diff mbox

[RESEND,3/8] clk: core: support clocks which requires parents on (part 1)

Message ID 1461144880-8724-4-git-send-email-aisheng.dong@nxp.com (mailing list archive)
State Superseded, archived
Headers show

Commit Message

Dong Aisheng April 20, 2016, 9:34 a.m. UTC
On Freescale i.MX7D platform, all clocks operations, including
enable/disable, rate change and re-parent, requires its parent
clock on. Current clock core can not support it well.
This patch introduce a new flag CLK_OPS_PARENT_ON to handle this
special case in clock core that enable its parent clock firstly for
each operation and disable it later after operation complete.

The patch part 1 fixes the possible disabling clocks while its parent
is off during kernel booting phase in clk_disable_unused_subtree().

Before the completion of kernel booting, clock tree is still not built
completely, there may be a case that the child clock is on but its
parent is off which could be caused by either HW initial reset state
or bootloader initialization.

Taking bootloader as an example, we may enable all clocks in HW by default.
And during kernel booting time, the parent clock could be disabled in its
driver probe due to calling clk_prepare_enable and clk_disable_unprepare.
Because it's child clock is only enabled in HW while its SW usecount
in clock tree is still 0, so clk_disable of parent clock will gate
the parent clock in both HW and SW usecount ultimately. Then there will
be a child clock is still on in HW but its parent is already off.

Later in clk_disable_unused(), this clock disable accessing while its
parent off will cause system hang due to the limitation of HW which
must require its parent on.

This patch simply enables the parent clock first before disabling
if flag CLK_OPS_PARENT_ON is set in clk_disable_unused_subtree().
This is a simple solution and only affects booting time.

After kernel booting up the clock tree is already created, there will
be no case that child is off but its parent is off.
So no need do this checking for normal clk_disable() later.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@codeaurora.org>
Cc: Shawn Guo <shawnguo@kernel.org>
Signed-off-by: Dong Aisheng <aisheng.dong@nxp.com>
---
 drivers/clk/clk.c            | 5 +++++
 include/linux/clk-provider.h | 2 ++
 2 files changed, 7 insertions(+)

Comments

Michael Turquette June 24, 2016, 11:52 p.m. UTC | #1
Quoting Dong Aisheng (2016-04-20 02:34:35)
> On Freescale i.MX7D platform, all clocks operations, including
> enable/disable, rate change and re-parent, requires its parent
> clock on. Current clock core can not support it well.
> This patch introduce a new flag CLK_OPS_PARENT_ON to handle this
> special case in clock core that enable its parent clock firstly for
> each operation and disable it later after operation complete.
> 
> The patch part 1 fixes the possible disabling clocks while its parent
> is off during kernel booting phase in clk_disable_unused_subtree().
> 
> Before the completion of kernel booting, clock tree is still not built
> completely, there may be a case that the child clock is on but its
> parent is off which could be caused by either HW initial reset state
> or bootloader initialization.
> 
> Taking bootloader as an example, we may enable all clocks in HW by default.
> And during kernel booting time, the parent clock could be disabled in its
> driver probe due to calling clk_prepare_enable and clk_disable_unprepare.
> Because it's child clock is only enabled in HW while its SW usecount
> in clock tree is still 0, so clk_disable of parent clock will gate
> the parent clock in both HW and SW usecount ultimately. Then there will
> be a child clock is still on in HW but its parent is already off.
> 
> Later in clk_disable_unused(), this clock disable accessing while its
> parent off will cause system hang due to the limitation of HW which
> must require its parent on.
> 
> This patch simply enables the parent clock first before disabling
> if flag CLK_OPS_PARENT_ON is set in clk_disable_unused_subtree().
> This is a simple solution and only affects booting time.
> 
> After kernel booting up the clock tree is already created, there will
> be no case that child is off but its parent is off.
> So no need do this checking for normal clk_disable() later.
> 
> Cc: Michael Turquette <mturquette@baylibre.com>
> Cc: Stephen Boyd <sboyd@codeaurora.org>
> Cc: Shawn Guo <shawnguo@kernel.org>
> Signed-off-by: Dong Aisheng <aisheng.dong@nxp.com>
> ---
>  drivers/clk/clk.c            | 5 +++++
>  include/linux/clk-provider.h | 2 ++
>  2 files changed, 7 insertions(+)
> 
> diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> index 9f944d4..1f054d1a 100644
> --- a/drivers/clk/clk.c
> +++ b/drivers/clk/clk.c
> @@ -765,6 +765,9 @@ static void clk_disable_unused_subtree(struct clk_core *core)
>         hlist_for_each_entry(child, &core->children, child_node)
>                 clk_disable_unused_subtree(child);
>  
> +       if (core->flags & CLK_OPS_PARENT_ON)
> +               clk_core_prepare_enable(core->parent);
> +
>         flags = clk_enable_lock();
>  
>         if (core->enable_count)
> @@ -789,6 +792,8 @@ static void clk_disable_unused_subtree(struct clk_core *core)
>  
>  unlock_out:
>         clk_enable_unlock(flags);
> +       if (core->flags & CLK_OPS_PARENT_ON)
> +               clk_core_disable_unprepare(core->parent);
>  }
>  
>  static bool clk_ignore_unused;
> diff --git a/include/linux/clk-provider.h b/include/linux/clk-provider.h
> index 15628644..c878f9c 100644
> --- a/include/linux/clk-provider.h
> +++ b/include/linux/clk-provider.h
> @@ -33,6 +33,8 @@
>  #define CLK_RECALC_NEW_RATES   BIT(9) /* recalc rates after notifications */
>  #define CLK_SET_RATE_UNGATE    BIT(10) /* clock needs to run to set rate */
>  #define CLK_IS_CRITICAL                BIT(11) /* do not gate, ever */
> +/* parents need on during gate/ungate, set rate and re-parent */
> +#define CLK_OPS_PARENT_ON      BIT(12)

