This page is a snapshot from the LWG issues list, see the Library Active Issues List for more information and the meaning of New status.
Section: 31.5.2.6 [ios.base.storage] Status: New Submitter: Jiang An Opened: 2022-02-14 Last modified: 2022-03-04
Priority: 4
View all other issues in [ios.base.storage].
View all issues with New status.
Discussion:
Currently 31.5.2.6 [ios.base.storage] p5 and p8 say "On failure, a valid long&/void*& initialized to 0". Such wording seems wrong, because a long&/void*& variable or return value can't be initialized with 0. And the values of referenced objects may be underspecified, because an implementation may reuse the same long/void* objects on failure, and thus it's insufficient to specify the initial values of these objects only.
[2022-03-04; Reflector poll]
Set priority to 4 after reflector poll.
Proposed resolution:
This wording is relative to N4901.
Modify 31.5.2.6 [ios.base.storage] as indicated:
long& iword(int idx);-3- […]
-4- […] -5- Returns: On success iarray[idx]. On failure, anvalidlvalue of type long&with value 0Linitialized to 0.void*& pword(int idx);-6- […]
-7- […] -8- Returns: On success parray[idx]. On failure, anvalidlvalue of type void*&with a null pointer valueinitialized to 0. -9- Remarks: After a subsequent call to pword(int) for the same object, the earlier return value may no longer be valid.