[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180706103541.GC112168@atomide.com>
Date: Fri, 6 Jul 2018 03:35:41 -0700
From: Tony Lindgren <tony@...mide.com>
To: Faiz Abbas <faiz_abbas@...com>
Cc: linux-kernel@...r.kernel.org, devicetree@...r.kernel.org,
linux-omap@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
linux-clk@...r.kernel.org, robh+dt@...nel.org,
bcousson@...libre.com, paul@...an.com, t-kristo@...com,
mark.rutland@....com
Subject: Re: [PATCH v4 4/6] bus: ti-sysc: Add support for software reset
* Faiz Abbas <faiz_abbas@...com> [180706 10:05]:
> Hi,
>
> On Thursday 05 July 2018 09:17 PM, Tony Lindgren wrote:
> > * Faiz Abbas <faiz_abbas@...com> [180705 14:24]:
> >> +static int sysc_reset(struct sysc *ddata)
> >> +{
> >> + int offset = ddata->offsets[SYSC_SYSCONFIG];
> >> + int val = sysc_read(ddata, offset);
> >
> > Can you please just warn and return early for now if no
> > syss_mask is specified? Otherwise we'll have mysterious
> > errors if somebody leaves out "ti,hwmods" for module types
> > we do not yet support reset for.
>
> RESET write to sysconfig can still happen even if there's no syss_mask
> right? Not all modules need to poll on reset status.
It can but we currently don't have anything similar to
srst_udelay like we have in _ocp_softreset(). And we don't
currently have suport for sysc reset done bit.
So we want to warn and return error except in the known
working case for now.
Regards,
Tony
Powered by blists - more mailing lists