| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Sun, 01 Feb 2009 23:50:10 +0100
From: Andrea Righi <righi.andrea@...il.com>
To: Andrey Borzenkov <arvidjaar@...l.ru>
CC: Geert Uytterhoeven <geert@...ux-m68k.org>,
Linux Frame Buffer Device Development
<linux-fbdev-devel@...ts.sourceforge.net>,
"Antonino A. Daplas" <adaplas@...il.com>,
linux-pm@...ts.linux-foundation.org,
Linux Kernel Development <linux-kernel@...r.kernel.org>
Subject: Re: [Linux-fbdev-devel] [2.6.29-rc2] fb_mmap: circular locking dependency
on hibernation
On 2009-02-01 08:15, Andrey Borzenkov wrote:
> On 31 of January 2009 20:09:52 Andrea Righi wrote:
>> On 2009-01-30 05:15, Andrey Borzenkov wrote:
>>> On 29 of January 2009 12:10:11 Geert Uytterhoeven wrote:
>>>> On Tue, 27 Jan 2009, Andrey Borzenkov wrote:
>>>>> I do not know if this is new; I usually do not hibernate at all
>>>>> and was testing another driver actually. I do not remember such
>>>>> issues in 2.6.28 though. Dmesg follows; config attached.
>>>>>
>>>>> [11715.652498]
>>>>> =======================================================
>>>>> [11715.652520] [ INFO: possible circular locking dependency
>>>>> detected ] [11715.652534] 2.6.29-rc2-1avb #11
>>>>> [11715.652541]
>>>>> -------------------------------------------------------
>>>>> [11715.652550] s2disk/25519 is trying to acquire lock:
>>>>> [11715.652562] (&fb_info->lock){--..}, at: [<c0223017>]
>>>>> fb_mmap+0x97/0x170 [11715.652623]
>>>>> [11715.652625] but task is already holding lock:
>>>>> [11715.652632] (&mm->mmap_sem){----}, at: [<c010639e>]
>>>>> sys_mmap2+0x8e/0xc0 [11715.652669]
>>>>> [11715.652671] which lock already depends on the new lock.
>>>>> [11715.652675]
>>>>> [11715.652681]
>>>>> [11715.652683] the existing dependency chain (in reverse order)
>>>>> is: [11715.652691]
>>>>> [11715.652693] -> #1 (&mm->mmap_sem){----}:
>>>>> [11715.652707] [<c0144ecf>] __lock_acquire+0x129f/0x1930
>>>>> [11715.652730] [<c01455bc>] lock_acquire+0x5c/0x80
>>>>> [11715.652742] [<c0181e57>] might_fault+0x77/0xa0
>>>>> [11715.652764] [<c0209325>] copy_from_user+0x35/0x130
>>>>> [11715.652782] [<c0223ef5>] fb_ioctl+0x1d5/0x3f0
>>>>> [11715.652797] [<c01a51d0>] vfs_ioctl+0x20/0x80
>>>>> [11715.652832] [<c01a53e2>] do_vfs_ioctl+0x72/0x570
>>>>> [11715.652845] [<c01a5919>] sys_ioctl+0x39/0x70
>>>>> [11715.652857] [<c0103371>] sysenter_do_call+0x12/0x31
>>>>> [11715.652870] [<ffffffff>] 0xffffffff
>>>>> [11715.652973]
>>>>> [11715.652975] -> #0 (&fb_info->lock){--..}:
>>>>> [11715.652988] [<c014503c>] __lock_acquire+0x140c/0x1930
>>>>> [11715.653000] [<c01455bc>] lock_acquire+0x5c/0x80
