linux-kernel - Re: rcu self-detected stall messages on OMAP3, 4 boards

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 for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:	Wed, 19 Sep 2012 17:03:51 -0700
From:	"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>
To:	Paul Walmsley <paul@...an.com>
Cc:	"Paul E. McKenney" <paul.mckenney@...aro.org>,
	linux-kernel@...r.kernel.org, bbruce@...com,
	linux-omap@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
	khilman@...com, santosh.shilimkar@...com, jon-hunter@...com,
	snijsure@...d-net.com
Subject: Re: rcu self-detected stall messages on OMAP3, 4 boards

On Thu, Sep 13, 2012 at 06:52:10PM +0000, Paul Walmsley wrote:
> Hi Paul,
> 
> thanks for the reply,
> 
> On Wed, 12 Sep 2012, Paul E. McKenney wrote:
> 
> > Interesting.  I am assuming that the interrupt in the stack below came
> > from idle, if not, please let me know what.
> 
> According to the exception stack section in the original traceback, it
> appears that the serial interrupt took the SoC out of idle.
> 
> > Could you please reproduce with CONFIG_RCU_CPU_STALL_INFO=y?  That would
> > give me a bit more information about why RCU thought that there was
> > a stall.  (CCing Becky Bruce, who saw something similar recently.)
> 
> At the bottom of this mail is a series of tracebacks with
> CONFIG_RCU_CPU_STALL_INFO=y.  Unlike the traceback that was sent in
> the last message, these were not triggered by serial activity.  These
> appeared every 300 seconds.
> 
> > Subodh Nijsure (also CCed) reported something that might be similar on
> > ARM, and also reported that setting the following got rid of the stalls:
> > 
> > 	CONFIG_CPU_IDLE=y
> > 	CONFIG_CPU_IDLE_GOV_LADDER=y
> > 	CONFIG_CPU_IDLE_GOV_MENU=y
> > 
> > At which point he was happy, which was good, but which also left the
> > underlying problem unsolved.  Do these affect your system?  If so,
> > do they cause a different ARM idle loop to be executed?
> 
> Will give this a try.  What board was Subodh using?

Hello, Paul,

Any news on trying the above settings?

