[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <f7tzgfwmobr.fsf@redhat.com>
Date: Mon, 22 Aug 2022 08:53:44 -0400
From: Aaron Conole <aconole@...hat.com>
To: Andrey Zhadchenko <andrey.zhadchenko@...tuozzo.com>
Cc: netdev@...r.kernel.org, dev@...nvswitch.org, pshelar@....org,
davem@...emloft.net, edumazet@...gle.com, kuba@...nel.org,
pabeni@...hat.com, ptikhomirov@...tuozzo.com,
alexander.mikhalitsyn@...tuozzo.com, avagin@...gle.com,
brauner@...nel.org, mark.d.gray@...hat.com, i.maximets@....org
Subject: Re: [PATCH net-next v2 2/3] openvswitch: fix memory leak at failed
datapath creation
Andrey Zhadchenko <andrey.zhadchenko@...tuozzo.com> writes:
> ovs_dp_cmd_new()->ovs_dp_change()->ovs_dp_set_upcall_portids()
> allocates array via kmalloc.
> If for some reason new_vport() fails during ovs_dp_cmd_new()
> dp->upcall_portids must be freed.
> Add missing kfree.
>
> Kmemleak example:
> unreferenced object 0xffff88800c382500 (size 64):
> comm "dump_state", pid 323, jiffies 4294955418 (age 104.347s)
> hex dump (first 32 bytes):
> 5e c2 79 e4 1f 7a 38 c7 09 21 38 0c 80 88 ff ff ^.y..z8..!8.....
> 03 00 00 00 0a 00 00 00 14 00 00 00 28 00 00 00 ............(...
> backtrace:
> [<0000000071bebc9f>] ovs_dp_set_upcall_portids+0x38/0xa0
> [<000000000187d8bd>] ovs_dp_change+0x63/0xe0
> [<000000002397e446>] ovs_dp_cmd_new+0x1f0/0x380
> [<00000000aa06f36e>] genl_family_rcv_msg_doit+0xea/0x150
> [<000000008f583bc4>] genl_rcv_msg+0xdc/0x1e0
> [<00000000fa10e377>] netlink_rcv_skb+0x50/0x100
> [<000000004959cece>] genl_rcv+0x24/0x40
> [<000000004699ac7f>] netlink_unicast+0x23e/0x360
> [<00000000c153573e>] netlink_sendmsg+0x24e/0x4b0
> [<000000006f4aa380>] sock_sendmsg+0x62/0x70
> [<00000000d0068654>] ____sys_sendmsg+0x230/0x270
> [<0000000012dacf7d>] ___sys_sendmsg+0x88/0xd0
> [<0000000011776020>] __sys_sendmsg+0x59/0xa0
> [<000000002e8f2dc1>] do_syscall_64+0x3b/0x90
> [<000000003243e7cb>] entry_SYSCALL_64_after_hwframe+0x63/0xcd
>
> Fixes: b83d23a2a38b ("openvswitch: Introduce per-cpu upcall dispatch")
> Signed-off-by: Andrey Zhadchenko <andrey.zhadchenko@...tuozzo.com>
> ---
Thanks for this patch. I guess independent of this series, this patch
should be applied to the net tree as well - it fixes an existing issue.
Acked-by: Aaron Conole <aconole@...hat.com>
I will try as well the other patches in the series.
Powered by blists - more mailing lists