Message-ID: <50462190.7090702@grid-net.com>
Date:	Tue, 4 Sep 2012 08:43:12 -0700
From:	Subodh Nijsure <snijsure@...d-net.com>
To:	<paulmck@...ux.vnet.ibm.com>
CC:	Artem Bityutskiy <dedekind1@...il.com>,
	<linux-mtd@...ts.infradead.org>,
	Linux Kernel Maling List <linux-kernel@...r.kernel.org>,
	<linux-arm-kernel@...ts.infradead.org>
Subject: Re: RCU warning during UBI file system mount

Hello Paul,

On 09/04/2012 07:48 AM, Paul E. McKenney wrote:
> On Tue, Sep 04, 2012 at 10:46:36AM +0300, Artem Bityutskiy wrote:
>> Let's CC LKML and Paul.
>>
>>   
>> On Sat, 2012-09-01 at 22:19 -0700, Subodh Nijsure wrote:
>>> Hello,
>>>
>>> I am running 3.6-rc4 kernel on my MX28 based board and every time I
>>> mount UBI file system I see following warning.
>>> Anybody else seen this errors, particularly those running it on MXS
>>> platform?
>>>
>>> I haven't debugged RCU warnings before any clues as to what is the
>>> source of this warning is, should I be looking at
>>> start_dma_without_bch_irq() code? Would appreciate any pointers in
>>> fixing this...
> RCU is upset because what it believes to be a non-idle task is telling
> RCU that it is entering the idle loop.
>
> Does the MX28 kernel have any unusual idle features?  (This is a subarch
> of ARM, correct?)
>
> 							Thanx, Paul

MX28 is ARM device. I removed following from my config and these 
warnings went away.

CONFIG_CPU_IDLE=y
CONFIG_CPU_IDLE_GOV_LADDER=y
CONFIG_CPU_IDLE_GOV_MENU=y

I have attached the kernel configuration file I used.

It doesn't affect specific hw I am working with as device is not 
required to enter idle mode.
I am Cc-ing linux-arm folks,  in case someone on arm platform requires 
CPU idele &, RCU to work.

