 Ben Sander
|
1a6727e2fc
|
Merge pull request #246 from AlexVlx/feature_use_module_based_dispatch_instead_of_pfe
Use natural dispatch syntax
|
2017-12-01 14:21:25 -06:00 |
|
 Alex Voicu
|
33bb425013
|
Fix legacy mode detection of the address of an agent allocated variable. In this mode, there exist two executables per each code object, one created by HCC and one created by HIP. Since we dispatch through HCC in legacy mode, we should obtain the address for an agent allocated variable from the latter's executable. Also add two omitted validity checks, whose absence could lead to segfaults when the current process had no .kernel section and / or when an invalid or empty blob was extracted from the latter.
|
2017-11-30 03:29:04 +00:00 |
|
|
Alex Voicu
|
2e39534377
|
Add missing space (the final frontier).
|
2017-11-29 21:50:43 +00:00 |
|
|
Alex Voicu
|
4966518846
|
Revert "Revert adoption of CUDA indexing in general - this can only work with later versions of the compiler, just like module based dispatch, and thus must be guarded against usage in earlier (e.g. 1.6) versions."
This reverts commit 4792475
|
2017-11-29 21:49:10 +00:00 |
|
|
Alex Voicu
|
92e80f0943
|
Use a much simpler guard for version 1.6, which allows for direct CUDA indexing to be used.
|
2017-11-29 21:47:04 +00:00 |
|
|
Alex Voicu
|
30e4d445b1
|
Merge branch 'feature_use_module_based_dispatch_instead_of_pfe' of https://github.com/AlexVlx/HIP into feature_use_module_based_dispatch_instead_of_pfe
|
2017-11-29 21:45:56 +00:00 |
|
|
Alex Voicu
|
2557000b56
|
Revert "Revert adoption of CUDA indexing in general - this can only work with later versions of the compiler, just like module based dispatch, and thus must be guarded against usage in earlier (e.g. 1.6) versions."
This reverts commit d2fd1f5
|
2017-11-29 21:36:29 +00:00 |
|
|
Ben Sander
|
2f542c46c8
|
Merge pull request #275 from mangupta/bump_version_1.5
Bumped HIP version to 1.5 and updated summary in HIP release notes
|
2017-11-29 15:06:02 -06:00 |
|
|
Alex Voicu
|
4792475d01
|
Revert adoption of CUDA indexing in general - this can only work with later versions of the compiler, just like module based dispatch, and thus must be guarded against usage in earlier (e.g. 1.6) versions.
|
2017-11-29 21:01:28 +00:00 |
|
 Maneesh Gupta
|
2d5e0274ef
|
Bumped HIP version to 1.5 and updated summary in HIP release notes
Change-Id: I4b612da88236674fac2c585ab6891bdd495511b0
|
2017-11-29 13:56:17 +05:30 |
|
|
Alex Voicu
|
93e595c267
|
Fix compiler version check.
|
2017-11-29 03:05:53 +00:00 |
|
|
Alex Voicu
|
e305d3fc94
|
Add missing file.
|
2017-11-29 02:16:44 +00:00 |
|
|
Alex Voicu
|
37c1811b2a
|
Fix oversight in selection mechanism which led to erroneous code to be compiled for the grid_launch_GGL component.
|
2017-11-29 01:37:52 +00:00 |
|
|
Alex Voicu
|
c6ab6f292b
|
Choose whether or not to use functional grid_launch based on the version of HCC used to compile.
|
2017-11-29 00:17:44 +00:00 |
|
|
Alex Voicu
|
45a9f4f7b1
|
Remove leftover agent allocated globals.
|
2017-11-28 19:56:04 +00:00 |
|
|
Alex Voicu
|
d5c8de3f41
|
Change memset kernel to use memcpy instead of placement new. Simplify indexers.
|
2017-11-28 19:45:47 +00:00 |
|
|
Alex Voicu
|
6a0efb7ed2
|
Re-sync with upstream and re-factor platform global management for texture references.
