Wykres commitów

203 Commity

Autor SHA1 Wiadomość Data
agodavar d10124e1d7 cmake hip-rocclr as default build,change HIP_PLATFORM to amd|nvidia
Change-Id: I59d2e15b297b6ca2d5eee1a6dee642bdb01a3839
2020-10-21 07:26:14 -04:00
sumanthtg 543dc8cea5 [dtest] Tests for atomicAdd/atomicAddnoRet
Positive and negative test cases for atomicAdd and
AtomicAddnoRet device functions

SWDEV-238517 for enhancing hip unit tests

Change-Id: Id20ba2550d20f224004f105cdcd087002cb80e56
2020-10-07 13:20:12 -04:00
Sarbojit Sarkar a2f9a2d469 SWDEV:253548 : remove gfx-arch macros
Change-Id: I1ef0b4e080c15a9c6228f48ecd6a65b134d0a72d
2020-10-06 01:15:58 -04:00
Jatin Chaudhary e7eaafe165 Fixing/Disabling tests on nvcc path
Change-Id: I7ed4a1246943867482ff595d5f98bb25bd79074b
2020-07-30 00:39:07 -04:00
sushantkumar14 e590f20cbd [dtest] Half2 comparison related tests
Tests heq2, hne2, hle2, hge2, hlt2, hgt2 APIs for functionality
and NaN tests

SWDEV-238517 for enhancing hip unit tests

Change-Id: I88a9a8ead0d00a1261f3d650361d655f2f397e48
2020-07-14 23:53:55 -04:00
mshivama 27e237ef48 Add support for few CUDA compatible reduction operations like __syncthreads_count()
Change-Id: I0917820058f0edad5c0bdd42dea37f06353fc63c
2020-07-14 12:03:14 -04:00
Yaxun (Sam) Liu 13316e2919 Add pow(float/double/_Float16,int)
Change-Id: Ie65d15cd3df9853a3bbd613d8c7188ae39c327c7
2020-07-06 07:38:57 -04:00
Siu Chi Chan 784ca6f43c add constexpr constructor for vector types
Change-Id: I45bb0537d6a24ee50b548c2fd8b4f20518764813
2020-06-04 01:57:03 -04:00
Jatin 2d517fdcc6 Adding changes for hipExtLaunchKernel for rocCLR
Change-Id: Iba52bc3bde7c37f3fb375a55ba0947e87b3cdc9b
2020-06-02 14:16:41 -04:00
Matt Arsenault d2dd307c7d Remove some asm declarations for intrinsics
This technique should never be used, and only accessed through
__builtins.

There's currently no builtin for groupstaticsize. I left ds_swizzle
since for some reason it switches to the builtin based on __HCC__ or
not.

Change-Id: If1e1394221dba83ea4add6db5e94d6b715552044
2020-05-11 15:20:58 -04:00
agodavar 1468982f8a find_package(hip) changes for clang
Change-Id: I9dc1be1dd5f9424e26ebcd2961207e1c4467550f
2020-05-01 20:48:32 +00:00
Yaxun (Sam) Liu 808dae6813 Enable template max and min for HIP-Clang (#2028)
It was for HCC only. HIP-Clang also needs it for __fp16 since AMDMIGraphX uses it.

Change-Id: Id49322b7b89ef799accdf6b47627a6fce51d1ab5
2020-04-24 12:30:28 -07:00
Yaxun (Sam) Liu 8d83e95457 Disable device side malloc (#2009)
* Disable device side malloc

Currently device side malloc is not working and takes excessive
device memory.

Disable it for now until a working malloc is implemented.

