127f7d0320
Make HIP functional again with HCC from ROCm 1.9.x
...
Change-Id: I214acdfd0b79dcf783993e44fe31baee64fd4dc3
[ROCm/clr commit: b01ac26948
2018-10-24 10:41:56 +05:30
c505c2be38
Re-sync with upstream.
...
[ROCm/clr commit: 602280eb8b
2018-10-18 12:27:03 +01:00
cf95358003
Merge pull request #692 from whchung/hip-reinit-take2
...
HIP program state re-initialization logic (take 2)
[ROCm/clr commit: 9143ae6bdb
2018-10-18 12:06:41 +05:30
bb3706b6f6
Merge pull request #703 from mangupta/stream_create_with_priority
...
Implementation for stream priority
[ROCm/clr commit: 95be669f4a
2018-10-17 10:53:43 +05:30
50388a3bd1
Re-sync with upstream.
...
[ROCm/clr commit: 9ec697c620
2018-10-10 11:43:49 +01:00
c4ed6b18fb
Implementation for stream priority
...
- Requires ROCm 1.9.x or higher
- Requires HCC with PR#886 merged
Change-Id: Id7c95ea091ee610e80c9ad815f1cb989cba570ca
[ROCm/clr commit: 07ee1f07d8
2018-10-05 16:27:46 +05:30
ed5c4bca1d
Corrected the width calculation logic to accomodate multi channels
...
[ROCm/clr commit: a78d68d639
2018-10-03 12:07:38 +05:30
8981438983
Fixed image width for linear resource type texture
...
[ROCm/clr commit: 167265c279
2018-10-01 15:28:34 +05:30
73fc21f717
Keep the map which tracks GPU kernel symbols to grow monotonically
...
[ROCm/clr commit: e257de95f3
2018-09-26 19:49:02 +00:00
1e956fcd70
Improve performance of re-initialization logic
...
Keep track of shared libaries already discovered. Do not build HSA executables
for them.
[ROCm/clr commit: 060b3c0bf8
2018-09-26 19:48:56 +00:00
4e0f0b3e89
HIP program state re-initialization logic
...
This commit is to support kernels dynamically loaded thru means such as
dlopen() after HIP runtime initializes.
[ROCm/clr commit: 319f007bf1
2018-09-26 19:48:47 +00:00
e0efc8a09e
Return hipSuccess when sizeBytes=0 in hipMemset
...
[ROCm/clr commit: bd27310127
2018-09-26 12:47:36 +05:30
f5d204450c
Align with HC Next.
...
[ROCm/clr commit: 26c26f63cb
2018-09-17 11:50:29 +03:00
20abaa15c4
Merge branch 'master' into support-malloc
...
[ROCm/clr commit: b23609ad42
2018-09-17 10:17:25 +05:30
cbb7e4af8b
Add HIP_DB=fatbin for debugging fat binary issues
...
[ROCm/clr commit: ce13c2061a
2018-08-17 11:53:45 -04:00
bb9d95188a
Merge pull request #621 from ROCm-Developer-Tools/disable_startup_loader
...
Disable startup loader by default and guard with env var
[ROCm/clr commit: e10fa70173
2018-08-10 10:18:25 +05:30
3621ec1971
resolve a segfault bug when env var not set; remove startup_kernel_loader class
...
[ROCm/clr commit: 7c467220a8
2018-08-09 16:40:26 +00:00
e1d72394da
Add more check to ensure the startup loader only be enabled with the env var set to 1
...
[ROCm/clr commit: 7d776ae665
2018-08-04 01:52:27 +00:00
ef368c0f43
Add startup loader under HIP_STARTUP_LOADER env var, disable by default
...
[ROCm/clr commit: 9af862d364
2018-08-04 01:48:06 +00:00
d7b2bab183
Revert "HIP program state re-initialization logic"
...
This reverts commit 865fe01b371fe7f1be91
2018-08-03 17:03:04 -05:00
fbf97dddb5
Revert "Improve performance of re-initialization logic"
...
This reverts commit 1d1b402e5bf6178c2159
2018-08-03 17:02:58 -05:00
1494fc442f
Revert "Keep the map which tracks GPU kernel symbols to grow monotonically"
...
This reverts commit ca62af218c329f3b7a5e
2018-08-03 17:02:50 -05:00
57c15b815d
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 >
[ROCm/clr commit: 84aadb9274
2018-07-28 23:15:16 -07:00
9a4744566f
Support malloc/free for hip-clang
...
[ROCm/clr commit: d17dc21c9c
2018-07-27 16:24:51 -04:00
792b9ebe7c
Null check before setting offset
...
[ROCm/clr commit: f17c897751
2018-07-24 12:25:40 +05:30
99758b3064
Set offset in hipGetTextureAlignmentOffset
...
[ROCm/clr commit: 0ec64e8b24
2018-07-24 10:11:26 +05:30
653c7c24f4
It is unclear what I was thinking when authoring the original code...
...
[ROCm/clr commit: 633ce36ea7
2018-07-17 14:04:57 +01:00
c96a617b78
Merge branch 'master' into move-memcpy
...
[ROCm/clr commit: 1fc3ef3cd4
2018-07-17 10:51:42 +05:30
8eeb225c04
Merge pull request #515 from ROCm-Developer-Tools/hipclang-add-amdgcn-funcs
...