|
2017-11-28 19:15:29 +00:00 |
|
|
Alex Voicu
|
d37a5a6008
|
Merge remote-tracking branch 'origin/master' into feature_use_module_based_dispatch_instead_of_pfe
# Conflicts:
# src/hip_module.cpp
|
2017-11-28 17:29:11 +00:00 |
|
|
Maneesh Gupta
|
cc418da654
|
Fix float2int rounding functions
Change-Id: I67943859a6344c5eec0eaa23418c9b802ef72468
|
2017-11-28 17:23:43 +00:00 |
|
|
Alex Voicu
|
bb35299560
|
This actually (tries) to do the right thing all the way, by using memcpy for bitcasting, and not rely on undefined behaviour of a different flavour as a substitute for the original undefined behaviour. Note that the compiler will (should) optimise down to the same emitted code, since this is a pattern it understands.
|
2017-11-28 17:23:06 +00:00 |
|
|
Alex Voicu
|
4b7cb0624e
|
This fixes some outright quaint choices made when implementing HIP's bitwise conversion functions, by using simple reinterpret_casts, as is idiomatic. These functions are supposed to be re-entrant, correct and efficient. Sadly, they were neither: they hid a massive race condition against a value stored in global memory, which means that they were also unreasonably slow if they ever managed to be correct, and relied on union based type punning which is in a grey area of the standard. It is difficult to ascertain what may have been the reason for coming up with this quirky solution.
|
2017-11-28 17:23:06 +00:00 |
|
 Chris Kitching
|
7761f63bbc
|
Add hipify mappings for all CUDA headers that have HIP equivalents
I'm particularly running into issues with `device_types.h` in real
CUDA code...
|
2017-11-28 17:23:06 +00:00 |
|
|
Ben Sander
|
e93a24bdbe
|
Merge pull request #256 from gargrahul/texture_driver_api_support
Texture driver APIs support
|
2017-11-27 13:52:39 -06:00 |
|
 Rahul Garg
|
6c18df987d
|
Added note in Release notes for texture driver APIs
|
2017-11-27 22:06:50 +05:30 |
|
 Evgeny Mankov
|
b25d199111
|
Merge pull request #262 from ChrisKitching/frontendaction
[HIPIFY] Mostly fix preprocessor-or-template induced issues
|
2017-11-27 17:30:11 +03:00 |
|
|
Rahul Garg
|
3e9a4cfdd1
|
Porting guides update for texture APIs usage
|
2017-11-24 12:00:55 +05:30 |
|
|
Ben Sander
|
77de4b96cf
|
Merge pull request #273 from mangupta/swdev-129574
Fix float2int rounding functions
|
2017-11-23 12:04:36 -06:00 |
|
|
Maneesh Gupta
|
822743845c
|
Fix float2int rounding functions
Change-Id: I67943859a6344c5eec0eaa23418c9b802ef72468
|
2017-11-23 09:57:24 +05:30 |
|
|
Alex Voicu
|
dfa532db98
|
Remove leftover comment.
|
2017-11-22 19:37:03 +00:00 |
|
|
Evgeny Mankov
|
6bddac1cfa
|
Merge pull request #263 from ChrisKitching/headers
[HIPIFY] Add hipify mappings for all CUDA headers that have HIP equivalents
|
2017-11-22 21:24:21 +03:00 |
|
|
Alex Voicu
|
0755f1fc26
|
Modify the set component of the memcpy test (unclear why there is a memset component to begin with).
|
2017-11-21 17:52:01 +00:00 |
|
|
Rahul Garg
|
38029f2849
|
Fixed review comments
|
2017-11-21 21:19:06 +05:30 |
|
|
Alex Voicu
|
fb1021cc0a
|
This corrects how addresses are formed for symbols which reside in shared objects. For this case, the .value component of an ELF symbol holds the offset from the base VA where the shared object was loaded. Thus, to correctly obtain the VA of the object refered by the symbol, we must add the offset to the VA where the shared object is loaded. We were already doing this correctly for symbols denoting functions, but we were incorrect for those denoting objects.
