Re: netfilter: xt_bpf: Fix XT_BPF_MODE_FD_PINNED mode of 'xt_bpf_info_v1'

On 12/02/2017 07:48 PM, Al Viro wrote:

On Fri, Dec 01, 2017 at 09:47:00PM +0100, Daniel Borkmann wrote:

quoted

quoted

Might want to replace security_path_mknod() with something saner, while we are
at it.

Objections?

No, thanks for looking into this, and sorry for this fugly hack! :( Not
that this doesn't make it any better, but I think back then I took it
over from mqueue implementation ... should have known better and looking
into making this generic instead, sigh. The above looks good to me, so
no objections from my side and thanks for working on it!

quoted

PS: mqueue.c would also benefit from such primitive - do_create() there would
simply pass attr as callback's argument into vfs_mkobj(), with callback being
the guts of mqueue_create()...

OK...  See vfs.git#untested.mkobj; it really needs testing, though - mq_open(2)
passes LTP tests, but that's not saying much, and BPF side is completely
untested.

I pulled vfs.git#untested.mkobj into my local tree and ran tests for both
progs and maps on it, all went fine and the patch looks good to me.

For 'bpf_obj_do_pin(): switch to vfs_mkobj(), quit abusing ->mknod()' when
you push the fix to Linus, feel free to add:

Acked-by: Daniel Borkmann <daniel@iogearbox.net>

Thanks for your help, Al!