| commit | 4c35366cf7d08dc76b353d43e14e628586636592 | |
| author | Tobias Grosser <tobias@grosser.es> | |
| Sun, 18 Sep 2016 06:50:35 +0000 (18 06:50 +0000) | ||
| committer | Tobias Grosser <tobias@grosser.es> | |
| Sun, 18 Sep 2016 06:50:35 +0000 (18 06:50 +0000) | ||
| tree | 3ef5d6b88ecb233abecbd1dbceb08078b2d714b8 | treesnapshot (tar.gz zip) |
| parent | 081a29fdd844ad83b6ca4b60b194274dc0fec896 | commitdiff |
GPGPU: Dynamically ensure 'sufficient compute'
Offloading to a GPU is only beneficial if there is a sufficient amount of
compute that can be accelerated. Many kernels just have a very small number
of dynamic compute, which means GPU acceleration is not beneficial. We
compute at run-time an approximation of how many dynamic instructions will be
executed and fall back to CPU code in case this number is not sufficiently
large. To keep the run-time checking code simple, we over-approximate the
number of instructions executed in each statement by computing the volume of
the rectangular hull of its iteration space.
git-svn-id: https://llvm.org/svn/llvm-project/polly/trunk@281848 91177308-0d34-0410-b5e6-96231b3b80d8
Offloading to a GPU is only beneficial if there is a sufficient amount of
compute that can be accelerated. Many kernels just have a very small number
of dynamic compute, which means GPU acceleration is not beneficial. We
compute at run-time an approximation of how many dynamic instructions will be
executed and fall back to CPU code in case this number is not sufficiently
large. To keep the run-time checking code simple, we over-approximate the
number of instructions executed in each statement by computing the volume of
the rectangular hull of its iteration space.
git-svn-id: https://llvm.org/svn/llvm-project/polly/trunk@281848 91177308-0d34-0410-b5e6-96231b3b80d8
| lib/CodeGen/PPCGCodeGeneration.cpp | diffblobblamehistory | |
| test/GPGPU/double-parallel-loop.ll | diffblobblamehistory |