[PATCH] clk: check for invalid parent index of orphans in __clk_init()

[PATCH] clk: check for invalid parent index of orphans in __clk_init()

[Mans Rullgard]

If a mux clock is initialised (by hardware or firmware) with an
invalid parent, its ->get_parent() can return an out of range
index.  For example, the generic mux clock attempts to return
-EINVAL, which due to the u8 return type ends up a rather large
number.  Using this index with the parent_names[] array results
in an invalid pointer and (usually) a crash in the following
strcmp().

This patch adds a check for the parent index being in range,
ignoring clocks reporting invalid values.

Signed-off-by: Mans Rullgard <mans@mansr.com>
Cc: Rhyland Klein <rklein@nvidia.com>
---
 drivers/clk/clk.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
index d48ac71..bc0662b 100644
--- a/drivers/clk/clk.c
+++ b/drivers/clk/clk.c
@@ -1950,7 +1950,8 @@ int __clk_init(struct device *dev, struct clk *clk)
 	hlist_for_each_entry_safe(orphan, tmp2, &clk_orphan_list, child_node) {
 		if (orphan->num_parents && orphan->ops->get_parent) {
 			i = orphan->ops->get_parent(orphan->hw);
-			if (!strcmp(clk->name, orphan->parent_names[i]))
+			if (i >= 0 && i < orphan->num_parents &&
+			    !strcmp(clk->name, orphan->parent_names[i]))
 				__clk_reparent(orphan, clk);
 			continue;
 		}
-- 
2.3.0

Re: [PATCH] clk: check for invalid parent index of orphans in __clk_init()

[Rhyland Klein]

On 2/15/2015 7:33 AM, Mans Rullgard wrote:
> If a mux clock is initialised (by hardware or firmware) with an
> invalid parent, its ->get_parent() can return an out of range
> index.  For example, the generic mux clock attempts to return
> -EINVAL, which due to the u8 return type ends up a rather large
> number.  Using this index with the parent_names[] array results
> in an invalid pointer and (usually) a crash in the following
> strcmp().
> 
> This patch adds a check for the parent index being in range,
> ignoring clocks reporting invalid values.
> 
> Signed-off-by: Mans Rullgard <mans@mansr.com>
> Cc: Rhyland Klein <rklein@nvidia.com>
> ---
>  drivers/clk/clk.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> index d48ac71..bc0662b 100644
> --- a/drivers/clk/clk.c
> +++ b/drivers/clk/clk.c
> @@ -1950,7 +1950,8 @@ int __clk_init(struct device *dev, struct clk *clk)
>  	hlist_for_each_entry_safe(orphan, tmp2, &clk_orphan_list, child_node) {
>  		if (orphan->num_parents && orphan->ops->get_parent) {
>  			i = orphan->ops->get_parent(orphan->hw);
> -			if (!strcmp(clk->name, orphan->parent_names[i]))
> +			if (i >= 0 && i < orphan->num_parents &&
> +			    !strcmp(clk->name, orphan->parent_names[i]))
>  				__clk_reparent(orphan, clk);
>  			continue;
>  		}
> 

This works for me and is less invasive than the original patch series.

Tested-by: Rhyland Klein <rklein@nvidia.com>

-rhyland

-- 
nvpublic

Re: [PATCH] clk: check for invalid parent index of orphans in __clk_init()

[Michael Turquette]

Quoting Rhyland Klein (2015-02-17 08:58:29)
> On 2/15/2015 7:33 AM, Mans Rullgard wrote:
> > If a mux clock is initialised (by hardware or firmware) with an
> > invalid parent, its ->get_parent() can return an out of range
> > index.  For example, the generic mux clock attempts to return
> > -EINVAL, which due to the u8 return type ends up a rather large
> > number.  Using this index with the parent_names[] array results
> > in an invalid pointer and (usually) a crash in the following
> > strcmp().
> > 
> > This patch adds a check for the parent index being in range,
> > ignoring clocks reporting invalid values.
> > 
> > Signed-off-by: Mans Rullgard <mans@mansr.com>
> > Cc: Rhyland Klein <rklein@nvidia.com>
> > ---
> >  drivers/clk/clk.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> > index d48ac71..bc0662b 100644
> > --- a/drivers/clk/clk.c
> > +++ b/drivers/clk/clk.c
> > @@ -1950,7 +1950,8 @@ int __clk_init(struct device *dev, struct clk *clk)
> >       hlist_for_each_entry_safe(orphan, tmp2, &clk_orphan_list, child_node) {
> >               if (orphan->num_parents && orphan->ops->get_parent) {
> >                       i = orphan->ops->get_parent(orphan->hw);
> > -                     if (!strcmp(clk->name, orphan->parent_names[i]))
> > +                     if (i >= 0 && i < orphan->num_parents &&
> > +                         !strcmp(clk->name, orphan->parent_names[i]))
> >                               __clk_reparent(orphan, clk);
> >                       continue;
> >               }
> > 
> 
> This works for me and is less invasive than the original patch series.
> 
> Tested-by: Rhyland Klein <rklein@nvidia.com>

Applied.

Thanks,
Mike

> 
> -rhyland
> 
> -- 
> nvpublic

Re: [PATCH] clk: check for invalid parent index of orphans in __clk_init()

[Måns Rullgård]

Michael Turquette <mturquette@linaro.org> writes:

> Quoting Rhyland Klein (2015-02-17 08:58:29)
>> On 2/15/2015 7:33 AM, Mans Rullgard wrote:
>> > If a mux clock is initialised (by hardware or firmware) with an
>> > invalid parent, its ->get_parent() can return an out of range
>> > index.  For example, the generic mux clock attempts to return
>> > -EINVAL, which due to the u8 return type ends up a rather large
>> > number.  Using this index with the parent_names[] array results
>> > in an invalid pointer and (usually) a crash in the following
>> > strcmp().
>> > 
>> > This patch adds a check for the parent index being in range,
>> > ignoring clocks reporting invalid values.
>> > 
>> > Signed-off-by: Mans Rullgard <mans@mansr.com>
>> > Cc: Rhyland Klein <rklein@nvidia.com>
>> > ---
>> >  drivers/clk/clk.c | 3 ++-
>> >  1 file changed, 2 insertions(+), 1 deletion(-)
>> > 
>> > diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
>> > index d48ac71..bc0662b 100644
>> > --- a/drivers/clk/clk.c
>> > +++ b/drivers/clk/clk.c
>> > @@ -1950,7 +1950,8 @@ int __clk_init(struct device *dev, struct clk *clk)
>> >       hlist_for_each_entry_safe(orphan, tmp2, &clk_orphan_list, child_node) {
>> >               if (orphan->num_parents && orphan->ops->get_parent) {
>> >                       i = orphan->ops->get_parent(orphan->hw);
>> > -                     if (!strcmp(clk->name, orphan->parent_names[i]))
>> > +                     if (i >= 0 && i < orphan->num_parents &&
>> > +                         !strcmp(clk->name, orphan->parent_names[i]))
>> >                               __clk_reparent(orphan, clk);
>> >                       continue;
>> >               }
>> > 
>> 
>> This works for me and is less invasive than the original patch series.
>> 
>> Tested-by: Rhyland Klein <rklein@nvidia.com>
>
> Applied.

Did this get lost somewhere?  It's not in mainline, and I can't find it
in the clk tree on kernel.org either.

-- 
Måns Rullgård
mans@mansr.com

Re: [PATCH] clk: check for invalid parent index of orphans in __clk_init()

[Stephen Boyd]

On 09/15, Måns Rullgård wrote:
> Michael Turquette <mturquette@linaro.org> writes:
> 
> > Quoting Rhyland Klein (2015-02-17 08:58:29)
> >> On 2/15/2015 7:33 AM, Mans Rullgard wrote:
> >> > If a mux clock is initialised (by hardware or firmware) with an
> >> > invalid parent, its ->get_parent() can return an out of range
> >> > index.  For example, the generic mux clock attempts to return
> >> > -EINVAL, which due to the u8 return type ends up a rather large
> >> > number.  Using this index with the parent_names[] array results
> >> > in an invalid pointer and (usually) a crash in the following
> >> > strcmp().
> >> > 
> >> > This patch adds a check for the parent index being in range,
> >> > ignoring clocks reporting invalid values.
> >> > 
> >> > Signed-off-by: Mans Rullgard <mans@mansr.com>
> >> > Cc: Rhyland Klein <rklein@nvidia.com>
> >> > ---
> >> >  drivers/clk/clk.c | 3 ++-
> >> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >> > 
> >> > diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> >> > index d48ac71..bc0662b 100644
> >> > --- a/drivers/clk/clk.c
> >> > +++ b/drivers/clk/clk.c
> >> > @@ -1950,7 +1950,8 @@ int __clk_init(struct device *dev, struct clk *clk)
> >> >       hlist_for_each_entry_safe(orphan, tmp2, &clk_orphan_list, child_node) {
> >> >               if (orphan->num_parents && orphan->ops->get_parent) {
> >> >                       i = orphan->ops->get_parent(orphan->hw);
> >> > -                     if (!strcmp(clk->name, orphan->parent_names[i]))
> >> > +                     if (i >= 0 && i < orphan->num_parents &&
> >> > +                         !strcmp(clk->name, orphan->parent_names[i]))
> >> >                               __clk_reparent(orphan, clk);
> >> >                       continue;
> >> >               }
> >> > 
> >> 
> >> This works for me and is less invasive than the original patch series.
> >> 
> >> Tested-by: Rhyland Klein <rklein@nvidia.com>
> >
> > Applied.
> 
> Did this get lost somewhere?  It's not in mainline, and I can't find it
> in the clk tree on kernel.org either.

I think it got lost. I've applied this to clk-fixes.

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project