>>>>> [11715.653011] [<c030761a>] mutex_lock_nested+0xba/0x2f0
>>>>> [11715.653032] [<c0223017>] fb_mmap+0x97/0x170
>>>>> [11715.653044] [<c0188996>] mmap_region+0x1d6/0x530
>>>>> [11715.653067] [<c0188ec3>] do_mmap_pgoff+0x1d3/0x2f0
>>>>> [11715.653079] [<c01063bd>] sys_mmap2+0xad/0xc0
>>>>> [11715.653091] [<c0103371>] sysenter_do_call+0x12/0x31
>>>>> [11715.653102] [<ffffffff>] 0xffffffff
>>>>> [11715.653116]
>>>>> [11715.653119] other info that might help us debug this:
>>>>> [11715.653122]
>>>>> [11715.653130] 1 lock held by s2disk/25519:
>>>>> [11715.653137] #0: (&mm->mmap_sem){----}, at: [<c010639e>]
>>>>> sys_mmap2+0x8e/0xc0
>>>>> [11715.653156]
>>>>> [11715.653158] stack backtrace:
>>>>> [11715.653170] Pid: 25519, comm: s2disk Not tainted
>>>>> 2.6.29-rc2-1avb #11 [11715.653179] Call Trace:
>>>>> [11715.653190] [<c0306248>] ? printk+0x18/0x20
>>>>> [11715.653202] [<c01437df>] print_circular_bug_tail+0xcf/0xe0
>>>>> [11715.653213] [<c014503c>] __lock_acquire+0x140c/0x1930
>>>>> [11715.653224] [<c014328b>] ? trace_hardirqs_on+0xb/0x10
>>>>> [11715.653242] [<c0309125>] ? _spin_unlock_irqrestore+0x35/0x60
>>>>> [11715.653254] [<c0140dc5>] ? lock_release_holdtime+0x35/0x210
>>>>> [11715.653265] [<c01455bc>] lock_acquire+0x5c/0x80
>>>>> [11715.653277] [<c0223017>] ? fb_mmap+0x97/0x170
>>>>> [11715.653287] [<c030761a>] mutex_lock_nested+0xba/0x2f0
>>>>> [11715.653298] [<c0223017>] ? fb_mmap+0x97/0x170
>>>>> [11715.653309] [<c0223017>] ? fb_mmap+0x97/0x170
>>>>> [11715.653335] [<c019514d>] ? kmem_cache_alloc+0xad/0x100
>>>>> [11715.653347] [<c0223017>] fb_mmap+0x97/0x170
>>>>> [11715.653358] [<c0188996>] mmap_region+0x1d6/0x530
>>>>> [11715.653371] [<c0186a9d>] ?
>>>>> arch_get_unmapped_area_topdown+0x10d/0x170 [11715.653383]
>>>>> [<c0188ec3>] do_mmap_pgoff+0x1d3/0x2f0
>>>>> [11715.653395] [<c01063bd>] sys_mmap2+0xad/0xc0
>>>>> [11715.653406] [<c0103371>] sysenter_do_call+0x12/0x31
>>>> Is this fixed by http://patchwork.kernel.org/patch/3259/?
>>> No. It just makes stack a bit more deep :)
>>>
>>> [ 408.267429]
>>> ======================================================= [
>>> 408.267448] [ INFO: possible circular locking dependency detected ]
>>> [ 408.267462] 2.6.29-rc3-1avb #14
>>> [ 408.267469]
>>> ------------------------------------------------------- [
>>> 408.267478] s2disk/5695 is trying to acquire lock:
>>> [ 408.267487] (&fb_info->lock){--..}, at: [<c02235e7>]
>>> fb_mmap+0x97/0x170
>>> [ 408.267544]
>>> [ 408.267546] but task is already holding lock:
>>> [ 408.267553] (&mm->mmap_sem){----}, at: [<c010639e>]
>>> sys_mmap2+0x8e/0xc0
>>> [ 408.267591]
>>> [ 408.267593] which lock already depends on the new lock.
