[Cfp-interest 2425] Re: HAS_SUBNORM paper

[Jim Thomas]

Another problem is that footnotes 27 and 28 seem to give inconsistent interpretations.  For example, with FTZ and not DAZ, footnote 27 implies type_HAS_SUBNORM is intended to be -1 (indeterminant) but footnote 28 implies type_HAS_SUBNORM is intended to be 0 (absent). Likewise for IEC 60559 arithmetic with abrupt underflow. 

Even with both DAZ and FTZ, footnote 27 might suggest type_HAS_SUBNORM is intended to be -1, because DAZ and FTZ don’t apply to bit operations absolute value. 

Footnotes are not normative, but here they’re the only clues for what it is meant by “support subnormal numbers”.

- Jim Thomas


> On Apr 20, 2022, at 7:59 AM, Fred J. Tydeman <tydeman at tybor.com> wrote:
> 
> http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2797.htm
> 
> ---
> Fred J. Tydeman        Tydeman Consulting
> tydeman at tybor.com      Testing, numerics, programming
> +1 (702) 608-6093      Vice-chair of PL22.11 (ANSI "C")
> Sample C99+FPCE tests: http://www.tybor.com
> Savers sleep well, investors eat well, spenders work forever.
> 
> _______________________________________________
> Cfp-interest mailing list
> Cfp-interest at oakapple.net
> http://mailman.oakapple.net/mailman/listinfo/cfp-interest