Re: [PATCH 1/3] auxdisplay: ht16k33: do not try to free fbdev

[Robin van der Gracht]

On Tue, 31 Jan 2017 12:54:36 -0800
Dmitry Torokhov <dmitry.torokhov@xxxxxxxxx> wrote:

> 'fbdev' is allocated as part of larger ht16k33_priv structure; trying to
> free it will cause troubles.
> 
> Signed-off-by: Dmitry Torokhov <dmitry.torokhov@xxxxxxxxx>
> ---
> 
> Note that the patches have not been tried on a real hardware.
> 
> I am pretty confident in #1 and #3, but #2 is larger and it would be great
> if someone with hardware tried running it before it gets applied.
> 
> Thanks!
> 
>  drivers/auxdisplay/ht16k33.c | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/drivers/auxdisplay/ht16k33.c b/drivers/auxdisplay/ht16k33.c
> index eeb323f56c07..f2f304b3f061 100644
> --- a/drivers/auxdisplay/ht16k33.c
> +++ b/drivers/auxdisplay/ht16k33.c
> @@ -378,7 +378,7 @@ static int ht16k33_probe(struct i2c_client *client,
>  	fbdev->buffer = (unsigned char *) get_zeroed_page(GFP_KERNEL);
>  	if (!fbdev->buffer) {
>  		err = -ENOMEM;
> -		goto err_free_fbdev;
> +		goto err_destroy_wq;
>  	}
>  
>  	fbdev->cache = devm_kmalloc(&client->dev, HT16K33_FB_SIZE, GFP_KERNEL);
> @@ -510,8 +510,6 @@ static int ht16k33_probe(struct i2c_client *client,
>  	framebuffer_release(fbdev->info);
>  err_fbdev_buffer:
>  	free_page((unsigned long) fbdev->buffer);
> -err_free_fbdev:
> -	kfree(fbdev);
>  err_destroy_wq:
>  	destroy_workqueue(priv->workqueue);
>  

Acked-by: Robin van der Gracht <robin@xxxxxxxxxxx>