Add indirect matrix multiplication (imatmul) benchmarking support
- Introduced operator flags to
kleidiai_benchmarkto run matmul and imatmul micro-kernels separately. - Added imatmul benchmark runner alongside 6 imatmul micro-kernels to the benchmarking suite.
- Updated README and main.cpp for usage instructions and examples.
Signed-off-by: Cathal Lawlor cathal.lawlor@arm.com