Hi Dong,

In the clk framework we use "enabled" instead of "on". How about
changing the flag to CLK_OPS_PARENT_ENABLE?

Regards,
Mike

>  
>  struct clk;
>  struct clk_hw;
> -- 
> 1.9.1
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-clk" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Dong Aisheng June 29, 2016, 1:57 p.m. UTC | #2
On Fri, Jun 24, 2016 at 04:52:42PM -0700, Michael Turquette wrote:
> Quoting Dong Aisheng (2016-04-20 02:34:35)
> > On Freescale i.MX7D platform, all clocks operations, including
> > enable/disable, rate change and re-parent, requires its parent
> > clock on. Current clock core can not support it well.
> > This patch introduce a new flag CLK_OPS_PARENT_ON to handle this
> > special case in clock core that enable its parent clock firstly for
> > each operation and disable it later after operation complete.
> > 
> > The patch part 1 fixes the possible disabling clocks while its parent
> > is off during kernel booting phase in clk_disable_unused_subtree().
> > 
> > Before the completion of kernel booting, clock tree is still not built
> > completely, there may be a case that the child clock is on but its
> > parent is off which could be caused by either HW initial reset state
> > or bootloader initialization.
> > 
> > Taking bootloader as an example, we may enable all clocks in HW by default.
> > And during kernel booting time, the parent clock could be disabled in its
> > driver probe due to calling clk_prepare_enable and clk_disable_unprepare.
> > Because it's child clock is only enabled in HW while its SW usecount
> > in clock tree is still 0, so clk_disable of parent clock will gate
> > the parent clock in both HW and SW usecount ultimately. Then there will
> > be a child clock is still on in HW but its parent is already off.
> > 
> > Later in clk_disable_unused(), this clock disable accessing while its
> > parent off will cause system hang due to the limitation of HW which
> > must require its parent on.
> > 
> > This patch simply enables the parent clock first before disabling
> > if flag CLK_OPS_PARENT_ON is set in clk_disable_unused_subtree().
> > This is a simple solution and only affects booting time.
> > 
> > After kernel booting up the clock tree is already created, there will
> > be no case that child is off but its parent is off.
> > So no need do this checking for normal clk_disable() later.
> > 
> > Cc: Michael Turquette <mturquette@baylibre.com>
> > Cc: Stephen Boyd <sboyd@codeaurora.org>
> > Cc: Shawn Guo <shawnguo@kernel.org>
> > Signed-off-by: Dong Aisheng <aisheng.dong@nxp.com>
> > ---
> >  drivers/clk/clk.c            | 5 +++++
> >  include/linux/clk-provider.h | 2 ++
> >  2 files changed, 7 insertions(+)
> > 
> > diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> > index 9f944d4..1f054d1a 100644
> > --- a/drivers/clk/clk.c
> > +++ b/drivers/clk/clk.c
> > @@ -765,6 +765,9 @@ static void clk_disable_unused_subtree(struct clk_core *core)
