wdev->valid_links is not cleared when upper layer disconnect from a
wdev->AP MLD. It has been observed that this would prevent offchannel
operations like remain-on-channel which would be needed for user space
operations with Public Action frame.
Clear the wdev->valid_links when STA disconnect.
Signed-off-by: Xin Deng <[email protected]>
---
net/wireless/sme.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/net/wireless/sme.c b/net/wireless/sme.c
index 82e3ce42206c..86e837f37f8c 100644
--- a/net/wireless/sme.c
+++ b/net/wireless/sme.c
@@ -492,6 +492,7 @@ void cfg80211_wdev_release_link_bsses(struct wireless_dev *wdev, u16 link_mask)
&wdev->links[link].client.current_bss->pub);
wdev->links[link].client.current_bss = NULL;
}
+ wdev->valid_links = 0;
}
static int cfg80211_sme_get_conn_ies(struct wireless_dev *wdev,
base-commit: 8027a466a77a288eccd2d11868f504e24231f3b7
--
2.17.1
On 3/26/2024 8:24 PM, Xin Deng wrote:
> wdev->valid_links is not cleared when upper layer disconnect from a
> wdev->AP MLD. It has been observed that this would prevent offchannel
> operations like remain-on-channel which would be needed for user space
> operations with Public Action frame.
> Clear the wdev->valid_links when STA disconnect.
>
> Signed-off-by: Xin Deng <[email protected]>
Reviewed-by: Jeff Johnson <[email protected]>
On Tue, 2024-03-26 at 20:24 -0700, Xin Deng wrote:
> wdev->valid_links is not cleared when upper layer disconnect from a
> wdev->AP MLD. It has been observed that this would prevent offchannel
> operations like remain-on-channel which would be needed for user space
> operations with Public Action frame.
I agree that's a problem, we shouldn't leave the valid_links set.
> Signed-off-by: Xin Deng <[email protected]>
> ---
> net/wireless/sme.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/net/wireless/sme.c b/net/wireless/sme.c
> index 82e3ce42206c..86e837f37f8c 100644
> --- a/net/wireless/sme.c
> +++ b/net/wireless/sme.c
> @@ -492,6 +492,7 @@ void cfg80211_wdev_release_link_bsses(struct wireless_dev *wdev, u16 link_mask)
> &wdev->links[link].client.current_bss->pub);
> wdev->links[link].client.current_bss = NULL;
> }
> + wdev->valid_links = 0;
>
But this is (very obviously, even with only the limited context shown!)
in the completely wrong place.
johannes
On 4/8/2024 11:41 AM, Johannes Berg wrote:
> On Tue, 2024-03-26 at 20:24 -0700, Xin Deng wrote:
>> wdev->valid_links is not cleared when upper layer disconnect from a
>> wdev->AP MLD. It has been observed that this would prevent offchannel
>> operations like remain-on-channel which would be needed for user space
>> operations with Public Action frame.
>
> I agree that's a problem, we shouldn't leave the valid_links set.
>
>> Signed-off-by: Xin Deng <[email protected]>
>> ---
>> net/wireless/sme.c | 1 +
>> 1 file changed, 1 insertion(+)
>>
>> diff --git a/net/wireless/sme.c b/net/wireless/sme.c
>> index 82e3ce42206c..86e837f37f8c 100644
>> --- a/net/wireless/sme.c
>> +++ b/net/wireless/sme.c
>> @@ -492,6 +492,7 @@ void cfg80211_wdev_release_link_bsses(struct wireless_dev *wdev, u16 link_mask)
>> &wdev->links[link].client.current_bss->pub);
>> wdev->links[link].client.current_bss = NULL;
>> }
>> + wdev->valid_links = 0;
>>
>
> But this is (very obviously, even with only the limited context shown!)
> in the completely wrong place.
Concur, so NAK my Reviewed-by:
That function is potentially only removing a subset of links, so it should not
be setting the valid_links to 0 since not all of the links may have been removed.
/jeff