Re: net: memory leak in mkiss_open

On Mon, Feb 22, 2016 at 2:05 AM, Dmitry Vyukov [off-list ref] wrote:

unreferenced object 0xffff8800628991d8 (size 4096):
  comm "a.out", pid 7081, jiffies 4294920662 (age 35.917s)
  hex dump (first 32 bytes):
    00 00 00 00 00 00 00 00 61 78 30 00 00 00 00 00  ........ax0.....
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff86680923>] kmemleak_alloc+0x63/0xa0 mm/kmemleak.c:916
    [<     inline     >] kmemleak_alloc_recursive include/linux/kmemleak.h:47
    [<     inline     >] slab_post_alloc_hook mm/slub.c:1337
    [<     inline     >] slab_alloc_node mm/slub.c:2596
    [<     inline     >] slab_alloc mm/slub.c:2604
    [<ffffffff8176b585>] __kmalloc+0x1c5/0x3a0 mm/slub.c:3564
    [<     inline     >] kmalloc include/linux/slab.h:468
    [<ffffffff8554c93e>] kzalloc+0x1e/0x30 include/linux/slab.h:607
    [<ffffffff85551442>] alloc_netdev_mqs+0x82/0xd20 net/core/dev.c:7387
    [<ffffffff83ef39da>] mkiss_open+0xaa/0xc30 drivers/net/hamradio/mkiss.c:707
    [<ffffffff82fbd258>] tty_ldisc_open.isra.2+0x78/0xd0
drivers/tty/tty_ldisc.c:454
    [<ffffffff82fbd842>] tty_set_ldisc+0x292/0x8a0 drivers/tty/tty_ldisc.c:561
    [<     inline     >] tiocsetd drivers/tty/tty_io.c:2655
    [<ffffffff82fa651e>] tty_ioctl+0xb2e/0x2160 drivers/tty/tty_io.c:2910
    [<     inline     >] vfs_ioctl fs/ioctl.c:43
    [<ffffffff8180597c>] do_vfs_ioctl+0x18c/0xfb0 fs/ioctl.c:674
    [<     inline     >] SYSC_ioctl fs/ioctl.c:689
    [<ffffffff8180682f>] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:680
    [<ffffffff8669ddb6>] entry_SYSCALL_64_fastpath+0x16/0x7a
arch/x86/entry/entry_64.S:185
    [<ffffffffffffffff>] 0xffffffffffffffff
unreferenced object 0xffff8800652b1aa0 (size 96):
  comm "a.out", pid 7081, jiffies 4294920662 (age 35.917s)
  hex dump (first 32 bytes):
    b8 94 89 62 00 88 ff ff b8 94 89 62 00 88 ff ff  ...b.......b....
    98 92 9c aa b0 40 02 00 00 00 00 00 00 00 00 00  .....@..........
  backtrace:
    [<ffffffff86680923>] kmemleak_alloc+0x63/0xa0 mm/kmemleak.c:916
    [<     inline     >] kmemleak_alloc_recursive include/linux/kmemleak.h:47
    [<     inline     >] slab_post_alloc_hook mm/slub.c:1337
    [<     inline     >] slab_alloc_node mm/slub.c:2596
    [<     inline     >] slab_alloc mm/slub.c:2604
    [<ffffffff8176ae6e>] kmem_cache_alloc_trace+0x1ee/0x350 mm/slub.c:2621
    [<     inline     >] kmalloc include/linux/slab.h:463
    [<ffffffff8557a4e2>] __hw_addr_create_ex+0x62/0x240
net/core/dev_addr_lists.c:34
    [<ffffffff8557a7bc>] __hw_addr_add_ex+0xfc/0x270
net/core/dev_addr_lists.c:80
    [<     inline     >] __hw_addr_add net/core/dev_addr_lists.c:88
    [<ffffffff8557ac35>] dev_addr_init+0x105/0x1c0 net/core/dev_addr_lists.c:371
    [<ffffffff855514f0>] alloc_netdev_mqs+0x130/0xd20 net/core/dev.c:7400
    [<ffffffff83ef39da>] mkiss_open+0xaa/0xc30 drivers/net/hamradio/mkiss.c:707
    [<ffffffff82fbd258>] tty_ldisc_open.isra.2+0x78/0xd0
drivers/tty/tty_ldisc.c:454
    [<ffffffff82fbd842>] tty_set_ldisc+0x292/0x8a0 drivers/tty/tty_ldisc.c:561
    [<     inline     >] tiocsetd drivers/tty/tty_io.c:2655
    [<ffffffff82fa651e>] tty_ioctl+0xb2e/0x2160 drivers/tty/tty_io.c:2910
    [<     inline     >] vfs_ioctl fs/ioctl.c:43
    [<ffffffff8180597c>] do_vfs_ioctl+0x18c/0xfb0 fs/ioctl.c:674
    [<     inline     >] SYSC_ioctl fs/ioctl.c:689
    [<ffffffff8180682f>] SyS_ioctl+0x8f/0xc0 fs/ioctl.c:680
    [<ffffffff8669ddb6>] entry_SYSCALL_64_fastpath+0x16/0x7a
arch/x86/entry/entry_64.S:185

This looks more like a tty layer leak than netdev layer.