[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: tmpfs status

From: Samuel Thibault
Subject: Re: tmpfs status
Date: Sat, 31 Mar 2012 14:31:32 +0200
User-agent: Mutt/1.5.21+34 (58baf7c9f32f) (2010-12-30)

Maksym Planeta, le Sat 31 Mar 2012 12:07:13 +0300, a écrit :
> >> > I don't see an "initial content" option in tmpfs. That would be very
> >> > useful, by e.g. providing a .tgz archive to be untarred at translator
> >> > creation.
> >> 
> >> Nice feature, but couldn't it be implemented just untarring archive to
> >> specified directory?
> >
> > I didn't actually mean that tmpfs has to do the untar itself, it'd be
> > for sure way more flexible that it's an external process just using the
> > tmpfs as its root FS.
> Under external process do you mean that when "initial content" option is
> specified tmpfs calls function "system ()", that extracts archive?

Or just by starting yet another process at boot, between tmpfs.static
and ld.so exec. About how to provide tar with the tarball, look in the
list archives for discussions about initial ramdisk support, in
particular the last discussion bits:

24 O L sept. 05 To bug-hurd@gnu (0,7K) Re: initrd support

> > I don't have the time to dig more, but I guess some statically linked
> > tar or some such in the multiboot sequence could manage to do the
> > work.
> But than why tar should be statically linked? 

Because before the content of / is unpacked, libraries are not


reply via email to

[Prev in Thread] Current Thread [Next in Thread]