Re: git: b44715d91b98 - main - net/wireguard-kmod: Update to 0.0.20220610

From: <decke_at_freebsd.org>
Date: Sat, 11 Jun 2022 18:30:24 UTC
 ---- On Sat, 11 Jun 2022 13:35:25 +0200 Guido Falsi <mad@madpilot.net> wrote ----
 > On 11/06/22 12:58, Bernhard Froehlich wrote:
 > > The branch main has been updated by decke:
 > > 
 > > URL: https://cgit.FreeBSD.org/ports/commit/?id=b44715d91b98beefed44294caf090243e3ee1214
 > > 
 > > commit b44715d91b98beefed44294caf090243e3ee1214
 > > Author:     Bernhard Froehlich <decke@FreeBSD.org>
 > > AuthorDate: 2022-06-11 10:52:01 +0000
 > > Commit:     Bernhard Froehlich <decke@FreeBSD.org>
 > > CommitDate: 2022-06-11 10:58:13 +0000
 > > 
 > >      net/wireguard-kmod: Update to 0.0.20220610
 > >      
 > >      == Changes ==
 > >      
 > >      * if_wg: wg_queue_len: remove locking
 > >      * if_wg: wg_queue_delist_staged: use more standard STAILQ_CONCAT
 > >      * if_wg: wgc_get/set: use M_WAITOK with malloc()
 > >      * if_wg: wg_clone_create: Use M_WAITOK with malloc
 > >      * wg_cookie: ratelimit_init: use callout_init_mtx
 > >      * if_wg: wg_mbuf_reset: don't free send tags
 > >      * if_wg: wg_module_init: clean up more if the self tests fail
 > >      
 > >      Numerous cleanups.
 > >      
 > >      * if_wg: avoid scheduling excessive tasks for encryption/decryption
 > >      
 > >      A nice performance improvement - John reports double on iperf3 TCP throughput.
 > >      
 > >      * crypto: return an error code from mbuf crypt routines
 > >      
 > >      Preparation for the OCF work John's been working on.
 > >      
 > >      * if_wg: account for input function returning a boolean
 > >      * if_wg: account for added argument to sbcreatecontrol
 > 
 > Hi!
 > 
 > Thanks for the update and all the work it entails.
 > 
 > Unluckily this is still failing on head.
 > 
 > Likely cause seems to be that upstream has a mistake or typo in compat.h:
 > 
 > #if __FreeBSD_version < 1400517
 > 
 > 
 > I think that should be 1400057
 > 
 > I can report it myself upstream later, but I guess you have a more 
 > direct channel.
 > 
 > Thanks!
 > 
 > -- 
 > Guido Falsi <mad@madpilot.net>
 > 

Thanks for the heads up! It's fixed in the portstree and patch has been send
upstream.

https://cgit.freebsd.org/ports/commit/?id=7ba61726af7c05faf63b04279084f95d1937e27e