[Cfp-interest 1277] Re: AI about default vs alternate exception handling

[Jim Thomas]

> On Mar 2, 2019, at 12:01 PM, Jim Thomas <jaswthomas at sbcglobal.net> wrote:
> 
> Action item:
>>     Jim: Look to see if there is a place to put a note to address the alternate exception handling in annex F to clarify how C specifies using default exception handling only.
>> 
> 
> 
> Currently we have
> 
> F.8 Floating-point environment
> 1 The floating-point environment defined in <fenv.h> includes the IEC 60559 floating-point exception
> status flags and directed-rounding control modes. It includes also IEC 60559 dynamic rounding
> precision and trap enablement modes, if the implementation supports them.383)
> 
> F.8.1 Environment management
> 1 IEC 60559 requires that floating-point operations implicitly raise floating-point exception status
> flags, and that rounding control modes can be set explicitly to affect result values of floating-point
> operations. These changes to the floating-point state are treated as side effects which respect
> sequence points.384)
> 
> 383)This specification does not require dynamic rounding precision nor trap enablement modes.
> 384)If the state for the FENV_ACCESS pragma is "off", the implementation is free to assume the dynamic floating-point control
> modes will be the default ones and the floating-point status flags will not be tested, which allows certain optimizations (see
> F.9).
> 
> F.8.1 and footnote 384 are included here for context.
> 
> We could add a second paragraph to F.8 to address default vs alternate exception handling, as done below.
> 
> “directed-rounding control modes” should be “rounding direction control modes”.

"rounding direction control modes" should be "rounding-direction control modes”.

> "directed-rounding" refers to rounding up, down, toward zero, and alway from zero, as opposed to rounding to nearest. This mistake also appears in footnote 211 (in all-20190211.pdf).

Footnote 211 also needs "directed-rounding control modes" replaced by "rounding-direction control modes".

> 
> F.8#1and footnote 383 below have been updated. The dynamic rounding precision and trap enablement modes mentioned in F.8 and footnote 383 were features of the original IEEE 754-1985 standard that were not carried over to IEEE 754-2008 (IEC 60559:2011) nor do they appear in IEEE 754-2019. However, some implementations still provide them.
> 
> Here’s a suggested replacement for F.8

and footnote 383
> :
> 
> F.8 Floating-point environment
> 1 The floating-point environment defined in <fenv.h> includes the IEC 60559 floating-point exception status flags and rounding direction

rounding-direction

> control modes. It may also include other floating-point status or modes that the implementation provides as extensions.383)
> 
> This annex does not include support for IEC 60559’s optional alternate exception handling. The specification in this annex assumes IEC 60559 default exception handling: the flag is set, a default result is delivered, and execution continues. Implementations might provide alternate exception handling as an extension.
> 
> F.8.1 ...
> 
> 383) Dynamic rounding precision and trap enablement modes are examples of such extensions.
> 384) ...
> 
> 
> - Jim Thomas
>  
> _______________________________________________
> Cfp-interest mailing list
> Cfp-interest at oakapple.net
> http://mailman.oakapple.net/mailman/listinfo/cfp-interest

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.oakapple.net/pipermail/cfp-interest/attachments/20190317/c740cb6c/attachment-0001.html