2022-04-18 12:38:05

by Hangyu Hua

Subject: [PATCH] rpmsg: virtio: fix possible double free in rpmsg_probe()

vch will be free in virtio_rpmsg_release_device() when
rpmsg_ns_register_device() fails. There is no need to call kfree() again.

Fix this by changing error path from free_vch to free_ctrldev.

Fixes: c486682ae1e2 ("rpmsg: virtio: Register the rpmsg_char device")
Signed-off-by: Hangyu Hua <[email protected]>
---
drivers/rpmsg/virtio_rpmsg_bus.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/rpmsg/virtio_rpmsg_bus.c b/drivers/rpmsg/virtio_rpmsg_bus.c
index 3ede25b1f2e4..603233f0686e 100644
--- a/drivers/rpmsg/virtio_rpmsg_bus.c
+++ b/drivers/rpmsg/virtio_rpmsg_bus.c
@@ -973,7 +973,7 @@ static int rpmsg_probe(struct virtio_device *vdev)

err = rpmsg_ns_register_device(rpdev_ns);
if (err)
- goto free_vch;
+ goto free_ctrldev;
}

/*
@@ -997,8 +997,6 @@ static int rpmsg_probe(struct virtio_device *vdev)

return 0;

-free_vch:
- kfree(vch);
free_ctrldev:
rpmsg_virtio_del_ctrl_dev(rpdev_ctrl);
free_coherent:
--
2.25.1

2022-04-22 20:50:20

by Arnaud Pouliquen

Subject: Re: [PATCH] rpmsg: virtio: fix possible double free in rpmsg_probe()

On 4/18/22 11:31, Hangyu Hua wrote:
> vch will be free in virtio_rpmsg_release_device() when
> rpmsg_ns_register_device() fails. There is no need to call kfree() again.
>
> Fix this by changing error path from free_vch to free_ctrldev.
>
> Fixes: c486682ae1e2 ("rpmsg: virtio: Register the rpmsg_char device")
> Signed-off-by: Hangyu Hua <[email protected]>

Tested-by: Arnaud Pouliquen <[email protected]>

Thanks,
Arnaud

> ---
> drivers/rpmsg/virtio_rpmsg_bus.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/rpmsg/virtio_rpmsg_bus.c b/drivers/rpmsg/virtio_rpmsg_bus.c
> index 3ede25b1f2e4..603233f0686e 100644
> --- a/drivers/rpmsg/virtio_rpmsg_bus.c
> +++ b/drivers/rpmsg/virtio_rpmsg_bus.c
> @@ -973,7 +973,7 @@ static int rpmsg_probe(struct virtio_device *vdev)
>
> err = rpmsg_ns_register_device(rpdev_ns);
> if (err)
> - goto free_vch;
> + goto free_ctrldev;
> }
>
> /*
> @@ -997,8 +997,6 @@ static int rpmsg_probe(struct virtio_device *vdev)
>
> return 0;
>
> -free_vch:
> - kfree(vch);
> free_ctrldev:
> rpmsg_virtio_del_ctrl_dev(rpdev_ctrl);
> free_coherent:

2022-04-25 22:24:04

by Mathieu Poirier

Subject: Re: [PATCH] rpmsg: virtio: fix possible double free in rpmsg_probe()

On Mon, Apr 18, 2022 at 05:31:44PM +0800, Hangyu Hua wrote:
> vch will be free in virtio_rpmsg_release_device() when
> rpmsg_ns_register_device() fails. There is no need to call kfree() again.
>
> Fix this by changing error path from free_vch to free_ctrldev.
>
> Fixes: c486682ae1e2 ("rpmsg: virtio: Register the rpmsg_char device")
> Signed-off-by: Hangyu Hua <[email protected]>
> ---
> drivers/rpmsg/virtio_rpmsg_bus.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/rpmsg/virtio_rpmsg_bus.c b/drivers/rpmsg/virtio_rpmsg_bus.c
> index 3ede25b1f2e4..603233f0686e 100644
> --- a/drivers/rpmsg/virtio_rpmsg_bus.c
> +++ b/drivers/rpmsg/virtio_rpmsg_bus.c
> @@ -973,7 +973,7 @@ static int rpmsg_probe(struct virtio_device *vdev)
>
> err = rpmsg_ns_register_device(rpdev_ns);
> if (err)
> - goto free_vch;

Please add a comment that highlights where 'vch' will be free'd to avoid
receiving patches that will introduce another kfree(). Same for your other
patch.

In the next revision please use a cover letter and add Arnaud's patches to it.

Thanks,
Mathieu

> + goto free_ctrldev;
> }
>
> /*
> @@ -997,8 +997,6 @@ static int rpmsg_probe(struct virtio_device *vdev)
>
> return 0;
>
> -free_vch:
> - kfree(vch);
> free_ctrldev:
> rpmsg_virtio_del_ctrl_dev(rpdev_ctrl);
> free_coherent:
> --
> 2.25.1
>

2022-04-26 09:03:40

by Hangyu Hua

Subject: Re: [PATCH] rpmsg: virtio: fix possible double free in rpmsg_probe()

On 2022/4/26 00:55, Mathieu Poirier wrote:
> On Mon, Apr 18, 2022 at 05:31:44PM +0800, Hangyu Hua wrote:
>> vch will be free in virtio_rpmsg_release_device() when
>> rpmsg_ns_register_device() fails. There is no need to call kfree() again.
>>
>> Fix this by changing error path from free_vch to free_ctrldev.
>>
>> Fixes: c486682ae1e2 ("rpmsg: virtio: Register the rpmsg_char device")
>> Signed-off-by: Hangyu Hua <[email protected]>
>> ---
>> drivers/rpmsg/virtio_rpmsg_bus.c | 4 +---
>> 1 file changed, 1 insertion(+), 3 deletions(-)
>>
>> diff --git a/drivers/rpmsg/virtio_rpmsg_bus.c b/drivers/rpmsg/virtio_rpmsg_bus.c
>> index 3ede25b1f2e4..603233f0686e 100644
>> --- a/drivers/rpmsg/virtio_rpmsg_bus.c
>> +++ b/drivers/rpmsg/virtio_rpmsg_bus.c
>> @@ -973,7 +973,7 @@ static int rpmsg_probe(struct virtio_device *vdev)
>>
>> err = rpmsg_ns_register_device(rpdev_ns);
>> if (err)
>> - goto free_vch;
>
> Please add a comment that highlights where 'vch' will be free'd to avoid
> receiving patches that will introduce another kfree(). Same for your other
> patch.
>
> In the next revision please use a cover letter and add Arnaud's patches to it.
>
> Thanks,
> Mathieu
>

Thanks! I will send a v2 later.

>> + goto free_ctrldev;
>> }
>>
>> /*
>> @@ -997,8 +997,6 @@ static int rpmsg_probe(struct virtio_device *vdev)
>>
>> return 0;
>>
>> -free_vch:
>> - kfree(vch);
>> free_ctrldev:
>> rpmsg_virtio_del_ctrl_dev(rpdev_ctrl);
>> free_coherent:
>> --
>> 2.25.1
>>