2022-08-08 04:18:30

by Lin Ma

[permalink] [raw]
Subject: [PATCH v0] ieee802154/adf7242: defer destroy_workqueue call

There is a possible race condition (use-after-free) like below

(FREE) | (USE)
adf7242_remove | adf7242_channel
cancel_delayed_work_sync |
destroy_workqueue (1) | adf7242_cmd_rx
| mod_delayed_work (2)
|

The root cause for this race is that the upper layer (ieee802154) is
unaware of this detaching event and the function adf7242_channel can
be called without any checks.

To fix this, we can add a flag write at the beginning of adf7242_remove
and add flag check in adf7242_channel. Or we can just defer the
destructive operation like other commit 3e0588c291d6 ("hamradio: defer
ax25 kfree after unregister_netdev") which let the
ieee802154_unregister_hw() to handle the synchronization. This patch
takes the second option.

Fixes: 58e9683d1475 ("net: ieee802154: adf7242: Fix OCL calibration
runs")
Signed-off-by: Lin Ma <[email protected]>
---
drivers/net/ieee802154/adf7242.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ieee802154/adf7242.c b/drivers/net/ieee802154/adf7242.c
index 6afdf1622944..efcc45aef508 100644
--- a/drivers/net/ieee802154/adf7242.c
+++ b/drivers/net/ieee802154/adf7242.c
@@ -1310,10 +1310,11 @@ static void adf7242_remove(struct spi_device *spi)

debugfs_remove_recursive(lp->debugfs_root);

+ ieee802154_unregister_hw(lp->hw);
+
cancel_delayed_work_sync(&lp->work);
destroy_workqueue(lp->wqueue);
-
- ieee802154_unregister_hw(lp->hw);
+
mutex_destroy(&lp->bmux);
ieee802154_free_hw(lp->hw);
}
--
2.36.1


2022-08-08 08:23:16

by Hennerich, Michael

[permalink] [raw]
Subject: RE: [PATCH v0] ieee802154/adf7242: defer destroy_workqueue call



> -----Original Message-----
> From: Lin Ma <[email protected]>
> Sent: Montag, 8. August 2022 05:42
> To: Hennerich, Michael <[email protected]>;
> [email protected]; [email protected]; [email protected];
> [email protected]; [email protected]; [email protected]; linux-
> [email protected]; [email protected]; linux-
> [email protected]
> Cc: Lin Ma <[email protected]>
> Subject: [PATCH v0] ieee802154/adf7242: defer destroy_workqueue call
>
>
> There is a possible race condition (use-after-free) like below
>
> (FREE) | (USE)
> adf7242_remove | adf7242_channel
> cancel_delayed_work_sync |
> destroy_workqueue (1) | adf7242_cmd_rx
> | mod_delayed_work (2)
> |
>
> The root cause for this race is that the upper layer (ieee802154) is unaware of
> this detaching event and the function adf7242_channel can be called without
> any checks.
>
> To fix this, we can add a flag write at the beginning of adf7242_remove and
> add flag check in adf7242_channel. Or we can just defer the destructive
> operation like other commit 3e0588c291d6 ("hamradio: defer
> ax25 kfree after unregister_netdev") which let the
> ieee802154_unregister_hw() to handle the synchronization. This patch takes
> the second option.
>
> Fixes: 58e9683d1475 ("net: ieee802154: adf7242: Fix OCL calibration
> runs")
> Signed-off-by: Lin Ma <[email protected]>

Acked-by: Michael Hennerich <[email protected]>

> ---
> drivers/net/ieee802154/adf7242.c | 5 +++--
> 1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/net/ieee802154/adf7242.c
> b/drivers/net/ieee802154/adf7242.c
> index 6afdf1622944..efcc45aef508 100644
> --- a/drivers/net/ieee802154/adf7242.c
> +++ b/drivers/net/ieee802154/adf7242.c
> @@ -1310,10 +1310,11 @@ static void adf7242_remove(struct spi_device
> *spi)
>
> debugfs_remove_recursive(lp->debugfs_root);
>
> + ieee802154_unregister_hw(lp->hw);
> +
> cancel_delayed_work_sync(&lp->work);
> destroy_workqueue(lp->wqueue);
> -
> - ieee802154_unregister_hw(lp->hw);
> +
> mutex_destroy(&lp->bmux);
> ieee802154_free_hw(lp->hw);
> }
> --
> 2.36.1

2022-08-23 09:53:55

by Stefan Schmidt

[permalink] [raw]
Subject: Re: [PATCH v0] ieee802154/adf7242: defer destroy_workqueue call

Hello.

On 08.08.22 05:42, Lin Ma wrote:
> There is a possible race condition (use-after-free) like below
>
> (FREE) | (USE)
> adf7242_remove | adf7242_channel
> cancel_delayed_work_sync |
> destroy_workqueue (1) | adf7242_cmd_rx
> | mod_delayed_work (2)
> |
>
> The root cause for this race is that the upper layer (ieee802154) is
> unaware of this detaching event and the function adf7242_channel can
> be called without any checks.
>
> To fix this, we can add a flag write at the beginning of adf7242_remove
> and add flag check in adf7242_channel. Or we can just defer the
> destructive operation like other commit 3e0588c291d6 ("hamradio: defer
> ax25 kfree after unregister_netdev") which let the
> ieee802154_unregister_hw() to handle the synchronization. This patch
> takes the second option.
>
> Fixes: 58e9683d1475 ("net: ieee802154: adf7242: Fix OCL calibration
> runs")
> Signed-off-by: Lin Ma <[email protected]>


This patch has been applied to the wpan tree and will be
part of the next pull request to net. Thanks!

regards
Stefan Schmidt