View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000979 | 1003.1(2013)/Issue7+TC1 | System Interfaces | public | 2015-08-24 07:14 | 2019-06-10 08:54 |
| Reporter | EdSchouten | Assigned To | ajosey | ||
| Priority | normal | Severity | Editorial | Type | Clarification Requested |
| Status | Closed | Resolution | Accepted As Marked | ||
| Name | Ed Schouten | ||||
| Organization | Nuxi | ||||
| User Reference | |||||
| Section | getaddrinfo() | ||||
| Page Number | 925 | ||||
| Line Number | 31185 | ||||
| Interp Status | --- | ||||
| Final Accepted Text | See 0000979:0002845. | ||||
| Summary | 0000979: AI_V4MAPPED: "ai_addrlen shall be 16", should it really? | ||||
| Description | If I understand the description of getaddrinfo() correctly, ai_addrlen corresponds with the size of the socket address structure returned. This allows you to call connect(), bind(), etc., using [ai_addr, ai_addrlen]. There is a sentence in the article that does contradict this: "If the AI_V4MAPPED flag is specified along with an ai_family of AF_INET6, then getaddrinfo() shall return IPv4-mapped IPv6 addresses on finding no matching IPv6 addresses (ai_addrlen shall be 16)" My objection is with the "ai_addrlen shall be 16". The intent would be that ai_addrlen is set to sizeof(struct sockaddr_in6) -- not 16 specifically. Having a struct sockaddr_in6 that is only 16 bytes in size would be impossible. | ||||
| Desired Action | This part of the sentence should either be removed or replaced to state that ai_addrlen shall be sizeof(struct sockaddr_in6). | ||||
| Tags | tc2-2008 | ||||
|
|
I concur; it appears that the intent of the 16 was to indicate all 16 bytes of the mapped sin6_addr are significant, and that only storing the 4 bytes of the V4 address in that location, as a way to save 12 bytes of allocation, was not to be done by conforming implementations. If the intent was the 16 be an arbitrary value indicating the address was a mapped form and not a regular IP6 address that was reported I'd think that would have been mentioned in the Application Usage section. |
|
|
Change:If the AI_V4MAPPED flag is specified along with an ai_family of AF_INET6 on P925, L31183-31187 to (keeping the IP6 shading): By default, with an ai_family of AF_INET6, getaddrinfo() shall |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2015-08-24 07:14 | EdSchouten | New Issue | |
| 2015-08-24 07:14 | EdSchouten | Status | New => Under Review |
| 2015-08-24 07:14 | EdSchouten | Assigned To | => ajosey |
| 2015-08-24 07:14 | EdSchouten | Name | => Ed Schouten |
| 2015-08-24 07:14 | EdSchouten | Organization | => Nuxi |
| 2015-08-24 07:14 | EdSchouten | Section | => getaddrinfo() |
| 2015-08-24 07:14 | EdSchouten | Page Number | => (page or range of pages) |
| 2015-08-24 07:14 | EdSchouten | Line Number | => (Line or range of lines) |
| 2015-08-25 07:31 | shware_systems | Note Added: 0002796 | |
| 2015-09-24 15:18 | nick | Page Number | (page or range of pages) => 925 |
| 2015-09-24 15:18 | nick | Line Number | (Line or range of lines) => 31185 |
| 2015-09-24 15:18 | nick | Interp Status | => --- |
| 2015-09-24 16:06 | Don Cragun | Note Added: 0002845 | |
| 2015-09-24 16:07 | Don Cragun | Note Edited: 0002845 | |
| 2015-09-24 16:08 | Don Cragun | Final Accepted Text | => See 0000979:0002845. |
| 2015-09-24 16:08 | Don Cragun | Status | Under Review => Resolved |
| 2015-09-24 16:08 | Don Cragun | Resolution | Open => Accepted As Marked |
| 2015-09-24 16:09 | Don Cragun | Tag Attached: tc2-2008 | |
| 2015-10-05 14:28 | geoffclare | Category | Base Definitions and Headers => System Interfaces |
| 2015-10-05 14:28 | geoffclare | Project | 1003.1(2008)/Issue 7 => 1003.1(2013)/Issue7+TC1 |
| 2019-06-10 08:54 | agadmin | Status | Resolved => Closed |