Change-Id: I1ad908c1c53a83752383b4be96688a848642c699
2020-04-14 16:07:14 +05:30
ansurya 770e76e752 Initial support for bfloat16 (#1980) 2020-04-06 15:35:43 +05:30
Jatin Chaudhary ee5fa8977c [dtest] Adding VectorTest (#1732)
Adding unit test for operations of device vector types
2020-03-26 17:03:43 +05:30
Jatin Chaudhary 16a6a94fbf Adding Half Abs APIs (#1902) 2020-03-17 14:13:19 +05:30
Yaxun (Sam) Liu 02848e0b80 Fix test hip_bitextract.cpp (#1784)
The randomly generated offset+width may exceeds 32, which causes
a left shift operation with 32-offset-width. As an unsigned number
that is greater than 32 and causes undefined behavior. When the
test is compiled without -mavx it is still OK. However when
the test is compiled with -mavx, the undefined behavior causes
wrong results and test failure.

This patch adjusts width so that offset+width<=32 always.
2020-01-10 13:46:57 +05:30
Sarbojit2019 aa4aea0754 Change to generate hipVersion.h (#1726)
HIP_VERSION_MAJOR, HIP_VERSION_MINOR, HIP_VERSION_PATCH and HIP_VERSION pre-processor macros are now defined in hipVersion.h instead of being set by hipcc.
2019-12-30 12:44:24 +05:30
Maneesh Gupta 32442c6506 Revert changes for atomic FADD support when address is in LDS (#1701)
This reverts PR #1591 and follow-on PR #1695
2019-11-29 11:58:12 +05:30
Alex Voicu d597e7ca20 Use native support for atomic FADD when address is in LDS (#1591) 2019-11-22 07:53:48 +05:30
Alex Voicu b5b3d1bbaa Extend vector type capabilities and add tests to reflect it. (#1656) 2019-11-20 21:32:32 +05:30
vsytch 2fb734ef5f Update hipMathFunctions, hipTestHalf and hipTestNativeHalf tests to support Navi10 and Navi14. (#1545) 2019-10-16 10:51:48 +05:30
ansurya ba9c6e13e4 Added new Memory API's (#1399)
Added new memory API's hipMemAllocPitch, hipMemAllocHost, hipMemsetD16, hipMemsetD16Async, hipMemsetD8Async
Modified to support all scenarios hipMemcpyParam2DAsync, hipMemcpyParam2D.
2019-10-04 13:36:31 +05:30
Yaxun (Sam) Liu 56193a7828 Fix cast of __half for HIP-clang (#1475) 2019-09-30 10:40:42 +05:30
AlexBinXie 5ed1f3e2c8 [hip]Skip test when hipHostMallocCoherent is not supported by implementation (#1380) 2019-09-16 08:31:43 +00:00
Sameer Sahasrabuddhe 70023c9075 remove obsolete test for OCKL Asynchronous Streams
The implementation for OCKL AS was recently removed from the device
library since that feature is now superseded by hostcall.
2019-08-30 20:41:29 +05:30
amd-lthakur d3ffad7c83 [dtests] Fixed build issues with hipAsynchronousStreams.cpp on windows (#1292)
Changed the third arg of the functions __hip_as_write_block and __ockl_as_write_block from ulong to uint64_t so as to fix the compilation error in windows
2019-08-09 11:49:31 +00:00
lthakur d18160920e Resubmitting the fix so as to address the changed format in the last submission 2019-08-08 11:26:26 +05:30
lthakur 435badcb0e Fixed compilation errors being observed on windows 2019-08-07 15:54:46 +05:30
Vladislav Sytchenko fd3b626386 Don't use a hardcoded warp size, since it can be dynamically changed. Query it from the runtime instead. 2019-07-31 17:04:31 -04:00
Aaron Enye Shi b754de37c1 Add GFX908 specific changes to HIP (#1229)
* Add GFX908 specific for HIP

* Fix missing __halfTest in hipTestNativeHalf
2019-07-24 07:51:17 +00:00
Alex Voicu 67abac1365 Put 3-wide vector types on a ketogenic diet. (#1180)
* Put 3-wide vector types on a ketogenic diet.

* Remove needless include.

