a8a8b2c888
Validity check of input arguments in Ipc Mem APIs
...
Change-Id: Ia48e949d19f354f10c7e44cc2457fd4154bf6d76
[ROCm/hip commit: ba51d7f676
2017-06-14 15:18:57 +05:30
ef7c9be9c6
Use amHostCoherentFlag. Requires new HCC version.
...
[ROCm/hip commit: 99e9c7cca5
2017-06-07 09:06:40 -05:00
5b9ce41633
Add hipHostMallocCoherent, hipHostMallocNonCoherent
...
Provide per-allocation control over coherent/non-coherent mem.
These overrid the default HIP_COHERENT_HOST_ALLOC setting.
[ROCm/hip commit: 75f691ec2f
2017-05-24 00:48:10 -05:00
07865c3a02
Remove HIP_MAX_QUEUES (replaced with HCC_MAX_QUEUES)
...
[ROCm/hip commit: d0ef9d8462
2017-05-23 23:48:01 -05:00
1e0046f737
Fix trace category for hipHostMalloc
...
[ROCm/hip commit: 7cfe07cff4
2017-05-23 23:15:45 -05:00
324dfb870f
Return precise address for hipHostGetDevicePointer.
...
[ROCm/hip commit: 46030bb2d2
2017-05-17 07:36:06 -05:00
a55ce5bee4
Add initial HIP_SYNC_NULL_STREAM=0 mode.
...
This eliminates host-synchronization for null stream. Instead, the
null-stream uses GPU-side events to wait for other streams.
Default is OFF pending additional testing.
Add enhanced null-stream test.
Also refine HIP_TRACE_API.
[ROCm/hip commit: 27877f8854
2017-05-16 19:04:25 -05:00
2cf05ad54b
Add HIP_TRACE_API=4. Only display memory allocation/free apis.
...
[ROCm/hip commit: 0edab14139
2017-05-16 19:04:25 -05:00
ff72a701b7
hipHostMalloc allocation are mapped to all devices by default.
...
Support hipHostMallocPortable flag.
Default flags are hipHostMallocPortable | hipHostMallocMapped.
Also:
-refactor tests to move addCount and addCountReverse into HipTest
namespace.
-test multi-GPU host memory.
[ROCm/hip commit: c3ccaa01e5
2017-05-10 17:34:36 -05:00
ef98415d0e
Fix some typos, add additional guidance for -BSymbolic
...
[ROCm/hip commit: 687809104b
2017-05-05 17:29:04 -05:00
340cd36e8d
Added support for hipMemcpy2DAsync in HIP/HCC
...
Change-Id: Ia4a8306f2dc1e33a81a7195ec29aef652fcccc4b
[ROCm/hip commit: ec27c695c4
2017-05-03 22:29:12 +05:30
ba2f2f56b8
Fix hipMalloc to return error code if allocation fails.
...
[ROCm/hip commit: 59df70662a
2017-04-24 22:30:54 -05:00
2b8fbf40a8
fixed build issues with hipPointerGetAttributes
...
Change-Id: I3f5fbc05bdaef720884ba949075928752a070377
[ROCm/hip commit: 59ab3659ee
2017-04-24 15:31:07 -05:00
c5840d81ad
changed arguments for hipPointerGetAttributes
...
Change-Id: Ia7a7c4722c1f7d0a23f0e5cc3dd6dea6c01c1fd8
[ROCm/hip commit: 72bcfa438b
2017-04-24 15:24:16 -05:00
06b32795c8
mgpu IPC support fix
...
Change-Id: I12e4b2fd189c3658efd3b07defa18ece3853b0eb
[ROCm/hip commit: fc61b793fe
2017-04-04 15:51:10 +05:30
ccf799c453
added new api hipHccModuleLaunchKernel
...
1. hipHccModuleLaunchKernel is same as hipModuleLaunchKernel with OpenCL workitem model
2. Added copy right
3. Fixed header naming
Change-Id: I6a7c35a3566e2f8d3f5056613e34193775d4b236
[ROCm/hip commit: b9091ba818
2017-03-31 12:11:34 -05:00
be8d979d08
Fix for MemcpyFromSymbol on HIP/NVCC path
...
Change-Id: Ice38307f72870ae468cbf0861e104f0fa46dfd56
[ROCm/hip commit: 1d18006ab4
2017-03-27 00:35:10 +05:30
1494cb9339
Fix for hipMemcpyFromSymbolAsync
...