-Subodh
>>> [    4.280000] WARNING: at /opt/git/kernel-mainline/kernel/rcutiny.c:75
>>> rcu_idle_enter_common.clone.7+0x90/0xac()
>>> [    4.280000] Current pid: 1 comm: swapper / Idle pid: 0 comm: swapper
>>> [    4.280000] Modules linked in:
>>> [    4.280000] [<c0014eec>] (unwind_backtrace+0x0/0x130) from
>>> [<c001dd94>] (warn_slowpath_common+0x4c/0x64)
>>> [    4.280000] [<c001dd94>] (warn_slowpath_common+0x4c/0x64) from
>>> [<c001de40>] (warn_slowpath_fmt+0x30/0x40)
>>> [    4.280000] [<c001de40>] (warn_slowpath_fmt+0x30/0x40) from
>>> [<c0061618>] (rcu_idle_enter_common.clone.7+0x90/0xac)
>>> [    4.280000] [<c0061618>] (rcu_idle_enter_common.clone.7+0x90/0xac)
>>> from [<c0061670>] (rcu_irq_exit+0x3c/0x78)
>>> [    4.280000] [<c0061670>] (rcu_irq_exit+0x3c/0x78) from [<c000fe90>]
>>> (handle_IRQ+0x34/0x84)
>>> [    4.280000] [<c000fe90>] (handle_IRQ+0x34/0x84) from [<c000ec34>]
>>> (__irq_svc+0x34/0x58)
>>> [    4.280000] [<c000ec34>] (__irq_svc+0x34/0x58) from [<c0394c4c>]
>>> (schedule_timeout+0xdc/0x16c)
>>> [    4.280000] [<c0394c4c>] (schedule_timeout+0xdc/0x16c) from
>>> [<c0395f34>] (wait_for_common+0xb4/0x160)
>>> [    4.280000] [<c0395f34>] (wait_for_common+0xb4/0x160) from
>>> [<c02579d0>] (start_dma_without_bch_irq+0x6c/0xa8)
>>> [    4.280000] [<c02579d0>] (start_dma_without_bch_irq+0x6c/0xa8) from
>>> [<c0258660>] (gpmi_send_command+0x120/0x154)
>>> [    4.280000] [<c0258660>] (gpmi_send_command+0x120/0x154) from
>>> [<c0256fb4>] (gpmi_cmd_ctrl+0x44/0x6c)
>>> [    4.280000] [<c0256fb4>] (gpmi_cmd_ctrl+0x44/0x6c) from [<c024f8e0>]
>>> (nand_command_lp+0xe8/0x270)
>>> [    4.280000] [<c024f8e0>] (nand_command_lp+0xe8/0x270) from
>>> [<c0251e00>] (nand_do_read_ops+0x124/0x444)
>>> [    4.280000] [<c0251e00>] (nand_do_read_ops+0x124/0x444) from
>>> [<c0252558>] (nand_read+0x54/0x78)
>>> [    4.280000] [<c0252558>] (nand_read+0x54/0x78) from [<c02499a4>]
>>> (part_read+0x48/0x80)
>>> [    4.280000] [<c02499a4>] (part_read+0x48/0x80) from [<c0246910>]
>>> (mtd_read+0x88/0xb4)
>>> [    4.280000] [<c0246910>] (mtd_read+0x88/0xb4) from [<c026b188>]
>>> (ubi_io_read+0xb0/0x3e8)
>>> [    4.280000] [<c026b188>] (ubi_io_read+0xb0/0x3e8) from [<c026740c>]
>>> (ubi_eba_read_leb+0x198/0x4bc)
>>> [    4.280000] [<c026740c>] (ubi_eba_read_leb+0x198/0x4bc) from
>>> [<c0266800>] (ubi_leb_read+0xf4/0x16c)
>>> [    4.280000] [<c0266800>] (ubi_leb_read+0xf4/0x16c) from [<c0155550>]
>>> (ubifs_leb_read+0x24/0x88)
>>> [    4.280000] [<c0155550>] (ubifs_leb_read+0x24/0x88) from [<c01572f8>]
>>> (ubifs_read_node+0xa8/0x2e4)
>>> [    4.280000] [<c01572f8>] (ubifs_read_node+0xa8/0x2e4) from
>>> [<c0153c50>] (ubifs_read_sb_node+0x54/0x78)
>>> [    4.280000] [<c0153c50>] (ubifs_read_sb_node+0x54/0x78) from
>>> [<c0154940>] (ubifs_read_superblock+0xc8c/0x133c)
>>> [    4.280000] [<c0154940>] (ubifs_read_superblock+0xc8c/0x133c) from
>>> [<c01520ac>] (ubifs_mount+0x73c/0x1bc0)
>>> [    4.280000] [<c01520ac>] (ubifs_mount+0x73c/0x1bc0) from [<c00a49ec>]
>>> (mount_fs+0x44/0x184)
>>> [    4.280000] [<c00a49ec>] (mount_fs+0x44/0x184) from [<c00baf90>]
>>> (vfs_kern_mount+0x4c/0xc0)
>>> [    4.280000] [<c00baf90>] (vfs_kern_mount+0x4c/0xc0) from [<c00bb1a0>]
>>> (do_kern_mount+0x34/0xd0)
>>> [    4.280000] [<c00bb1a0>] (do_kern_mount+0x34/0xd0) from [<c00bc93c>]
>>> (do_mount+0x15c/0x73c)
>>> [    4.280000] [<c00bc93c>] (do_mount+0x15c/0x73c) from [<c00bcfa0>]
>>> (sys_mount+0x84/0xc4)
>>> [    4.280000] [<c00bcfa0>] (sys_mount+0x84/0xc4) from [<c04e4a04>]
>>> (do_mount_root+0x24/0xb0)
>>> [    4.280000] [<c04e4a04>] (do_mount_root+0x24/0xb0) from [<c04e4b78>]
>>> (mount_block_root+0xe8/0x214)
>>> [    4.280000] [<c04e4b78>] (mount_block_root+0xe8/0x214) from
>>> [<c04e4e28>] (prepare_namespace+0x8c/0x180)
>>> [    4.280000] [<c04e4e28>] (prepare_namespace+0x8c/0x180) from
>>> [<c04e4894>] (kernel_init+0x16c/0x1b4)
>>> [    4.280000] [<c04e4894>] (kernel_init+0x16c/0x1b4) from [<c000ff70>]
>>> (kernel_thread_exit+0x0/0x8)
>>> [    4.280000] ---[ end trace 4d36b722759fdc4f ]---
>>> [    4.950000] UBIFS DBG (pid 58): ubifs_bg_thread: background thread
>>> "ubifs_bgt0_0" started, PID 58
>>>
>>> -Subodh
>>>
>>> ______________________________________________________
>>> Linux MTD discussion mailing list
>>> http://lists.infradead.org/mailman/listinfo/linux-mtd/
>> -- 
>> Best Regards,
>> Artem Bityutskiy
>


View attachment "config" of type "text/plain" (54946 bytes)

Powered by blists - more mailing lists