Bug 833105

Summary:	RFE: provide SIGSTOP way to signal READY state
Product:	[Fedora] Fedora	Reporter:	Enrico Scholz <rh-bugzilla>
Component:	systemd	Assignee:	systemd-maint
Status:	CLOSED WONTFIX	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	17	CC:	johannbg, lnykryn, lpoetter, metherid, msekleta, notting, plautrba, systemd-maint
Target Milestone:	---
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2012-06-20 12:27:53 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Enrico Scholz 2012-06-18 15:07:28 UTC

Description of problem:

It would be nice when systemd supports the upstart like SIGSTOP method to signal when a daemon is ready.  E.g.:

1. .service contains 'Type=sigstop' or similar entry and executes daemon with a special cmdline option

2. daemon initializes itself (e.g. goes into chroot, drops permissions)

3. daemon calls 'raise(SIGSTOP)'

4. systemd catches the SIGSTOP, sends SIGCONT and marks the service as ready


Although there exists already a more powerful sd_notify(3) function, this is often not applicable because:

* it is complicated (extra sources + ./configure checks  vs. a simple 'raise(SIGSTOP)')

* does not work when daemon goes into a chroot() at step 2

* requires additional SELinux permissions for working on the socket

Comment 1 Lennart Poettering 2012-06-20 12:27:53 UTC

SIGSTOP is a mechanism for stopping processes, not for notifying service readiness. We shouldn't attempt to overload OS functionality like this, as SIGSTOP might happen for it's real purpose too.

I also fail to see in which way:

#include <signal.h>
...
raise(SIGSTOP);
...

was any simpler than this:

#include <systemd/sd-daemon.h>
...
sd_notify("READY=1");
...

And people can just use the .pc file to add libsystemd-daemon to their build, so that's dead easy.

The chroot() issue is a real issue, but easily fixed, by simply using an abstract namespace socket for this, which the spec explicitly allows, and which we actually used before. I have noq added a todo list item to revert back to using an abstract namespace socket for this.

Another big problem with raise(SIGSTOP) is that if done on an init system that can't handle it will result in a freeze. OTOH sd_notify() handles this gracefully and just becomes a NOP.

But anyway, we really shouldn't overload OS functionality like this anyway. SIGSTOP has a useful meaning already, and use for anything else would conflict with this.