Prefetch instruction mechanism for processor

Abstract

Accordingly, a prefetch instruction mechanism is desired for implementing a prefetch instruction which is non-faulting, non-blocking, and non-modifying of architectural register state. Advantageously, a prefetch mechanism described herein is provided largely without the addition of substantial complexity to a load execution unit. In one embodiment, the non-faulting attribute of the prefetch mechanism is provided though use of the vector decode supplied Op sequence that activates an alternate exception handler. The non-modifying of architectural register state attribute is provided (in an exemplary embodiment) by first decoding a PREFETCH instruction to an Op sequence targeting a scratch register wherein the scratch register has scope limited to the Op sequence corresponding to the PREFETCH instruction. Although described in the context of a vector decode embodiment, the prefetch mechanism can be implemented with hardware decoders and suitable modifications to decode paths will be appreciated by those of skill in the art based on the description herein. Similarly, although in one particular embodiment such a scratch register is architecturally defined to read as a NULL (or zero) value, any target for the Op sequence that is not part of the architectural state of the processor would also be suitable. Finally, in one embodiment the non-blocking attribute is provided by the Op sequence completing (without waiting for return of fill data) upon posting of a cache fill request to load logic of a data cache. In this way, LdOps which follow in a load pipe are not stalled by a prefetch-related miss and can instead execute concurrently with the prefetch-related line fill.