A NOTE has been added to this issue. ====================================================================== https://www.austingroupbugs.net/view.php?id=1387 ====================================================================== Reported By: rhansen Assigned To: ajosey ====================================================================== Project: 1003.1(2008)/Issue 7 Issue ID: 1387 Category: System Interfaces Type: Clarification Requested Severity: Editorial Priority: normal Status: Under Review Name: Richard Hansen Organization: User Reference: Section: malloc Page Number: 1295 (Issue 7 2018 edition) Line Number: 43161 (Issue 7 2018 edition) Interp Status: --- Final Accepted Text: ====================================================================== Date Submitted: 2020-08-10 16:46 UTC Last Modified: 2020-08-20 17:58 UTC ====================================================================== Summary: Should EAGAIN be acceptable for malloc failure? ======================================================================
---------------------------------------------------------------------- (0004935) rhansen (manager) - 2020-08-20 17:58 https://www.austingroupbugs.net/view.php?id=1387#c4935 ---------------------------------------------------------------------- > But isn't this a distinction without a difference? I agree with your analysis. That's why it feels wrong to me that Solaris-based systems use EAGAIN—they should use ENOMEM instead. I see these options:<ul><li>Change Solaris-based systems to use ENOMEM instead of EAGAIN. This is unlikely to happen.</li><li>Declare Solaris-based systems to be non-conforming and live with that.</li><li>Try to convince ourselves that the Solaris EAGAIN error condition is completely different than "insufficient storage space is available", and therefore the behavior does not run afoul of "Implementations shall not generate a different error number from one required by this volume of POSIX.1-2017 for an error condition described in this volume of POSIX.1-2017." I don't like this option because I don't think the error condition is completely different, for the reasons you state in your note.</li><li>Change the standard to mention EAGAIN as a possible error with a subtly different meaning ("insufficient storage space right now, but it might succeed if you try again later" vs. "there will never be sufficient storage space no matter what").</li></ul> Issue History Date Modified Username Field Change ====================================================================== 2020-08-10 16:46 rhansen New Issue 2020-08-10 16:46 rhansen Status New => Under Review 2020-08-10 16:46 rhansen Assigned To => ajosey 2020-08-10 16:46 rhansen Name => Richard Hansen 2020-08-10 16:46 rhansen Section => malloc 2020-08-10 16:46 rhansen Page Number => 1295 (Issue 7 2018 edition) 2020-08-10 16:46 rhansen Line Number => 43161 (Issue 7 2018 edition) 2020-08-10 16:46 rhansen Interp Status => --- 2020-08-10 17:09 rhansen Note Added: 0004914 2020-08-10 17:14 rhansen Description Updated 2020-08-10 17:35 rhansen Desired Action Updated 2020-08-10 17:36 rhansen Desired Action Updated 2020-08-10 17:39 Don Cragun Note Added: 0004915 2020-08-10 17:45 alanc Note Added: 0004916 2020-08-10 17:45 alanc Note Edited: 0004916 2020-08-11 01:08 shware_systems Note Added: 0004918 2020-08-14 09:50 joerg Note Added: 0004923 2020-08-14 09:50 joerg Note Edited: 0004923 2020-08-14 09:51 joerg Note Edited: 0004923 2020-08-19 09:53 Konrad_Schwarz Note Added: 0004927 2020-08-19 19:57 rhansen Note Added: 0004932 2020-08-20 07:23 Konrad_Schwarz Note Added: 0004933 2020-08-20 17:58 rhansen Note Added: 0004935 ======================================================================
