Timo Teras

fix builds with gcc10

fix incomplete conversion to system libev

Prefer /dev/gpiomem in bcm2835gpio

No. The strongSwan patches at https://git.alpinelinux.org/cgit/user/tteras/strongswan/log/?h=tteras-release will enable writing opennhrp-script that would work mostly (with some restrictions). The exercise to do the script is left for the reader. I strongly recommend using quagga or frr nrhpd because it solves several issues opennhrp had and superior in almost any aspect. See also: http://git.savannah.gnu.org/cgit/quagga.git/tree/nhrpd/README.nhrpd

Quagga/NHRP and frr/nhrpd supercede opennhrp and integrate with strongswan. For further information see: - https://wiki.alpinelinux.org/wiki/Dynamic_Multipoint_VPN_(DMVPN)_Phase_3_with_Quagga_NHRPd - https://git.alpinelinux.org/cgit/dmvpn-tools/about/

To start off I recommend using Quagga/NHRP or FRR/NHRP if possible. I am not sure how IPsec is configured, but that likely is the cause. This is because NHRP does not detect liveliness but depends on IPsec to do it. If IPsec is not in use, this would cause the issue. If IPsec is in use, the racoon's phase1_dead hook is not likely configured, or the script is not working. On ipsec-tools/opennhrp the dead peer detection works so that ipsec-tools executes a dead peer hook which should be a script executing...

To start off I recommend using Quagga/NHRP or FRR/NHRP if possible. I am not sure how IPsec is configured, but that likely is the cause. This is because NHRP does not detect liveliness but depends on IPsec to do it. If IPsec is not in use, this would cause the issue. If IPsec is in use, the racoon's phase1_dead hook is not likely configured, or the script is working. On ipsec-tools/opennhrp the dead peer detection works so that ipsec-tools executes a dead peer hook which should be a script executing...

Very good. Unfortunately the kernel bugs cannot be worked around. The only solution...