Commit Graph

7 Commits

Author SHA1 Message Date
German Andryeyev d9397590de Add option to skip AQL barrier
The change reuses HSA signals for dispatches as a wait signal.
Skipping the barrier requires to  disable L2 cache for sysmem
allocations and extra tracking for HDP access with the large bar.
ROC_BARRIER_SYNC=0 activates the new logic. Barrier sync is
still used by default.
ROC_ACTIVE_WAIT=1 enables unconditional active wait in ROCr.
The change also consolidated ROCr wait logic under single function.

Change-Id: I6bd1be30aa88258da1b1f9de319ef5a45852afd8
2020-10-06 08:37:12 -04:00
German Andryeyev c5afd5d412 Initial HMM support
- Expose ROCclr interfaces for HIP usage
- ROCr interfaces aren't available in staging, thus control the
build with AMD_HMM_SUPPORT define

Change-Id: Iadc2bcc230e78d3b0dc22b235189c8cc80843446
2020-06-12 09:06:07 -04:00
Jason Tang b4f1239f34 device/rocm: split gfxVersion to major/minor/stepping
Change-Id: I1e437eaee30794147713d9516229211670f01d90
2020-05-12 12:17:13 -04:00
Saleel Kudchadker 5f64e6e7ad Add a threshold for forcing ROCr to take blit path
This workaround is to avoid performance penalty of SDMA engine
taking a while to clock up from a lower DPM state. Add env var
GPU_FORCE_BLIT_COPY_SIZE (1024 by default for HIP in KB). Forcing
Src and Dst agent to be amdgpu makes ROCr take blit copy path for
what otherwise should have been SDMA copy

Change-Id: I222f687155f86000d17d66d25182e490b6710463
2020-04-28 17:11:24 -04:00
German Andryeyev 481d526859 SWDEV-184709 - support hipLaunchCooperativeKernel()
- Enable cooperative groups support, based on ROCr capability

Change-Id: I975bcea0af7865009eaed24454ce71d897ea8fc4
2020-04-01 12:13:33 -04:00
Laurent Morichetti b4c6143a2f Update copyright info
Change-Id: Ia4f9ff0f5f873b4223a8cca154188bb0d2f1abba
2020-02-04 09:26:14 -08:00
Laurent Morichetti 20c7173849 Merge branch 'origin/pghafari/vdi-prototype' into lmoriche/amd-master
Change-Id: Id3b833d405596735becb3346f3b08c6da57033fe
2020-01-30 20:12:13 -08:00