Add hipclang amdgcn functions
[ROCm/clr commit: 7d8c1ae761
2018-07-17 09:25:09 +05:30
25df971097
Merge pull request #562 from ROCm-Developer-Tools/fix-build-failure
...
Fix build failure in code_object_bundle.cpp
[ROCm/clr commit: 4702a6c86e
2018-07-12 07:49:41 +05:30
0346cccd6a
Merge pull request #546 from gargrahul/fix_bindtex_offset_null_check
...
Fixed offset null check in bind texture functions
[ROCm/clr commit: d63db71b61
2018-07-11 12:52:31 +05:30
81fc2e03b4
Merge pull request #545 from ROCm-Developer-Tools/revert-521-temp_fixmemcpy2dasync_trsmissue
...
Revert "Use memcpy kernel for all pinned memory cases in hipMemcpy2DAsync"
[ROCm/clr commit: 404bd43a0e
2018-07-11 12:52:09 +05:30
007a005cf3
Move __hip_hc_memcpy and __hip_hc_memset from device_utils.cpp to device_functions.h as inline functions
...
[ROCm/clr commit: 903dfa94eb
2018-07-10 18:12:41 -04:00
733c5b98a0
Fix build failure in code_object_bundle.cpp
...
[ROCm/clr commit: e6d155bcd5
2018-07-10 16:49:59 -04:00
eab9cacef2
Implement hip_ldg Functions into HIP header
...
Move all the function definitions for hip_ldg.cpp into hip_ldg.h header and enable for HIP clang path.
[ROCm/clr commit: d39508b4fd
2018-07-05 20:38:46 +00:00
94cc27d4d6
Implement min/max functions in HIP header
...
Remove using hc::precise_math min and max. Instead we can use ocml directly for device and std:: for host.
[ROCm/clr commit: b362c53475
2018-07-05 20:15:41 +00:00
f78a8d7fd6
Implement Memory Fence Functions in header
...
Enabled __llvm_fence_* functions for seq_cst.
[ROCm/clr commit: 96ab7c7b25
2018-07-04 23:35:24 +00:00
4b941a0db8
Merge branch 'master' into hipclang-add-amdgcn-funcs
...
[ROCm/clr commit: f8109c5b06
2018-07-04 17:36:08 +00:00
70a7d3961e
Update hip_hcc_internal.h
...
Adding missing include for hip_hcc_internal in order to build with HCC
[ROCm/clr commit: d372d0eaad
2018-07-04 09:33:51 +05:30
676c289310
Fixed offset null check in bind texture functions
...
[ROCm/clr commit: 77cbd4476d
2018-07-03 08:54:17 +05:30
a796eedde9
Revert "Use memcpy kernel for all pinned memory cases in hipMemcpy2DAsync"
...
[ROCm/clr commit: f554e48db3
2018-07-02 14:32:11 +05:30
c60682797a
Implement __shfl_* funcs into HIP headers
...
[ROCm/clr commit: e44eaa1a1e
2018-06-26 18:32:11 +00:00
7cb3cec513
Implement __ballot, __any, __all into HIP headers
...
[ROCm/clr commit: 9a295a732d
2018-06-20 17:39:39 +00:00
d09a2f9237
Implement hip_hc.ll into HIP headers
...
Move all __hip_hc_ir_* functions from hip_hc.ll into HIP header as inline asm. Remove hip_hc.ll and build dependencies from HIP.
[ROCm/clr commit: 175c87f2bf
2018-06-20 17:39:31 +00:00
7f2693010b
Implement device_functions.cpp into HIP headers
...
Move all Integer Intrinsics, device_functions.cpp definitions and HIP specific device functions into HIP headers. Implement the device functions using llvm_intrinsics and device-libs functions instead of calling hc::__* functions. Remove device_functions.cpp since everything is now defined in header.
[ROCm/clr commit: ad5f1dcbf2
2018-06-20 17:39:23 +00:00
1e6f00bf67
Add hipclang amdgcn functions
...
These are moving from hipclang in device library to hip headers. These are required for the functionality of HIPclang project.
[ROCm/clr commit: f4399d57a3
2018-06-20 17:38:37 +00:00
bdc67ac1b4
Merge pull request #490 from ROCm-Developer-Tools/feature_decouple_atomics_from_hc
...
Switch the atomic implementation to use Clang builtins.
[ROCm/clr commit: d27991e1c8
2018-06-20 14:16:43 +05:30
721a3f60be
Merge pull request #457 from whchung/hip-reinit
...
HIP program state re-initialization logic
[ROCm/clr commit: 8ff33a83b1
2018-06-20 09:37:27 +05:30
ca62af218c
Keep the map which tracks GPU kernel symbols to grow monotonically
...
[ROCm/clr commit: 8f521edff1
2018-06-18 16:54:18 -05:00
1d1b402e5b
Improve performance of re-initialization logic
...
Keep track of shared libaries already discovered. Do not build HSA executables
for them.
[ROCm/clr commit: b883ea759d
2018-06-15 18:07:33 -05:00
Maneesh Gupta
Alex Voicu
Rahul Garg
Wen-Heng (Jack) Chung
Yaxun Sam Liu
sunway513
Sarunya Pumma
Aaron Enye Shi
Rahul Garg