Convolve over FIR

Hi,

I need to create an array of FIR filters with the same input block for all and the same filter state for all as well (but different coefficients for each filter).
I understand that the FIR module has an inner state buffer that it controls and probably utilize the HW's circular buffer support to prevent reorder of the state each new sample calculation. 
I don't want this state to be an inner function because I want to reuse it for all the other filters as well.
I understand I can use the Convolve module instead and manage the state buffer outside, but how can I utilize the HW's circular buffer support and prevent redundant state reordering and additional MIPS? 
And in addition, does the Convolve module also utilize the SIMD acceleration engine on the SHARC processor like in the FIR module?

Thanks,
Sharon.