netdev - Re: [PATCH v4 09/13] net: ethernet: ti: cpts: rework initialization/deinitialization

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]

Date:   Tue, 6 Dec 2016 19:04:21 +0100
From:   Richard Cochran <richardcochran@...il.com>
To:     Grygorii Strashko <grygorii.strashko@...com>
Cc:     "David S. Miller" <davem@...emloft.net>, netdev@...r.kernel.org,
        Mugunthan V N <mugunthanvnm@...com>,
        Sekhar Nori <nsekhar@...com>, linux-kernel@...r.kernel.org,
        linux-omap@...r.kernel.org, devicetree@...r.kernel.org,
        Murali Karicheri <m-karicheri2@...com>,
        Wingman Kwok <w-kwok2@...com>,
        Thomas Gleixner <tglx@...utronix.de>
Subject: Re: [PATCH v4 09/13] net: ethernet: ti: cpts: rework
 initialization/deinitialization

On Tue, Dec 06, 2016 at 11:49:14AM -0600, Grygorii Strashko wrote:
> But we do reset whole cpsw :( and that's required to support PM use cases as
> suspend/resume.

The code is resetting the clock unconditionally on ifup/down.  That
sucks.  If you reset the clock *only* after resume, that would be ok.

> There are also PM requirement to shutdown cpsw in case all interfaces are down.

Well, those requirements are not too smart.  As an end user, I expect
that ifdown/up does not change the time.  There isn't any reason to
reset the clock in this case.

> More over, there are requirement to minimize cpsw power consumption in case all links are
> disconnected (and cpts is special case here).
> 
> So, at least resetting of the timecounter still required.

Only if you follow that poorly conceived PM plan.  Anyhow, I agree
that it isn't the task of your present series to fix that.

> Ok. I'll try to optimize it following your directions.

What I would like to see is: initialize the cyclecounter fields
exactly once.

Thanks,
Richard