[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CA+h21hqw8Ed_k_E8KK7mzDTag7p8jTYFLLkJ_i6+8__BKCdEUQ@mail.gmail.com>
Date: Wed, 27 May 2020 19:43:29 +0300
From: Vladimir Oltean <olteanv@...il.com>
To: Dan Carpenter <dan.carpenter@...cle.com>
Cc: kbuild@...ts.01.org, kbuild test robot <lkp@...el.com>,
kbuild-all@...ts.01.org, lkml <linux-kernel@...r.kernel.org>,
Andrew Lunn <andrew@...n.ch>
Subject: Re: net/dsa/tag_8021q.c:109 dsa_8021q_restore_pvid() error:
uninitialized symbol 'pvid'.
Hi Dan,
On Tue, 26 May 2020 at 17:55, Dan Carpenter <dan.carpenter@...cle.com> wrote:
>
> tree: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> head: 9cb1fd0efd195590b828b9b865421ad345a4a145
> commit: c80ed84e76886487703bf04b38ce10e92e2d6e26 net: dsa: tag_8021q: Fix dsa_8021q_restore_pvid for an absent pvid
> date: 6 months ago
> config: x86_64-randconfig-m001-20200526 (attached as .config)
> compiler: gcc-7 (Ubuntu 7.5.0-6ubuntu2) 7.5.0
>
> If you fix the issue, kindly add following tag as appropriate
> Reported-by: kbuild test robot <lkp@...el.com>
> Reported-by: Dan Carpenter <dan.carpenter@...cle.com>
>
> smatch warnings:
> net/dsa/tag_8021q.c:109 dsa_8021q_restore_pvid() error: uninitialized symbol 'pvid'.
>
> # https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=c80ed84e76886487703bf04b38ce10e92e2d6e26
> git remote add linus https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
> git remote update linus
> git checkout c80ed84e76886487703bf04b38ce10e92e2d6e26
> vim +/pvid +109 net/dsa/tag_8021q.c
>
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 96 static int dsa_8021q_restore_pvid(struct dsa_switch *ds, int port)
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 97 {
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 98 struct bridge_vlan_info vinfo;
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 99 struct net_device *slave;
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 100 u16 pvid;
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 101 int err;
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 102
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 103 if (!dsa_is_user_port(ds, port))
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 104 return 0;
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 105
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 106 slave = ds->ports[port].slave;
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 107
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 108 err = br_vlan_get_pvid(slave, &pvid);
> c80ed84e768864 Vladimir Oltean 2019-11-16 @109 if (!pvid || err < 0)
> ^^^^^^^^^^^^^^^^
> I don't know why this warning is only showing up now in May 2020...
> Anyway, we could solve the problem by checking for errors first:
>
> if (err < 0 || !pvid) {
>
> The UBSan tool can detect uninitialized values used at runtime so it
> might generate a splat as well. Other than that, obviously it is
> harmless.
>
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 110 /* There is no pvid on the bridge for this port, which is
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 111 * perfectly valid. Nothing to restore, bye-bye!
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 112 */
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 113 return 0;
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 114
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 115 err = br_vlan_get_info(slave, pvid, &vinfo);
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 116 if (err < 0) {
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 117 dev_err(ds->dev, "Couldn't determine PVID attributes\n");
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 118 return err;
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 119 }
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 120
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 121 return dsa_port_vid_add(&ds->ports[port], pvid, vinfo.flags);
> 5f33183b7fdfeb Vladimir Oltean 2019-08-30 122 }
>
> ---
> 0-DAY CI Kernel Test Service, Intel Corporation
> https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
For net-next I just sent a patch that removes this function altogether:
https://patchwork.ozlabs.org/project/netdev/patch/20200527164134.1081548-1-olteanv@gmail.com/
Do you think a patch for the net tree is still necessary, that
reorders the checks? Will you send that patch or should I?
Thanks,
-Vladimir
Powered by blists - more mailing lists