package/keepalived: don't use --{en, dis}able-iptables
PATCH 1/1 package/keepalived: don't use --{en, dis}able-iptables
Commit 5fcee496df262fcc304ad22738441df35a960a67 wrongly replaced --{en,dis}able-libiptc by --{en,dis}able-iptables.
Indeed, this will raise the following build failure if --disable-iptables and --disable-ipset are both set:
configure: error: disable-libipset requires vrrp and iptables
The issue is that when iptables flag is set, it will first check for libiptc and then for ipset so setting --disable-iptables will disable libiptc and ipset (which is not what we wants)
As iptables flag is enabled by default, just drop
--{en,dis}able-iptables to keep it simple
Fixes:
- http://autobuild.buildroot.org/results/a1712b2cc3ad878e6876325ec7d4c434d0d9d11b
--- package/keepalived/keepalived.mk | 3 --- 1 file changed, 3 deletions(-)
package/keepalived/keepalived.mk +0 -3
@@ -41,9 +41,6 @@ endif ifeq ($(BR2_PACKAGE_IPTABLES),y) KEEPALIVED_DEPENDENCIES += iptables -KEEPALIVED_CONF_OPTS += --enable-iptables -else -KEEPALIVED_CONF_OPTS += --disable-iptables endif ifeq ($(BR2_PACKAGE_LIBNFTNL),y)
Signature · 1 line
2.27.0
PATCH 1/1 package/keepalived: don't use --{en, dis}able-iptables
Hello Fabrice,
Quoted 12 lines · On Tue, 14 Jul 2020 18:58:00 +0200 Fabrice Fontaine…
Commit
5fcee496df262fcc304ad22738441df35a960a67wrongly replaced --{en,dis}able-libiptc by --{en,dis}able-iptables.Indeed, this will raise the following build failure if --disable-iptables and --disable-ipset are both set:
configure: error: disable-libipset requires vrrp and iptables
The issue is that when iptables flag is set, it will first check for libiptc and then for ipset so setting --disable-iptables will disable libiptc and ipset (which is not what we wants)
Quoted 7 lines
As iptables flag is enabled by default, just drop
--{en,dis}able-iptables to keep it simpleFixes:
- http://autobuild.buildroot.org/results/a1712b2cc3ad878e6876325ec7d4c434d0d9d11bSigned-off-by Fabrice Fontaine <[email protected]>
I was not very convinced by this explanation, so I dug deeper into this issue. The configure.ac is far from being crystal clear, but I think I got to the bottom of this, and submitted what I consider a better patch:
https://patchwork.ozlabs.org/project/buildroot/patch/20200815195851.282863-1-thomas.petazzoni at bootlin.com/
The commit log contains an extensive description of how I came up with the alternate solution.
Thanks!
Thomas
Signature · 3 lines
Thomas Petazzoni, CTO, Bootlin Embedded Linux and Kernel engineering https://bootlin.com