Date:   Wed, 14 Feb 2018 13:15:23 +0100
From:   Maxime Ripard <maxime.ripard@...tlin.com>
To:     Philipp Rossak <embed3d@...il.com>
Cc:     Chen-Yu Tsai <wens@...e.org>,
        Alessandro Zummo <a.zummo@...ertech.it>,
        Alexandre Belloni <alexandre.belloni@...e-electrons.com>,
        linux-kernel <linux-kernel@...r.kernel.org>,
        linux-sunxi <linux-sunxi@...glegroups.com>,
        linux-rtc@...r.kernel.org
Subject: Re: [linux-sunxi] [PATCH v2] rtc: ac100: Fix ac100 determine rate bug

On Tue, Feb 13, 2018 at 04:23:00PM +0100, Philipp Rossak wrote:
> 
> 
> On 13.02.2018 14:44, Chen-Yu Tsai wrote:
> > On Tue, Feb 13, 2018 at 9:32 PM, Maxime Ripard
> > <maxime.ripard@...tlin.com> wrote:
> > > On Tue, Feb 13, 2018 at 01:14:14PM +0100, Philipp Rossak wrote:
> > > > This patch fixes a bug, that prevents the Allwinner A83T and the A80
> > > > from a successful boot. You can find the shortend trace below:
> > > 
> > > Since when is it there?
> > > 
> The bug is there since v4.16-rc1 and appeared after the clk branch was
> merged.
> 
> ^^ Should I add this info also in the commit message?

Yes

> > > > Unable to handle kernel NULL pointer dereference at virtual address
> > > > 00000000
> > > > pgd = (ptrval)
> > > > [00000000] *pgd=00000000
> > > > Internal error: Oops: 5 [#1] SMP ARM
> > > > Modules linked in:
> > > > CPU: 0 PID: 49 Comm: kworker/0:1 Not tainted 4.15.0-10190-gb89e32ccd1be #2
> > > > Hardware name: Allwinner sun8i Family
> > > > Workqueue: events deferred_probe_work_func
> > > > PC is at clk_hw_get_rate+0x0/0x34
> > > > LR is at ac100_clkout_determine_rate+0x48/0x19c
> > > > 
> > > > [ ... ]
> > > > 
> > > > (clk_hw_get_rate) from (ac100_clkout_determine_rate+0x48/0x19c)
> > > > (ac100_clkout_determine_rate) from  (clk_core_set_rate_nolock+0x3c/0x1a0)
> > > > (clk_core_set_rate_nolock) from (clk_set_rate+0x30/0x88)
> > > > (clk_set_rate) from (of_clk_set_defaults+0x200/0x364)
> > > > (of_clk_set_defaults) from (platform_drv_probe+0x18/0xb0)
> > > > 
> > > > To fix that bug, we first check if the return of the
> > > > clk_hw_get_parent_by_index is non zero. If it is zero we skip that
> > > > clock parent.
> > > > 
> > > > The BUG report could be found here: https://lkml.org/lkml/2018/2/10/198
> > > > 
> > > > Fixes: 04940631b8d2 ("rtc: ac100: Add clk output support")
> > > 
> > > Should it be sent to stable?
> > > 
> > > > Signed-off-by: Philipp Rossak <embed3d@...il.com>
> > > > ---
> > > > 
> > > > Changes in v2:
> > > >        * add tag Fixes: ... to commit message
> > > >        * add comment to if statement why we are doing this check
> > > > 
> > > >   drivers/rtc/rtc-ac100.c | 12 +++++++++++-
> > > >   1 file changed, 11 insertions(+), 1 deletion(-)
> > > > 
> > > > diff --git a/drivers/rtc/rtc-ac100.c b/drivers/rtc/rtc-ac100.c
> > > > index 8ff9dc3fe5bf..ba73201d8cc1 100644
> > > > --- a/drivers/rtc/rtc-ac100.c
> > > > +++ b/drivers/rtc/rtc-ac100.c
> > > > @@ -183,7 +183,17 @@ static int ac100_clkout_determine_rate(struct clk_hw *hw,
> > > > 
> > > >        for (i = 0; i < num_parents; i++) {
> > > >                struct clk_hw *parent = clk_hw_get_parent_by_index(hw, i);
> > > > -             unsigned long tmp, prate = clk_hw_get_rate(parent);
> > > > +             unsigned long tmp, prate;
> > > > +
> > > > +             /*
> > > > +              * We purposefully left open the possibility to use the clock
> > > > +              * from the codec side but it is not implemented right now.
> > > > +              * Thus we need to check if the parent exists.
> > > > +              */
> > > > +             if (!parent)
> > > > +                     continue;
> > > > +
> > > > +             prate = clk_hw_get_rate(parent);
> > > 
> > > clk_hw_get_num_parents should return the exact number of parents,
> > > which is going to be 1 if you only have one parent, like all DTS seems
> > > to have.
> > > 
> > > If not, then it should be explained in the comment and / or fixed
> > > properly.
> > 
> > The clock has two parents. One is a fixed clock internally registered
> > by the driver. This is actually an external crystal, and we should
> > probably add a device node and the works for it. The other parent
> > is a clock from the codec side, which we properly declare and
> > reference in the device tree. This clock, though defined, is not
> > implemented in any driver (because we don't have any ATM).
> > 
> > This second missing clock is what's causing issues here. The clk core
> > looks for the parent by name, can't find one that is registered, and
> > returns NULL.
> > 
> > I guess the comment above is still not clear enough?
> 
> I can get more detailed in the comment. I thought about this:
> 
> The clock has two parents, one is a fixed clock which is internally
> registered by the ac100 driver. The other parent is a clock from the codec
> side of the chip, which we properly declare and reference in the devicetree
> and is not implemented in any driver right now.
> If the clock core looks for the parent of that second missing clock, it
> can't one that is registered and returns NULL.
> Thus we need to check if the parent exists before we get the parent rate.
> 
> Is that ok for you?

Yep, much better thanks!
Maxime

-- 
Maxime Ripard, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com

Download attachment "signature.asc" of type "application/pgp-signature" (834 bytes)

Powered by blists - more mailing lists