Date: Wed, 12 Jan 2022 08:50:23 -0800 (PST)
On Mon, 6 Dec 2021 21:40:30 +0100 Jens Maurer wrote:
>
>On 02/12/2021 17.18, Fred J. Tydeman via Liaison wrote:
>> Issue for C++ liason group.
>>
>> The WG14 paper:
>> http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2797.htm
>> discusses several issues with the *_HAS_SUBNORM macros.
>
>The existing text tries to give semantics to the *_HAS_SUBNORM
>preprocessor macros.
Agreed.
>The "Proposal" removes any mention of these macros and just adds
>general text about subnormals.
Independent of what we do with the *_HAS_SUBNORM macros, we
should add the green text to remove undefined behaviour from C.
>That seems a step backwards; the values of the *_HAS_SUBNORM
>macros should continue to be related to subnormal behavior
>of the implementation.
Do you want to leave them as is? Or, try to expand them to cover the
implementations that flush either operands or results (but not both)?
>Put differently, are you suggesting to remove the *_HAS_SUBNORM
>macros from the C standard? If not, they appear to exist
>without any semantic definition given your proposal.
Yes, the suggestion was to remove.
>
>On 02/12/2021 17.18, Fred J. Tydeman via Liaison wrote:
>> Issue for C++ liason group.
>>
>> The WG14 paper:
>> http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2797.htm
>> discusses several issues with the *_HAS_SUBNORM macros.
>
>The existing text tries to give semantics to the *_HAS_SUBNORM
>preprocessor macros.
Agreed.
>The "Proposal" removes any mention of these macros and just adds
>general text about subnormals.
Independent of what we do with the *_HAS_SUBNORM macros, we
should add the green text to remove undefined behaviour from C.
>That seems a step backwards; the values of the *_HAS_SUBNORM
>macros should continue to be related to subnormal behavior
>of the implementation.
Do you want to leave them as is? Or, try to expand them to cover the
implementations that flush either operands or results (but not both)?
>Put differently, are you suggesting to remove the *_HAS_SUBNORM
>macros from the C standard? If not, they appear to exist
>without any semantic definition given your proposal.
Yes, the suggestion was to remove.
--- Fred J. Tydeman Tydeman Consulting tydeman_at_[hidden] 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.
Received on 2022-01-12 16:50:26