Messages in this thread |  | | Date | Thu, 28 Nov 2013 19:13:04 +0100 | From | Oleg Nesterov <> | Subject | Re: [PATCH]: exec: avoid propagating PF_NO_SETAFFINITY into userspace child |
| |
On 11/28, Peter Zijlstra wrote: > > So in specific, you say that piping a core file into a usermode helper > is currently busted in pid-namespaces and that fixing that would indeed > introduce such pid-namespace awareness to the usermode-helper stuff?
Perhaps yes. People want to run the core handler in the namespace of the crashed task. And (perhaps) even dumping to the file should use per-namespace ->root/etc.
(and of course core_pattern[] should not be global, but this is another story).
And there are other potential users. Say, nfsd user-mode helpers wants to have the "right" mnt (at least) namespace.
But again, this all needs a lot of discussions. Even the API is not actually clear. As for implementation, we could probably even force ->child_reaper (the subnamespace's init) to fork/exec the usermode helper in this case. Or a namespace should explicitly create a thread which should do this.
Oleg.
|  |