>>> [ 408.267596]
>>> [ 408.267603]
>>> [ 408.267604] the existing dependency chain (in reverse order) is:
>>> [ 408.267612]
>>> [ 408.267614] -> #3 (&mm->mmap_sem){----}:
>>> [ 408.267628] [<c014520f>] __lock_acquire+0x129f/0x1930
>>> [ 408.267650] [<c01458fc>] lock_acquire+0x5c/0x80
>>> [ 408.267661] [<c0182117>] might_fault+0x77/0xa0
>>> [ 408.267683] [<c0209916>] copy_to_user+0x36/0x120
>>> [ 408.267699] [<c01a60f7>] filldir+0x97/0xe0
>>> [ 408.267735] [<c01de339>] sysfs_readdir+0x129/0x220
>>> [ 408.267751] [<c01a62a6>] vfs_readdir+0x86/0xa0
>>> [ 408.267763] [<c01a63e8>] sys_getdents+0x68/0xc0
>>> [ 408.267775] [<c010344a>] syscall_call+0x7/0xb
>>> [ 408.267787] [<ffffffff>] 0xffffffff
>>> [ 408.267862]
>>> [ 408.267864] -> #2 (sysfs_mutex){--..}:
>>> [ 408.267877] [<c014520f>] __lock_acquire+0x129f/0x1930
>>> [ 408.267889] [<c01458fc>] lock_acquire+0x5c/0x80
>>> [ 408.267899] [<c0307f3a>] mutex_lock_nested+0xba/0x2f0
>>> [ 408.267921] [<c01de6cc>] sysfs_addrm_start+0x2c/0xc0
>>> [ 408.267933] [<c01dec80>] create_dir+0x40/0x90
>>> [ 408.267944] [<c01decfb>] sysfs_create_dir+0x2b/0x50
>>> [ 408.267955] [<c02036bc>] kobject_add_internal+0xbc/0x1b0
>>> [ 408.267984] [<c0203881>] kobject_add_varg+0x31/0x50
>>> [ 408.267996] [<c02038fc>] kobject_add+0x2c/0x60
>>> [ 408.268008] [<c0273448>] device_add+0xa8/0x550
>>> [ 408.268041] [<c0273902>] device_register+0x12/0x20
>>> [ 408.268053] [<c02739bb>] device_create_vargs+0xab/0xc0
>>> [ 408.268065] [<c02739f8>] device_create+0x28/0x30
>>> [ 408.268077] [<c026cc6d>] register_con_driver+0xed/0x130
>>> [ 408.268090] [<c026e18b>] take_over_console+0x1b/0x50
>>> [ 408.268102] [<c023052d>] fbcon_takeover+0x5d/0xb0
>>> [ 408.268125] [<c02311a0>] fbcon_event_notify+0x820/0x900
>>> [ 408.268137] [<c030c3b3>] notifier_call_chain+0x53/0xa0
>>> [ 408.268161] [<c0138ab4>]
>>> __blocking_notifier_call_chain+0x44/0x60
>>> [ 408.268184] [<c0138aea>]
>>> blocking_notifier_call_chain+0x1a/0x20
>>> [ 408.268197] [<c0222f81>] fb_notifier_call_chain+0x11/0x20
>>> [ 408.268209] [<c0223f08>] register_framebuffer+0x168/0x220
>>> [ 408.268223] [<c0414ba2>] vesafb_probe+0x542/0x783
>>> [ 408.268250] [<c02765af>] platform_drv_probe+0xf/0x20
>>> [ 408.268269] [<c02757c7>] driver_probe_device+0x87/0x1b0
>>> [ 408.268281] [<c0275988>] __device_attach+0x8/0x10
>>> [ 408.268291] [<c0274e0b>] bus_for_each_drv+0x5b/0x80
>>> [ 408.268308] [<c0275a36>] device_attach+0x76/0x80
>>> [ 408.268319] [<c0274c27>] bus_attach_device+0x47/0x70
>>> [ 408.268331] [<c02736c3>] device_add+0x323/0x550
>>> [ 408.268342] [<c0276f85>] platform_device_add+0x175/0x1c0
>>> [ 408.268358] [<c0414e7d>] vesafb_init+0x9a/0x1ec
>>> [ 408.268370] [<c010111a>] do_one_initcall+0x2a/0x160
>>> [ 408.268382] [<c04004d5>] kernel_init+0x83/0xd5
