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
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Tue, 6 May 2008 15:52:54 -0400
From:	Daniel Hazelton <dhazelton@...er.net>
To:	Stephen Rothwell <sfr@...b.auug.org.au>
Cc:	linux-kernel@...r.kernel.org, linux-next@...r.kernel.org,
	Ingo Molnar <mingo@...e.hu>
Subject: Re: linux-next 5/5 does not boot

On Tuesday 06 May 2008 01:37:46 Stephen Rothwell wrote:
> cc'ing Ingo ...

About this error - should I bother trying the 5/6 tree? If I'm going to be 
stuck with a lock-up during boot I'm wondering if I should try a test before 
seeing any action on this bug. (I realize I could just turn off the ftrace 
code, but I am attempting to do some "regular use" testing of the -next tree 
with new options (well, those that I think might become defaulted to "on") 
enabled)

DRH

> On Mon, 5 May 2008 20:04:33 -0400 Daniel Hazelton <dhazelton@...er.net> 
wrote:
> > This may be related to one of the stack dumps I reported earlier as
> > occurring when booting the 4/30 linux-next tree...
> >
> > During two attempted boots of the 5/5 tree using the same config as the
> > 4/30 tree the system locked solid at "testing tracer ftrace".
> >
> > The 4/30 tree produced the following dump (the 5/5 tree locks solid as
> > soon as the test is started):
> > [   10.238230] Testing tracer ftrace: <4>------------[ cut here
> > ]------------ [   10.356558] WARNING: at kernel/lockdep.c:2712
> > check_flags+0xfb/0x180() [   10.356726] Modules linked in:
> > [   10.356915] Pid: 1, comm: swapper Not tainted 2.6.25-next-20080430 #2
> > [   10.357718]
> > [   10.357722] Call Trace:
> > [   10.357939]  <IRQ>  [<ffffffff8023d5f4>] warn_on_slowpath+0x64/0x90
> > [   10.358174]  [<ffffffff8025f789>] ? tick_get_tick_sched+0x9/0x30
> > [   10.358337]  [<ffffffff8022f048>] ? __update_rq_clock+0x68/0x140
> > [   10.358501]  [<ffffffff8026173d>] ? trace_hardirqs_off+0xd/0x10
> > [   10.358665]  [<ffffffff8026173d>] ? trace_hardirqs_off+0xd/0x10
> > [   10.358827]  [<ffffffff802953ce>] ? function_trace_call+0x9e/0xa0
> > [   10.358992]  [<ffffffff8020c4b4>] ? ftrace_call+0x5/0x2b
> > [   10.359148]  [<ffffffff8023177a>] ? cpu_clock+0xda/0x190
> > [   10.359302]  [<ffffffff80262d7b>] check_flags+0xfb/0x180
> > [   10.359457]  [<ffffffff80265635>] lock_acquire+0x65/0xd0
> > [   10.359614]  [<ffffffff80231799>] cpu_clock+0xf9/0x190
> > [   10.359769]  [<ffffffff80293de9>] ftrace_now+0x9/0x10
> > [   10.359920]  [<ffffffff80295194>] trace_function+0xc4/0x260
> > [   10.360079]  [<ffffffff80243723>] ? _local_bh_enable+0x63/0xc0
> > [   10.360240]  [<ffffffff804aeec9>] ? sub_preempt_count+0x9/0x90
> > [   10.360400]  [<ffffffff80243723>] ? _local_bh_enable+0x63/0xc0
> > [   10.360561]  [<ffffffff804aeec9>] ? sub_preempt_count+0x9/0x90
> > [   10.360721]  [<ffffffff8029539c>] function_trace_call+0x6c/0xa0
> > [   10.360883]  [<ffffffff8020c4b4>] ftrace_call+0x5/0x2b
> > [   10.361038]  [<ffffffff804aeec9>] ? sub_preempt_count+0x9/0x90
> > [   10.361199]  [<ffffffff80243723>] _local_bh_enable+0x63/0xc0
> > [   10.361357]  [<ffffffff8024399e>] __do_softirq+0xfe/0x150
> > [   10.361514]  [<ffffffff8020d93c>] call_softirq+0x1c/0x30
> > [   10.361669]  [<ffffffff8020f76a>] do_softirq+0x6a/0xb0
> > [   10.361823]  [<ffffffff80244155>] irq_exit+0xa5/0xb0
> > [   10.361975]  [<ffffffff802204ee>] smp_apic_timer_interrupt+0x9e/0xf0
> > [   10.362143]  [<ffffffff8020d3fb>] apic_timer_interrupt+0x6b/0x70
> > [   10.362301]  <EOI>  [<ffffffff804ab808>] ?
> > _spin_unlock_irqrestore+0x68/0x80