* Do not be narrow-minded.

* Do not be narrow-minded.

* Put the C people on a diet too.
2019-06-25 06:32:09 +05:30
Maneesh Gupta 9f2d1453fb [hit] Rename RUN -> TEST & RUN_NAMED -> TEST_NAMED
Change-Id: I75e24f15129973cee15fc9dac65d678bd2172074
2019-05-09 09:59:18 +05:30
Maneesh Gupta f931152280 Merge pull request #1075 from mhbliao/hliao/master/test_fix2
[test] Add device variant of `std::declval`.
2019-05-07 16:15:01 +05:30
Maneesh Gupta 1d4941e487 Merge pull request #1074 from mhbliao/hliao/master/test_fix
[test] Use explicit cast for address space cast.
2019-05-07 16:09:15 +05:30
Maneesh Gupta 730763c817 Merge pull request #1069 from mhbliao/hliao/master/test_cleanup
[test] Remove unused common routines.
2019-05-07 16:02:57 +05:30
Michael LIAO 4fbfb646f6 [test] Add device variant of std::declval.
- Current clang disallows any invocation of wrong-side functions even
  under context with type-inspection only. Work around that by adding a
  variant of `std::decl` with `__device__` attribute.
2019-05-03 15:58:31 -04:00
Michael LIAO 2746d95906 [test] Use explicit cast for address space cast. 2019-05-03 14:56:00 -04:00
Michael LIAO 16a27e06ee [test] Remove unused common routines.
- Remove used common routines, which variadic functions, not supported
  feature, are used.
2019-05-02 12:34:07 -04:00
Michael LIAO 9bd2d5746d [Device Function] Fix implementation of __bitinsert_u64
- It's a common mistake by assuming 1 << shamt would be promoted to
  64-bit, if shamt is a 64-bit integer. That's not the case. Replace
  that left shift to a 64-bit one to ensure it won't fall into undefined
  behavior.
- Fix the host-side implementation as well for device function testing.
2019-04-30 08:59:13 -04:00
Aaron Enye Shi ede494bd5f Fix mixed dot for Jenkins rocm_2_0 2019-01-24 22:48:35 +00:00
Aaron Enye Shi 7e46f0639c Exclude mixed dot functions from nvcc path 2019-01-24 21:00:45 +00:00
Aaron Enye Shi f31aa946e7 Fix mixed dot for Jenkins rocm_head 2019-01-23 16:39:25 +00:00
Aaron Enye Shi 1d70f4217e Add tests for dot functions 2019-01-22 22:32:28 +00:00
Maneesh Gupta 580c808400 Update hipAsynchronousStreams.cpp
This test needs to be excluded from nvcc platform
2019-01-07 09:26:21 +05:30
Sameer Sahasrabuddhe 8fe2a4b167 introduce test for asynchronous streams
This test is identical to the HCC tests under
tests/Unit/OCKL/AsyncStreams, but only for synchronous operation.
2019-01-04 05:51:50 -05:00
Aaron Enye Shi 58dfeff27a Add *_rn functions back into HIP intrinsics
Add back the round-to-nearest-even intrinsics back to HIP math
intrinsics as it caused regression.
2018-12-18 19:31:54 +00:00
Michael Kuron 8610128c3e Merge branch 'master' into getsymboladdress 2018-11-20 12:03:22 +01:00
Aaron Enye Shi 0cfaa52d15 Guard rcp rounded implementation as well
Since rcp implementations of non-default rounded versions are not correct or supported in OCML, guard them using the same macro OCML_BASIC_ROUNDED_OPERATIONS. Also update the docs and tests.
2018-11-06 19:53:28 +00:00
Aaron Enye Shi e59c33250a Guard the OCML rounded operations instead
Instead of commenting all these functions out, guard the functions with a macro OCML_BASIC_ROUNDED_OPERATIONS.
2018-11-06 16:32:14 +00:00