 Alex Voicu
|
abe3c22cf1
|
This is primarily intended as an additional cleanup of the module functionality, in the aftermath of adopting module based dispatch. The main effort was associated with refactoring the questionable ihipModuleGetSymbol. It was quaintly written and misleading, in that it had little to do with getting symbols, and was exactly retrieving a kernel object. Error handling is modified so as to reduce branching depth. Functions which serve as interfaces to the HSA RT are moved in a separate helper header. Code object readers are properly deleted. Some leftover dead functionality pertaining to associating namespace scope variables with their allocated memory is removed. Executable loading is changed to use a string which holds the ELF image of the code object being loaded, thus avoiding some corner cases where using a istream would fail.
[ROCm/hip commit: 5127ce67e8]
|
2017-12-03 23:09:06 +00:00 |
|
|
Alex Voicu
|
08a0d96448
|
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.
[ROCm/hip commit: 7c0b9a005b]
|
2017-11-30 03:29:04 +00:00 |
|
|
Alex Voicu
|
b70bcb5f60
|
Add missing space (the final frontier).
[ROCm/hip commit: 20fc68c9a1]
|
2017-11-29 21:50:43 +00:00 |
|
|
Alex Voicu
|
8d51eaafb6
|
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 1c50968
[ROCm/hip commit: 32e11e7dc6]
|
2017-11-29 21:49:10 +00:00 |
|
|
Alex Voicu
|
cd61d38118
|
Use a much simpler guard for version 1.6, which allows for direct CUDA indexing to be used.
[ROCm/hip commit: 7acb1e6ff6]
|
2017-11-29 21:47:04 +00:00 |
|
|
Alex Voicu
|
1c50968051
|
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.
[ROCm/hip commit: d2fd1f5544]
|
2017-11-29 21:01:28 +00:00 |
|
|
Alex Voicu
|
c58a083e96
|
Fix compiler version check.
[ROCm/hip commit: b881cf713c]
|
2017-11-29 03:05:53 +00:00 |
|
|
Alex Voicu
|
b996310710
|
Fix oversight in selection mechanism which led to erroneous code to be compiled for the grid_launch_GGL component.
[ROCm/hip commit: faa546d194]
|
2017-11-29 01:37:52 +00:00 |
|
|
Alex Voicu
|
dd8a589893
|
Choose whether or not to use functional grid_launch based on the version of HCC used to compile.
[ROCm/hip commit: 89e9399427]
|
2017-11-29 00:17:44 +00:00 |
|
|
Alex Voicu
|
9668003fe3
|
Change memset kernel to use memcpy instead of placement new. Simplify indexers.
[ROCm/hip commit: 6e4ca3fbb4]
|
2017-11-28 19:45:47 +00:00 |
|
|
Alex Voicu
|
aef26d3477
|
Re-sync with upstream and re-factor platform global management for texture references.
[ROCm/hip commit: 02c2bfc7ef]
|
2017-11-28 19:15:29 +00:00 |
|
|
Alex Voicu
|
11f7d895f4
|
Merge remote-tracking branch 'origin/master' into feature_use_module_based_dispatch_instead_of_pfe
# Conflicts:
# src/hip_module.cpp
[ROCm/hip commit: dc67ca3feb]
|
2017-11-28 17:29:11 +00:00 |
|
 Ben Sander
|
d31ef2b462
|
Merge pull request #256 from gargrahul/texture_driver_api_support
Texture driver APIs support
[ROCm/hip commit: 0da0426f94]
|
2017-11-27 13:52:39 -06:00 |
|
|
Alex Voicu
|
acc05cdff7
|
Remove leftover comment.
[ROCm/hip commit: 08f252e4bf]
|
2017-11-22 19:37:03 +00:00 |
|
 Rahul Garg
|
13879387e5
|
Fixed review comments
[ROCm/hip commit: 56862b1c35]
|
2017-11-21 21:19:06 +05:30 |
|
|
Rahul Garg
|
03552f2e94
|
Changed function hipMemcpy_2D to hipMemcpyParam2D
[ROCm/hip commit: 9866fa250d]
|
2017-11-21 12:36:24 +05:30 |
|
|
Alex Voicu
|
45ff3c31c4
|
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.
