On 2015-09-25 17:09, Paul Jakma wrote:
On Fri, 25 Sep 2015, Kaloyan Kovachev wrote:

Hello, both calls are inside ifdef checking for SUPPORT_REALMS, which in turn is only possible if opsys is gnu-linux. I don't understand how it can be enabled at all?! Is the patch changed?

Ah, but you changed rib_add_ipv4? That's a key internal zebra API
which is also called by the BSD/Solaris kernel_socket.c code - which
wasn't updated. ;)


It was in the original patch, but it seems it was modified afterwards or an older version of the patch was used. See https://github.com/KaloNK/quagga/blob/realms_patchset/zebra/kernel_socket.c and other calls to rib_add_ipv4 - they are all modified to include one more parameter (as 0 ) for the realms.

When i remove conditional compilation from the rest of the code (except rt_name.c), will prepare a new diff with all the changes as single file.

As for the different approach of the patch, i will need much more time learning the code ... not sure if it is reasonable at all to include this one then, but also not sure if i will be able to do it the other way either.

FreeBSD 10: (see https://ci1.netdef.org/browse/QUAGGA-QMASTER1-CI003BUILD-1 [2] ) kernel_socket.c:966:66: error: too few arguments to function call, expected 12, have 11
NULL, 0, VRF_DEFAULT, 0, 0, 0, SAFI_UNICAST);
^
./rib.h:506:1: note: 'rib_add_ipv4' declared here
extern int rib_add_ipv4 (int type, int flags, struct prefix_ipv4 *p,
^
1 error generated.


regards,

_______________________________________________
Quagga-dev mailing list
[email protected]
https://lists.quagga.net/mailman/listinfo/quagga-dev

Reply via email to