AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
13,691 Commits 3 Branches 0 Tags
499e2ee044e15ebdf2c5cae1ee3ea282736ed66a
Commit Graph

13691 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Aditya Atluri 48b8b426d5 moved cpp specific device code from headers to source 2016-03-31 04:45:00 -05:00
Aditya Atluri 8d406b68d8 Added directory for stress testing 2016-03-31 02:14:12 -05:00
foreman 93730fbae0 P4 to Git Change 1253157 by johtaylo@johtaylo-JTBUILDER03-increment on 2016/03/31 03:00:10 
SWDEV-2 - Change OpenCL version number from 2079 to 2080.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1826 edit
 2016-03-31 03:07:46 -04:00
Aditya Atluri ed17e3489e Fixed failing tests 2016-03-30 13:00:44 -05:00
foreman 4b7a3bd5b4 P4 to Git Change 1252764 by jatang@jatang-opencl-hsa-stg1 on 2016/03/30 10:16:09 
SWDEV-86378 -  Apply the sDMA L2T work-around to T2L as well.

	HW confirms T2L also has issues.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gslbe/src/rt/GSLDevice.cpp#164 edit
 2016-03-30 10:21:54 -04:00
Aditya Atluri b2b3eec95c Revert to multiple output files than single .a lib 2016-03-30 07:54:18 -05:00
Aditya Atluri c4d93cf3db C guarding the apis 2016-03-30 06:08:50 -05:00
Aditya Atluri e476d4e8f0 Moved flag to hip_runtime.h 2016-03-30 03:32:39 -05:00
Maneesh Gupta 80b176d540 Implement more complex HCC, HSA directory selection logic 2016-03-30 13:58:38 +05:30
foreman f9492c32c8 P4 to Git Change 1252646 by johtaylo@johtaylo-JTBUILDER03-increment on 2016/03/30 03:00:11 
SWDEV-2 - Change OpenCL version number from 2078 to 2079.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1825 edit
 2016-03-30 03:11:23 -04:00
Aditya Atluri 00bf37d28d Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging 2016-03-29 11:18:09 -05:00
Ben Sander 3b1cd2d626 rename to hipMultiThreadDevice for consistency 2016-03-29 17:33:29 -05:00
Ben Sander 98a766ba75 Test improvements 
- partition hipThreadSafeDevice into smaller pieces.
- Add debug to hipMultiThreadStream.
- print more precision when mismatch detected.
- enable more tests in CMakeFiles.txt.
 2016-03-29 17:29:31 -05:00
Ben Sander e7b15a53da Inline some new string functions. 2016-03-29 17:28:27 -05:00
Ben Sander e22925be22 Add runtime switch to control HIP_ATP_MARKER 
Only generate the function strings if requested at
compile-time && runtime.
 2016-03-29 17:27:30 -05:00
Ben Sander 57783aad2d include codexl marker path, if found 2016-03-29 17:26:44 -05:00
Ben Sander de3391637f Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging 2016-03-29 17:12:36 -05:00
Ben Sander 03c0fcbfec Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging 2016-03-29 14:43:35 -05:00
Aditya Atluri ae2b136bfe Update README.md 2016-03-29 14:23:41 -05:00
Aditya Atluri 573d331ab8 Update README.md 2016-03-29 14:11:10 -05:00
Aditya Atluri 3b3db60c25 Calling intrinsics from global kernel 2016-03-29 11:17:55 -05:00
Aditya Atluri e652426715 added host flags to hipDeviceBody test 2016-03-29 09:27:57 -05:00
Aditya Atluri 4320e5c574 Added tests for running device code from a difference source file 2016-03-29 09:22:31 -05:00
foreman 511c6a5de8 P4 to Git Change 1252154 by emankov@em-hsa-amd on 2016/03/29 07:23:56 
SWDEV-79309 - Compiler Lib memory consumption decreasing by aclQueryInfo refactoring.

	[Problem] Memcpy on every aclQueryInfo for metadata. As a result an unreasonable memory consumption on every compiled binary (on Runtime).

	[Solution] Once extracted kernel Metadata is cached, secondary extraction doesn't happen. Blits are also affected.

	[Testing] precheckin
	http://ocltc.amd.com:8111/viewModification.html?modId=69016&personal=true&buildTypeId=&tab=vcsModificationBuilds&show_all_builds=true

	[Reviewers] Stanislav Mekhanoshin, Nikolay Haustov

Affected files ...

