linux compat kernel version (was: CVS commit: syssrc/sys/compat/linux)

[Bernd Ernesti <netbsd%arresum.inka.de@localhost>]

On Fri, Feb 15, 2002 at 06:48:04PM +0200, Christos Zoulas wrote:
> 
> Module Name:  syssrc
> Committed By: christos
> Date:         Fri Feb 15 16:48:03 UTC 2002
> 
> Modified Files:
[..]
>           linux_file64.c linux_machdep.h linux_misc.c linux_misc.h
>           linux_mmap.h linux_oldmmap.c linux_sig_notalpha.c linux_sigaction.c
>           linux_signal.c linux_signal.h linux_types.h
> 
> Log Message:
> - add a real linux_sys___sysctl because the mib numbers are different.
>   implement only the kernel version getting entries.
> - make the signal conversion functions consistent; dest is arg 1.
> - make linux_fakedev deal with block and char devices differently.
> - add linux_sys_ugetrlimit, and flesh out the regular get/setrlimit so
>   that they work properly.
> - add linux_mmap2 [untested].
> - bump kernel to 2.4.18, and make the date be valentine's day :-)
                   ^^^^^^
Is that a good idea?

I thought that some programs check for the kernel version and want to
use a new glibc, which has some support which we don't have and that
fails on NetBSD. Is that problem gone?

See rev 1.83 and 1.82 of linux_misc.c:

revision 1.83
date: 2001/01/05 15:31:15;  author: fvdl;  state: Exp;  lines: +2 -2
branches:  1.83.2;
Set Linux version to 2.0.38 for now, until the clone + "real time signal"
problems are fixed.

revision 1.82
date: 2000/12/29 21:07:16;  author: fvdl;  state: Exp;  lines: +8 -17
Hardcode what the linux uname() calls return (Linux 2.2.5 is the current
value). A few major applications check for this (like VMware), and
there is no other way to trick them. Needs a more generic solution
eventually.

Bernd