Re: [PATCH bpf-next 02/16] bpf: array: move checks out of alloc function

2018-01-05 Thread Jakub Kicinski
On Fri, 5 Jan 2018 09:21:02 -0800, Alexei Starovoitov wrote:
> On Thu, Jan 04, 2018 at 10:09:17PM -0800, Jakub Kicinski wrote:
> > Use the new callback to perform allocation checks for array maps.
> > The fd maps don't need a special allocation callback, they only
> > need a special check callback.
> > 
> > Signed-off-by: Jakub Kicinski 
> > Reviewed-by: Quentin Monnet   
> 
> LGTM
> but I would like to hold on this set until we fix the fallout
> from cpu bug. See my patch: https://patchwork.ozlabs.org/patch/855911/
> If we merge your set to bpf-next today we'll have very nasty
> conflicts between bpf and bpf-next.
> I'd like to fix everything necessary for these CVEs in bpf tree first
> and get that merged all the way to bpf-next through Linus tree
> before proceeding with this refactoring.
> Ok?

Makes perfect sense, I wasn't sure what's the plan about the CVEs.

Thanks!


Re: [PATCH bpf-next 02/16] bpf: array: move checks out of alloc function

2018-01-05 Thread Alexei Starovoitov
On Thu, Jan 04, 2018 at 10:09:17PM -0800, Jakub Kicinski wrote:
> Use the new callback to perform allocation checks for array maps.
> The fd maps don't need a special allocation callback, they only
> need a special check callback.
> 
> Signed-off-by: Jakub Kicinski 
> Reviewed-by: Quentin Monnet 

LGTM
but I would like to hold on this set until we fix the fallout
from cpu bug. See my patch: https://patchwork.ozlabs.org/patch/855911/
If we merge your set to bpf-next today we'll have very nasty
conflicts between bpf and bpf-next.
I'd like to fix everything necessary for these CVEs in bpf tree first
and get that merged all the way to bpf-next through Linus tree
before proceeding with this refactoring.
Ok?