>>> [ 408.268411] [<c0103a77>] kernel_thread_helper+0x7/0x10
>>> [ 408.268423] [<ffffffff>] 0xffffffff
>>> [ 408.268436]
>>> [ 408.268438] -> #1 ((fb_notifier_list).rwsem){----}:
>>> [ 408.268451] [<c014520f>] __lock_acquire+0x129f/0x1930
>>> [ 408.268463] [<c01458fc>] lock_acquire+0x5c/0x80
>>> [ 408.268474] [<c0308879>] down_read+0x49/0x90
>>> [ 408.268484] [<c0138a9a>]
>>> __blocking_notifier_call_chain+0x2a/0x60
>>> [ 408.268497] [<c0138aea>]
>>> blocking_notifier_call_chain+0x1a/0x20
>>> [ 408.268509] [<c0222f81>] fb_notifier_call_chain+0x11/0x20
>>> [ 408.268521] [<c022459e>] do_fb_ioctl+0x2ae/0x5d0
>>> [ 408.268533] [<c02248dd>] fb_ioctl+0x1d/0x20
>>> [ 408.268545] [<c01a54b0>] vfs_ioctl+0x20/0x80
>>> [ 408.268558] [<c01a56c2>] do_vfs_ioctl+0x72/0x570
>>> [ 408.268570] [<c01a5bf9>] sys_ioctl+0x39/0x70
>>> [ 408.268581] [<c0103371>] sysenter_do_call+0x12/0x31
>>> [ 408.268592] [<ffffffff>] 0xffffffff
>>> [ 408.268623]
>>> [ 408.268625] -> #0 (&fb_info->lock){--..}:
>>> [ 408.268638] [<c014537c>] __lock_acquire+0x140c/0x1930
>>> [ 408.268649] [<c01458fc>] lock_acquire+0x5c/0x80
>>> [ 408.268661] [<c0307f3a>] mutex_lock_nested+0xba/0x2f0
>>> [ 408.268672] [<c02235e7>] fb_mmap+0x97/0x170
>>> [ 408.268684] [<c0188c76>] mmap_region+0x1d6/0x530
>>> [ 408.268706] [<c01891a3>] do_mmap_pgoff+0x1d3/0x2f0
>>> [ 408.268718] [<c01063bd>] sys_mmap2+0xad/0xc0
>>> [ 408.268729] [<c0103371>] sysenter_do_call+0x12/0x31
>>> [ 408.268740] [<ffffffff>] 0xffffffff
>>> [ 408.268752]
>>> [ 408.268754] other info that might help us debug this:
>>> [ 408.268758]
>>> [ 408.268767] 1 lock held by s2disk/5695:
>>> [ 408.268773] #0: (&mm->mmap_sem){----}, at: [<c010639e>]
>>> sys_mmap2+0x8e/0xc0
>>> [ 408.268792]
>>> [ 408.268794] stack backtrace:
>>> [ 408.268805] Pid: 5695, comm: s2disk Not tainted 2.6.29-rc3-1avb
>>> #14 [ 408.268814] Call Trace:
>>> [ 408.268825] [<c0306b68>] ? printk+0x18/0x20
>>> [ 408.268836] [<c0143b1f>] print_circular_bug_tail+0xcf/0xe0
>>> [ 408.268847] [<c014537c>] __lock_acquire+0x140c/0x1930
>>> [ 408.268862] [<c01487f0>] ? futex_wait+0x170/0x450
>>> [ 408.268881] [<c011826c>] ? try_to_wake_up+0x10c/0x120
>>> [ 408.268894] [<c0309a45>] ? _spin_unlock_irqrestore+0x35/0x60
>>> [ 408.268906] [<c0141105>] ? lock_release_holdtime+0x35/0x210
>>> [ 408.268917] [<c01458fc>] lock_acquire+0x5c/0x80
>>> [ 408.268928] [<c02235e7>] ? fb_mmap+0x97/0x170
>>> [ 408.268939] [<c0307f3a>] mutex_lock_nested+0xba/0x2f0
>>> [ 408.268950] [<c02235e7>] ? fb_mmap+0x97/0x170
>>> [ 408.268961] [<c02235e7>] ? fb_mmap+0x97/0x170
>>> [ 408.268987] [<c019542d>] ? kmem_cache_alloc+0xad/0x100
>>> [ 408.268999] [<c02235e7>] fb_mmap+0x97/0x170
>>> [ 408.269010] [<c0188c76>] mmap_region+0x1d6/0x530
>>> [ 408.269023] [<c0186d7d>] ?