							Thanx, Paul

> - Paul
> 
> 
> Debian GNU/Linux wheezy/sid armel ttyO2
> 
> armel login: [  305.942108] INFO: rcu_sched self-detected stall on CPU
> [  305.944946]  1: (7 GPs behind) idle=57b/1/0 
> [  305.947265]   (t=22811 jiffies)
> [  305.949066] [<c001b7cc>] (unwind_backtrace+0x0/0xf0) from [<c00acc28>] (rcu_check_callbacks+0x1b0/0x678)
> [  305.954223] [<c00acc28>] (rcu_check_callbacks+0x1b0/0x678) from [<c00529e0>] (update_process_times+0x38/0x68)
> [  305.959625] [<c00529e0>] (update_process_times+0x38/0x68) from [<c008bf14>] (tick_sched_timer+0x80/0xec)
> [  305.964813] [<c008bf14>] (tick_sched_timer+0x80/0xec) from [<c006840c>] (__run_hrtimer+0x7c/0x1e0)
> [  305.969696] [<c006840c>] (__run_hrtimer+0x7c/0x1e0) from [<c00691f0>] (hrtimer_interrupt+0x11c/0x2d0)
> [  305.974731] [<c00691f0>] (hrtimer_interrupt+0x11c/0x2d0) from [<c001a04c>] (twd_handler+0x30/0x44)
> [  305.979644] [<c001a04c>] (twd_handler+0x30/0x44) from [<c00a7068>] (handle_percpu_devid_irq+0x90/0x13c)
> [  305.984741] [<c00a7068>] (handle_percpu_devid_irq+0x90/0x13c) from [<c00a37dc>] (generic_handle_irq+0x30/0x48)
> [  305.990234] [<c00a37dc>] (generic_handle_irq+0x30/0x48) from [<c0014c58>] (handle_IRQ+0x4c/0xac)
> [  305.995025] [<c0014c58>] (handle_IRQ+0x4c/0xac) from [<c0008478>] (gic_handle_irq+0x28/0x5c)
> [  305.999633] [<c0008478>] (gic_handle_irq+0x28/0x5c) from [<c04f8ca4>] (__irq_svc+0x44/0x5c)
> [  306.004180] Exception stack(0xde86ff88 to 0xde86ffd0)
> [  306.006927] ff80:                   0003c6d0 00000001 00000000 de8660c0 de86e000 c07c23c8
> [  306.011383] ffa0: c0504590 c0749e20 00000000 411fc092 c074a040 00000000 00000001 de86ffd0
> [  306.015838] ffc0: 0003c6d1 c0014f50 20000113 ffffffff
> [  306.018585] [<c04f8ca4>] (__irq_svc+0x44/0x5c) from [<c0014f50>] (default_idle+0x20/0x44)
> [  306.023040] [<c0014f50>] (default_idle+0x20/0x44) from [<c001517c>] (cpu_idle+0x9c/0x114)
> [  306.027526] [<c001517c>] (cpu_idle+0x9c/0x114) from [<804f1af4>] (0x804f1af4)
> [  602.004486] INFO: rcu_sched detected stalls on CPUs/tasks:
> [  602.007476]  (detected by 0, t=60707 jiffies)
> [  602.009857] INFO: Stall ended before state dump start
> [  906.027893] INFO: rcu_sched self-detected stall on CPU
> [  906.030700]  1: (6 GPs behind) idle=647/1/0 
> [  906.033020]   (t=38379 jiffies)
> [  906.034790] [<c001b7cc>] (unwind_backtrace+0x0/0xf0) from [<c00acc28>] (rcu_check_callbacks+0x1b0/0x678)
> [  906.039947] [<c00acc28>] (rcu_check_callbacks+0x1b0/0x678) from [<c00529e0>] (update_process_times+0x38/0x68)
> [  906.045349] [<c00529e0>] (update_process_times+0x38/0x68) from [<c008bf14>] (tick_sched_timer+0x80/0xec)
> [  906.050537] [<c008bf14>] (tick_sched_timer+0x80/0xec) from [<c006840c>] (__run_hrtimer+0x7c/0x1e0)
> [  906.055419] [<c006840c>] (__run_hrtimer+0x7c/0x1e0) from [<c00691f0>] (hrtimer_interrupt+0x11c/0x2d0)
> [  906.060424] [<c00691f0>] (hrtimer_interrupt+0x11c/0x2d0) from [<c001a04c>] (twd_handler+0x30/0x44)
> [  906.065307] [<c001a04c>] (twd_handler+0x30/0x44) from [<c00a7068>] (handle_percpu_devid_irq+0x90/0x13c)
> [  906.070434] [<c00a7068>] (handle_percpu_devid_irq+0x90/0x13c) from [<c00a37dc>] (generic_handle_irq+0x30/0x48)
> [  906.075897] [<c00a37dc>] (generic_handle_irq+0x30/0x48) from [<c0014c58>] (handle_IRQ+0x4c/0xac)
> [  906.080688] [<c0014c58>] (handle_IRQ+0x4c/0xac) from [<c0008478>] (gic_handle_irq+0x28/0x5c)
> [  906.085296] [<c0008478>] (gic_handle_irq+0x28/0x5c) from [<c04f8ca4>] (__irq_svc+0x44/0x5c)
> [  906.089843] Exception stack(0xde86ff88 to 0xde86ffd0)
> [  906.092590] ff80:                   0003cb06 00000001 00000000 de8660c0 de86e000 c07c23c8
> [  906.097045] ffa0: c0504590 c0749e20 00000000 411fc092 c074a040 00000000 00000001 de86ffd0
> [  906.101501] ffc0: 0003cb07 c0014f50 20000113 ffffffff
> [  906.104278] [<c04f8ca4>] (__irq_svc+0x44/0x5c) from [<c0014f50>] (default_idle+0x20/0x44)
> [  906.108734] [<c0014f50>] (default_idle+0x20/0x44) from [<c001517c>] (cpu_idle+0x9c/0x114)
> [  906.113189] [<c001517c>] (cpu_idle+0x9c/0x114) from [<804f1af4>] (0x804f1af4)
> 

--
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/