47725baf8c
static inline in a header, just like excess sugar in a diet, causes bloat (#1692 )
...
[ROCm/clr commit: cf526f5484
2019-12-23 19:09:38 +05:30
48faf7423a
Revert - Changes related to hipMemcpyWithStream ( #1718 )
...
Reverting #1673 , #1697 and #1707 .
Support for hipMemcpyWithStream and memcpy optimizations, will be brought in again once issues seen with these are resolved independently.
[ROCm/clr commit: a369bd4418
2019-12-06 09:51:53 +05:30
6c7a338471
Fix hipMemcpyWithStream ( #1707 )
...
* Fix hipMemcpyWithStream
* Add a dtest for the same.
[ROCm/clr commit: 261a04580f
2019-12-04 11:56:23 +05:30
2471c1950c
[hip] refactoring hipExtLaunchMultiKernelMultiDevice API ( #1698 )
...
[Background] it was found that if lazy linking used for a library that calls hipExtLaunchMultiKernelMultiDevice API then this API can get the wrong program_state object for looking up device kernels leading to a "No device code available" error in this API.
To fix this issue, the API was refactored to be inline and get and pass the correct program_state to an internal hip API to request a multi-device kernel launch.
[ROCm/clr commit: 8eaea4d114
2019-12-04 11:50:51 +05:30
22c07e2331
fix hipMempcy precondition, test for zero size first ( #1697 )
...
[ROCm/clr commit: 6df73e1f12
2019-11-25 19:03:03 -08:00
e2a0a547ff
fixed directed tests fail when hcc bumped to 3.0 ( #1678 )
...
Handled the HCC version check appropriately as few of the directed tests (SWDEV-212161) were failing when hcc was bumped to 3.0.
[ROCm/clr commit: d4dde7a27d
2019-11-20 21:37:52 +05:30
553a7c2ce3
General sync memcpy improvements. Add hipMemcpyWithStream ( #1673 )
...
* General sync memcpy improvements. Add `hipMemcpyWithStream`
* Update hip_memory.cpp
[ROCm/clr commit: 022ac3cb0a
2019-11-20 21:36:37 +05:30
81dd3751a6
Update error codes for hipGetDevice for doxygen and move up null check ( #1668 )
...
* [docs] Update error codes for hipGetDevice
* Move up out ptr check
[ROCm/clr commit: 13c2a31d7e
2019-11-20 21:35:27 +05:30
2c00d66995
Revert [HIP] Fixed hipStreamAddCallback ( #1674 )
...
This reverts commit 4d0b7e65bb8a3c36a424
2019-11-20 11:55:46 +05:30
58d1ae32ee
hipStreamSynchronize can skip marker if stream is empty ( #1667 )
...
[ROCm/clr commit: dfaaf04611
2019-11-19 09:42:43 -08:00
7d5bc56ab7
Added null check in hipEventSynchronize ( #1627 )
...
* Added missing null check in hipEventSynchronize
* Minor correction in the Event API description
[ROCm/clr commit: 7985ad218f
2019-11-18 12:18:55 +05:30
bc5596cc74
General hipMemset improvements ( #1495 )
...
* hipMemset et al can use HSA API directly for synchronous cases
* lock and flush stream in hipMemset, hold lock until complete
* move hipMemset async check to front of conditional
* use hsa_amd_memory_fill for additional sync memset cases
code cleanup/review for all memset calls
* Fix inversion of execution mutating value.
* ihipMemsetSync fall back to kernel if HSA memset fails
* Never fallback, never surrender.
* Allow NULL stream.
* Optimise memset kernel. Remove deadwood.
* Update hip_memory.cpp
* Clean up stream logic in sync memset
* Revert "Clean up stream logic in sync memset"
This reverts commit 6117dedf673367f44cc704192573a117a3d92477.
[ROCm/clr commit: 854147e911
2019-11-07 13:19:54 +05:30
4d0b7e65bb
[HIP] Fixed hipStreamAddCallback [SWDEV#165185] ( #1425 )
...
Fixed hipStreamAddCallback() as requested in SWDEV#165185
Added unit test to test the behavior
[ROCm/clr commit: 97fca3439d
2019-11-07 13:18:12 +05:30
f5a92e5cb6
Fixed texture 2D mapping for pitched arrays & 3D Texture read ( #1415 )
...
