On Thu, Sep 29, 2011 at 12:19 PM, Volker Armin Hemmann <volkerar...@googlemail.com> wrote: > Am Mittwoch 28 September 2011, 17:15:34 schrieb Grant Edwards: > >> >> Regardless, my point was that Linus's statement that it's unacceptable >> to break things seemed rather disingenuous given the API churn that >> Linux has compared with the BSD kernels. > > Linux has zero userland visible API 'churn'. > > You can't have less than zero.
Uh, that can't be right. Largely, libc masks things. Several kernel options explicitly state in their description that they require new-enough versions of this or that userland tool to function properly. Randomizing module base addresses is one of those, IIRC. Some things related to sysfs. sysfs itself. I think some network filesystems. modutils. If there's no API churn, it should be pretty trivial to run a current userland on top of, e.g. 2.6.0-pre1, or even 2.6.0. I also STR 2.6.9 being a common pin point where a bunch of userland tools required that-or-newer. And that's ignoring dropping things like A.OUT support. I'm not arguing whether or not it's reasonable (it almost certainly is), but there certainly is churn. -- :wq