|
2017-11-21 13:15:13 +00:00 |
|
|
Rahul Garg
|
24307fe5c4
|
Changed function hipMemcpy_2D to hipMemcpyParam2D
|
2017-11-21 12:36:24 +05:30 |
|
|
Alex Voicu
|
f8c1c1b38e
|
Refactor the __device__ versions of memset and memcpy to be less awkward i.e. not return nullptr as opposed to the destination pointer (it can only be assumed it was done for maximum confusion) and actually unroll as they claim to. Change all of the {to, from}Symbol functions to use hipModuleGetGlobal, as opposed to hc::accelerator::get_symbol_address which is no longer valid with module based dispatch.
|
2017-11-21 02:40:34 +00:00 |
|
|
Alex Voicu
|
24f8a93ff7
|
Clean-up some remaining noise in program_state.cpp.
|
2017-11-20 22:41:46 +00:00 |
|
|
Alex Voicu
|
27bc487635
|
Correct ill-formed merge in earlier commit and adjust for differences with the new CUDA natural indexing mechanism.
|
2017-11-20 16:33:52 +00:00 |
|
|
Alex Voicu
|
30d90dab38
|
Re-sync with upstream.
|
2017-11-20 15:34:50 +00:00 |
|
|
Ben Sander
|
a1e0c96689
|
Merge pull request #264 from pzins/missing_end_marker
Fix missing MARKER_END
|
2017-11-20 06:08:01 -06:00 |
|
|
Rahul Garg
|
eff09fcd86
|
-Moved coGlobals in hipModule class (takes care of multi module case)
-Used mutex scope for updating coGlobals
|
2017-11-20 16:23:18 +05:30 |
|
|
Maneesh Gupta
|
4477d3d314
|
Merge pull request #266 from gargrahul/fix_half2_gfx900
Fixed half2 issue on gfx900
|
2017-11-20 07:28:41 +05:30 |
|
|
Maneesh Gupta
|
29c0ab8401
|
Merge pull request #265 from phani544/nvccTests
[nvccTests]Enabled inline_asm_vadd on nvcc
|
2017-11-20 07:28:29 +05:30 |
|
|
Ben Sander
|
aeadc3f18f
|
Fix test on cuda
|
2017-11-19 15:31:02 -06:00 |
|
|
Ben Sander
|
a43262e699
|
Merge branch 'feature_natural_indexing' of https://github.com/AlexVlx/HIP
|
2017-11-19 15:25:17 -06:00 |
|
|
Ben Sander
|
fc34fd6f03
|
Temporarily disable P2P on nvidia (fails on dual GPU)
|
2017-11-19 15:21:37 -06:00 |
|
|
Rahul Garg
|
76f174b536
|
Update hipModuleGetTexRef API
|
2017-11-19 22:10:46 +05:30 |
|
|
Alex Voicu
|
0cc921f103
|
This implements the trivial change needed to move back from the hip{Something}_{x, y, z} macros to the natural CUDA syntax of Something.{x, y, z}. This is contained in lines 384-404 in hip_runtime.h. All of the other changes have to do with changing unit tests to use this syntax. The macros are retained for backwards compatibility.
|
2017-11-19 01:54:12 +00:00 |
|
|
Alex Voicu
|
45ac5b70df
|
Merge remote-tracking branch 'origin/master' into feature_use_module_based_dispatch_instead_of_pfe
|
2017-11-16 23:20:15 +00:00 |
|
|
Rahul Garg
|
cfc9f87465
|
Fixed test case for GFX900
|
2017-11-16 09:34:52 +05:30 |
|
|
Rahul Garg
|
20947f80fb
|
Fixed half2 issue on gfx900
|
2017-11-15 18:52:59 +05:30 |
|