linux-kernel - Re: [PATCH 08/13] phy: qcom-qmp-pcie: drop power-down delay config

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

Message-ID: <Y0V1TIBySTPhYqX/@hovoldconsulting.com>
Date:   Tue, 11 Oct 2022 15:53:16 +0200
From:   Johan Hovold <johan@...nel.org>
To:     Dmitry Baryshkov <dmitry.baryshkov@...aro.org>
Cc:     Johan Hovold <johan+linaro@...nel.org>,
        Vinod Koul <vkoul@...nel.org>, Andy Gross <agross@...nel.org>,
        Bjorn Andersson <andersson@...nel.org>,
        Konrad Dybcio <konrad.dybcio@...ainline.org>,
        Kishon Vijay Abraham I <kishon@...com>,
        linux-arm-msm@...r.kernel.org, linux-phy@...ts.infradead.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 08/13] phy: qcom-qmp-pcie: drop power-down delay config

On Tue, Oct 11, 2022 at 04:46:53PM +0300, Dmitry Baryshkov wrote:
> On 11/10/2022 16:14, Johan Hovold wrote:
> > The power-down delay was included in the first version of the QMP driver
> > as an optional delay after powering on the PHY (using
> > POWER_DOWN_CONTROL) and just before starting it. Later changes modified
> > this sequence by powering on before initialising the PHY, but the
> > optional delay stayed where it was (i.e. before starting the PHY).
> > 
> > The vendor driver does not use a delay before starting the PHY and this
> > is likely not needed on any platform unless there is a corresponding
> > delay in the vendor kernel init sequence tables (i.e. in devicetree).
> > 
> > Let's keep the delay for now, but drop the redundant delay period
> > configuration while increasing the unnecessarily low timer slack
> > somewhat.
> 
> Actually, the vendor driver does this 995..1005 sleep. But contrary to 
> our driver it does that after programming whole PHY init sequence, which 
> includes SW_RESET / START_CTL, but before programming the pipe clocks.

Right, it does it after starting the PHY which means that you don't have
to poll for as long for the PHY status.

It's a different delay entirely.

> I think we can either drop this delay completely, or move it before 
> read_poll_timeout().

It definitely shouldn't be used for any new platforms, but I opted for
the conservative route of keeping it in case some of the older platforms
actually do need it.

My bet is that this is all copy-paste cruft that could be removed, but
I'd rather do that as a separate follow-on change. Perhaps after testing
some more SoC after removing the delay.

SC8280XP certainly doesn't need it.

Johan