Texture 2D image mapping for pitched arrays:
github issue: Texture Object's Buffer seems to be Misaligned #886
JIRA ticket: SWDEV-199313
SWDEV-151670 : Fixed issue with 3D texture with 4 components
SWDEV-151671 : Issue with 2D layered texture with 4 components
[ROCm/clr commit: dc8f556460
2019-11-07 13:17:46 +05:30
cf648f7dbe
Rename hip/hip_hcc.h to hip/hip_ext.h ( #1341 )
...
* Rename hip/hip_hcc.h to hip/hip_ext.h
* Deprecate hip_hcc.h
[ROCm/clr commit: dfee3ae279
2019-11-07 13:17:10 +05:30
39a98df790
hipEventRecord only needs one lock; remove locked_eventIsReady
...
[ROCm/clr commit: b2af1820f5
2019-11-06 15:56:32 +00:00
539d0c2ae3
remove ihipEvent_t::refreshEventStatus(), new hipEventElapsedTime
...
This fixes possible races in hipEventElapsedTime.
[ROCm/clr commit: 444122dfb9
2019-11-06 15:46:46 +00:00
f7e255c3c4
fix race condition in hipEventRecord
...
[ROCm/clr commit: 6c3479a155
2019-11-04 17:45:11 +00:00
e61ff0408c
Merge pull request #1598 from lmoriche/master
...
Fix a code object memory corruption
[ROCm/clr commit: 2199f8a6c6
2019-10-31 17:12:24 -07:00
138b0327da
Add stream
...
[ROCm/clr commit: f556e15361
2019-10-31 12:15:56 -04:00
440a172f0e
Fix HIP init calls in hipMemcpy2DFromArray
...
[ROCm/clr commit: 0718ba0f00
2019-10-31 12:15:56 -04:00
04785f2d54
Merge pull request #1515 from ansurya/tex_unbind_issue_fix
...
Fix undefined ref to hipUnbindTexture for texture types
[ROCm/clr commit: aeb7cebbad
2019-10-30 17:54:15 -07:00
f1f649e57c
Addressed review comments
...
Change comment "must exceed" to "must be no shorter than"
move the std::string instead of creating a copy
[ROCm/clr commit: 3243f06eef
2019-10-30 13:14:41 -07:00
72c686ed67
Revert "Fix occupany APIs ( #1560 )"
...
This reverts commit ad1e409a2427221bc823
2019-10-29 11:41:08 -07:00
0113961674
Fix a code object memory corruption
...
The lifetime of the buffer given to
hsa_code_object_reader_create_from_memory must exceed that of the
code object reader. We need to create a copy of the code object
binary memory (file) that is kept allocated until the code object
reader is destroyed.
[ROCm/clr commit: 66c91b42e6
2019-10-29 08:23:57 -07:00
dfa019bdf6
Merge branch 'master' into tex_unbind_issue_fix
...
[ROCm/clr commit: 9332a39838
2019-10-25 15:54:25 +05:30
0140ea8e1a
merge from master
...
[ROCm/clr commit: ae838f8cee
2019-10-25 15:52:09 +05:30
7ea7a9c3b7
Add hipMemcpy2DfromArray ( #1510 )
...
Adds hipMemcpy2DFromArray and hipMemcpy2DFromArrayAsync equivalent to cudaMemcpy2DFromArray and cudaMemcpy2DFromArrayAsync.
[ROCm/clr commit: 14b870d1ce
2019-10-25 15:43:33 +05:30
2d538b702d
Merge branch 'master' into tex_unbind_issue_fix
...
[ROCm/clr commit: c0fc5e718c
2019-10-25 15:36:55 +05:30
ad1e409a24
Fix occupany APIs ( #1560 )
...
Addresses SWDEV-205006
[ROCm/clr commit: 6c5fbf9b4a
2019-10-24 17:44:47 +05:30
510be4b5dc
Improve performance of v2 arg handling ( #1539 )
...
* Improve performance of v2 arg handling
* Missing change to `std::string`
[ROCm/clr commit: 15a699688e
2019-10-24 17:44:05 +05:30
9e0eaef846
[hip] add support for implicit kernel argument for multi-grid sync ( #1456 )
...
* [hip] add support for implicit kernel argument for multi-grid sync
* modified code for calculating the prev_sum
* change the impCoopArg type to size_t
* add memory clean up
* launch init_gws and main kernels into two separate loops
[ROCm/clr commit: 93c688a0c9
2019-10-24 17:43:30 +05:30
685a4cd182
Merge pull request #1544 from vsytch/master
...
QoL changes to the hipMemset family
[ROCm/clr commit: a21fe1443b
2019-10-16 18:54:20 -07:00
a7d6c03e17
Guard against division by zero for no VGPR usage (e.g., in an empty kernel) ( #1528 )
...
