netdev - Re: [PATCH v4 net-next 2/2] net: introduce budget

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 for Android: free password hash cracker in your pocket

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

Date:   Fri, 31 Mar 2023 10:33:47 +0800
From:   Jason Xing <kerneljasonxing@...il.com>
To:     Jakub Kicinski <kuba@...nel.org>
Cc:     jbrouer@...hat.com, davem@...emloft.net, edumazet@...gle.com,
        pabeni@...hat.com, ast@...nel.org, daniel@...earbox.net,
        hawk@...nel.org, john.fastabend@...il.com,
        stephen@...workplumber.org, simon.horman@...igine.com,
        sinquersw@...il.com, bpf@...r.kernel.org, netdev@...r.kernel.org,
        Jason Xing <kernelxing@...cent.com>
Subject: Re: [PATCH v4 net-next 2/2] net: introduce budget_squeeze to help us
 tune rx behavior

On Fri, Mar 31, 2023 at 10:20 AM Jakub Kicinski <kuba@...nel.org> wrote:
>
> On Fri, 31 Mar 2023 08:48:07 +0800 Jason Xing wrote:
> > On Fri, Mar 31, 2023 at 12:23 AM Jakub Kicinski <kuba@...nel.org> wrote:
> > > On Thu, 30 Mar 2023 17:59:46 +0800 Jason Xing wrote:
> > > > I'm wondering for now if I can update and resend this patch to have a
> > > > better monitor (actually we do need one) on this part since we have
> > > > touched the net_rx_action() in the rps optimization patch series?
> > > > Also, just like Jesper mentioned before, it can be considered as one
> > > > 'fix' to a old problem but targetting to net-next is just fine. What
> > > > do you think about it ?
> > >
> > > Sorry, I don't understand what you're trying to say :(
> >
> > Previously this patch was not accepted because we do not want to touch
> > softirqs (actually which is net_rx_action()). Since it is touched in
> > the commit [1] in recent days, I would like to ask your permission:
> > could I resend this patch to the mailing list? I hope we can get it
> > merged.
> >
> > This patch can be considered as a 'fix' to the old problem. It's
> > beneficial and harmless, I think :)
>
> The not touching part was about softirqs which is kernel/softirq.c,
> this patch was rejected because it's not useful.

But...but time_squeeze here is really vague and it doesn't provide any
useful information to those user-side tools, which means we cannot
conclude anything from this output if we have to trace back to the
history of servers.

Right now, time_squeeze looks abandoned but many applications still
rely on it. It's not proper :(

Thanks,
Jason