On Tue, Jun 12, 2018 at 12:10:36PM +0000, Aaron Komisar wrote: > > On 6/12/2018 4:50, Chi-hsien Lin wrote: > >On 06/11/2018 8:25, Johannes Berg wrote: > >> On Mon, 2018-06-11 at 05:29 -0500, Wan-Hua Chang wrote: > >>> Hi, > >>> > >>> I see the patch causes compiling error becasue family->family is not > >>> found when LINUX_VERSION >= 3.13 > >>> > >>> backports/backport/backport-include/genetlink.h > >>> struct backport_genl_fmaily is only defined when LINUX_VERSION < 3.13 > >>> > > I built backports-output for recent kernels (4.4.92, for example) and didn't get any compilation errors. > Which version are you using? I am using target kernel version 4.1.15. In this case, I can see compiling error that family->family is not found. > > >>> > >>> And in backports/backport/compat/backport-3.13.c line 127 > >>> > >>> family->id = family->family.id > >>> > >>> it looks like this line has done what the patch want to do. Is the > >>> comprehension correct? > > This line does update family id, but it gets **copy**->family. > copy->family.id in backport-4.12.c is changed, but family->family.id remains unchanged. > > >> > >> I guess we have to put an ifdef around this ... > >> > >> Aaron? > > > >Hi Aaron, > > > >As Wan-Hua mentioned, it looks like backport-3.13.c line 127 syncs back > >the id for backport_genl_family structure for kernel < 3.13 already. > > > >Could you please share more details about the failure you saw without > >this change? What's the kernel version you used? > > > > I am using target kernel version 3.10.104. > The problem I saw was that a driver using family id after calling genl_register_family would always get 0 instead of the actual value. > > > > >Regards, > >Chi-hsien Lin > > > >> > >> johannes > > . > > > > Aaron > -- To unsubscribe from this list: send the line "unsubscribe backports" in