[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
Re: [Stds-754] IEEE 754 Revision, comments from IFIP WG 2.5
I'm sure the issue of trying to standardise NaNs has come up a few times
already.
Based on those conversations since 2001,
I'm sure the present members of 754R will not approve anything that
renders any of their existing implementations
non-upward-performance-compatible.
I note that PowerPC records the distinction between some of the classes
mentioned in 9.2 by setting separate bits in the FPSCR
The Motorola 68K did this partially too.
I believe 754R has addressed this issue. It is of course practical only
when hardware instructions exist that incorporate rounding information
in the opcode.
Any implementation is free to add such operations. Alpha had them in
hardware.
The real issue is getting language standardizers to agree on a syntax for
accessing such operations, and application programmers to come up with realistic
meaningful application benchmarks that exploit such features enough that
hardware implementation becomes attractive. This is a long-term
process.