... //depot/stg/opencl/drivers/opencl/compiler/legacy-lib/include/v0_8/aclStructs.h#6 edit
... //depot/stg/opencl/drivers/opencl/compiler/legacy-lib/include/v0_8/aclTypes.h#3 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/api/v0_8/acl.cpp#41 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/common/v0_8/if_acl.cpp#92 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/backends/gpu/hsail_be.cpp#66 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/include/v0_8/aclStructs.h#23 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/include/v0_8/aclTypes.h#8 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/v0_8/libUtils.cpp#15 edit
... //depot/stg/opencl/drivers/opencl/compiler/lib/utils/v0_8/libUtils.h#24 edit
... //depot/stg/opencl/drivers/opencl/tests/ocltst/module/complib/CLAssumptionCheck.cpp#49 edit
 2016-03-29 07:35:41 -04:00
Ben Sander ba114a238f Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging 
Conflicts:
	include/hcc_detail/trace_helper.h
 2016-03-29 05:56:03 -05:00
Aditya Atluri 73ecbf5c95 Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging 2016-03-29 02:28:20 -05:00
streamhsa a629d4975f change makefile for samples 2016-03-29 16:02:09 +08:00
Aditya Atluri 8e437d0565 Not using classes inside headers 2016-03-29 02:27:51 -05:00
foreman 31bc8d0dbe P4 to Git Change 1252028 by johtaylo@johtaylo-JTBUILDER03-increment on 2016/03/29 03:00:11 
SWDEV-2 - Change OpenCL version number from 2077 to 2078.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1824 edit
 2016-03-29 03:11:36 -04:00
Ben Sander 1b2ab173c1 Tweak thread-safe implementation. 
introduce LockedAccessor option so destructor does not unlock.
Allows locks to exist across function boundaries, required
for hipLaunchKernel macro which has several unusual requirements.
(including C comppatibility, must use variadic macro, more).
 2016-03-28 21:41:47 -05:00
Siu Chi Chan e417fe8536 Merge pull request #19 from AMDComputeLibraries/fix_static_lib_link 
Various fixes in HIP exposed by linking to static library
 2016-03-28 22:40:04 -04:00
scchan 5e76b13cec remove the --hipcc_explicit_lib 2016-03-28 13:56:23 -05:00
scchan 3aa542e21b turn off ONE_OBJECT_FILE since we are building a static library 2016-03-28 12:56:15 -05:00
Maneesh Gupta f59b9e0aaf Fixed cmake not honoring user specified HIP_INSTALL_DIR 2016-03-28 22:23:27 +05:30
Maneesh Gupta 72751fa87c Improve cmake settings 2016-03-28 21:55:03 +05:30
Ben Sander 6cab7862ae Stream thread-safe checkpoint. 
Moving data structures to critical / protected section.
 2016-03-28 09:46:40 -05:00
scchan 5afa5d23fc add inline keyword to all the template functions 2016-03-28 09:18:38 -05:00
Ben Sander 4f8786c66a fix ihipLogStatus so status arg only evaluated once 2016-03-28 05:23:53 -05:00
Ben Sander ecd56e1400 Stream thread-safe checkpoint. 2016-03-28 04:22:20 -05:00
foreman 45f9afde22 P4 to Git Change 1251574 by johtaylo@johtaylo-JTBUILDER03-increment on 2016/03/28 03:00:10 
SWDEV-2 - Change OpenCL version number from 2076 to 2077.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1823 edit
 2016-03-28 03:12:29 -04:00
foreman 92fb88b700 P4 to Git Change 1251523 by johtaylo@johtaylo-JTBUILDER03-increment on 2016/03/27 03:00:11 
SWDEV-2 - Change OpenCL version number from 2075 to 2076.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1822 edit
 2016-03-27 03:06:16 -04:00
Ben Sander 82f57ca610 Improve debug print messages. 
- Remove "call-to-call" for hipStreamCreate and hipEventCreate.
  These now call an internal functions rather than calling through
  hipStreamCreateWithFalgs and hipEventCreateWithFlags.

- Add HIP_INIT_API for more functions so they trace correctly.

- Use stream#DEVICE.STREAMID in debug messages via new specialization in
  tace_helper.
 2016-03-26 12:35:04 -05:00
Ben Sander c47b5b04ef Protect _stream_id as well. 
- move lockedaccessor
- clean up device class.
- add simple ihipDevice constructor.
 2016-03-26 11:45:25 -05:00
Ben Sander 4dd77c6612 Make ihipDevice_t thread-safe. 
Move critical data into separate class and protect with LockAccessor
wrapper class.

For device, the streams list is the critical data since it is modified when
streams are created or destroyed.   The streams list is accessed in
several places including when synchronizing across all streams on the
device (ie from the default stream).
Other device data is set once by the device cosntructor and is not critical
so

All functions which acquire the LockAccessor now named with "locked_" prefix.
 2016-03-26 10:46:20 -05:00
dfukalov 666a887a1e removed FileCheck dependency & significantly improved test coverage 2016-03-26 16:01:49 +03:00
foreman 7b06fe12ff P4 to Git Change 1251492 by johtaylo@johtaylo-JTBUILDER03-increment on 2016/03/26 03:00:15 
SWDEV-2 - Change OpenCL version number from 2074 to 2075.

Affected files ...

... //depot/stg/opencl/drivers/opencl/runtime/utils/versions.hpp#1821 edit
 2016-03-26 03:11:41 -04:00
Ben Sander 18d3dc606b Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging 2016-03-25 17:12:13 -05:00
Ben Sander f8f839d9dd describe HIP_PLATFORM 2016-03-25 17:11:49 -05:00
Ben Sander 6d9eafa18a Improve automated platform detection. 
If AMD GPU is installed and detected by driver, default
HIP_PLATFORM to hcc.
 2016-03-25 17:08:34 -05:00
Aditya Atluri 268b316491 Merge branch 'privatestaging' of https://github.com/AMDComputeLibraries/HIP-privatestaging into privatestaging 2016-03-25 13:45:46 -05:00