Change-Id: I449c669c8f0ef041deaf0a1bc812a71b2f0cc5a6
[ROCm/hip commit: f649c225a7
2017-03-24 10:30:33 +05:30
aa59d4d580
Fix for hipMemcpyFromSymbol (sync)
...
Change-Id: I66afec5443ce904a63ced1fafece5144ca59393e
[ROCm/hip commit: 4395582810
2017-03-21 23:48:04 +05:30
0bfd692b59
Added hipMemsetD8
...
Change-Id: I6a230a036c9c46c72a77d5f93c16ce8a00c3f837
[ROCm/hip commit: 1aba3c4375
2017-03-14 22:11:34 +05:30
71fc982cbe
Refactor registered memory calls.
...
[ROCm/hip commit: 524e007db5
2017-03-11 09:18:27 -06:00
a735ffa947
Update hipHostRegister debug and pointerTracker debug and notes
...
[ROCm/hip commit: 0a554f4dc1
2017-03-11 09:18:27 -06:00
3297a65b8f
Fix copying of registered memory.
...
Set device properly so copying can recover context.
Enhance test to catch this case.
[ROCm/hip commit: 94c85fd4fc
2017-03-11 09:18:27 -06:00
4954d140d4
IPC supported using ROCR APIs
...
Change-Id: I0a353b1240098f4b20fa266a871f5f5826290af9
[ROCm/hip commit: 32d8a58f18
2017-03-10 23:45:28 +05:30
fe81d08987
Added new API, hipMemPtrGetInfo
...
1. This API returns memory allocation size of pointer
2. Added test to check its functionality
Change-Id: I87976d817b5a6ca5530336c09e7cb0420601cb2c
[ROCm/hip commit: 1546732604
2017-03-07 13:46:29 -06:00
c6969c157a
Added initial support for hipMemcpyFromSymbol. But not working!
...
Change-Id: I48d8c7de4ec9f85c6c942be995fb488a3931f5d7
[ROCm/hip commit: 2e245ae58c
2017-02-23 11:29:06 -06:00
d24435ea9b
added runtime api hipMemcpyFromSymbolAsync
...
Change-Id: Ibaf925faf0ba464dd0ed6c5ea74c224c2ce38889
[ROCm/hip commit: 639fd4dd5e
2017-02-22 19:16:35 -06:00
afb38e15de
fixed symbol memcpy issue
...
Change-Id: I89d7401be51d194bcbf771020ba66e3d3b6a18f8
[ROCm/hip commit: 2790e9a448
2017-02-01 17:54:59 -06:00
f9ecf383e3
Add debug tips to docs
...
[ROCm/hip commit: 96eac67929
2017-01-23 22:34:41 -06:00
7a402a7fb1
Add HIP_SYNC_HOST_ALLOC, HipReadEnv
...
[ROCm/hip commit: 927ac3d81c
2017-01-19 23:55:24 -06:00
5de29029cd
changed copyright year from 2016 to 2017 in src directory
...
Change-Id: Idb97db509b2b4b1656b2df7a14a62ade38c9d574
[ROCm/hip commit: 73fcce26f9
2017-01-11 18:05:41 -06:00
1908d9504b
Add HIP_MAX_QUEUES feature.
...
Includes some tricky manipulation of the locks for contexts and streams.
issue is that stealing a stream requires we lock the context to
walk the streams to find a victim. To avoid deadlock, we can't
have a stream locked when we lock the context. This implementation
releases the stream lock, then acquires the context and selects the
victim.
A more stable implemenation might be to copy the stream list
from a context so that a lock is not required to walk all streams.
Smart shared_ptr could be used to prevent the streams from being
deallocated during the walk.
[ROCm/hip commit: a3e0012567
2017-01-09 21:02:56 -06:00
bbb1485a83
First pass at virtualized queue support.
...
Also updated stream debug messages to consistently use trace_helper.
[ROCm/hip commit: 93fbc9cf7b
2017-01-09 21:02:53 -06:00
2e9defbef8
Added state for hipDevice.
...
Change-Id: Idbc3c04cd054a01b634856a1e0a23ff172e991aa
[ROCm/hip commit: 5fb09879c7
2017-01-09 23:54:01 +05:30
ec4f4a643d
Support size_t in memset kernel.
...
Add disable for HSA_AMD_AGENT_INFO_MAX_WAVES_PER_CU
Remove one copy of completion_future in memset.