* guard against division by zero for no VGPR usage (e.g., in an empty kernel)
* fix bracket format
* clean up parenthesis
[ROCm/clr commit: d16963c9d5
2019-10-16 10:49:56 +05:30
e2c2025e3e
Update the declarations of hipMemsetD8, hipMemsetD8Async, hipMemsetD16, hipMemsetD16Async. These functions are type aware and take in as their third argument the number of elements in the buffer, not the buffer size. Change the name of this argument from sizeBytes to count to align with the above description.
...
[ROCm/clr commit: 0200aa3a21
2019-10-15 14:18:42 -04:00
c5fc879606
Fix undefine ref to hipUbindTexture for texture types
...
[ROCm/clr commit: a5d9f56bef
2019-10-09 16:26:06 +05:30
266d2fd9c8
tracing layer unifying with hcc: removing API_ID_ANY
...
[ROCm/clr commit: ed29c10394
2019-10-08 19:09:43 -05:00
09bb83cf2c
Fix PCI Domain ID query ( #1424 )
...
* Fix PCI Domain ID query
* Update BDF comment
[ROCm/clr commit: e2260d82a6
2019-10-07 14:11:52 +05:30
85640091d5
[hip] Add the missing header <sstream>. ( #1498 )
...
[ROCm/clr commit: 5fbb09fb39
2019-10-04 13:39:29 +05:30
8d500c1b59
Add texref get APIs support ( #1471 )
...
Added support for -
hipTexRefGetArray
hipTexRefGetAddressMode
hipTexRefGetAddress
[ROCm/clr commit: d5a61736d8
2019-10-04 13:38:45 +05:30
efc64a4f8a
Added new Memory API's ( #1399 )
...
Added new memory API's hipMemAllocPitch, hipMemAllocHost, hipMemsetD16, hipMemsetD16Async, hipMemsetD8Async
Modified to support all scenarios hipMemcpyParam2DAsync, hipMemcpyParam2D.
[ROCm/clr commit: 96890792b0
2019-10-04 13:36:31 +05:30
1a7fc4e5c1
fix kernel descriptor bug with code object v3
...
Change-Id: I9306b2baf36d338e36c5ab1226f74373a61a5ae0
[ROCm/clr commit: d8e09c4b70
2019-10-03 10:56:35 -04:00
1c9dc1331e
Reimplement hipMemGetInfo ( #1447 )
...
Addresses SWDEV-136570. hipMemGetInfo changed to compute free memory based on information from kfd instead of relying on hc::am_tracker.
[ROCm/clr commit: 1ae4cbed4d
2019-10-01 12:40:36 +05:30
280212353c
hipModuleUnload should remove global variables from memtracker ( #1464 )
...
[ROCm/clr commit: 2a53299f07
2019-09-30 10:41:20 +05:30
3ef2e123bb
[HIP] Add tccDriver info in hipDeviceProp
...
Fixes #1433 .
[ROCm/clr commit: db4c0c0811
2019-09-26 13:53:33 +05:30
ca6b4bbc0d
Fix checks in hipMemcpy[D/H]to[D/H]() ( #1397 )
...
* SWDEV-202043 hipMemcpyDtoD() issue
[ROCm/clr commit: 842e304a9c
2019-09-16 08:32:38 +00:00
4436a38e65
Added new device attributes ( #1377 )
...
* Added new device attributes
* updated comment
* updated with new device attributes supported
[ROCm/clr commit: e2a934f377
2019-09-16 08:31:30 +00:00
f68652cc5f
[hip] add initial support for hipLaunchCooperativeKernelMultiDevice API ( #1368 )
...
* [hip] add initial support for hipLaunchCooperativeKernelMultiDevice API
* fix formatting
[ROCm/clr commit: 6c7da60e28
2019-09-16 08:31:17 +00:00
f79ffd587e
Fix typo and colon in comment
...
[ROCm/clr commit: d9179cbe70
2019-09-13 03:09:01 +05:30
b7a06efb6c
[HACK] Temporary fix for hipFree for hipManagedMalloc
...
[ROCm/clr commit: 2e9b8e9d0d
2019-09-13 02:10:21 +05:30
Alex Voicu
Rahul Garg
Aryan Salmanpour
Jeff Daily
satyanveshd
Sarbojit2019
ansurya
Rahul Garg
Laurent Morichetti
Anusha Godavarthy Surya
searlmc1
Nick Curtis
Vladislav Sytchenko
Evgeny
mhbliao
Siu Chi Chan