Configurable generic filter hardware block and methods
A configurable generic filter hardware block and corresponding methods are provided. A configurable generic filter hardware block includes a plurality of multipliers; a plurality of adders; and one or more multiplexers. The, configurable generic filter hardware block is configured using a header data structure, and the header data structure includes a pointer to a memory location storing a plurality of input samples, a pointer to a memory location storing a plurality of output samples and a coefficient selection control value. The configurable generic filter hardware block is optionally invoked by a convolution instruction in one or more of a vector processor and a state machine. An exemplary Generic Filter Iteration loads input samples; selects coefficients; convolves the input samples and the selected coefficients and stores output samples. The header data structures are optionally stored sequentially in memory and processed in a single loop.
1. A configurable filter hardware block, comprising:
a plurality of multipliers;
a plurality of adders connected to the plurality of multipliers; and
one or more multiplexers connected to at least one of the plurality of adders, wherein said configurable filter hardware block is configured using a header data structure, said header data structure comprises a pointer to a memory location storing a plurality of input samples, a pointer to a memory location storing a plurality of output samples and a coefficient selection control value.
2. The configurable filter hardware block of claim 1 , wherein said header data structure further comprises one or more of an accumulation control value and an input/output data selection value.
3. The configurable filter hardware block of claim 1 , wherein said configurable filter hardware block is invoked by a convolution instruction in one or more of a vector processor and a state machine.
4. The configurable filter hardware block of claim 1 , further comprising a Generic Filter Iteration to perform the following steps: loading said plurality of input samples; selecting a plurality of coefficients; convolving said plurality of input samples and said plurality of selected coefficients and storing a plurality of output samples.
5. The configurable filter hardware block of claim 4 , wherein said configurable filter hardware block is invoked by said convolving step in said Generic Filter Iteration a plurality of times in a single loop for a plurality of filters.
6. The configurable filter hardware block of claim 4 , wherein each of said Generic Filter Iterations has a corresponding one of said header data structures.
7. The configurable filter hardware block of claim 6 , wherein said header data structures are stored sequentially in memory and processed in a single loop.
8. The configurable filter hardware block of claim 7 , wherein said header data structures are sequenced to reduce dependencies.
9. The configurable filter hardware block of claim 6 , wherein said header data structures are precomputed off-line.
10. A method for performing a plurality of filter operations, said method comprising:
providing a configurable filter hardware block comprising a plurality of multipliers, a plurality of adders and one or more multiplexers, wherein the plurality of multipliers are connected to the plurality of adders and the one or more multiplexers are connected to at least one of the plurality of adders, wherein said configurable filter hardware block is configured using a header data structure, wherein said header data structure comprises a pointer to a memory location storing a plurality of input samples, a pointer to a memory location storing a plurality of output samples and a coefficient selection control value; and
performing a Generic Filter Iteration routine comprising the following steps to configure said configurable filter hardware block based on one of said header data structures for a given one of said plurality of filter operations:
loading a plurality of input samples based on said one of said header data structures;
selecting a plurality of coefficients based on said one of said header data structures;
convolving said plurality of input samples and said plurality of selected coefficients; and
storing a plurality of output samples based on said pointer to a memory location storing a plurality of output samples in said one of said header data structures.
11. The method of claim 10 , wherein said step of selecting a plurality of coefficients further comprises the step of loading said plurality of coefficients from a memory based on said one of said header data structures.
12. The method of claim 10 , wherein said convolving step is performed in response to a vector convolve software instruction keyword to invoke said configurable filter hardware block.
13. The method of claim 12 , wherein said vector convolve software instruction keyword is part of an instruction set of one or more of a vector processor and a state machine.
14. The method of claim 10 , wherein said method is invoked by said convolving step in said Generic Filter Iteration a plurality of times in a single loop for a plurality of filters.
15. The method of claim 10 , wherein each of said Generic Filter Iterations has a corresponding one of said header data structures.
16. The method of claim 15 , wherein said header data structures are stored sequentially in memory and processed in a single loop.
17. An apparatus that performs a plurality of filter operations, comprising:
a configurable filter hardware block comprising a plurality of multipliers, a plurality of adders and one or more multiplexers, wherein the plurality of multipliers are connected to the plurality of adders and the one or more multiplexers are connected to at least one of the plurality of adders, wherein said configurable filter hardware block is configured using a header data structure; and
at least one hardware device operative to:
perform a Generic Filter Iteration routine comprising the following steps to configure said configurable filter hardware block based on one of said header data structures for a given one of said plurality of filter operations:
load a plurality of input samples based on said one of said header data structures; select a plurality of coefficients based on said one of said header data structures; convolve said plurality of input samples and said plurality of selected coefficients; and
store a plurality of output samples based on said pointer to a memory location storing a plurality of output samples in said one of said header data structures.
18. The apparatus of claim 17 , wherein said step of selecting a plurality of coefficients further comprises the step of loading said plurality of coefficients from a memory based on said one of said header data structures.
19. The apparatus of claim 17 , wherein said convolve operation is performed in response to a vector convolve software instruction keyword to invoke said configurable filter hardware block.
20. The apparatus of claim 17 , wherein said at least one hardware device comprises one or more of a vector processor and a state machine.
21. The apparatus of claim 17 , wherein each of said Generic Filter Iterations has a corresponding one of said header data structures.
22. The apparatus of claim 21 , wherein said header data structures are stored sequentially in memory and processed in a single loop.