>>> arch_get_unmapped_area_topdown+0x10d/0x170
>>> [ 408.269035] [<c01891a3>] do_mmap_pgoff+0x1d3/0x2f0
>>> [ 408.269047] [<c01063bd>] sys_mmap2+0xad/0xc0
>>> [ 408.269058] [<c0103371>] sysenter_do_call+0x12/0x31
>> Anyway, looking at this trace in detail, it seems that in the fbcon
>> notifier the FB_EVENT_SET_CONSOLE_MAP handler doesn't use fb_info at
>> all.
>>
>> So, just for this particular case, we could unlock fb_info->lock
>> before calling the routine that might acquire mm->mmap_sem, and avoid
>> the circular locking dependency.
>>
>> Any chance to do a test with the following patch?
>>
>
> Step by step ... :)
>
> With both patches applied:
>
> [ 113.949848] =======================================================
> [ 113.950054] [ INFO: possible circular locking dependency detected ]
> [ 113.950054] 2.6.29-rc3-1avb #16
> [ 113.950054] -------------------------------------------------------
> [ 113.950054] s2disk/4168 is trying to acquire lock:
> [ 113.950054] (&fb_info->lock){--..}, at: [<c0230c56>]
> fbcon_event_notify+0x2d6/0x950
> [ 113.950054]
> [ 113.950054] but task is already holding lock:
> [ 113.950054] ((fb_notifier_list).rwsem){----}, at: [<c0138a9a>]
> __blocking_notifier_call_chain+0x2a/0x60
> [ 113.950054]
> [ 113.950054] which lock already depends on the new lock.
> [ 113.950054]
> [ 113.950054]
> [ 113.950054] the existing dependency chain (in reverse order) is:
> [ 113.950054]
> [ 113.950054] -> #1 ((fb_notifier_list).rwsem){----}:
> [ 113.950054] [<c014520f>] __lock_acquire+0x129f/0x1930
> [ 113.950054] [<c01458fc>] lock_acquire+0x5c/0x80
> [ 113.950054] [<c03088c9>] down_read+0x49/0x90
> [ 113.950054] [<c0138a9a>] __blocking_notifier_call_chain+0x2a/0x60
> [ 113.950054] [<c0138aea>] blocking_notifier_call_chain+0x1a/0x20
> [ 113.950054] [<c0222f81>] fb_notifier_call_chain+0x11/0x20
> [ 113.950054] [<c022459e>] do_fb_ioctl+0x2ae/0x5d0
> [ 113.950054] [<c02248dd>] fb_ioctl+0x1d/0x20
> [ 113.950054] [<c01a54b0>] vfs_ioctl+0x20/0x80
> [ 113.950054] [<c01a56c2>] do_vfs_ioctl+0x72/0x570
> [ 113.950054] [<c01a5bf9>] sys_ioctl+0x39/0x70
> [ 113.950054] [<c0103371>] sysenter_do_call+0x12/0x31
> [ 113.950054] [<ffffffff>] 0xffffffff
> [ 113.950054]
> [ 113.950054] -> #0 (&fb_info->lock){--..}:
> [ 113.950054] [<c014537c>] __lock_acquire+0x140c/0x1930
> [ 113.950054] [<c01458fc>] lock_acquire+0x5c/0x80
> [ 113.950054] [<c0307f8a>] mutex_lock_nested+0xba/0x2f0
> [ 113.950054] [<c0230c56>] fbcon_event_notify+0x2d6/0x950
> [ 113.950054] [<c030c403>] notifier_call_chain+0x53/0xa0
> [ 113.950054] [<c0138ab4>] __blocking_notifier_call_chain+0x44/0x60
> [ 113.950054] [<c0138aea>] blocking_notifier_call_chain+0x1a/0x20
> [ 113.950054] [<c0222f81>] fb_notifier_call_chain+0x11/0x20
> [ 113.950054] [<c02246ca>] do_fb_ioctl+0x3da/0x5d0
> [ 113.950054] [<c02248dd>] fb_ioctl+0x1d/0x20
