Austin Group Defect Tracker

Aardvark Mark III

Viewing Issue Simple Details Jump to Notes ] Issue History ] Print ]
ID Category Severity Type Date Submitted Last Update
0001201 [1003.1(2013)/Issue7+TC1] System Interfaces Editorial Clarification Requested 2018-08-22 14:10 2018-08-24 20:23
Reporter Florian Weimer View Status public  
Assigned To
Priority normal Resolution Open  
Status New  
Name Florian Weimer
Organization Red Hat
User Reference
Section sigsuspend
Page Number unknown
Line Number unknown
Interp Status ---
Final Accepted Text
Summary 0001201: Atomicity requirements for sigsuspend
Description The description says that sigsuspend changes the signal mask and then suspends the thread until a signal arrives, but does not indicate whether these two steps are atomic with regards to signal delivery. Based on the current wording, the following can happen:

1. The requested signal is unblocked.
2. The signal handler runs, so there is no pending signal anymore.
3. sigsuspend blocks indefinitely if no further signal arrives.

This is somewhat counter-intuitive and makes the function harder to use than necessary.
Desired Action Clarify whether implementations are required to provide atomicity in the described sense.
Tags No tags attached.
Attached Files

- Relationships

-  Notes
carlos (reporter)
2018-08-24 20:23

The same question should be raised for sigpause, which does an unmask and wait also.

- Issue History
Date Modified Username Field Change
2018-08-22 14:10 Florian Weimer New Issue
2018-08-22 14:10 Florian Weimer Name => Florian Weimer
2018-08-22 14:10 Florian Weimer Organization => Red Hat
2018-08-22 14:10 Florian Weimer Section => sigsuspend
2018-08-22 14:10 Florian Weimer Page Number => unknown
2018-08-22 14:10 Florian Weimer Line Number => unknown
2018-08-24 20:23 carlos Note Added: 0004091

Mantis 1.1.6[^]
Copyright © 2000 - 2008 Mantis Group
Powered by Mantis Bugtracker