4dc24498af
hipDeviceReset(): make sure to reinitialize the printf buffer in hcc RT
...
[ROCm/hip commit: ccef1cbd6e
2017-10-18 16:26:13 -04:00
8c67dd3b2e
Fixed hipTexture fetch issue
...
[ROCm/hip commit: 9929e8e717
2017-09-25 06:54:29 +05:30
9c0dfd3538
Bump device major version from 2 to 3
...
This would significantly improve performance for certain apps in kernel
selection logic.
[ROCm/hip commit: c74d3fe2cb
2017-09-15 15:47:39 +00:00
ab4128a53f
Merge pull request #179 from gargrahul/fix_hipmallocarray
...
Fixed hipMallocArray for 1D cases
[ROCm/hip commit: 3ed5f42dad
2017-09-14 12:59:33 -05:00
b446d8ebdf
Merge branch 'master' into hip_init_alloc
...
[ROCm/hip commit: fbd22c3e49
2017-09-14 11:53:33 -05:00
5418e3de45
Add HIP_INIT_ALLOC to init allocated memory.
...
[ROCm/hip commit: cea80cd8b3
2017-09-13 23:31:48 +00:00
3178dc04cd
Fixed hipMallocArray for 1D cases
...
[ROCm/hip commit: 2a915b2790
2017-09-12 21:52:11 +05:30
cbbb0826f2
hipStreamQuery uses av::is_empty. Add HIP_FORCE_NULL_STREAM.
...
[ROCm/hip commit: 4ac6d643c1
2017-08-31 03:00:14 +00:00
2a23d6dbe7
Refactor hipStreamWaitEvent
...
- Null streams use same flow as non-null.
- Add HIP_SYNC_STREAM_WAIT
- Resolve null stream.
[ROCm/hip commit: 882dab4536
2017-08-31 03:00:14 +00:00
f66a03fa9c
Merge pull request #167 from bensander/event_safety
...
Event safety
[ROCm/hip commit: cc08b36647
2017-08-29 13:55:46 -05:00
bd434502db
Lock streams when waiting on event completion or querying event safety.
...
[ROCm/hip commit: 6ff74d0e97
2017-08-28 18:40:16 -05:00
870228006e
Null check on input pointer arguments
...
[ROCm/hip commit: 38c9febe21
2017-08-25 08:46:34 +05:30
e673304653
Device count check on input device id
...
[ROCm/hip commit: 38ecfb9384
2017-08-22 14:05:44 +05:30
542bd863a2
Add some new HIP_TRACE_API options.
...
[ROCm/hip commit: e3161bb40e
2017-08-16 04:00:33 +00:00
c26e1d8c3d
Merge pull request #148 from weixingzhang/vmaddr1
...
[HIP Texture] The GPU virtual address for texture memory needs to be
[ROCm/hip commit: 8b7dde6519
2017-08-15 14:19:45 -05:00
c036a1b25c
[texture] guard new HCC APIs under workweek
...
Change-Id: I4f60a64fb0b0496ca1eb01ffe6ddda121c25d976
[ROCm/hip commit: e40047f2bf
2017-08-15 15:51:38 +05:30
0a25b3e972
Merge pull request #138 from scchan/threadfence_system
...
implement __threadfence_system
[ROCm/hip commit: f14e31d9d7
2017-08-10 12:51:05 +05:30
d4d9836a96
Merge pull request #145 from gargrahul/context_mgmt_changes
...
Updated context management logic:
[ROCm/hip commit: 660061c3b4
2017-08-09 17:47:45 -05:00
5bf177340a
[HIP Texture] The GPU virtual address for texture memory needs to be
...
aligned.
In hcc_am, a bigger buffer will be allocated for alignment purpose
and _unalignedDevicePointer is added in struct AmPointerInfo for
original allocated address.
[ROCm/hip commit: 4264a4ce56
2017-08-08 11:18:00 -04:00
c050396510
Bump min hcc_workweek required for named kernel dispatch to 17312
...
Change-Id: I8c7b58306b279ed113d03260e4bc6086bb8b4e68
[ROCm/hip commit: 8d6acec135
2017-08-08 11:08:55 +05:30
8d254a7113
Updated context management logic:
...
1) hipSetDevice sets a flag so that next call to hipCtxGetCurrent returns primary context on current device
2) hipCtxGetCurrent returns primary context on current device if TLS context stack is empty
3) hipCtxPopCurrent falls back to primary context on current device as default
4) hipCtxPushCurrent, hipCtxSetCurrent and hipCtxCreate reset the flag set in hipSetDevice
[ROCm/hip commit: 3e84cf4aba
2017-08-08 07:02:22 +05:30
21c1d69e77
implement __threadfence_system
...
[ROCm/hip commit: 96e79f97e2
2017-08-02 08:50:18 +00:00
e7c823b4ed
Merge pull request #134 from gargrahul/fix_hipMemcpy2DAsync
...
fix hipMemcpy2DAsync
[ROCm/hip commit: fce1d0d7f0
2017-07-31 10:27:18 +05:30
1be0719063
Merge pull request #135 from bensander/fix_tracing
...
Some fixes to tracing.
[ROCm/hip commit: 48573a037e
2017-07-31 10:24:41 +05:30
4a234edea9
Some fixes to tracing.
...
[ROCm/hip commit: ecbb494a58
2017-07-28 22:13:43 -05:00
7e9c13f8de
fix hipMemcpy2DAsync
...
[ROCm/hip commit: 0053e4db03
2017-07-29 06:50:56 +05:30
b763034c65
Merge pull request #127 from bensander/pass_module_kname
...
