Anonymous | Login | 2024-04-27 03:56 UTC |
Main | My View | View Issues | Change Log | Docs |
Viewing Issue Simple Details [ Jump to Notes ] | [ Issue History ] [ Print ] | ||||||
ID | Category | Severity | Type | Date Submitted | Last Update | ||
0000543 | [1003.1(2008)/Issue 7] Base Definitions and Headers | Objection | Omission | 2012-02-17 10:25 | 2019-06-10 08:55 | ||
Reporter | geoffclare | View Status | public | ||||
Assigned To | ajosey | ||||||
Priority | normal | Resolution | Accepted | ||||
Status | Closed | ||||||
Name | Geoff Clare | ||||||
Organization | The Open Group | ||||||
User Reference | |||||||
Section | 4.19 | ||||||
Page Number | 117 | ||||||
Line Number | 3205 | ||||||
Interp Status | --- | ||||||
Final Accepted Text | |||||||
Summary | 0000543: Missing C99 text about undeserved underflow | ||||||
Description |
Section 4.19 Treatment of Error Conditions for Mathematical Functions and its corresponding rationale are missing an equivalent for the following text from C99 F.9 para 9 and footnote 321 (as amended by a technical corrigendum): Whether or when library functions raise an undeserved "underflow" floating-point exception is unspecified.321) Otherwise, as implied by F.7.6, the <math.h> functions do not raise spurious floating-point exceptions (detectable by the user), other than the "inexact" floating-point exception. 321) It is intended that undeserved "underflow" and "inexact" floating-point exceptions are raised only if avoiding them would be too costly. Also, the feraiseexcept() page is missing an equivalent of the text in F.7.6 referred to above, which is: If the argument to the feraiseexcept function in <fenv.h> represents IEC 60559 valid coincident floating-point exceptions for atomic operations (namely "overflow" and "inexact", or "underflow" and "inexact"), then "overflow" or "underflow" is raised before "inexact". Since POSIX.1 defers to C99, implementations are already required to behave in this way and no interpretation is needed when adding the missing text. Note that the added text in 4.19 should not be shaded MX because the introduction to chapter 4 says "No shading to denote extensions or options occurs in this chapter". |
||||||
Desired Action |
Add a new paragraph in 4.19 before the last paragraph: On implementations that support the IEC 60559 Floating-Point option, whether or when functions in the <math.h> header raise an undeserved underflow floating-point exception is unspecified. Otherwise, as implied by [xref to XSH feraiseexcept()] the <math.h> functions do not raise spurious floating-point exceptions (detectable by the user), other than the inexact floating-point exception. Cross-volume change to XSH... On page 834 line 27835 section feraiseexcept() change: The order in which these floating-point exceptions are raised is unspecified. to: The order in which these floating-point exceptions are raised is unspecified, [MX]except that if the excepts argument represents IEC 60559 valid coincident floating-point exceptions for atomic operations (namely overflow and inexact, or underflow and inexact), then overflow or underflow shall be raised before inexact.[/MX] Cross-volume change to XRAT... On page 3452 line 115876 section A.4.19 change: There is no additional rationale provided for this section. to: It is intended that undeserved underflow and inexact floating-point exceptions are raised only if avoiding them would be too costly. |
||||||
Tags | tc2-2008 | ||||||
Attached Files | |||||||
|
There are no notes attached to this issue. |
Mantis 1.1.6[^] Copyright © 2000 - 2008 Mantis Group |