[bug report] clk: ti: divider: add driver internal API for parsing divider data

[bug report] clk: ti: divider: add driver internal API for parsing divider data

[Dan Carpenter]

Hello Tero Kristo,

The patch 4f6be5655dc9: "clk: ti: divider: add driver internal API
for parsing divider data" from Feb 9, 2017, leads to the following
static checker warning:

	drivers/clk/ti/divider.c:491 ti_clk_register_divider()
	warn: 'table' isn't an ERR_PTR

drivers/clk/ti/divider.c
   468  struct clk *ti_clk_register_divider(struct ti_clk *setup)
   469  {
   470          struct ti_clk_divider *div = setup->data;
   471          struct clk_omap_reg reg = {
   472                  .index = div->module,
   473                  .offset = div->reg,
   474          };
   475          u8 width;
   476          u32 flags = 0;
   477          u8 div_flags = 0;
   478          const struct clk_div_table *table;
   479          struct clk *clk;
   480  
   481          if (div->flags & CLKF_INDEX_STARTS_AT_ONE)
   482                  div_flags |= CLK_DIVIDER_ONE_BASED;
   483  
   484          if (div->flags & CLKF_INDEX_POWER_OF_TWO)
   485                  div_flags |= CLK_DIVIDER_POWER_OF_TWO;
   486  
   487          if (div->flags & CLKF_SET_RATE_PARENT)
   488                  flags |= CLK_SET_RATE_PARENT;
   489  
   490          table = _get_div_table_from_setup(div, &width);
   491          if (IS_ERR(table))
                           ^^^^^

NULL is actually allowed here so we can't just change this to a check
for NULL.  Prior to this commit if the:

	tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);

allocation failed then table was PTR_ERR(-ENOMEM).  I guess we should
change it back.  We could probably do sothing like the diff below?

   492                  return (struct clk *)table;
   493  
   494          clk = _register_divider(NULL, setup->name, div->parent,
   495                                  flags, &reg, div->bit_shift,
   496                                  width, -EINVAL, div_flags, table);
   497  
   498          if (IS_ERR(clk))
   499                  kfree(table);
   500  
   501          return clk;
   502  }


diff --git a/drivers/clk/ti/divider.c b/drivers/clk/ti/divider.c
index 8d77090ad94a..c4335eba2b2e 100644
--- a/drivers/clk/ti/divider.c
+++ b/drivers/clk/ti/divider.c
@@ -403,8 +403,10 @@ int ti_clk_parse_divider_data(int *div_table, int num_dividers, int max_div,
 	num_dividers = i;
 
 	tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);
-	if (!tmp)
+	if (!tmp) {
+		*table = PTR_ERR(-ENOMEM);
 		return -ENOMEM;
+	}
 
 	valid_div = 0;
 	*width = 0;

Re: [bug report] clk: ti: divider: add driver internal API for parsing divider data

[Stephen Boyd]

Quoting Dan Carpenter (2018-12-17 06:45:53)
> Hello Tero Kristo,
> 
> The patch 4f6be5655dc9: "clk: ti: divider: add driver internal API
> for parsing divider data" from Feb 9, 2017, leads to the following
> static checker warning:
> 
>         drivers/clk/ti/divider.c:491 ti_clk_register_divider()
>         warn: 'table' isn't an ERR_PTR
> 
> drivers/clk/ti/divider.c
>    468  struct clk *ti_clk_register_divider(struct ti_clk *setup)
>    469  {
>    470          struct ti_clk_divider *div = setup->data;
>    471          struct clk_omap_reg reg = {
>    472                  .index = div->module,
>    473                  .offset = div->reg,
>    474          };
>    475          u8 width;
>    476          u32 flags = 0;
>    477          u8 div_flags = 0;
>    478          const struct clk_div_table *table;
>    479          struct clk *clk;
>    480  
>    481          if (div->flags & CLKF_INDEX_STARTS_AT_ONE)
>    482                  div_flags |= CLK_DIVIDER_ONE_BASED;
>    483  
>    484          if (div->flags & CLKF_INDEX_POWER_OF_TWO)
>    485                  div_flags |= CLK_DIVIDER_POWER_OF_TWO;
>    486  
>    487          if (div->flags & CLKF_SET_RATE_PARENT)
>    488                  flags |= CLK_SET_RATE_PARENT;
>    489  
>    490          table = _get_div_table_from_setup(div, &width);
>    491          if (IS_ERR(table))
>                            ^^^^^
> 
> NULL is actually allowed here so we can't just change this to a check
> for NULL.  Prior to this commit if the:
> 
>         tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);
> 
> allocation failed then table was PTR_ERR(-ENOMEM).  I guess we should
> change it back.  We could probably do sothing like the diff below?