> > [   10.362543]  [<ffffffff8023540b>] ? try_to_wake_up+0xab/0x270
> > [   10.362704]  [<ffffffff8020c4b4>] ? ftrace_call+0x5/0x2b
> > [   10.362860]  [<ffffffff80235625>] ? wake_up_process+0x15/0x20
> > [   10.362979]  [<ffffffff80255264>] ? kthread_create+0xa4/0x120
> > [   10.362979]  [<ffffffff8027a390>] ? do_stop+0x0/0x190
> > [   10.362979]  [<ffffffff80265a6e>] ? mark_held_locks+0x3e/0x80
> > [   10.362979]  [<ffffffff80265d4d>] ? trace_hardirqs_on+0xd/0x10
> > [   10.362979]  [<ffffffff80265c85>] ?
> > trace_hardirqs_on_caller+0xf5/0x1b0 [   10.362979]  [<ffffffff80263f2b>]
> > ? lockdep_init_map+0x4b/0x140 [   10.362979]  [<ffffffff8037d6f8>] ?
> > __spin_lock_init+0x38/0x70 [   10.362979]  [<ffffffff80255714>] ?
> > init_waitqueue_head+0x24/0x40 [   10.362979]  [<ffffffff8028ffc0>] ?
> > __ftrace_modify_code+0x0/0xa0 [   10.362979]  [<ffffffff8027a185>] ?
> > __stop_machine_run+0x85/0xf0 [   10.362979]  [<ffffffff8028ffc0>] ?
> > __ftrace_modify_code+0x0/0xa0 [   10.362979]  [<ffffffff8028ffc0>] ?
> > __ftrace_modify_code+0x0/0xa0 [   10.362979]  [<ffffffff8028ffc0>] ?
> > __ftrace_modify_code+0x0/0xa0 [   10.362979]  [<ffffffff8027a223>] ?
> > stop_machine_run+0x33/0x60 [   10.362979]  [<ffffffff8028f3a0>] ?
> > ftrace_run_update_code+0x20/0x30 [   10.362979]  [<ffffffff8028f7ea>] ?
> > unregister_ftrace_function+0x14a/0x190 [   10.362979] 
> > [<ffffffff80292e70>] ? tracing_stop_function_trace+0x10/0x20 [  
> > 10.362979]  [<ffffffff80296759>] ? stop_function_trace+0x9/0x20 [  
> > 10.362979]  [<ffffffff80296787>] ? function_trace_ctrl_update+0x17/0x20 [
> >   10.362979]  [<ffffffff80292d5e>] ?
> > trace_selftest_startup_function+0xae/0x140
> > [   10.362979]  [<ffffffff8029220d>] ? register_tracer+0x11d/0x1f0
> > [   10.362979]  [<ffffffff80665d50>] ? init_function_trace+0x10/0x20
> > [   10.362979]  [<ffffffff8064eee4>] ? kernel_init+0x164/0x2f0
> > [   10.362979]  [<ffffffff804aeec9>] ? sub_preempt_count+0x9/0x90
> > [   10.362979]  [<ffffffff804ab77b>] ? _spin_unlock_irq+0x3b/0x60
> > [   10.362979]  [<ffffffff804aab1d>] ? trace_hardirqs_on_thunk+0x3a/0x3f
> > [   10.362979]  [<ffffffff80265c85>] ?
> > trace_hardirqs_on_caller+0xf5/0x1b0 [   10.362979]  [<ffffffff8020d5d8>]
> > ? child_rip+0xa/0x12
> > [   10.362979]  [<ffffffff8020cc63>] ? restore_args+0x0/0x30
> > [   10.362979]  [<ffffffff8064ed80>] ? kernel_init+0x0/0x2f0
> > [   10.362979]  [<ffffffff8020d5ce>] ? child_rip+0x0/0x12
> > [   10.362979]
> > [   10.362979] ---[ end trace 5a5d197966b56a2e ]---
> > [   10.362979] irq event stamp: 3391956
> > [   10.362979] hardirqs last  enabled at (3391954): [<ffffffff80265d4d>]
> > trace_hardirqs_on+0xd/0x10
> > [   10.362979] hardirqs last disabled at (3391955): [<ffffffff8026173d>]
> > trace_hardirqs_off+0xd/0x10
> > [   10.362979] softirqs last  enabled at (3391956): [<ffffffff8024399e>]
> > __do_softirq+0xfe/0x150
> > [   10.362979] softirqs last disabled at (3391933): [<ffffffff8020d93c>]
> > call_softirq+0x1c/0x30
> > [   10.376978] kstopmachine used greatest stack depth: 6000 bytes left
> > [   10.377650] PASSED
> >
> > config for 5/5 tree will be provided on request. Dmesg from 4/30 tree is
> > also available. System had not mounted the root drive before the ftrace
> > crash, so I have no available logs or dmesg from the 5/5 tree. If asked,
> > I will attempt to setup a netconsole and try to capture the logs there.
> >
> > DRH



-- 
Dialup is like pissing through a pipette. Slow and excruciatingly painful.
--
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

Powered by Openwall GNU/*/Linux Powered by OpenVZ