Bug 67451 - memory leak in tty driver
Summary: memory leak in tty driver
Status: NEW
Alias: None
Product: Drivers
Classification: Unclassified
Component: Console/Framebuffers (show other bugs)
Hardware: x86-64 Linux
: P1 high
Assignee: Alan
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-12-20 17:47 UTC by Filipe David Manana
Modified: 2013-12-21 16:59 UTC (History)
2 users (show)

See Also:
Kernel Version: 3.11.0 3.12.5 3.13-rc4
Subsystem:
Regression: No
Bisected commit-id:


Attachments

Description Filipe David Manana 2013-12-20 17:47:10 UTC
Often, while running a vanilla kernel from kernel.org (tested with 3.11.0, 3.12.5 and 3.13-rc4) while using gnome's terminal app, kmemleak reports the following leak:

unreferenced object 0xffff8807f74d7b10 (size 512):
  comm "setfont", pid 819, jiffies 4294895521 (age 1420.336s)
  hex dump (first 32 bytes):
    9c fe b1 f1 9b 9d 00 00 92 00 91 00 00 00 00 00  ................
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<ffffffff8175f606>] kmemleak_alloc+0x26/0x50
    [<ffffffff8118dd44>] kmem_cache_alloc_trace+0x114/0x200
    [<ffffffff814b31b9>] set_inverse_transl+0x109/0x150
    [<ffffffff814b357f>] con_set_unimap+0x1ef/0x2a0
    [<ffffffff814ad0f2>] vt_ioctl+0x10e2/0x12a0
    [<ffffffff8149fde8>] tty_ioctl+0x2a8/0xc80
    [<ffffffff811b2946>] do_vfs_ioctl+0x96/0x570
    [<ffffffff811b2eb1>] SyS_ioctl+0x91/0xb0
    [<ffffffff8177ef12>] system_call_fastpath+0x16/0x1b
    [<ffffffffffffffff>] 0xffffffffffffffff
unreferenced object 0xffff8807f74d4d20 (size 512):
  comm "setfont", pid 819, jiffies 4294895521 (age 1420.336s)
  hex dump (first 32 bytes):
    00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f  ................
    10 11 12 13 14 15 16 17 18 19 1a 1b 1c 1d 1e 1f  ................
  backtrace:
    [<ffffffff8175f606>] kmemleak_alloc+0x26/0x50
    [<ffffffff8118dd44>] kmem_cache_alloc_trace+0x114/0x200
    [<ffffffff814b31b9>] set_inverse_transl+0x109/0x150
    [<ffffffff814b357f>] con_set_unimap+0x1ef/0x2a0
    [<ffffffff814ad0f2>] vt_ioctl+0x10e2/0x12a0
    [<ffffffff8149fde8>] tty_ioctl+0x2a8/0xc80
    [<ffffffff811b2946>] do_vfs_ioctl+0x96/0x570
    [<ffffffff811b2eb1>] SyS_ioctl+0x91/0xb0
    [<ffffffff8177ef12>] system_call_fastpath+0x16/0x1b
    [<ffffffffffffffff>] 0xffffffffffffffff
unreferenced object 0xffff8807e90f44a8 (size 1024):
  comm "setfont", pid 819, jiffies 4294895521 (age 1420.336s)
  hex dump (first 32 bytes):
    a3 00 a0 25 92 25 b1 00 a2 00 a5 00 a9 00 ae 00  ...%.%..........
    c6 00 dd 00 e6 00 52 01 53 01 78 01 14 20 20 20  ......R.S.x..   
  backtrace:
    [<ffffffff8175f606>] kmemleak_alloc+0x26/0x50
    [<ffffffff8118dd44>] kmem_cache_alloc_trace+0x114/0x200
    [<ffffffff814b2ab0>] set_inverse_trans_unicode.isra.2+0x120/0x140
    [<ffffffff814b358c>] con_set_unimap+0x1fc/0x2a0
    [<ffffffff814ad0f2>] vt_ioctl+0x10e2/0x12a0
    [<ffffffff8149fde8>] tty_ioctl+0x2a8/0xc80
    [<ffffffff811b2946>] do_vfs_ioctl+0x96/0x570
    [<ffffffff811b2eb1>] SyS_ioctl+0x91/0xb0
    [<ffffffff8177ef12>] system_call_fastpath+0x16/0x1b
    [<ffffffffffffffff>] 0xffffffffffffffff


I don't know exactly what the steps are to reproduce this, it happens not too often, but when it happens, I end up getting 50+ such leaks reported.
Comment 1 Alan 2013-12-21 01:20:20 UTC
Thanks, that ought to be enough information to trace it back as it comes from 'setfont' so I can trace the path that is apparently causing the leak. Added to the queue but it's not I think particularly urgent
Comment 2 Filipe David Manana 2013-12-21 01:24:34 UTC
Ok, thank you Alan.

Btw, if it leaks too much, can it freeze my machine?
I've had a few freezes, and they seemed to (by coincidence maybe) have happened during a boot where kmemleak reported those leaks.
Comment 3 Alan 2013-12-21 16:59:18 UTC
I would not expect so although it is possible both are symptoms of some other bug of course.

Note You need to log in before you can comment on or make changes to this bug.