View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000148 | 1003.1(2008)/Issue 7 | System Interfaces | public | 2009-09-14 23:28 | 2013-04-16 13:06 |
| Reporter | drepper | Assigned To | ajosey | ||
| Priority | normal | Severity | Objection | Type | Clarification Requested |
| Status | Closed | Resolution | Accepted As Marked | ||
| Name | Ulrich Drepper | ||||
| Organization | Red Hat, Inc. | ||||
| User Reference | ud-pow-pole | ||||
| Section | pow | ||||
| Page Number | 1518 | ||||
| Line Number | 49005-49008 | ||||
| Interp Status | --- | ||||
| Final Accepted Text | 0000148:0000225 | ||||
| Summary | 0000148: unclear return value description | ||||
| Description | The pow man page currently says: On systems that support the IEC 60559 Floating-Point option, a pole error shall occur and pow(), powf(), and powl() shall return ±HUGE_VAL, ±HUGE_VALF, and ±HUGE_VALL, respectively if y is an odd integer, or HUGE_VAL, HUGE_VALF, and HUGE_VALL, respectively if y is not an odd integer. The problem is that it is not described when the + or - in case of odd integers is used. In fact, I think the + is wrong. | ||||
| Desired Action | Replace in the above paragraph ±HUGE_VAL, ±HUGE_VALF, and ±HUGE_VALL with -HUGE_VAL, -HUGE_VALF, and -HUGE_VALL | ||||
| Tags | c99, tc1-2008 | ||||
|
|
The suggested change would create a conflict with the C Standard. C99 (n1256) says in F.9.4.4: -- pow(±0, y) returns ±inf and raises the ``divide-by-zero'' floating-point exception for y an odd integer < 0. -- pow(±0, y) returns +inf and raises the ``divide-by-zero'' floating-point exception for y < 0 and not an odd integer. Note that "pow(±0, y) returns ±inf" is shorthand for "pow(+0, y) returns +inf and pow(-0, y) returns -inf". (See F.9 para 12). These requirements in C99 seem right to me. POSIX doesn't quite match C99, in that it just says "zero" instead of "±0". I think we should change: "On systems that support the IEC 60559 Floating-Point option, a pole error shall occur ..." to: "On systems that support the IEC 60559 Floating-Point option, if x is ±0, a pole error shall occur ..." |
|
|
I agree with Geoff's change, that's what's missing. The ± in the new text is the anchor for the ± in the rest of the sentence. Without this change there are problems. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2009-09-14 23:28 | drepper | New Issue | |
| 2009-09-14 23:28 | drepper | Status | New => Under Review |
| 2009-09-14 23:28 | drepper | Assigned To | => ajosey |
| 2009-09-14 23:28 | drepper | Name | => Ulrich Drepper |
| 2009-09-14 23:28 | drepper | Organization | => Red Hat, Inc. |
| 2009-09-14 23:28 | drepper | User Reference | => ud-pow-pole |
| 2009-09-14 23:28 | drepper | Section | => pow |
| 2009-09-14 23:28 | drepper | Page Number | => 1518 |
| 2009-09-14 23:28 | drepper | Line Number | => 49005-49008 |
| 2009-09-15 08:36 | geoffclare | Note Added: 0000225 | |
| 2009-09-15 14:52 | drepper | Note Added: 0000226 | |
| 2009-09-17 00:43 | nick | Tag Attached: c99 | |
| 2009-09-24 15:42 | msbrown | Interp Status | => --- |
| 2009-09-24 15:42 | msbrown | Final Accepted Text | => 0000148:0000225 |
| 2009-09-24 15:42 | msbrown | Status | Under Review => Resolved |
| 2009-09-24 15:42 | msbrown | Resolution | Open => Accepted As Marked |
| 2010-08-27 13:12 | ajosey | Tag Attached: tc1-2008 | |
| 2013-04-16 13:06 | ajosey | Status | Resolved => Closed |
