Summary: | net-misc/connman-? with net-firewall/iptables-1.6.0 - ? | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | dsreyes1014 |
Component: | Current packages | Assignee: | nvinson234 |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | dlan, proxy-maint, Sander.Sweers |
Priority: | Normal | ||
Version: | autobuilds | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: | connman-1.31-iptables-1.6.patch |
Description
dsreyes1014
2016-01-28 05:59:34 UTC
Looked at their git log and it needs a patch [1] to fix the crash. I can confirm it no longer crashes. [1] http://git.kernel.org/cgit/network/connman/connman.git/commit/?id=acea08a0e4234a4c1a87bedc087c73ff36de0c7b Sander Sweers do you wish to consider being a substitute proxy maintainer? After 6 weeks, nothing form the assignee here. See https://wiki.gentoo.org/wiki/Project:Proxy_Maintainers (In reply to Ian Delaney from comment #2) > Sander Sweers do you wish to consider being a substitute proxy maintainer? If I used it fulltime I would consider it but I only installed to implement rfkill (powered) handling through connman for blueman. > After 6 weeks, nothing form the assignee here. I'd be happy to post an ebuild patch and the patch to this bug if you want :) Very well, but you don't want to be set as a maintainer. I sha;; settle with attributing you with the contribution then Very well, but you don't want to be set as a maintainer. I shall settle with attributing you with the contribution then (In reply to Sander Sweers from comment #3) > (In reply to Ian Delaney from comment #2) > > Sander Sweers do you wish to consider being a substitute proxy maintainer? > > If I used it fulltime I would consider it but I only installed to implement > rfkill (powered) handling through connman for blueman. > If you're sure you don't want to be the maintainer, I'll pick up the package. > > After 6 weeks, nothing form the assignee here. > > I'd be happy to post an ebuild patch and the patch to this bug if you want :) Assuming I become the maintainer, I'm OK with accepting the patch from you or writing it my own. As long as this issue is fixed, I don't care who gets the credit. Created attachment 428834 [details, diff]
connman-1.31-iptables-1.6.patch
This is the patch that add support for iptables 1.6
--- connman-1.31.ebuild 2016-03-23 14:56:44.586878247 +0100
+++ connman-1.31-r1.ebuild 2016-03-23 15:25:11.972796783 +0100
@@ -31,7 +31,10 @@
DEPEND="${RDEPEND}
>=sys-kernel/linux-headers-2.6.39"
-PATCHES=( "${FILESDIR}/${PN}-1.16-execinfo-assumptions.patch" )
+PATCHES=(
+ "${FILESDIR}/${PN}-1.16-execinfo-assumptions.patch"
+ "${FILESDIR}/${PN}-1.31-iptables-1.6.patch"
+)
src_configure() {
econf \
At the moment connman fails to build with or without the patch. Not sure why now and not before but it also seems related to the new iptables 1.6. log, https://dpaste.de/3ypv No, it's 577584. I have a patch for it. (In reply to Sander Sweers from comment #8) > At the moment connman fails to build with or without the patch. Not sure why > now and not before but it also seems related to the new iptables 1.6. log, > https://dpaste.de/3ypv I completely overlooked that you used a pastebin, so my last comment won't make any sense now. For anyone looking at this bug, the build errors that the above pastebin showed (https://dpaste.de/3yp) were the same errors as reported via bug 577584. I'm not sure what the root trigger for 577584 is exactly, but it's clear it's an incompatibility between glibc and the Linux kernel. commit 9df5a7e55212f98ea227493da7ff244aebca4e65 Author: Nicholas Vinson <nvinson234@gmail.com> Date: Sat Mar 26 15:41:16 2016 -0700 net-misc/connman: Fix iptables and xtables issues Fixes issues with iptables, xtables, and glibc when building and running connan. Specifically, the patch, connman-1.31-iptables-1.6.0.patch, fixes the icompatibilities between iptables and connman (due to an ABI break in iptables-1.6.0). The patch, connman-1.31-xtables.patch, uses some preprocessor tricks to route around incompatibilies seen between newer Linux kernels and glibc when both net/if.h and linu/if.h are included in the same source. Gentoo-bug: #573174, #577584 All well here now, thanks. |