCm/hip commit: 3aa764d5eb
2016-03-05 15:57:56 -06:00
8e0fc269d7
Added hipHostAlloc feature for CUDA
...
[ROCm/hip commit: f479531be5
2016-03-05 13:58:56 -06:00
5e9d9cbabf
v2 Added canHostMapMemory
...
[ROCm/hip commit: a5408ed7b6
2016-03-05 13:15:07 -06:00
6bfbe0483a
Revert "Added canMapHostMemory feature"
...
This reverts commit 8b585536ef2ebbdd6ec5
2016-03-05 13:08:57 -06:00
8b585536ef
Added canMapHostMemory feature
...
[ROCm/hip commit: af4edd277f
2016-03-05 13:06:37 -06:00
29c423a22b
Added canMapHostMemory to hipDeviceProp
...
[ROCm/hip commit: 4b271ec013
2016-03-05 19:30:29 -06:00
e0497a9ae1
Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging
...
[ROCm/hip commit: 6d4c830613
2016-03-04 08:22:56 -06:00
7aaa474ef0
Add FORCE_SAMEDIR_DEP for copies.
...
[ROCm/hip commit: ed71cded1c
2016-03-04 08:21:15 -06:00
52049a5927
v2: Fixed implementation of hipMemcpyDefault flag
...
[ROCm/hip commit: bd23279271
2016-03-04 06:00:04 -06:00
e60db0c25f
Add define to disable STREAM thread-safety
...
[ROCm/hip commit: 5e51ff177a
2016-03-03 10:49:25 -06:00
cc9a9cc5db
Added support for hipMemcpyDefault
...
[ROCm/hip commit: c154e1f4e4
2016-03-03 10:30:06 -06:00
12e7f550b7
Update hip_hcc.cpp
...
[ROCm/hip commit: 40eefc1cde
2016-03-03 13:59:43 -06:00
057bcc73f1
Fix output of hipPointerGetAttributes
...
The output of hipPointerGetAttributes is fixed to match CUDA counterpart.
[ROCm/hip commit: b6e34a44b0
2016-03-03 13:58:18 -06:00
Ben Sander
Aditya Atluri
Maneesh Gupta
Aditya Atluri