Pass kernel name to HCC dispatch_hsa_kernel, for debug/profile
[ROCm/hip commit: 375ec84bc6
2017-07-28 10:19:15 +05:30
1ff23db17b
Add workweek check to make sure we have a new enough compiler
...
[ROCm/hip commit: 4980a6d3ab
2017-07-27 23:00:58 -05:00
90fa6b18eb
Merge pull request #122 from bensander/enable_async_null_stream
...
Set HIP_SYNC_NULL_STREAM=0.
[ROCm/hip commit: fcc294ed41
2017-07-28 09:15:56 +05:30
3ab70ad042
Pass kernel name to HCC dispatch_hsa_kernel, for debug/profile
...
[ROCm/hip commit: 3a4dfc0f85
2017-07-27 22:00:15 -05:00
cba82d9e24
Set HIP_SYNC_NULL_STREAM=0.
...
Optimizes null stream synchronization so it uses GPU-side dependency
resolution. Requires HCC __hcc_workweek__ > 17300.
[ROCm/hip commit: 77fb9893b4
2017-07-27 11:11:54 -05:00
da4042554f
Merge pull request #116 from bensander/enable_coh_host
...
Make host memory allocations coherent by default.
[ROCm/hip commit: 526fcb1223
2017-07-27 10:45:32 +05:30
9ec4b0b66f
Merge pull request #100 from weixingzhang/texture
...
HIP Texture Support
[ROCm/hip commit: 4f93b99f23
2017-07-26 19:28:15 -05:00
6834270f15
Make host memory allocations coherent by default.
...
Associated change is to optimize event recording so it uses
agent-scope releaes (since it was only using system-scope release
to support non-coherent host mem).
Flags and environment variables exist to obtain previous behavior
if desired. Options are documented in new performance guide.
[ROCm/hip commit: 8b8e97ff28
2017-07-26 19:20:34 -05:00
149ec02e84
Enable HCC_OPT_FLUSH=1 (if HCC compiler new enough)
...
[ROCm/hip commit: 67f5d2de3d
2017-07-24 18:57:19 -05:00
f4ca1c96c4
Don't crash for hipEventDestroy(NULL)
...
Change-Id: I3f4de1ad2712dbe8af81e10c5d3bef16d468c582
[ROCm/hip commit: 7ddb6b6b76
2017-07-18 11:01:02 +05:30
65bec11393
HIP Texture Support
...
[ROCm/hip commit: e671cec367
2017-07-17 15:16:12 -04:00
ee697e8c4c
remove unsupported erf(c)inv related host functionality
...
Change-Id: I665c33616359a0124b5552076359d2f8faa54930
[ROCm/hip commit: 7066ff4478
2017-07-14 11:31:40 +05:30
a80bc4c3b6
Temporarily disables HCC_OPT_FLUSH
...
Change-Id: I290791e58dd52ab3823f6c3315e42b0d386e9d64
[ROCm/hip commit: 17555eabc0
2017-07-12 16:08:36 +00:00
a4d39f50a1
Fix hip_db header inclusion path
...
Change-Id: I8a95798301de73887dde3316dc593a1847002712
[ROCm/hip commit: 35ebfda9d4
2017-07-12 16:08:20 +00:00
f28a1365f8
Remove explicit address spaces declarations
...
[ROCm/hip commit: 30ee797dd4
2017-07-07 12:46:29 -05:00
042ed2c38f
Add hipdbPrintMem - wrapper for hcc memory tracker.
...
[ROCm/hip commit: 0459f3db72
2017-07-03 15:05:30 -05:00
9498754cf6
Fixed build error in hipDeviceGetByPCIBusId
...
Change-Id: I33be362910c4c5dca7435fb9c41f427b54f0e9a7
[ROCm/hip commit: 34fcaa23bb
2017-07-01 07:56:30 +05:30
ba0215ad9f
Set default HIP_SYNC_NULL_STREAM=1.
...
[ROCm/hip commit: 626b5191e8
2017-06-30 19:01:14 -05:00
83f3ed16de
Fixed hipDeviceGetByPCIBusId
...
Change-Id: Ia36bb9425671ef7659541c9aeedae4098456a31b
[ROCm/hip commit: 65cb024b61
2017-06-30 10:11:41 +05:30
f7bfec2252
automate gcnarch detection
...
Change-Id: Ibbad22db136f7f5e2be84c82e9169298a144cc77
[ROCm/hip commit: 3e610ab0fe
2017-06-29 12:01:40 -05:00
ff48608355
Fixed hipDeviceGetPCIBusId for HIP/HCC
...
Change-Id: I3688fa2476e1baada2d3c5fc3735cec3f15a1e21
[ROCm/hip commit: f4ea3b76c8
2017-06-28 23:48:27 +05:30
3ba979d2b7
Set default HIP_HIDDEN_FREE_MEM
...
[ROCm/hip commit: 148dbc1027
2017-06-27 12:17:12 -05:00
84fbc3dfaa
Add support of HIP_HIDDEN_FREE_MEM, to deduct the returned available
...
memory from hipMemGetInfo API, measured in MB.
Change-Id: I7a8260c12e032e04e26611db4c38c893a29f2653
[ROCm/hip commit: 1df08626c8
2017-06-26 15:29:38 -05:00
0f2fd1372f
Add option to pass names to HCC dispatch API (for debug)
...
[ROCm/hip commit: 176ff824d1
2017-06-23 18:05:30 -05:00
Siu Chi Chan
Rahul Garg
Wen-Heng (Jack) Chung
Ben Sander
Maneesh Gupta
Weixing Zhang
Aditya Atluri
Sun, Peng