netdev - Re: [PATCH net-next v2] Do not invoke addrconf_verify

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-Id: <20241218065744.4063286-1-gnaaman@drivenets.com>
Date: Wed, 18 Dec 2024 06:57:44 +0000
From: Gilad Naaman <gnaaman@...venets.com>
To: pabeni@...hat.com
Cc: davem@...emloft.net,
	dsahern@...nel.org,
	edumazet@...gle.com,
	gnaaman@...venets.com,
	horms@...nel.org,
	kuba@...nel.org,
	netdev@...r.kernel.org
Subject: Re: [PATCH net-next v2] Do not invoke addrconf_verify_rtnl unnecessarily

> >>> But calling "verify" immediately will result in a performance hit when
> >>> deleting many addresses.
> >>
> >> Since this is about (control plane) performances, please include the
> >> relevant test details (or even better, please add a small/fast self-test
> >> covering the use-case).
> > 
> > Is it common to add scale-test to selftests?
> 
> AFAIK, not common at all. Note that the argument "so self-test for this
> kind of thing" is actually a very good argument to add a self-tests.

Sorry, I didn't mean to present it as an argument.
I wanted a clarification if you inteded me to write a self-test to check
functionality, or performance.

And if the latter, what are we going to set as the pass-criteria for the test,
seeing that the test may run on variety of hardwares/VMs.

No objection to writing it, of course.

> > (In our original bug the VLANs were deleted, it is just easier to perf
> > one iproute command if it's a flush)
> 
> Nice, so you already have the test infra ready :)
Wouldn't call it a test infra ^^', I have a script that calls `ip` in a loop.

Thanks,
Gilad