On Thu, 2019-05-02 at 15:10 +0200, Michal Kubecek wrote: > On Thu, May 02, 2019 at 02:51:33PM +0200, Johannes Berg wrote: > > On Thu, 2019-05-02 at 12:48 +0000, Michal Kubecek wrote: > > > Unlike do requests, dump genetlink requests now perform strict validation > > > by default even if the genetlink family does not set policy and maxtype > > > because it does validation and parsing on its own (e.g. because it wants > > > to > > > allow different message format for different commands). While the null > > > policy will be ignored, maxtype (which would be zero) is still checked so > > > that any attribute will fail validation. > > > > > > The solution is to only call __nla_validate() from genl_family_rcv_msg() > > > if family->maxtype is set. > > > > D'oh. Which family was it that you found this on? I checked only ones > > with policy I guess. > > It was with my ethtool netlink series (still work in progress).
Then you should probably *have* a policy to get all the other goodies like automatic policy export (once I repost those patches) johannes