> [ 113.950054] [<c01a54b0>] vfs_ioctl+0x20/0x80
> [ 113.950054] [<c01a56c2>] do_vfs_ioctl+0x72/0x570
> [ 113.950054] [<c01a5bf9>] sys_ioctl+0x39/0x70
> [ 113.950054] [<c0103371>] sysenter_do_call+0x12/0x31
> [ 113.950054] [<ffffffff>] 0xffffffff
> [ 113.950054]
> [ 113.950054] other info that might help us debug this:
> [ 113.950054]
> [ 113.950054] 1 lock held by s2disk/4168:
> [ 113.950054] #0: ((fb_notifier_list).rwsem){----}, at: [<c0138a9a>]
> __blocking_notifier_call_chain+0x2a/0x60
> [ 113.950054]
> [ 113.950054] stack backtrace:
> [ 113.950054] Pid: 4168, comm: s2disk Not tainted 2.6.29-rc3-1avb #16
> [ 113.950054] Call Trace:
> [ 113.950054] [<c0306bb8>] ? printk+0x18/0x20
> [ 113.950054] [<c0143b1f>] print_circular_bug_tail+0xcf/0xe0
> [ 113.950054] [<c014537c>] __lock_acquire+0x140c/0x1930
> [ 113.950054] [<c0141105>] ? lock_release_holdtime+0x35/0x210
> [ 113.950054] [<c0143992>] ? check_noncircular+0x92/0x150
> [ 113.950054] [<c01463ce>] ? lock_release_non_nested+0x16e/0x2a0
> [ 113.950054] [<c0138a9a>] ? __blocking_notifier_call_chain+0x2a/0x60
> [ 113.950054] [<c01458fc>] lock_acquire+0x5c/0x80
> [ 113.950054] [<c0230c56>] ? fbcon_event_notify+0x2d6/0x950
> [ 113.950054] [<c0307f8a>] mutex_lock_nested+0xba/0x2f0
> [ 113.950054] [<c0230c56>] ? fbcon_event_notify+0x2d6/0x950
> [ 113.950054] [<c0230c56>] ? fbcon_event_notify+0x2d6/0x950
> [ 113.950054] [<c0230c56>] fbcon_event_notify+0x2d6/0x950
> [ 113.950054] [<c010b9ab>] ? save_stack_trace+0x2b/0x50
> [ 113.950054] [<c014185c>] ? save_trace+0x3c/0xa0
> [ 113.950054] [<c01418fd>] ? add_lock_to_list+0x3d/0xb0
> [ 113.950054] [<c0141105>] ? lock_release_holdtime+0x35/0x210
> [ 113.950054] [<c01462fd>] ? lock_release_non_nested+0x9d/0x2a0
> [ 113.950054] [<c0138a9a>] ? __blocking_notifier_call_chain+0x2a/0x60
> [ 113.950054] [<c030c403>] notifier_call_chain+0x53/0xa0
> [ 113.950054] [<c03088e5>] ? down_read+0x65/0x90
> [ 113.950054] [<c0138ab4>] __blocking_notifier_call_chain+0x44/0x60
> [ 113.950054] [<c0138aea>] blocking_notifier_call_chain+0x1a/0x20
> [ 113.950054] [<c0222f81>] fb_notifier_call_chain+0x11/0x20
> [ 113.950054] [<c02246ca>] do_fb_ioctl+0x3da/0x5d0
> [ 113.950054] [<c01820e6>] ? might_fault+0x46/0xa0
> [ 113.950054] [<c01820e6>] ? might_fault+0x46/0xa0
> [ 113.950054] [<c0182131>] ? might_fault+0x91/0xa0
> [ 113.950054] [<c01820e6>] ? might_fault+0x46/0xa0
> [ 113.950054] [<c0141105>] ? lock_release_holdtime+0x35/0x210
> [ 113.950054] [<c01462fd>] ? lock_release_non_nested+0x9d/0x2a0
> [ 113.950054] [<c01820e6>] ? might_fault+0x46/0xa0
> [ 113.950054] [<c01820e6>] ? might_fault+0x46/0xa0
> [ 113.950054] [<c0182131>] ? might_fault+0x91/0xa0
> [ 113.950054] [<c01820e6>] ? might_fault+0x46/0xa0
> [ 113.950054] [<c0209916>] ? copy_to_user+0x36/0x120
> [ 113.950054] [<c02248c0>] ? fb_ioctl+0x0/0x20
> [ 113.950054] [<c02248dd>] fb_ioctl+0x1d/0x20
