|
Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com |
Subject: Re: chroot vs namespaces (Was: Sys(k)logd running as non-root)
From: Solar Designer (solar
false.com)Date: Mon Sep 25 2000 - 20:49:14 CDT
- Next message: Solar Designer: "Re: Some possible format string errors"
- Previous message: Olaf Kirch: "Re: yptools (possible) buffer overflow"
- In reply to: Mitchell Blank Jr: "Re: chroot vs namespaces (Was: Sys(k)logd running as non-root)"
- Next in thread: Andreas Schuldei: "Re: chroot vs namespaces (Was: Sys(k)logd running as non-root)"
- Reply: Solar Designer: "Re: chroot vs namespaces (Was: Sys(k)logd running as non-root)"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
> (nit-pick) since the "locked pid" is just an attribute of a process, it
> should be set using prctl(), not a new syscall.
OK, so it should be even easier to get this into the kernel.
> Instead, you should keep the processes which are using lockpid in a linked
> list - then at fork-time you only need to search those, not for_each_task().
I was planning this (actually, a hash) for future updates of the
code. The patch adding lockpid should be simple to be accepted. ;-)
Signed,
Solar Designer
- Next message: Solar Designer: "Re: Some possible format string errors"
- Previous message: Olaf Kirch: "Re: yptools (possible) buffer overflow"
- In reply to: Mitchell Blank Jr: "Re: chroot vs namespaces (Was: Sys(k)logd running as non-root)"
- Next in thread: Andreas Schuldei: "Re: chroot vs namespaces (Was: Sys(k)logd running as non-root)"
- Reply: Solar Designer: "Re: chroot vs namespaces (Was: Sys(k)logd running as non-root)"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]