From: Herbert Poetzl (herbert_at_13thfloor.at)
Date: Sun 30 Nov 2003 - 15:16:07 GMT
On Mon, Dec 01, 2003 at 12:43:45AM +1100, Charles Dale wrote:
> 1. include/linux/sys.h: vanilla 2.4.21 had
> #define NR_syscalls 270
> but RHEL 2.4.21 has
> #define NR_syscalls 271
> So should I make NR_syscalls now 274 or 275?
it won't matter, but I would keep the 'default'
value used for vserver ...
(by the way, you have to change this for all
architectures, if RHEL did this, and in any
case I would advise to use the split patches,
as they group by functionality ...)
> 2. include/linux/sched.h: any ideas what to do with this hunk (following)?
> There's nothing even vaguely like it that I can see.
> *** 956,961 ****
> --- 968,975 ----
> + if (p->s_info)
> + p->s_info->nr_threads--;
how do you 'try' to apply the patches?
I assume that a change in whitespace is the
reason, just try with -l option for patch
> 3. forget_original_parent in exit.c seems to have changed completely - looks
> like the vanilla 2.4.21 used a read_lock while the RH one uses
> list_for_each_safe which I assume locks as it goes. But that means there's
> nowhere left to put the additional bit from the vserver patch (not sure what
> it does exactly, it tries to find a new parent for vchild_reaper I think).
> Boy I'm tired, that's way too much kernel hacking in one day for me. 10
> rejects down, 20 to go... Any help appreciated.
as I said, if you need help, let me know ...
by the way, where is the 'RHEL' aptch for 2.4.21
or even better 2.4.22?
> BTW, there are 266 patches applied inside the RHEL3 kernel SRPM... Grr.
> Have a good day =)
> Vserver mailing list
Vserver mailing list