In D55558#1271460, @ae wrote:

In D55558#1271208, @bms wrote:

Just my 2c: I think we should be discouraging users from implementing CARP at all.

But currently it is widely used by FreeBSD users. I personally have several hundreds of freebsd routers/firewalls that use CARP. And also we still have several patches to FreeBSD/bird/radvd to behave better.

In D55558#1271271, @kp wrote:

I haven't debugged it any depth (and won't be able to before early next week), but the new test case fails for me:

(kp@nut)  /usr/tests/sys/netinet % sudo kyua debug carp:vrrp_preempt                                                                                                               [14:00]
net.inet.carp.preempt: 0 -> 1
net.inet.carp.preempt: 0 -> 1
	vrrp: MASTER vrid 2 prio 10 interval 100
Files left in work directory after failure: created_interfaces.lst, created_jails.lst
ifconfig: interface epair0b does not exist
ifconfig: interface epair1b does not exist
carp:vrrp_preempt  ->  failed: preemption did not affect the second interface

In D55558#1271209, @kp wrote:

One small issue is that this patch claims to move these files. That's clearly unintentional. Perhaps an artefact of how it was uploaded?

Any news about comitting this fix ?

And how does a "-metal" (ie: Bare metal) Google Compute Engine behave with this patch?

For data, here is pmc while iperf3 limited to 1Gb/s in place of the expected 2.5Gb/s:

Detected on:

rge0@pci0:2:0:0:        class=0x020000 rev=0x05 hdr=0x00 vendor=0x10ec device=0x8125 subvendor=0x10ec subdevice=0
x0123
    vendor     = 'Realtek Semiconductor Co., Ltd.'
    device     = 'RTL8125 2.5GbE Controller'                                                                         class      = network
    subclass   = ethernet

Need to manually load the module (not auto-loaded when detected).

In D53413#1219714, @cperciva wrote:

Does this work with cross-builds? I suspect that WSTAGEDIR is getting set to "" and then you're setting PKG_ABI based on /usr/bin/uname in the host system.

Why are we using PKG_ABI with NOPKGBASE?

(adding context)

we are using this patch since months now, and it works great (because it solve the problem mentionned)

Rename MD_LEN to DEVICE_IO_LEN

In D52914#1209913, @asomers wrote:

The MD_LEN variable is now badly named. I think it's ok to leave its value the same. Most test cases don't require 1 MB of I/O. But you should at least change that name.

Added new version more simpler: Just increase the MD size

Increase the MD device size to support larger (non default) vfs.aio.max_buf_aio

In D48680#1201178, @vladlenpopolitov_list.ru wrote:

Why you do not use Weblate to store and process translations? It is very comfortable to track all changes. If you like, I could try to import this adoc file to French version of Weblate (make file for import and you can import it)? I am from Russian translators team.