Re: [vserver] linux 3.10.(4[789]|50) and patch-3.10.43-vs2.3.6.8.diff

On Fri, 18 Jul 2014 00:02:24 -0500
Corey Wright <undefined@pobox.com> wrote:

> On Wed, 9 Jul 2014 23:43:24 -0500
> Corey Wright <undefined@pobox.com> wrote:
>
> > On Mon, 7 Jul 2014 16:09:21 -0500
> > Corey Wright <undefined@pobox.com> wrote:
> >
> > > applying patch-3.10.43-vs2.3.6.8.diff to linux 3.10.47 fails.
> >
> > and with linux 3.10.48.
>
> and with linux 3.10.49

and with linux 3.10.50

> > > the upstream change of calling syscall_tracepoint_update() within copy_process
> > > () [1] invalidates the patch context and causes the patch hunk to fail:
> > >
> > > patching file kernel/fork.c
> > > Hunk #9 FAILED at 1495.
> > > 1 out of 9 hunks FAILED -- saving rejects to file kernel/fork.c.rej
> > >
> > > [1]
> > > http://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/kernel/fork.c?h=linux-3.10.y&id=e6bc60b8fb412b38db86bc1351f2ce40ad31d0e0
> > >
> > > the attached patch adjusts the patch context and fixes Makefile's
> > > EXTRAVERSION, too.
> >
> > the previously attached patch adjusts the patch context but does not fix the
> > Makefile's EXTRAVERSION for 3.10.48.
> >
> > > from my novice kernel analysis it doesn't appear to
> > > matter which comes first, the call to syscall_tracepoint_update() or the
> > > failed hunk (assuming CONFIG_TRACEPOINTS & CONFIG_HAVE_SYSCALL_TRACEPOINTS
> > > are defined, otherwise it for sure doesn't matter as syscall_tracepoint_update
> > > () is an empty function and should get optimized out; see
> > > include/trace/syscall.h), but verification would be appreciated (and i placed
> > > syscall_tracepoint_update() before the failed hunk).
> >
> > verification still appreciated (though i've been running 3.10.47 with the
> > previously attached patch in production for 12 hours now).
>
> i've appreciated confirmation from Florian and Guy-.
>
> > > instructions:
> > > 1. acquire linux-3.10.47
> > > 2. apply patch-3.10.43-vs2.3.6.8.diff
> > > 3. apply the attached patch-3.10.43-47-vs2.3.6.8.diff
> >
> > 1. acquire linux-3.10.48
> > 2. apply patch-3.10.43-vs2.3.6.8.diff
> > 3. apply the attached patch-3.10.43-47-vs2.3.6.8.diff
> > 4. manually fix the Makefile reject (EXTRAVERSION)
>
> 1. acquire linux-3.10.49
> 2. apply patch-3.10.43-vs2.3.6.8.diff
> 3. apply the attached patch-3.10.43-47-vs2.3.6.8.diff
> 4. manually fix the Makefile reject (EXTRAVERSION)

1. acquire linux-3.10.50
2. apply patch-3.10.43-vs2.3.6.8.diff
3. apply the previously attached patch-3.10.43-47-vs2.3.6.8.diff
4. manually fix the Makefile reject (EXTRAVERSION)

> > > i've successfully built the resulting kernel and tested it in virtualbox
> > > against testme.sh and testfs.sh and started a vserver.
> >
> > same for linux 3.10.48-vs2.3.6.8 (with delta-proc-feat01.diff).
>
> same for linux 3.10.49-vs2.3.6.8 (with delta-proc-feat01.diff).

i'm currently unable to test the resulting kernel due to logistical issues
(*not* due to technical issues with the new kernel). i should have the kernel
tested by the end of this week and will provide feedback then. any feedback
before then is appreciated.

corey

--
undefined@pobox.com
> corey
> --
> undefined@pobox.com
> 
> > > corey
> > > --
> > > undefined@pobox.com