Re: [Xen-devel] [PATCH 2/2] xen/privcmd: add PRIVCMD_MMAPBATCH_V2 ioctl

[Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>]

On Thu, Sep 06, 2012 at 09:41:44AM -0400, Andres Lagar-Cavilla wrote:
> On Sep 5, 2012, at 1:40 PM, Konrad Rzeszutek Wilk wrote:
> 
> > On Wed, Sep 05, 2012 at 01:09:32PM -0400, Andres Lagar-Cavilla wrote:
> >> Super. To which branch are you applying these now? (n00b question but have 
> >> to ask)
> > 
> > They will show up on stable/for-linus-3.7 once the overnight tests pass.
> 
> I would be very surprised if this passed your nighties. This is because the 
> following hunk is necessary:

It did :-) I guess the Xen 4.1 isn't using this that much.
> 
> diff --git a/drivers/xen/privcmd.c b/drivers/xen/privcmd.c
> index 5386f20..e4dfa3b 100644
> --- a/drivers/xen/privcmd.c
> +++ b/drivers/xen/privcmd.c
> @@ -389,7 +389,7 @@ static long privcmd_ioctl_mmap_batch(void __user *udata, 
> int version)
>                 state.err      = err_array;
>                 ret = traverse_pages(m.num, sizeof(xen_pfn_t),
>                                          &pagelist, mmap_return_errors_v1, 
> &state);
> -       } else
> +       } else if (version == 2)
>                 ret = __copy_to_user(m.err, err_array, m.num * sizeof(int));
>  
>      /* If we have not had any EFAULT-like global errors then set the global
> 
> I can either resubmit the original patch with this squashed in, or send this 
> stand-alone. Nightlies may have passed if your libxc never exercises v1 in 
> favor of v2. But this is broken for v1 as it will unconditionally attempt a 
> copy to user on a NULL target and this set rc to EFAULT.
> 

Send it stand alone pls.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel