hd44780_probe() allocates a memory chunk for hd with kzalloc() and
makes "lcd->drvdata->hd44780" point to it. When we call hd44780_remove(),
we should release all relevant memory and resource. But "lcd->drvdata
->hd44780" is not released, which will lead to a memory leak.
We should release the "lcd->drvdata->hd44780" in hd44780_remove() to fix
the memory leak bug.
Fixes: 41c8d0adf3c4 ("auxdisplay: hd44780: Fix memory leak on ->remove()")
Reported-by: kernel test robot <[email protected]>
Signed-off-by: Jianglei Nie <[email protected]>
---
drivers/auxdisplay/hd44780.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/auxdisplay/hd44780.c b/drivers/auxdisplay/hd44780.c
index 8b2a0eb3f32a..f4800a946bd8 100644
--- a/drivers/auxdisplay/hd44780.c
+++ b/drivers/auxdisplay/hd44780.c
@@ -322,8 +322,10 @@ static int hd44780_probe(struct platform_device *pdev)
static int hd44780_remove(struct platform_device *pdev)
{
struct charlcd *lcd = platform_get_drvdata(pdev);
+ struct hd44780_common *hdc = (struct hd44780_common *)lcd->drvdata;
charlcd_unregister(lcd);
+ kfree(hdc->hd44780);
kfree(lcd->drvdata);
kfree(lcd);
--
2.25.1
On Mon, Nov 28, 2022 at 04:44:21PM +0800, Jianglei Nie wrote:
> hd44780_probe() allocates a memory chunk for hd with kzalloc() and
> makes "lcd->drvdata->hd44780" point to it. When we call hd44780_remove(),
> we should release all relevant memory and resource. But "lcd->drvdata
> ->hd44780" is not released, which will lead to a memory leak.
>
> We should release the "lcd->drvdata->hd44780" in hd44780_remove() to fix
> the memory leak bug.
Better now! See my comments below. After addressing them,
Reviewed-by: Andy Shevchenko <[email protected]>
> Fixes: 41c8d0adf3c4 ("auxdisplay: hd44780: Fix memory leak on ->remove()")
Fixes: 718e05ed92ec ("auxdisplay: Introduce hd44780_common.[ch]")
What you found has nothing to do with the issue. Issue has been introduced
later on.
> Reported-by: kernel test robot <[email protected]>
> Signed-off-by: Jianglei Nie <[email protected]>
> ---
> drivers/auxdisplay/hd44780.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/auxdisplay/hd44780.c b/drivers/auxdisplay/hd44780.c
> index 8b2a0eb3f32a..f4800a946bd8 100644
> --- a/drivers/auxdisplay/hd44780.c
> +++ b/drivers/auxdisplay/hd44780.c
> @@ -322,8 +322,10 @@ static int hd44780_probe(struct platform_device *pdev)
> static int hd44780_remove(struct platform_device *pdev)
> {
> struct charlcd *lcd = platform_get_drvdata(pdev);
> + struct hd44780_common *hdc = (struct hd44780_common *)lcd->drvdata;
drvdata member is type of void *, no need to do an explicit casting as
per C standard.
> charlcd_unregister(lcd);
> + kfree(hdc->hd44780);
> kfree(lcd->drvdata);
>
> kfree(lcd);
> --
> 2.25.1
>
--
With Best Regards,
Andy Shevchenko
On Mon, Nov 28, 2022 at 12:45:39PM +0200, Andy Shevchenko wrote:
> On Mon, Nov 28, 2022 at 04:44:21PM +0800, Jianglei Nie wrote:
...
> Fixes: 718e05ed92ec ("auxdisplay: Introduce hd44780_common.[ch]")
>
> What you found has nothing to do with the issue. Issue has been introduced
> later on.
Side note (mostly for Miguel): That series by Lars was indeed problematic.
And I see now that he didn't get the parameter to the charlcd_alloc(). Now
we have problem that your patch solves and dangling parameter in the struct
charlcd_priv. So, I will restore charlcd_alloc() before his series (after
this patch has been applied, because of the backport needs) for a new
kernel development cycle.
--
With Best Regards,
Andy Shevchenko