26.13 Mapping of LISP Addresses to C addresses

This paragraph applies for some architectures only, e.g. the IBM/RS, HP PA-Risc and MIPS where the system architecture uses ’high’ bits in the addresses which conflict with the PSL tagging mechanism.(See the paragraph on the Size of the Address space above). The consequence of this is that using the inf operation for calculating the C equivalent from a LISP address does not work here. The easiest way (without explaining the reason) of mapping for the architectures is:

On Convex systems the ’high’ bit is the leftmost one. Therefore the tag bits start at bit 1 and the inf range is limited by 26 bits. The inf operation works as expected.