2023-01-04 14:31:08

by Jiasheng Jiang

Subject: [PATCH] wifi: rtw89: Add missing check for alloc_workqueue

On Wed, Jan 04, 2023 at 07:41:36PM +0800, Leon Romanovsky wrote:
> On Wed, Jan 04, 2023 at 05:33:53PM +0800, Jiasheng Jiang wrote:
>> Add check for the return value of alloc_workqueue since it may return
>> NULL pointer.
>>
>> Fixes: e3ec7017f6a2 ("rtw89: add Realtek 802.11ax driver")
>> Signed-off-by: Jiasheng Jiang <[email protected]>
>> ---
>> drivers/net/wireless/realtek/rtw89/core.c | 2 ++
>> 1 file changed, 2 insertions(+)
>>
>> diff --git a/drivers/net/wireless/realtek/rtw89/core.c b/drivers/net/wireless/realtek/rtw89/core.c
>> index 931aff8b5dc9..006fe0499f81 100644
>> --- a/drivers/net/wireless/realtek/rtw89/core.c
>> +++ b/drivers/net/wireless/realtek/rtw89/core.c
>> @@ -3124,6 +3124,8 @@ int rtw89_core_init(struct rtw89_dev *rtwdev)
>> INIT_DELAYED_WORK(&rtwdev->cfo_track_work, rtw89_phy_cfo_track_work);
>> INIT_DELAYED_WORK(&rtwdev->forbid_ba_work, rtw89_forbid_ba_work);
>> rtwdev->txq_wq = alloc_workqueue("rtw89_tx_wq", WQ_UNBOUND | WQ_HIGHPRI, 0);
>> + if (!rtwdev->txq_wq)
>> + return -ENOMEM;
>
> While the change is fixing one issue, you are adding another one.
> There is no destroy of this workqueue if rtw89_load_firmware fails.

Actually, I do not think the missing of destroy_workqueue is introduced by me.
Even without my patch, the destroy_workqueue is still missing.
Anyway, I will submit a v2 that adds the missing destroy_workqueue.

Thanks,
Jiang

2023-01-05 09:21:49

by Leon Romanovsky

Subject: Re: [PATCH] wifi: rtw89: Add missing check for alloc_workqueue

On Wed, Jan 04, 2023 at 10:23:53PM +0800, Jiasheng Jiang wrote:
> On Wed, Jan 04, 2023 at 07:41:36PM +0800, Leon Romanovsky wrote:
> > On Wed, Jan 04, 2023 at 05:33:53PM +0800, Jiasheng Jiang wrote:
> >> Add check for the return value of alloc_workqueue since it may return
> >> NULL pointer.
> >>
> >> Fixes: e3ec7017f6a2 ("rtw89: add Realtek 802.11ax driver")
> >> Signed-off-by: Jiasheng Jiang <[email protected]>
> >> ---
> >> drivers/net/wireless/realtek/rtw89/core.c | 2 ++
> >> 1 file changed, 2 insertions(+)
> >>
> >> diff --git a/drivers/net/wireless/realtek/rtw89/core.c b/drivers/net/wireless/realtek/rtw89/core.c
> >> index 931aff8b5dc9..006fe0499f81 100644
> >> --- a/drivers/net/wireless/realtek/rtw89/core.c
> >> +++ b/drivers/net/wireless/realtek/rtw89/core.c
> >> @@ -3124,6 +3124,8 @@ int rtw89_core_init(struct rtw89_dev *rtwdev)
> >> INIT_DELAYED_WORK(&rtwdev->cfo_track_work, rtw89_phy_cfo_track_work);
> >> INIT_DELAYED_WORK(&rtwdev->forbid_ba_work, rtw89_forbid_ba_work);
> >> rtwdev->txq_wq = alloc_workqueue("rtw89_tx_wq", WQ_UNBOUND | WQ_HIGHPRI, 0);
> >> + if (!rtwdev->txq_wq)
> >> + return -ENOMEM;
> >
> > While the change is fixing one issue, you are adding another one.
> > There is no destroy of this workqueue if rtw89_load_firmware fails.
>
> Actually, I do not think the missing of destroy_workqueue is introduced by me.
> Even without my patch, the destroy_workqueue is still missing.

Not really, without your change, rtw89_core_init() never failed.

> Anyway, I will submit a v2 that adds the missing destroy_workqueue.

Thanks

>
> Thanks,
> Jiang
>