> >         hlist_for_each_entry(child, &core->children, child_node)
> >                 clk_disable_unused_subtree(child);
> >  
> > +       if (core->flags & CLK_OPS_PARENT_ON)
> > +               clk_core_prepare_enable(core->parent);
> > +
> >         flags = clk_enable_lock();
> >  
> >         if (core->enable_count)
> > @@ -789,6 +792,8 @@ static void clk_disable_unused_subtree(struct clk_core *core)
> >  
> >  unlock_out:
> >         clk_enable_unlock(flags);
> > +       if (core->flags & CLK_OPS_PARENT_ON)
> > +               clk_core_disable_unprepare(core->parent);
> >  }
> >  
> >  static bool clk_ignore_unused;
> > diff --git a/include/linux/clk-provider.h b/include/linux/clk-provider.h
> > index 15628644..c878f9c 100644
> > --- a/include/linux/clk-provider.h
> > +++ b/include/linux/clk-provider.h
> > @@ -33,6 +33,8 @@
> >  #define CLK_RECALC_NEW_RATES   BIT(9) /* recalc rates after notifications */
> >  #define CLK_SET_RATE_UNGATE    BIT(10) /* clock needs to run to set rate */
> >  #define CLK_IS_CRITICAL                BIT(11) /* do not gate, ever */
> > +/* parents need on during gate/ungate, set rate and re-parent */
> > +#define CLK_OPS_PARENT_ON      BIT(12)
> 
> Hi Dong,
> 
> In the clk framework we use "enabled" instead of "on". How about
> changing the flag to CLK_OPS_PARENT_ENABLE?
> 

Yes, of course i'm fine with it.
I could update the patch title too.
Thanks for the review.

Regards
Dong Aisheng

> Regards,
> Mike
> 
> >  
> >  struct clk;
> >  struct clk_hw;
> > -- 
> > 1.9.1
> > 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-clk" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-clk" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
index 9f944d4..1f054d1a 100644
--- a/drivers/clk/clk.c
+++ b/drivers/clk/clk.c
@@ -765,6 +765,9 @@  static void clk_disable_unused_subtree(struct clk_core *core)
 	hlist_for_each_entry(child, &core->children, child_node)
 		clk_disable_unused_subtree(child);
 
+	if (core->flags & CLK_OPS_PARENT_ON)
+		clk_core_prepare_enable(core->parent);
+
 	flags = clk_enable_lock();
 
 	if (core->enable_count)
@@ -789,6 +792,8 @@  static void clk_disable_unused_subtree(struct clk_core *core)
 
 unlock_out:
 	clk_enable_unlock(flags);
+	if (core->flags & CLK_OPS_PARENT_ON)
+		clk_core_disable_unprepare(core->parent);
 }
 
 static bool clk_ignore_unused;
diff --git a/include/linux/clk-provider.h b/include/linux/clk-provider.h
index 15628644..c878f9c 100644
--- a/include/linux/clk-provider.h
+++ b/include/linux/clk-provider.h
@@ -33,6 +33,8 @@ 
 #define CLK_RECALC_NEW_RATES	BIT(9) /* recalc rates after notifications */
 #define CLK_SET_RATE_UNGATE	BIT(10) /* clock needs to run to set rate */
 #define CLK_IS_CRITICAL		BIT(11) /* do not gate, ever */
+/* parents need on during gate/ungate, set rate and re-parent */
+#define CLK_OPS_PARENT_ON	BIT(12)
 
 struct clk;
 struct clk_hw;