[ROCm/hip commit: 9d088d2283]
|
2017-11-21 02:40:34 +00:00 |
|
|
Alex Voicu
|
69f8043d12
|
Correct ill-formed merge in earlier commit and adjust for differences with the new CUDA natural indexing mechanism.
[ROCm/hip commit: 7d5a45ac1a]
|
2017-11-20 16:33:52 +00:00 |
|
|
Alex Voicu
|
3c936777aa
|
Re-sync with upstream.
[ROCm/hip commit: c5f2b22d0d]
|
2017-11-20 15:34:50 +00:00 |
|
|
Rahul Garg
|
62f20c6dc8
|
Update hipModuleGetTexRef API
[ROCm/hip commit: c7d60a7a75]
|
2017-11-19 22:10:46 +05:30 |
|
|
Alex Voicu
|
962bf7bfda
|
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.
[ROCm/hip commit: cffd0e14eb]
|
2017-11-19 01:54:12 +00:00 |
|
|
Rahul Garg
|
58b37dce5f
|
-Fixed texture driver API sample
-Added hipTexRefSetAddress and hipTexRefSetAddress2D APIs
[ROCm/hip commit: 4b19c2aa0c]
|
2017-11-15 18:23:28 +05:30 |
|
|
Rahul Garg
|
0263292087
|
Texture code reorganized
[ROCm/hip commit: 63680edd30]
|
2017-11-14 11:09:35 +05:30 |
|
|
Rahul Garg
|
49a2c6468e
|
Texture driver APIs support
[ROCm/hip commit: ef09c4918d]
|
2017-11-09 22:10:55 +05:30 |
|
|
Alex Voicu
|
35d8af2001
|
Clean up trailing whitespace so as to reduce noise in #246.
[ROCm/hip commit: d8e323d4b5]
|
2017-11-08 00:08:55 +00:00 |
|
|
Alex Voicu
|
70a41e7dac
|
This switches HIP from its currently convoluted macro + pfe based dispatch mechanism to a more natural one partially based on the existing module API. The basic idea is that HCC will always correctly emit __global__ functions: as empty-bodied stubs, on host, and as kernels, on device. It then becomes trivial to obtain the mangled name on host, at dispatch, from the function's address, and then to use the mangled name to retrieve the kernel. This should address all problems stemming from serialisation, dubious mismatches due to the manufactured functor, macro-isms et al. It also immediately enables support for generalised globals as a consequence of that being available in the module API. Finally, it will make debug much easier, since the actual names of the __global__ functions will automatically be used in traces etc. One detail is that due to how dispatch works now (hipLaunchKernel and hipLaunchKernelGGL are themselves variadic function templates which deduce the function type of the callee), in certain cases it may be necesssary to insert explicit casts to ensure that the variadic argument list selects a viable overload - this can be observed in some unit tests. Eventually we may be able to remove this limitation, but for now it does not appear terribly onerous. The code is not extremely HIPpie, nor is it fully optimised, but rather is intended as a starting point for the HIP team to make its own.
[ROCm/hip commit: c2482d1255]
|
2017-11-01 15:09:59 +00:00 |
|
|
Ben Sander
|
51ee7807db
|
Merge pull request #222 from bensander/fix_device_prop
Fix device prop
[ROCm/hip commit: 2e8ec71e40]
|
2017-10-30 17:58:48 +01:00 |
|
 Siu Chi Chan
|
6cc7f10e84
|
Merge remote-tracking branch 'origin/master' into HEAD
[ROCm/hip commit: a9789ddcda]
|
2017-10-27 01:18:28 -04:00 |
|
|
Siu Chi Chan
|
ccc193cf98
|
add a printf device function
[ROCm/hip commit: f8bdb9f688]
|
2017-10-23 18:30:08 -04:00 |
|
|
Ben Sander
|
e249e3c30e
|
Update device properties.
- clear properties to defined initial state.
- enable some property flags which are now supported.
[ROCm/hip commit: acf89b43d4]
|
2017-10-20 15:52:13 +00:00 |
|
 Chris Kitching
|
49feb372f5
|
Add [[nodiscard]] attribute to hipError_t in C++17 mode
This will yield a compiler warning any time you discard an error
return value from an API call.