> [ 113.950054] [<c01a54b0>] vfs_ioctl+0x20/0x80
> [ 113.950054] [<c01a56c2>] do_vfs_ioctl+0x72/0x570
> [ 113.950054] [<c019c762>] ? sys_fstat64+0x22/0x30
> [ 113.950054] [<c01a2180>] ? __link_path_walk+0x240/0xd40
> [ 113.950054] [<c010339f>] ? sysenter_exit+0xf/0x16
> [ 113.950054] [<c01a5bf9>] sys_ioctl+0x39/0x70
> [ 113.950054] [<c0103371>] sysenter_do_call+0x12/0x31
> [ 117.846986] Syncing filesystems ... done.
>
> In case it helps - suspend to RAM does not exhibit this issue.
Another attempt... ignore the previous patch and try to apply this one
on top of "fbmem: copy_from/to_user() with mutex held (v3)".
Thanks a lot for the tests!
-Andrea
---
fbmem: fix circular locking dependency pushing down fb_info->lock
Push down fb_info->lock into each framebuffer notifier callback to avoid
a circular locking dependency between fb_info->lock and
fb_notifier_list->rwsem.
Reported-by: Andrey Borzenkov <arvidjaar@...l.ru>
Signed-off-by: Andrea Righi <righi.andrea@...il.com>
---
drivers/video/backlight/backlight.c | 3 ++
drivers/video/backlight/lcd.c | 3 ++
drivers/video/console/fbcon.c | 62 +++++++++++++++++++++++++++++++----
drivers/video/fbmem.c | 7 ----
4 files changed, 61 insertions(+), 14 deletions(-)
diff --git a/drivers/video/backlight/backlight.c b/drivers/video/backlight/backlight.c
index 157057c..dd37cbc 100644
--- a/drivers/video/backlight/backlight.c
+++ b/drivers/video/backlight/backlight.c
@@ -35,6 +35,8 @@ static int fb_notifier_callback(struct notifier_block *self,
return 0;
bd = container_of(self, struct backlight_device, fb_notif);
+ if (!lock_fb_info(evdata->info))
+ return -ENODEV;
mutex_lock(&bd->ops_lock);
if (bd->ops)
if (!bd->ops->check_fb ||
@@ -47,6 +49,7 @@ static int fb_notifier_callback(struct notifier_block *self,
backlight_update_status(bd);
}
mutex_unlock(&bd->ops_lock);
+ unlock_fb_info(evdata->info);
return 0;
}
diff --git a/drivers/video/backlight/lcd.c b/drivers/video/backlight/lcd.c
index b644947..0bb13df 100644
--- a/drivers/video/backlight/lcd.c
+++ b/drivers/video/backlight/lcd.c
@@ -40,6 +40,8 @@ static int fb_notifier_callback(struct notifier_block *self,
if (!ld->ops)
return 0;
+ if (!lock_fb_info(evdata->info))
+ return -ENODEV;
mutex_lock(&ld->ops_lock);
if (!ld->ops->check_fb || ld->ops->check_fb(ld, evdata->info)) {
if (event == FB_EVENT_BLANK) {
@@ -51,6 +53,7 @@ static int fb_notifier_callback(struct notifier_block *self,
}
}
mutex_unlock(&ld->ops_lock);
+ unlock_fb_info(evdata->info);
return 0;
}
diff --git a/drivers/video/console/fbcon.c b/drivers/video/console/fbcon.c
index 81fc2f7..7bc2a26 100644
--- a/drivers/video/console/fbcon.c
+++ b/drivers/video/console/fbcon.c
@@ -3144,57 +3144,105 @@ static int fbcon_event_notify(struct notifier_block *self,
switch(action) {
case FB_EVENT_SUSPEND:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
fbcon_suspended(info);