Patch looks fine to me. Are you going to send a proper change?

[PATCH] clk: ti: Fix error handling in ti_clk_parse_divider_data()

[Dan Carpenter]

The ti_clk_parse_divider_data() function is only called from
_get_div_table_from_setup().  That function doesn't look at the return
value but instead looks at the "*table" pointer.  In this case, if the
kcalloc() fails then *table is NULL (which means success).  It should
instead be an error pointer.

The ti_clk_parse_divider_data() function has two callers.  One checks
for errors and the other doesn't.  I have fixed it so now both handle
errors.

Fixes: 4f6be5655dc9 ("clk: ti: divider: add driver internal API for parsing divider data")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
---
 drivers/clk/ti/divider.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/drivers/clk/ti/divider.c b/drivers/clk/ti/divider.c
index 8d77090ad94a..4c48ef424ad5 100644
--- a/drivers/clk/ti/divider.c
+++ b/drivers/clk/ti/divider.c
@@ -403,8 +403,10 @@ int ti_clk_parse_divider_data(int *div_table, int num_dividers, int max_div,
 	num_dividers = i;
 
 	tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);
-	if (!tmp)
+	if (!tmp) {
+		*table = PTR_ERR(-ENOMEM);
 		return -ENOMEM;
+	}
 
 	valid_div = 0;
 	*width = 0;
@@ -439,6 +441,7 @@ struct clk_hw *ti_clk_build_component_div(struct ti_clk_divider *setup)
 {
 	struct clk_omap_divider *div;
 	struct clk_omap_reg *reg;
+	int ret;
 
 	if (!setup)
 		return NULL;
@@ -458,6 +461,12 @@ struct clk_hw *ti_clk_build_component_div(struct ti_clk_divider *setup)
 		div->flags |= CLK_DIVIDER_POWER_OF_TWO;
 
 	div->table = _get_div_table_from_setup(setup, &div->width);
+	if (IS_ERR(div->table)) {
+		ret = PTR_ERR(div->table);
+		kfree(div);
+		return ret;
+	}
+
 
 	div->shift = setup->bit_shift;
 	div->latch = -EINVAL;
-- 
2.11.0

Re: [PATCH] clk: ti: Fix error handling in ti_clk_parse_divider_data()

[Tero Kristo]

On 15/01/2019 09:01, Dan Carpenter wrote:
> The ti_clk_parse_divider_data() function is only called from
> _get_div_table_from_setup().  That function doesn't look at the return
> value but instead looks at the "*table" pointer.  In this case, if the
> kcalloc() fails then *table is NULL (which means success).  It should
> instead be an error pointer.
> 
> The ti_clk_parse_divider_data() function has two callers.  One checks
> for errors and the other doesn't.  I have fixed it so now both handle
> errors.
> 
> Fixes: 4f6be5655dc9 ("clk: ti: divider: add driver internal API for parsing divider data")
> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>

Looks fine to me:

Acked-by: Tero Kristo <t-kristo@ti.com>

> ---
>   drivers/clk/ti/divider.c | 11 ++++++++++-
>   1 file changed, 10 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/clk/ti/divider.c b/drivers/clk/ti/divider.c
> index 8d77090ad94a..4c48ef424ad5 100644
> --- a/drivers/clk/ti/divider.c
> +++ b/drivers/clk/ti/divider.c
> @@ -403,8 +403,10 @@ int ti_clk_parse_divider_data(int *div_table, int num_dividers, int max_div,
>   	num_dividers = i;
>   
>   	tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);
> -	if (!tmp)
> +	if (!tmp) {
> +		*table = PTR_ERR(-ENOMEM);
>   		return -ENOMEM;
> +	}
>   
>   	valid_div = 0;
>   	*width = 0;
> @@ -439,6 +441,7 @@ struct clk_hw *ti_clk_build_component_div(struct ti_clk_divider *setup)
>   {
>   	struct clk_omap_divider *div;
>   	struct clk_omap_reg *reg;
> +	int ret;
>   
>   	if (!setup)
>   		return NULL;
> @@ -458,6 +461,12 @@ struct clk_hw *ti_clk_build_component_div(struct ti_clk_divider *setup)
>   		div->flags |= CLK_DIVIDER_POWER_OF_TWO;
>   
>   	div->table = _get_div_table_from_setup(setup, &div->width);
> +	if (IS_ERR(div->table)) {
> +		ret = PTR_ERR(div->table);
> +		kfree(div);
> +		return ret;
> +	}
> +
>   
>   	div->shift = setup->bit_shift;
>   	div->latch = -EINVAL;
> 

--
Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki. Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki

Re: [PATCH] clk: ti: Fix error handling in ti_clk_parse_divider_data()

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 4983 bytes --]

Hi Dan,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on clk/clk-next]
[also build test WARNING on v5.0-rc2 next-20190115]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Dan-Carpenter/clk-ti-Fix-error-handling-in-ti_clk_parse_divider_data/20190115-175541
base:   https://git.kernel.org/pub/scm/linux/kernel/git/clk/linux.git clk-next
config: arm-allmodconfig (attached as .config)
compiler: arm-linux-gnueabi-gcc (Debian 7.2.0-11) 7.2.0
reproduce:
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        GCC_VERSION=7.2.0 make.cross ARCH=arm 

All warnings (new ones prefixed by >>):

   drivers/clk/ti/divider.c: In function 'ti_clk_parse_divider_data':
>> drivers/clk/ti/divider.c:407:20: warning: passing argument 1 of 'PTR_ERR' makes pointer from integer without a cast [-Wint-conversion]
      *table = PTR_ERR(-ENOMEM);
                       ^
   In file included from include/linux/io.h:24:0,
                    from include/linux/clk-provider.h:9,
                    from drivers/clk/ti/divider.c:18:
   include/linux/err.h:29:33: note: expected 'const void *' but argument is of type 'int'
    static inline long __must_check PTR_ERR(__force const void *ptr)
                                    ^~~~~~~
>> drivers/clk/ti/divider.c:407:10: warning: assignment makes pointer from integer without a cast [-Wint-conversion]
      *table = PTR_ERR(-ENOMEM);
             ^
   drivers/clk/ti/divider.c: In function 'ti_clk_build_component_div':
>> drivers/clk/ti/divider.c:467:10: warning: return makes pointer from integer without a cast [-Wint-conversion]
      return ret;
             ^~~

vim +/PTR_ERR +407 drivers/clk/ti/divider.c

   360	
   361	int ti_clk_parse_divider_data(int *div_table, int num_dividers, int max_div,
   362				      u8 flags, u8 *width,
   363				      const struct clk_div_table **table)
   364	{
   365		int valid_div = 0;
   366		u32 val;
   367		int div;
   368		int i;
   369		struct clk_div_table *tmp;
   370	
   371		if (!div_table) {
   372			if (flags & CLKF_INDEX_STARTS_AT_ONE)
   373				val = 1;
   374			else
   375				val = 0;
   376	
   377			div = 1;
   378	
   379			while (div < max_div) {
   380				if (flags & CLKF_INDEX_POWER_OF_TWO)
   381					div <<= 1;
   382				else
   383					div++;
   384				val++;
   385			}
   386	
   387			*width = fls(val);
   388			*table = NULL;
   389	
   390			return 0;
   391		}
   392	
   393		i = 0;
   394	
   395		while (!num_dividers || i < num_dividers) {
   396			if (div_table[i] == -1)
   397				break;
   398			if (div_table[i])
   399				valid_div++;
   400			i++;
   401		}
   402	
   403		num_dividers = i;
   404	
   405		tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);
   406		if (!tmp) {
 > 407			*table = PTR_ERR(-ENOMEM);
   408			return -ENOMEM;
   409		}
   410	
   411		valid_div = 0;
   412		*width = 0;
   413	
   414		for (i = 0; i < num_dividers; i++)
   415			if (div_table[i] > 0) {
   416				tmp[valid_div].div = div_table[i];
   417				tmp[valid_div].val = i;
   418				valid_div++;
   419				*width = i;
   420			}
   421	
   422		*width = fls(*width);
   423		*table = tmp;
   424	
   425		return 0;
   426	}
   427	
   428	static const struct clk_div_table *
   429	_get_div_table_from_setup(struct ti_clk_divider *setup, u8 *width)
   430	{
   431		const struct clk_div_table *table = NULL;
   432	
   433		ti_clk_parse_divider_data(setup->dividers, setup->num_dividers,
   434					  setup->max_div, setup->flags, width,
   435					  &table);
   436	
   437		return table;
   438	}
   439	
   440	struct clk_hw *ti_clk_build_component_div(struct ti_clk_divider *setup)
   441	{
   442		struct clk_omap_divider *div;
   443		struct clk_omap_reg *reg;
   444		int ret;
   445	
   446		if (!setup)
   447			return NULL;
   448	
   449		div = kzalloc(sizeof(*div), GFP_KERNEL);
   450		if (!div)
   451			return ERR_PTR(-ENOMEM);
   452	
   453		reg = (struct clk_omap_reg *)&div->reg;
   454		reg->index = setup->module;
   455		reg->offset = setup->reg;
   456	
   457		if (setup->flags & CLKF_INDEX_STARTS_AT_ONE)
   458			div->flags |= CLK_DIVIDER_ONE_BASED;
   459	
   460		if (setup->flags & CLKF_INDEX_POWER_OF_TWO)
   461			div->flags |= CLK_DIVIDER_POWER_OF_TWO;
   462	
   463		div->table = _get_div_table_from_setup(setup, &div->width);
   464		if (IS_ERR(div->table)) {
   465			ret = PTR_ERR(div->table);
   466			kfree(div);
 > 467			return ret;
   468		}
   469	
   470	
   471		div->shift = setup->bit_shift;
   472		div->latch = -EINVAL;
   473	
   474		return &div->hw;
   475	}
   476	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 68457 bytes --]

Re: [PATCH] clk: ti: Fix error handling in ti_clk_parse_divider_data()

[Dan Carpenter]

On Tue, Jan 15, 2019 at 10:01:49AM +0300, Dan Carpenter wrote:
> The ti_clk_parse_divider_data() function is only called from
> _get_div_table_from_setup().  That function doesn't look at the return
> value but instead looks at the "*table" pointer.  In this case, if the
> kcalloc() fails then *table is NULL (which means success).  It should
> instead be an error pointer.
> 
> The ti_clk_parse_divider_data() function has two callers.  One checks
> for errors and the other doesn't.  I have fixed it so now both handle
> errors.
> 
> Fixes: 4f6be5655dc9 ("clk: ti: divider: add driver internal API for parsing divider data")
> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> ---
>  drivers/clk/ti/divider.c | 11 ++++++++++-
>  1 file changed, 10 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/clk/ti/divider.c b/drivers/clk/ti/divider.c
> index 8d77090ad94a..4c48ef424ad5 100644
> --- a/drivers/clk/ti/divider.c
> +++ b/drivers/clk/ti/divider.c
> @@ -403,8 +403,10 @@ int ti_clk_parse_divider_data(int *div_table, int num_dividers, int max_div,
>  	num_dividers = i;
>  
>  	tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);
> -	if (!tmp)
> +	if (!tmp) {
> +		*table = PTR_ERR(-ENOMEM);

Oh wow...  I don't know how I screwed that up.  :(

Let me resend.

regards,
dan carpenter

Re: [PATCH] clk: ti: Fix error handling in ti_clk_parse_divider_data()

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 5080 bytes --]

Hi Dan,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on clk/clk-next]
[also build test WARNING on v5.0-rc2 next-20190115]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Dan-Carpenter/clk-ti-Fix-error-handling-in-ti_clk_parse_divider_data/20190115-175541
base:   https://git.kernel.org/pub/scm/linux/kernel/git/clk/linux.git clk-next
config: arm-allmodconfig (attached as .config)
compiler: arm-linux-gnueabi-gcc (Debian 8.2.0-11) 8.2.0
reproduce:
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        GCC_VERSION=8.2.0 make.cross ARCH=arm 

