diff options
author | Viet-Hoa Do <viet-hoa.do@arm.com> | 2022-12-14 14:49:56 +0000 |
---|---|---|
committer | Viet-Hoa Do <viet-hoa.do@arm.com> | 2022-12-23 14:11:34 +0000 |
commit | 04f4620cf999846a44089c81720aa920edec6993 (patch) | |
tree | 1c0080ac59d5b2aa500cd2b2ceffe0575e22a4b6 /src/dynamic_fusion/sketch/gpu/GpuOperatorGroup.h | |
parent | 81fdaddaf36cb4c7ff0d2c52a370dd977a13dc72 (diff) | |
download | ComputeLibrary-04f4620cf999846a44089c81720aa920edec6993.tar.gz |
Add multiple output support for dynamic fusion
* The dependency graph now can schedule any acyclic graph into
a sequential list of operators. This is needed as the output
operators now form branches in the graph.
* Fix the definition of input, output and intermediate tensors
in GpuKernelComponentGroup to support non-linear but sequential
list of operators.
* Add constraint on GpuOperatorGroup to enforce strictly linear
fusion style, but allow output operator as the only form of
branch.
Resolves: COMPMID-5771
Signed-off-by: Viet-Hoa Do <viet-hoa.do@arm.com>
Change-Id: I68de3a31a2456145081f0a397e4e61dd66327682
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/8823
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/dynamic_fusion/sketch/gpu/GpuOperatorGroup.h')
-rw-r--r-- | src/dynamic_fusion/sketch/gpu/GpuOperatorGroup.h | 10 |
1 files changed, 6 insertions, 4 deletions
diff --git a/src/dynamic_fusion/sketch/gpu/GpuOperatorGroup.h b/src/dynamic_fusion/sketch/gpu/GpuOperatorGroup.h index 35abe6c543..308a9d796a 100644 --- a/src/dynamic_fusion/sketch/gpu/GpuOperatorGroup.h +++ b/src/dynamic_fusion/sketch/gpu/GpuOperatorGroup.h @@ -77,17 +77,19 @@ public: static constexpr size_t max_fused_operators = 32; /** Try adding (without actually adding) an operator to the group * - * @param[in] op Operator to be added + * @param[in] op Operator to be added + * @param[in] is_output Whether this operator is the output operator. * * @return true If @p op can be added while maintaining the invariants * @return false Otherwise */ - bool try_add_operator(const Operator &op) const; + bool try_add_operator(const Operator &op, bool is_output = false) const; /** Add an operator to the group * - * @param[in] op Operator to be added + * @param[in] op Operator to be added + * @param[in] is_output Whether this operator is the output operator. */ - void add_operator(const Operator &op); + void add_operator(const Operator &op, bool is_output = false); /** Create a new operator * * @param[in] operator_type @ref GpuOperatorType of the new operator |