AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
125 Commits 3 Branches 0 Tags
369e0d7b5be2b879fe59d42870636a2abd33836d
Commit Graph

8 Commits

Author SHA1 Message Date
Ben Sander ecec7e36d9 Add abstraction for pinned/unpinned, and sync/async mem copies selection in tests 2016-02-25 04:21:24 -06:00
Ben Sander 549b18ce77 Improve async copy implementation. 
- Add device-side signal waits when transitioning between command classes
(Kernel, H2D copy, D2H copy).
- Support waiting in staged memory copies as well.
- Add several chicken bits to control implementation:
    - HIP_DISABLE_ENQ_BARRIER
    - HIP_DISABLE_BIDIR_MEMCPY
    - HIP_ONESHOT_COPY_DEP
- Refactor signal pool to support efficient deallocation based on
signsequnm.
- Deallocate copy signals on eventSynchronize.
- Improve copy tests, add pingpong.
 2016-02-22 23:15:24 -06:00
Ben Sander d33d806a5b Track last command to a stream. 
Passing simple tests.
 2016-02-20 11:02:07 -06:00
Ben Sander 134d7975ce Improve copy testing implementation. 
- add tests for (unpinned/pinned) x H2H x D2D.
- Free memory at end of test.
 2016-02-12 18:24:08 -06:00
Ben Sander 4ee2a5229b Create address tracker for am_alloc. 
Tracks device where memory is allocated, pinned-host or device, and
more.

Uses memory-range-based lookups - so pointers that exist anywhere in

the range of hostPtr + size will find the associated AmPointerInfo.

The insertions and lookups use a self-balancing binary tree and
should support O(logN) lookup speed.
 2016-02-12 18:24:08 -06:00
Ben Sander 0c74d3fc44 Test fixes: 
- Remove reference to missing test.
- Add hipMemset back.
- Parse --gpu option to specify default starting GPU.
 2016-02-08 22:55:23 -06:00
Ben Sander b8f7126662 in HIPCHECK, only run command once even if error occurs 2016-02-08 21:45:49 -06:00
Ben Sander f38e63ff18 Initial commit for GPUOpen Launch 2016-01-26 20:14:33 -06:00