[ROCm/hip commit: c325c988b1
2016-12-22 12:25:09 -06:00
1a7ecbcd04
Add USE_IPC to disable use of IPC APIs. Set to 0.
...
[ROCm/hip commit: b30e4b4781
2016-12-13 15:07:04 -06:00
f6d6861ad4
HIP IPC implementation on ROCr IPC APIs
...
Change-Id: I1ca9d520f5d0b1b56694211471b81eb7c6c23d16
[ROCm/hip commit: a53d35fd6c
2016-12-07 15:38:36 -06:00
6d2bf52f9f
IPC prototyps and part of the implementation included
...
Change-Id: Id88c7f155d23ec63f57a6ef05098fba43f8af336
[ROCm/hip commit: 17b98d59b8
2016-12-06 14:24:09 -06:00
2b557e949e
local changes for hipnccl
...
Change-Id: I05a1f0381ce2914a800f573342cc954eb5ff82d9
[ROCm/hip commit: eba2595611
2016-12-06 14:22:02 -06:00
ce01d26a87
Add additional controls for forcing serialization and blocking.
...
Move HIP_COHERENT_HOST_ALLOC so it is read once at init time.
Add HIP_LAUNCH_BLOCKING_KERNELS, HIP_API_BLOCKING.
Update docs on debug and chicken bits.
Conflicts:
src/hip_hcc.cpp
[ROCm/hip commit: ff2f54c1bf
2016-12-02 18:03:59 -06:00
a523bc5632
Added support for hipMemGetAddressRange
...
Change-Id: I99a796a4eb765152cf15a12d6a86b58684d34f50
[ROCm/hip commit: fe6ba656c9
2016-11-29 22:04:09 +05:30
2234b57513
Refactor for building HIP as dynamic library
...
Change-Id: I65a3d9d589c4fdbbdcf1611e5427224253be8260
[ROCm/hip commit: c0419cc749
2016-11-18 14:33:20 +05:30
7192577e38
added texture header to memory api source
...
Change-Id: I1af6d60aca5a9a9ef1cadf8c304bea892acbe061
[ROCm/hip commit: c20c524400
2016-11-17 11:57:53 -06:00
a446102fcc
Add direct test case for threadfence_system workaround
...
Change-Id: I5b21b590e957c901044741ac94e816cd8b1426f9
[ROCm/hip commit: 50867efa10
2016-11-11 15:09:43 -06:00
e0ef3bb319
Refactor copy and P2P logic.
...
Prefer use of source-engine for DMA copies, even if user submits copy
in a stream attached to a different device.
The stream is now used only for synchronization, and HIP
makes the most optimal decision for which engine to perform the
copy - typically the source copy engine.
HIP now makes decision on which engine should perform the copy
and passes this to HCC using new apis.
HIP has additional information about peer
visibility and will make a decision which agent should perform
the copy .
Change-Id: I0cf4cfebeae256e6ca795f08a7ed7130f4857d1f
[ROCm/hip commit: ced9d72d94
2016-11-10 23:12:48 -06:00
67a305ec9e
Improve memory debug
...
Change-Id: I0f033139aa4e4b47039eb016e404009127bd0a44
[ROCm/hip commit: 2dea3a0b1a
2016-11-10 23:12:48 -06:00
bbd526b7e5
resolve conflicts for git pull
...
Change-Id: Ie353b831e2241bc28042069b6cc7405257e871e1
[ROCm/hip commit: 4d7ac1e091
2016-11-09 21:38:43 -06:00
74b0840cc6
Add option to alloc fingrained system memory
...
Change-Id: Ia13c8e058cb988b5857e75a590a4d67411362ae1
[ROCm/hip commit: f7e9f12bf1
2016-11-09 21:36:30 -06:00
2a9aa89491
Set forceHostCopyEngine for other copy dirs. Support HIP_FORCE_P2P_HOST
...
Also: more debug for copy and P2p.
Change-Id: I87030c525410e041b2a00baaf6c68e6c0977ff42
[ROCm/hip commit: 092b3dacda
2016-11-04 19:53:23 -05:00
04691cdff8
Don't call allow-access if allocating device's only peer is self.
...
Change-Id: Iac58e6c3e460675833f10b1e8b2e393de223654d
[ROCm/hip commit: 354091f357
2016-10-27 23:09:31 -05:00
Rahul Garg
Ben Sander
Aditya Atluri
pensun
Maneesh Gupta