Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
From:   Joel Fernandes <joel@joelfernandes.org>
Mime-Version: 1.0 (1.0)
Subject: Re: [PATCH] mac802154: Rename kfree_rcu() to kvfree_rcu_mightsleep()
Date:   Thu, 16 Mar 2023 13:58:28 -0400
Message-Id: <6EAB6B07-60C0-4489-9281-E01264E86DA5@joelfernandes.org>
References: <99dccb18-d16e-0b5b-586d-59a7649f68c4@datenfreihafen.org>
Cc:     linux-kernel@vger.kernel.org,
        Alexander Aring <alex.aring@gmail.com>,
        "David S. Miller" <davem@davemloft.net>,
        Eric Dumazet <edumazet@google.com>,
        Jakub Kicinski <kuba@kernel.org>, linux-wpan@vger.kernel.org,
        Miquel Raynal <miquel.raynal@bootlin.com>,
        netdev@vger.kernel.org, Paolo Abeni <pabeni@redhat.com>,
        boqun.feng@gmail.com, paulmck@kernel.org, urezki@gmail.com
In-Reply-To: <99dccb18-d16e-0b5b-586d-59a7649f68c4@datenfreihafen.org>
To:     Stefan Schmidt <stefan@datenfreihafen.org>
Precedence: bulk


> On Mar 16, 2023, at 12:36 PM, Stefan Schmidt <stefan@datenfreihafen.org> w=
rote:
>=20
> =EF=BB=BFHello Joel.
>=20
>> On 10.03.23 02:31, Joel Fernandes (Google) wrote:
>> The k[v]free_rcu() macro's single-argument form is deprecated.
>> Therefore switch to the new k[v]free_rcu_mightsleep() variant. The goal
>> is to avoid accidental use of the single-argument forms, which can
>> introduce functionality bugs in atomic contexts and latency bugs in
>> non-atomic contexts.
>> The callers are holding a mutex so the context allows blocking. Hence
>> using the API with a single argument will be fine, but use its new name.
>> There is no functionality change with this patch.
>> Fixes: 57588c71177f ("mac802154: Handle passive scanning")
>> Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
>> ---
>> Please Ack the patch but we can carry it through the RCU tree as well if
>> needed, as it is not a bug per-se and we are not dropping the old API bef=
ore
>> the next release.
>=20
> The "but we can carry it" part throws me off here. Not sure if you want th=
is through the RCU tree (I suppose). In that case see my ack below.
>=20
> If you want me to take it through my wpan tree instead let me know.

We will take this with your Ack below, thank you!

 - Joel


>>  net/mac802154/scan.c | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>> diff --git a/net/mac802154/scan.c b/net/mac802154/scan.c
>> index 9b0933a185eb..5c191bedd72c 100644
>> --- a/net/mac802154/scan.c
>> +++ b/net/mac802154/scan.c
>> @@ -52,7 +52,7 @@ static int mac802154_scan_cleanup_locked(struct ieee802=
154_local *local,
>>      request =3D rcu_replace_pointer(local->scan_req, NULL, 1);
>>      if (!request)
>>          return 0;
>> -    kfree_rcu(request);
>> +    kvfree_rcu_mightsleep(request);
>>        /* Advertize first, while we know the devices cannot be removed */=

>>      if (aborted)
>> @@ -403,7 +403,7 @@ int mac802154_stop_beacons_locked(struct ieee802154_l=
ocal *local,
>>      request =3D rcu_replace_pointer(local->beacon_req, NULL, 1);
>>      if (!request)
>>          return 0;
>> -    kfree_rcu(request);
>> +    kvfree_rcu_mightsleep(request);
>>        nl802154_beaconing_done(wpan_dev);
>> =20
>=20
>=20
> Acked-by: Stefan Schmidt <stefan@datenfreihafen.org>
>=20
> regards
> Stefan Schmidt