+ unlock_fb_info(info);
break;
case FB_EVENT_RESUME:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
fbcon_resumed(info);
+ unlock_fb_info(info);
break;
case FB_EVENT_MODE_CHANGE:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
fbcon_modechanged(info);
+ unlock_fb_info(info);
break;
case FB_EVENT_MODE_CHANGE_ALL:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
fbcon_set_all_vcs(info);
+ unlock_fb_info(info);
break;
case FB_EVENT_MODE_DELETE:
mode = event->data;
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
ret = fbcon_mode_deleted(info, mode);
+ unlock_fb_info(info);
break;
case FB_EVENT_FB_UNBIND:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
ret = fbcon_fb_unbind(info->node);
+ unlock_fb_info(info);
break;
case FB_EVENT_FB_REGISTERED:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
ret = fbcon_fb_registered(info);
+ unlock_fb_info(info);
break;
case FB_EVENT_FB_UNREGISTERED:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
ret = fbcon_fb_unregistered(info);
+ unlock_fb_info(info);
break;
case FB_EVENT_SET_CONSOLE_MAP:
- /*
- * XXX: info is not used here, we can unlock it to avoid
- * circular locking with ¤t->mm->mmap_sem in fb_mmap().
- */
- unlock_fb_info(info);
con2fb = event->data;
ret = set_con2fb_map(con2fb->console - 1,
con2fb->framebuffer, 1);
- lock_fb_info(info);
break;
case FB_EVENT_GET_CONSOLE_MAP:
con2fb = event->data;
con2fb->framebuffer = con2fb_map[con2fb->console - 1];
break;
case FB_EVENT_BLANK:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
fbcon_fb_blanked(info, *(int *)event->data);
+ unlock_fb_info(info);
break;
case FB_EVENT_NEW_MODELIST:
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
fbcon_new_modelist(info);
+ unlock_fb_info(info);
break;
case FB_EVENT_GET_REQ:
caps = event->data;
+ if (!lock_fb_info(info)) {
+ ret = -ENODEV;
+ goto done;
+ }
fbcon_get_requirement(info, caps);
+ unlock_fb_info(info);
break;
}
-
done:
return ret;
}
diff --git a/drivers/video/fbmem.c b/drivers/video/fbmem.c
index cfd9dce..5cdee72 100644
--- a/drivers/video/fbmem.c
+++ b/drivers/video/fbmem.c
@@ -1086,12 +1086,8 @@ static long do_fb_ioctl(struct fb_info *info, unsigned int cmd,
return -EINVAL;
con2fb.framebuffer = -1;
event.data = &con2fb;
-
- if (!lock_fb_info(info))
- return -ENODEV;
event.info = info;
fb_notifier_call_chain(FB_EVENT_GET_CONSOLE_MAP, &event);
- unlock_fb_info(info);
ret = copy_to_user(argp, &con2fb, sizeof(con2fb)) ? -EFAULT : 0;
break;
@@ -1109,12 +1105,9 @@ static long do_fb_ioctl(struct fb_info *info, unsigned int cmd,
break;
}
event.data = &con2fb;
- if (!lock_fb_info(info))
- return -ENODEV;
event.info = info;
ret = fb_notifier_call_chain(FB_EVENT_SET_CONSOLE_MAP,
&event);
- unlock_fb_info(info);
break;
case FBIOBLANK:
if (!lock_fb_info(info))
--
1.5.6.3
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists