diff mbox series

clk: Make clk_bulk_get_all() return a valid "id"

Message ID 20190913024029.2640-1-bjorn.andersson@linaro.org (mailing list archive)
State Accepted, archived
Headers show
Series clk: Make clk_bulk_get_all() return a valid "id" | expand

Commit Message

Bjorn Andersson Sept. 13, 2019, 2:40 a.m. UTC
The adreno driver expects the "id" field of the returned clk_bulk_data
to be filled in with strings from the clock-names property.

But due to the use of kmalloc_array() in of_clk_bulk_get_all() it
receives a list of bogus pointers instead.

Zero-initialize the "id" field and attempt to populate with strings from
the clock-names property to resolve both these issues.

Fixes: 616e45df7c4a ("clk: add new APIs to operate on all available clocks")
Fixes: 8e3e791d20d2 ("drm/msm: Use generic bulk clock function")
Cc: Dong Aisheng <aisheng.dong@nxp.com>
Cc: Jordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
---
 drivers/clk/clk-bulk.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

Comments

Jordan Crouse Sept. 13, 2019, 2:44 p.m. UTC | #1
On Thu, Sep 12, 2019 at 07:40:29PM -0700, Bjorn Andersson wrote:
> The adreno driver expects the "id" field of the returned clk_bulk_data
> to be filled in with strings from the clock-names property.
> 
> But due to the use of kmalloc_array() in of_clk_bulk_get_all() it
> receives a list of bogus pointers instead.
> 
> Zero-initialize the "id" field and attempt to populate with strings from
> the clock-names property to resolve both these issues.

This looks great to me.  Thanks for fixing that so quickly.

Reviewed-by: Jordan Crouse <jcrouse@codeaurora.org>

> Fixes: 616e45df7c4a ("clk: add new APIs to operate on all available clocks")
> Fixes: 8e3e791d20d2 ("drm/msm: Use generic bulk clock function")
> Cc: Dong Aisheng <aisheng.dong@nxp.com>
> Cc: Jordan Crouse <jcrouse@codeaurora.org>
> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
> ---
>  drivers/clk/clk-bulk.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/clk/clk-bulk.c b/drivers/clk/clk-bulk.c
> index 524bf9a53098..e9e16425c739 100644
> --- a/drivers/clk/clk-bulk.c
> +++ b/drivers/clk/clk-bulk.c
> @@ -18,10 +18,13 @@ static int __must_check of_clk_bulk_get(struct device_node *np, int num_clks,
>  	int ret;
>  	int i;
>  
> -	for (i = 0; i < num_clks; i++)
> +	for (i = 0; i < num_clks; i++) {
> +		clks[i].id = NULL;
>  		clks[i].clk = NULL;
> +	}
>  
>  	for (i = 0; i < num_clks; i++) {
> +		of_property_read_string_index(np, "clock-names", i, &clks[i].id);
>  		clks[i].clk = of_clk_get(np, i);
>  		if (IS_ERR(clks[i].clk)) {
>  			ret = PTR_ERR(clks[i].clk);
> -- 
> 2.18.0
>
Stephen Boyd Sept. 17, 2019, 8:33 p.m. UTC | #2
Quoting Bjorn Andersson (2019-09-12 19:40:29)
> The adreno driver expects the "id" field of the returned clk_bulk_data
> to be filled in with strings from the clock-names property.
> 
> But due to the use of kmalloc_array() in of_clk_bulk_get_all() it
> receives a list of bogus pointers instead.
> 
> Zero-initialize the "id" field and attempt to populate with strings from
> the clock-names property to resolve both these issues.
> 
> Fixes: 616e45df7c4a ("clk: add new APIs to operate on all available clocks")
> Fixes: 8e3e791d20d2 ("drm/msm: Use generic bulk clock function")
> Cc: Dong Aisheng <aisheng.dong@nxp.com>
> Cc: Jordan Crouse <jcrouse@codeaurora.org>
> Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
> ---

Applied to clk-next

And now I see that this whole thing needs to be inlined to the one call
site and should use the struct device instead of calling of_clk_get()...
I'll have to fix it later.
Bjorn Andersson Sept. 17, 2019, 8:42 p.m. UTC | #3
On Tue 17 Sep 13:33 PDT 2019, Stephen Boyd wrote:

> Quoting Bjorn Andersson (2019-09-12 19:40:29)
> > The adreno driver expects the "id" field of the returned clk_bulk_data
> > to be filled in with strings from the clock-names property.
> > 
> > But due to the use of kmalloc_array() in of_clk_bulk_get_all() it
> > receives a list of bogus pointers instead.
> > 
> > Zero-initialize the "id" field and attempt to populate with strings from
> > the clock-names property to resolve both these issues.
> > 
> > Fixes: 616e45df7c4a ("clk: add new APIs to operate on all available clocks")
> > Fixes: 8e3e791d20d2 ("drm/msm: Use generic bulk clock function")
> > Cc: Dong Aisheng <aisheng.dong@nxp.com>
> > Cc: Jordan Crouse <jcrouse@codeaurora.org>
> > Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
> > ---
> 
> Applied to clk-next
> 

Thanks

> And now I see that this whole thing needs to be inlined to the one call
> site and should use the struct device instead of calling of_clk_get()...
> I'll have to fix it later.
> 

I concluded the same, sorry for not mentioning it.

Regards,
Bjorn
diff mbox series

Patch

diff --git a/drivers/clk/clk-bulk.c b/drivers/clk/clk-bulk.c
index 524bf9a53098..e9e16425c739 100644
--- a/drivers/clk/clk-bulk.c
+++ b/drivers/clk/clk-bulk.c
@@ -18,10 +18,13 @@  static int __must_check of_clk_bulk_get(struct device_node *np, int num_clks,
 	int ret;
 	int i;
 
-	for (i = 0; i < num_clks; i++)
+	for (i = 0; i < num_clks; i++) {
+		clks[i].id = NULL;
 		clks[i].clk = NULL;
+	}
 
 	for (i = 0; i < num_clks; i++) {
+		of_property_read_string_index(np, "clock-names", i, &clks[i].id);
 		clks[i].clk = of_clk_get(np, i);
 		if (IS_ERR(clks[i].clk)) {
 			ret = PTR_ERR(clks[i].clk);