[ROCm/hip commit: 1d205ab130]
|
2017-10-19 13:34:02 +01:00 |
|
 Maneesh Gupta
|
710d095874
|
Merge pull request #194 from bensander/fix_genco_math_path
Fix math ordering for --genco mode.
[ROCm/hip commit: c1803c1951]
|
2017-10-03 16:17:34 +05:30 |
|
|
Ben Sander
|
d1be93bbd9
|
Fix math ordering for --genco mode.
[ROCm/hip commit: 68c2146dc3]
|
2017-10-02 21:52:16 +00:00 |
|
|
Rahul Garg
|
be4b7099bb
|
Fixed review comment
[ROCm/hip commit: b8aae66424]
|
2017-10-02 17:29:08 +05:30 |
|
|
Rahul Garg
|
28b92618b7
|
Fixed review comments
[ROCm/hip commit: 19d574d939]
|
2017-10-02 10:57:25 +05:30 |
|
|
Rahul Garg
|
71552c63ac
|
Fixed build issue
[ROCm/hip commit: f62254a8d8]
|
2017-10-02 10:40:01 +05:30 |
|
|
Rahul Garg
|
69508f52cc
|
Fix texture support on HIP/NVCC path
[ROCm/hip commit: 470224cf25]
|
2017-10-02 10:40:01 +05:30 |
|
|
Siu Chi Chan
|
6884635ae0
|
ensure the condition is evaluated before applying the negation operator
[ROCm/hip commit: 83d9e889a9]
|
2017-09-20 01:38:28 -04:00 |
|
|
Maneesh Gupta
|
ed0f94c051
|
Merge pull request #155 from wsttiger/master
Modify hip headers such that they are more compatible with strict C compiler
[ROCm/hip commit: 8c92d28ff3]
|
2017-09-15 10:49:55 +05:30 |
|
|
Maneesh Gupta
|
569ec04d0c
|
Fix typo in nvcc_detail/hip_runtime_api.h
[ROCm/hip commit: c3c94dcd8d]
|
2017-09-14 16:19:57 +05:30 |
|
|
Maneesh Gupta
|
0ba15f6faf
|
Merge pull request #175 from gargrahul/assert_support
Device side assert support
[ROCm/hip commit: c7e5fd24ac]
|
2017-09-14 16:08:00 +05:30 |
|
 sriharikarnam
|
ba8fd5cbee
|
Add hipDevice[Get/Set]SharedMemConfig API on nvcc details
[ROCm/hip commit: 3e69b2ed5a]
|
2017-09-14 16:03:19 +05:30 |
|
|
Rahul Garg
|
d1425f3cf5
|
Device side assert support
[ROCm/hip commit: 40ff42e555]
|
2017-09-07 20:45:27 +05:30 |
|
|
Ben Sander
|
a497059d78
|
Update docs.
- Review unsupported comments and refresh.
- Review experimental tag in API.
[ROCm/hip commit: 1a9e588932]
|
2017-08-28 15:44:38 -05:00 |
|
 wsttiger
|
09fb0fce42
|
Modify header such that it's compatible with strict C compiler
[ROCm/hip commit: b6a3d8c946]
|
2017-08-15 13:23:09 -05:00 |
|
|
wsttiger
|
db3ebbcbae
|
Add missing CUDA tags
[ROCm/hip commit: 512214e930]
|
2017-08-15 13:22:20 -05:00 |
|
|
Siu Chi Chan
|
21c1d69e77
|
implement __threadfence_system
[ROCm/hip commit: 96e79f97e2]
|
2017-08-02 08:50:18 +00:00 |
|
|
Ben Sander
|
9ec4b0b66f
|
Merge pull request #100 from weixingzhang/texture
HIP Texture Support
[ROCm/hip commit: 4f93b99f23]
|
2017-07-26 19:28:15 -05:00 |
|
|
Maneesh Gupta
|
9ce8abb1ed
|
Renable frexp(f) device math function
Change-Id: I53c022b8ddf38cd17ddb42eba457b9020db66395
[ROCm/hip commit: 7a446dcbc5]
|
2017-07-20 14:41:30 +05:30 |
|
 Weixing Zhang
|
65bec11393
|
HIP Texture Support
[ROCm/hip commit: e671cec367]
|
2017-07-17 15:16:12 -04:00 |
|