Re: [PATCH v3] LoongArch: Add support to clone a time namespace

[Youling Tang]

Hi, Tiezhu

On 05/31/2023 06:04 PM, Tiezhu Yang wrote:
> On 05/31/2023 05:21 PM, Thomas Gleixner wrote:
> > On Wed, May 31 2023 at 15:56, Tiezhu Yang wrote:
> > > When execute the following command to test clone3 on LoongArch:
> > >
> > >   # cd tools/testing/selftests/clone3 && make && ./clone3
> > >
> > > we can see the following error info:
> > >
> > >   # [5719] Trying clone3() with flags 0x80 (size 0)
> > >   # Invalid argument - Failed to create new process
> > >   # [5719] clone3() with flags says: -22 expected 0
> > >   not ok 18 [5719] Result (-22) is different than expected (0)
> > >
> > > This is because if CONFIG_TIME_NS is not set, but the flag
> > > CLONE_NEWTIME (0x80) is used to clone a time namespace, it
> > > will return -EINVAL in copy_time_ns().
> >
> > Which means that test case is broken.
> >
> > > With this patch, the following failed test can be fixed too:
> >
> > You cannot fix a broken test case by enabling functionality which
> > is not guaranteed to be available on all platforms.
> >
> > Adding time name spaces on loongarch is a completely separate issue and
> > has absolutely nothing to do with broken test cases.
> >
> > They are still broken when CONFIG_TIME_NS is disabled in Kconfig, no?
> >
> > So this wants to be two separate things:
> >
> >    1) Fix the broken test cases so they handle CONFIG_TIME_NS=n
> >       situations correctly
> >
> >    2) Add support for TIME_NS to loongarch

With this patch, we should also modify the stack_top() implementation,
because the vvar size is not PAGE_SIZE at this time.

'maps' before the patch,
 # sudo cat /proc/1/maps
 ...
 7ffffbec8000-7ffffbeec000 rw-p 00000000 00:00 0 
  [stack]
 7ffffc7d8000-7ffffc7dc000 r--p 00000000 00:00 0 
  [vvar]
 7ffffc7dc000-7ffffc7e0000 r-xp 00000000 00:00 0 
  [vdso]

STACK_TOP = TASK_SIZE - PAGE_ALIGN(current->thread.vdso->size) - VVAR_SIZE

Thanks,
Youling.

> Thanks for your reply and detailed suggestions,
> I will do it in the next version.
>
> Thanks,
> Tiezhu