Re: Undelete in user space

H. Peter Anvin (hpa@freya.yggdrasil.com)
20 Jun 1996 20:04:19 GMT


Followup to: <31C95F95.2087@et.tudelft.nl>
By author: Rogier Wolff <r.e.wolff@ET.TUDelft.NL>
In newsgroup: linux.dev.kernel
>=20
> Linus' opinion on "what goes into the kernel and what doesn't" is=20
> that what CAN be done in userspace, should go in userspace.=20
>=20

Not so. Linus has on several occations argued *against* moving a
certain facility to user space. Linus' opinion is that what can be
done WELL in user space, should go in user space.

Oh, by the way, for whomever was compaining about the lack of a "pure"
interface to the syscall, I suggest reading the rationale to POSIX.1,
which explains why the user shouldn't care about whether or not the
implementation in question uses a system call or a library call.

On Linux, for example, creat() is a library function that calls
SYS_open. On UNIX, it is a system call. Noone cares.

-hpa
--=20
PGP public key available - finger hpa@zytor.com
I don't work for Yggdrasil, but they sponsor the linux.* hierarchy.
"The earth is but one country, and mankind its citizens." -- Bah=E1'u=
'll=E1h
Just Say No to Morden * Save Babylon 5: http://www.babylon5.com/cmp/sup=
port/