From: mturquette@linaro.org (Mike Turquette)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 2/3] clk: composite: allow fixed rates & fixed dividers
Date: Thu, 11 Apr 2013 10:56:13 -0700 [thread overview]
Message-ID: <20130411175613.7915.25443@quantum> (raw)
In-Reply-To: <5166A2E6.70907@nvidia.com>
Quoting Prashant Gaikwad (2013-04-11 04:47:50)
> On Thursday 11 April 2013 03:32 AM, Mike Turquette wrote:
> > The composite clock assumes that any clock implementing the .recalc_rate
> > callback will also implement .round_rate and .set_rate. This is not
> > always true; the basic fixed-rate clock will only implement .recalc_rate
> > and a fixed-divider clock may choose to implement .recalc_rate and
> > .round_rate but not .set_rate.
> >
> > Fix this by conditionally registering .round_rate and .set_rate
> > callbacks based on the rate_ops passed in to clk_composite_register.
> >
> > Signed-off-by: Mike Turquette <mturquette@linaro.org>
> > Cc: Prashant Gaikwad <pgaikwad@nvidia.com>
> > Cc: Emilio L?pez <emilio@elopez.com.ar>
> > Cc: Gregory CLEMENT <gregory.clement@free-electrons.com>
> > ---
> > drivers/clk/clk-composite.c | 13 +++++++++----
> > 1 file changed, 9 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/clk/clk-composite.c b/drivers/clk/clk-composite.c
> > index 6f4728c..2779fb3 100644
> > --- a/drivers/clk/clk-composite.c
> > +++ b/drivers/clk/clk-composite.c
> > @@ -150,17 +150,22 @@ struct clk *clk_register_composite(struct device *dev, const char *name,
> > }
> >
> > if (rate_hw && rate_ops) {
> > - if (!rate_ops->recalc_rate || !rate_ops->round_rate ||
> > - !rate_ops->set_rate) {
> > + if (!rate_ops->recalc_rate) {
> > clk = ERR_PTR(-EINVAL);
> > goto err;
> > }
> >
> > + /* .round_rate is a prerequisite for .set_rate */
> > + if (rate_ops->round_rate) {
> > + clk_composite_ops->round_rate = clk_composite_round_rate;
> > + if (rate_ops->set_rate) {
> > + clk_composite_ops->set_rate = clk_composite_set_rate;
> > + }
> > + }
> > +
>
> What if rate_ops has only recalc_rate and set_rate? clk_composite_ops
> will have only recalc_rate and it will pass the sanity check in clk_init
> and caller of clk_register_composite will expect set_rate to work.
> We should check here if rate_ops has set_rate but not round_rate then
> return error.
>
I think I will instead put a WARN here instead of completely bailing out
of the clock registration.
Thanks,
Mike
> > composite->rate_hw = rate_hw;
> > composite->rate_ops = rate_ops;
> > clk_composite_ops->recalc_rate = clk_composite_recalc_rate;
> > - clk_composite_ops->round_rate = clk_composite_round_rate;
> > - clk_composite_ops->set_rate = clk_composite_set_rate;
> > }
> >
> > if (gate_hw && gate_ops) {
next prev parent reply other threads:[~2013-04-11 17:56 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-04-09 13:48 [PATCH 0/3] sunxi: unify main oscillator and its gate Emilio López
2013-04-09 13:48 ` [PATCH 1/3] clk: composite: Add fixed-rate support Emilio López
2013-04-09 15:41 ` Gregory CLEMENT
2013-04-09 16:04 ` Emilio López
2013-04-10 22:02 ` [PATCH 0/3] fixed-rate and fixed-factor clocks in composite clock Mike Turquette
2013-04-10 22:02 ` [PATCH 1/3] clk: composite: rename 'div' references to 'rate' Mike Turquette
2013-04-11 11:40 ` Prashant Gaikwad
2013-04-11 17:56 ` Mike Turquette
2013-04-10 22:02 ` [PATCH 2/3] clk: composite: allow fixed rates & fixed dividers Mike Turquette
2013-04-11 11:47 ` Prashant Gaikwad
2013-04-11 17:56 ` Mike Turquette [this message]
2013-04-10 22:02 ` [PATCH v2 3/3] clk: sunxi: Unify oscillator clock Mike Turquette
2013-04-11 0:07 ` Emilio López
2013-04-12 18:40 ` Mike Turquette
2013-04-11 18:31 ` [PATCH v2 1/2] clk: composite: rename 'div' references to 'rate' Mike Turquette
2013-04-11 18:31 ` [PATCH v2 2/2] clk: composite: allow fixed rates & fixed dividers Mike Turquette
2013-04-09 13:48 ` [PATCH 2/3] clk: sunxi: Unify oscillator clock Emilio López
2013-04-09 13:48 ` [PATCH 3/3] ARM: sunxi: unify osc24M_fixed and osc24M Emilio López
2013-04-15 8:45 ` Maxime Ripard
2013-04-09 14:56 ` [PATCH 0/3] sunxi: unify main oscillator and its gate Gregory CLEMENT
2013-04-09 15:04 ` Emilio López
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20130411175613.7915.25443@quantum \
--to=mturquette@linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox