From: Colin Ian King <[email protected]>
Currently the kfree of pointer qfprom_cdata is kfreeing an
error value that has been cast to a pointer rather than a
valid address. Fix this by removing the kfree.
Fixes: 95ededc17e4e ("thermal: qcom: tsens-v1: Add support for MSM8956 and MSM8976")
Signed-off-by: Colin Ian King <[email protected]>
---
drivers/thermal/qcom/tsens-v1.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/thermal/qcom/tsens-v1.c b/drivers/thermal/qcom/tsens-v1.c
index 2d1077b64887..bd2ddb684a45 100644
--- a/drivers/thermal/qcom/tsens-v1.c
+++ b/drivers/thermal/qcom/tsens-v1.c
@@ -240,10 +240,8 @@ static int calibrate_8976(struct tsens_priv *priv)
u32 *qfprom_cdata;
qfprom_cdata = (u32 *)qfprom_read(priv->dev, "calib");
- if (IS_ERR(qfprom_cdata)) {
- kfree(qfprom_cdata);
+ if (IS_ERR(qfprom_cdata))
return PTR_ERR(qfprom_cdata);
- }
mode = (qfprom_cdata[4] & MSM8976_CAL_SEL_MASK);
dev_dbg(priv->dev, "calibration mode is %d\n", mode);
--
2.20.1
On Tue, Oct 22, 2019 at 4:48 PM Colin King <[email protected]> wrote:
>
> From: Colin Ian King <[email protected]>
>
> Currently the kfree of pointer qfprom_cdata is kfreeing an
> error value that has been cast to a pointer rather than a
> valid address. Fix this by removing the kfree.
Hmm, we just added this to other places[1] as a fix for mem leaks
using the nvmem api.
I think we need to fix up the qfprom_read wrapper. Srini?
[1] https://lore.kernel.org/stable/[email protected]/
> Fixes: 95ededc17e4e ("thermal: qcom: tsens-v1: Add support for MSM8956 and MSM8976")
> Signed-off-by: Colin Ian King <[email protected]>
> ---
> drivers/thermal/qcom/tsens-v1.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/thermal/qcom/tsens-v1.c b/drivers/thermal/qcom/tsens-v1.c
> index 2d1077b64887..bd2ddb684a45 100644
> --- a/drivers/thermal/qcom/tsens-v1.c
> +++ b/drivers/thermal/qcom/tsens-v1.c
> @@ -240,10 +240,8 @@ static int calibrate_8976(struct tsens_priv *priv)
> u32 *qfprom_cdata;
>
> qfprom_cdata = (u32 *)qfprom_read(priv->dev, "calib");
> - if (IS_ERR(qfprom_cdata)) {
> - kfree(qfprom_cdata);
> + if (IS_ERR(qfprom_cdata))
> return PTR_ERR(qfprom_cdata);
> - }
>
> mode = (qfprom_cdata[4] & MSM8976_CAL_SEL_MASK);
> dev_dbg(priv->dev, "calibration mode is %d\n", mode);
> --
> 2.20.1
>
Il giorno mar 22 ott 2019 alle ore 13:18 Colin King
<[email protected]> ha scritto:
>
> From: Colin Ian King <[email protected]>
>
> Currently the kfree of pointer qfprom_cdata is kfreeing an
> error value that has been cast to a pointer rather than a
> valid address. Fix this by removing the kfree.
>
> Fixes: 95ededc17e4e ("thermal: qcom: tsens-v1: Add support for MSM8956 and MSM8976")
> Signed-off-by: Colin Ian King <[email protected]>
> ---
> drivers/thermal/qcom/tsens-v1.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/thermal/qcom/tsens-v1.c b/drivers/thermal/qcom/tsens-v1.c
> index 2d1077b64887..bd2ddb684a45 100644
> --- a/drivers/thermal/qcom/tsens-v1.c
> +++ b/drivers/thermal/qcom/tsens-v1.c
> @@ -240,10 +240,8 @@ static int calibrate_8976(struct tsens_priv *priv)
> u32 *qfprom_cdata;
>
> qfprom_cdata = (u32 *)qfprom_read(priv->dev, "calib");
> - if (IS_ERR(qfprom_cdata)) {
> - kfree(qfprom_cdata);
> + if (IS_ERR(qfprom_cdata))
> return PTR_ERR(qfprom_cdata);
> - }
>
> mode = (qfprom_cdata[4] & MSM8976_CAL_SEL_MASK);
> dev_dbg(priv->dev, "calibration mode is %d\n", mode);
> --
> 2.20.1
>
I confirm that was one stupid mistake. I was about to send the same patch, and I
can confirm that this fix is working. Tested on my Xperia X Compact.
Tested-by: AngeloGioacchino Del Regno <[email protected]>
On 22/10/2019 14:06, Amit Kucheria wrote:
> On Tue, Oct 22, 2019 at 4:48 PM Colin King <[email protected]> wrote:
>>
>> From: Colin Ian King <[email protected]>
>>
>> Currently the kfree of pointer qfprom_cdata is kfreeing an
>> error value that has been cast to a pointer rather than a
>> valid address. Fix this by removing the kfree.
>
> Hmm, we just added this to other places[1] as a fix for mem leaks
> using the nvmem api. >
This patch has nothing to do with the memleak fix.
> I think we need to fix up the qfprom_read wrapper. Srini?
not sure how we can fix that, as the pointer returned from read is needs
to be freed by the caller after its done with it!
--srini
>
> [1] https://lore.kernel.org/stable/[email protected]/
>
>> Fixes: 95ededc17e4e ("thermal: qcom: tsens-v1: Add support for MSM8956 and MSM8976")
>> Signed-off-by: Colin Ian King <[email protected]>
>> ---
>> drivers/thermal/qcom/tsens-v1.c | 4 +---
>> 1 file changed, 1 insertion(+), 3 deletions(-)
>>
>> diff --git a/drivers/thermal/qcom/tsens-v1.c b/drivers/thermal/qcom/tsens-v1.c
>> index 2d1077b64887..bd2ddb684a45 100644
>> --- a/drivers/thermal/qcom/tsens-v1.c
>> +++ b/drivers/thermal/qcom/tsens-v1.c
>> @@ -240,10 +240,8 @@ static int calibrate_8976(struct tsens_priv *priv)
>> u32 *qfprom_cdata;
>>
>> qfprom_cdata = (u32 *)qfprom_read(priv->dev, "calib");
>> - if (IS_ERR(qfprom_cdata)) {
>> - kfree(qfprom_cdata);
>> + if (IS_ERR(qfprom_cdata))
>> return PTR_ERR(qfprom_cdata);
>> - }
>>
>> mode = (qfprom_cdata[4] & MSM8976_CAL_SEL_MASK);
>> dev_dbg(priv->dev, "calibration mode is %d\n", mode);
>> --
>> 2.20.1
>>