Re: [PATCH] ppc64 utmp changes

[Jakub Jelinek <jakub at redhat dot com>]

On Wed, Oct 02, 2002 at 04:47:55PM +0200, Martin Schwidefsky wrote:
> 
> > I haven't changed s390x, but IMHO it is better to recompile the affected
> > apps now in the /usr/lib -> /usr/lib64 transition which changes most
> > of the things anyway, than to suffer forever.
> 
> I came to the conclusion that s390x will need the __WORDSIZE_COMPAT32
> version
> of utmp.h/utmpx.h as well. Fact is there are userland programs that access

Ok.

> /var/log/wtmp directly. To get this kind of program going under the 31 bit
> emulation the system /var/log/wtmp file needs to have the same format for
> 31 bit and 64 bit. Period.
> This leaves the question about "old" 64 bit programs that rely on the
> existing
> format of /var/log/wtmp. Apps that directly access /var/log/wtmp with the
> existing format can't be helped if we want to achive coexistence (the old
> question about how you want to die...). Apps that use the glibc to get/pass
> information from/to wtmp could be helped with versioning. Now the question:
> are there any functions other then getutent, getutid, getutline, pututline,
> getutxent, getutxid, getutxline, pututxline and login that get/pass a
> structure
> from utmp.h or utmpx.h?

login, updwtmp, getutent, getutid, getutline, pututline, getutent_r,
getutid_r, getutline_r, getutxent, getutxid, getutxline, pututxline,
updwtmpx, getutmp, getutmpx

at least.

	Jakub