All warnings (new ones prefixed by >>):

   drivers/clk/ti/divider.c: In function 'ti_clk_parse_divider_data':
   drivers/clk/ti/divider.c:407:20: warning: passing argument 1 of 'PTR_ERR' makes pointer from integer without a cast [-Wint-conversion]
      *table = PTR_ERR(-ENOMEM);
   In file included from include/linux/io.h:24,
                    from include/linux/clk-provider.h:9,
                    from drivers/clk/ti/divider.c:18:
   include/linux/err.h:29:61: note: expected 'const void *' but argument is of type 'int'
    static inline long __must_check PTR_ERR(__force const void *ptr)
                                                    ~~~~~~~~~~~~^~~
>> drivers/clk/ti/divider.c:407:10: warning: assignment to 'const struct clk_div_table *' from 'long int' makes pointer from integer without a cast [-Wint-conversion]
      *table = PTR_ERR(-ENOMEM);
             ^
   drivers/clk/ti/divider.c: In function 'ti_clk_build_component_div':
>> drivers/clk/ti/divider.c:467:10: warning: returning 'int' from a function with return type 'struct clk_hw *' makes pointer from integer without a cast [-Wint-conversion]
      return ret;
             ^~~

vim +407 drivers/clk/ti/divider.c

   360	
   361	int ti_clk_parse_divider_data(int *div_table, int num_dividers, int max_div,
   362				      u8 flags, u8 *width,
   363				      const struct clk_div_table **table)
   364	{
   365		int valid_div = 0;
   366		u32 val;
   367		int div;
   368		int i;
   369		struct clk_div_table *tmp;
   370	
   371		if (!div_table) {
   372			if (flags & CLKF_INDEX_STARTS_AT_ONE)
   373				val = 1;
   374			else
   375				val = 0;
   376	
   377			div = 1;
   378	
   379			while (div < max_div) {
   380				if (flags & CLKF_INDEX_POWER_OF_TWO)
   381					div <<= 1;
   382				else
   383					div++;
   384				val++;
   385			}
   386	
   387			*width = fls(val);
   388			*table = NULL;
   389	
   390			return 0;
   391		}
   392	
   393		i = 0;
   394	
   395		while (!num_dividers || i < num_dividers) {
   396			if (div_table[i] == -1)
   397				break;
   398			if (div_table[i])
   399				valid_div++;
   400			i++;
   401		}
   402	
   403		num_dividers = i;
   404	
   405		tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);
   406		if (!tmp) {
 > 407			*table = PTR_ERR(-ENOMEM);
   408			return -ENOMEM;
   409		}
   410	
   411		valid_div = 0;
   412		*width = 0;
   413	
   414		for (i = 0; i < num_dividers; i++)
   415			if (div_table[i] > 0) {
   416				tmp[valid_div].div = div_table[i];
   417				tmp[valid_div].val = i;
   418				valid_div++;
   419				*width = i;
   420			}
   421	
   422		*width = fls(*width);
   423		*table = tmp;
   424	
   425		return 0;
   426	}
   427	
   428	static const struct clk_div_table *
   429	_get_div_table_from_setup(struct ti_clk_divider *setup, u8 *width)
   430	{
   431		const struct clk_div_table *table = NULL;
   432	
   433		ti_clk_parse_divider_data(setup->dividers, setup->num_dividers,
   434					  setup->max_div, setup->flags, width,
   435					  &table);
   436	
   437		return table;
   438	}
   439	
   440	struct clk_hw *ti_clk_build_component_div(struct ti_clk_divider *setup)
   441	{
   442		struct clk_omap_divider *div;
   443		struct clk_omap_reg *reg;
   444		int ret;
   445	
   446		if (!setup)
   447			return NULL;
   448	
   449		div = kzalloc(sizeof(*div), GFP_KERNEL);
   450		if (!div)
   451			return ERR_PTR(-ENOMEM);
   452	
   453		reg = (struct clk_omap_reg *)&div->reg;
   454		reg->index = setup->module;
   455		reg->offset = setup->reg;
   456	
   457		if (setup->flags & CLKF_INDEX_STARTS_AT_ONE)
   458			div->flags |= CLK_DIVIDER_ONE_BASED;
   459	
   460		if (setup->flags & CLKF_INDEX_POWER_OF_TWO)
   461			div->flags |= CLK_DIVIDER_POWER_OF_TWO;
   462	
   463		div->table = _get_div_table_from_setup(setup, &div->width);
   464		if (IS_ERR(div->table)) {
   465			ret = PTR_ERR(div->table);
   466			kfree(div);
 > 467			return ret;
   468		}
   469	
   470	
   471		div->shift = setup->bit_shift;
   472		div->latch = -EINVAL;
   473	
   474		return &div->hw;
   475	}
   476	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 68400 bytes --]

[PATCH v2] clk: ti: Fix error handling in ti_clk_parse_divider_data()

[Dan Carpenter]

The ti_clk_parse_divider_data() function is only called from
_get_div_table_from_setup().  That function doesn't look at the return
value but instead looks at the "*table" pointer.  In this case, if the
kcalloc() fails then *table is NULL (which means success).  It should
instead be an error pointer.

The ti_clk_parse_divider_data() function has two callers.  One checks
for errors and the other doesn't.  I have fixed it so now both handle
errors.

Fixes: 4f6be5655dc9 ("clk: ti: divider: add driver internal API for parsing divider data")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
---
v2:  I somehow didn't compile the v1 of this patch so it had a couple
     stupid bugs.  I'm very sorry.

 drivers/clk/ti/divider.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

diff --git a/drivers/clk/ti/divider.c b/drivers/clk/ti/divider.c
index 8d77090ad94a..0241450f3eb3 100644
--- a/drivers/clk/ti/divider.c
+++ b/drivers/clk/ti/divider.c
@@ -403,8 +403,10 @@ int ti_clk_parse_divider_data(int *div_table, int num_dividers, int max_div,
 	num_dividers = i;
 
 	tmp = kcalloc(valid_div + 1, sizeof(*tmp), GFP_KERNEL);
-	if (!tmp)
+	if (!tmp) {
+		*table = ERR_PTR(-ENOMEM);
 		return -ENOMEM;
+	}
 
 	valid_div = 0;
 	*width = 0;
@@ -439,6 +441,7 @@ struct clk_hw *ti_clk_build_component_div(struct ti_clk_divider *setup)
 {
 	struct clk_omap_divider *div;
 	struct clk_omap_reg *reg;
+	int ret;
 
 	if (!setup)
 		return NULL;
@@ -458,6 +461,12 @@ struct clk_hw *ti_clk_build_component_div(struct ti_clk_divider *setup)
 		div->flags |= CLK_DIVIDER_POWER_OF_TWO;
 
 	div->table = _get_div_table_from_setup(setup, &div->width);
+	if (IS_ERR(div->table)) {
+		ret = PTR_ERR(div->table);
+		kfree(div);
+		return ERR_PTR(ret);
+	}
+
 
 	div->shift = setup->bit_shift;
 	div->latch = -EINVAL;
-- 
2.17.1

Re: [PATCH v2] clk: ti: Fix error handling in ti_clk_parse_divider_data()

[Stephen Boyd]

Quoting Dan Carpenter (2019-01-15 11:46:25)
> The ti_clk_parse_divider_data() function is only called from
> _get_div_table_from_setup().  That function doesn't look at the return
> value but instead looks at the "*table" pointer.  In this case, if the
> kcalloc() fails then *table is NULL (which means success).  It should
> instead be an error pointer.
> 
> The ti_clk_parse_divider_data() function has two callers.  One checks
> for errors and the other doesn't.  I have fixed it so now both handle
> errors.
> 
> Fixes: 4f6be5655dc9 ("clk: ti: divider: add driver internal API for parsing divider data")
> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> ---

Applied to clk-fixes

I'm going to add Tero's ack to this because it isn't really that
different.

Re: [PATCH v2] clk: ti: Fix error handling in ti_clk_parse_divider_data()

[Tero Kristo]

On 24/01/2019 21:24, Stephen Boyd wrote:
> Quoting Dan Carpenter (2019-01-15 11:46:25)
>> The ti_clk_parse_divider_data() function is only called from
>> _get_div_table_from_setup().  That function doesn't look at the return
>> value but instead looks at the "*table" pointer.  In this case, if the
>> kcalloc() fails then *table is NULL (which means success).  It should
>> instead be an error pointer.
>>
>> The ti_clk_parse_divider_data() function has two callers.  One checks
>> for errors and the other doesn't.  I have fixed it so now both handle
>> errors.
>>
>> Fixes: 4f6be5655dc9 ("clk: ti: divider: add driver internal API for parsing divider data")
>> Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
>> ---
> 
> Applied to clk-fixes
> 
> I'm going to add Tero's ack to this because it isn't really that
> different.
> 

Yeah, thats ok. Thanks Stephen.

-Tero
--
Texas Instruments Finland Oy, Porkkalankatu 22, 00180 Helsinki. Y-tunnus/Business ID: 0615521-4. Kotipaikka/Domicile: Helsinki