Re: Consensus around kernel syscall wrappers?

["Carlos O'Donell" <carlos at redhat dot com>]

On 08/13/2015 04:29 PM, Joseph Myers wrote:
> On Thu, 13 Aug 2015, Carlos O'Donell wrote:
> 
>> Have we achieved consensus on the kernel syscall wrappers?
> 
> I don't think there was consensus on any cases other than those where the 
> syscall is added to the OS-independent GNU API.  In 
> <https://sourceware.org/ml/libc-alpha/2015-05/msg00764.html> I suggested 
> treating Linux as a source of APIs like BSD and SysV were in the past but 
> don't think that part of the discussion got any further.
> 

If I had to summarize the present situation it would be:

(a) All Linux syscalls that are considered useful for the OS-independent
    GNU API should be added immediately.

(b) All Linux syscalls that are not considered useful for the OS-independent
    GNU API should be added to an AS_NEEDED library named appropriately
    e.g. libinux-syscalls.so.N, where this library is part of the implementation
    and must be updated in lock-step with the implementation to ensure that
    cancellation and other things are kept synchronized. It should not be a static
    library to avoid security issues and to allow it to be updated.

Nobody opposed (b), and (a) was universally accepted.

Therefore I think we have a path forward.

(1) Add sensible syscalls to libc.so.6.

(2) Add a new library and put everything else in there.

    - Following the WIP consensus about new headers to make
      it clear the new functions are in another library.

    - Minimally talk to Michael Kerrisk to get it documented
      how to use these new features.

Cheers,
Carlos.