Re: sysfs lockdep complaint

["Américo Wang" <xiyou.wangcong@gmail.com>]

On Sat, Feb 20, 2010 at 03:01:33PM +0100, Johannes Berg wrote:
>On Sat, 2010-02-20 at 21:51 +0800, Américo Wang wrote:
>
>> Please send us the backtraces you got, we could expect this.
>
>The first one is fun:
>
>[    1.576226] windfarm: initializing for dual-core desktop G5
>[    1.577182] BUG: key 6b6b6b6b6b6b6b6b not in .data!
>[    1.577190] ------------[ cut here ]------------
>[    1.577192] Badness at kernel/lockdep.c:2706
>[    1.577196] NIP: c0000000000b6a68 LR: c0000000000b6a4c CTR: c0000000003afedc
>[    1.577200] REGS: c000000216197620 TRAP: 0700   Not tainted  (2.6.33-rc8-wl-64022-gcdee415-dirty)
>[    1.577203] MSR: 9000000000029032 <EE,ME,CE,IR,DR>  CR: 28000082  XER: 200fffff
>[    1.577214] TASK = c000000216198000[1] 'swapper' THREAD: c000000216194000 CPU: 2
>[    1.577217] GPR00: 0000000000000000 c0000002161978a0 c000000000b10328 0000000000000001 
>[    1.577223] GPR04: 0000000000000001 c000000000071638 0000000000000000 0000000000000002 
>[    1.577229] GPR08: 0000000000000000 c0000000014320f0 0000000000000001 c000000000b27638 
>[    1.577235] GPR12: 0000000000000041 c000000000b21880 0000000000000056 0000000001400000 
>[    1.577241] GPR16: 0000000000240ee0 c000000000718923 c0000000007ff6ec 000000000000005e 
>[    1.577247] GPR20: 00000000ffb30f40 c0000000007188b9 0000000000000000 0000000000240ed0 
>[    1.577253] GPR24: 0000000000000108 0000000000000000 fffffffffffffff4 c00000021488b000 
>[    1.577259] GPR28: 6b6b6b6b6b6b6b6b c00000021488bcc8 c000000000a77780 c0000002161978a0 
>[    1.577271] NIP [c0000000000b6a68] .lockdep_init_map+0x148/0x66c
>[    1.577276] LR [c0000000000b6a4c] .lockdep_init_map+0x12c/0x66c
>[    1.577278] Call Trace:
>[    1.577282] [c0000002161978a0] [c0000000000b6a34] .lockdep_init_map+0x114/0x66c (unreliable)
>[    1.577289] [c0000002161979a0] [c0000000001efd88] .sysfs_add_file_mode+0x90/0x124
>[    1.577294] [c000000216197a60] [c0000000001efe58] .sysfs_add_file+0x3c/0x50
>[    1.577299] [c000000216197b00] [c0000000001effb4] .sysfs_create_file+0x5c/0x74
>[    1.577305] [c000000216197ba0] [c0000000004292b0] .device_create_file+0x40/0x5c
>[    1.577311] [c000000216197c40] [c000000000441178] .wf_register_control+0x108/0x178
>[    1.577316] [c000000216197cf0] [c0000000004440a4] .smu_fan_create+0x548/0x598
>[    1.577323] [c000000216197da0] [c0000000007e5e6c] .smu_controls_init+0xb0/0x1f4
>[    1.577329] [c000000216197e40] [c0000000000084e4] .do_one_initcall+0xac/0x204
>[    1.577335] [c000000216197ef0] [c0000000007b84b8] .kernel_init+0x1f0/0x270
>[    1.577340] [c000000216197f90] [c000000000028f8c] .kernel_thread+0x54/0x70

This should be fixed by the following patch, it is based on the
4 patches from Eric.

Thanks.


Reported-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: WANG Cong <xiyou.wangcong@gmail.com>


---
diff --git a/drivers/macintosh/windfarm_smu_controls.c b/drivers/macintosh/windfarm_smu_controls.c
index 6c68b9e..5f4e5a0 100644
--- a/drivers/macintosh/windfarm_smu_controls.c
+++ b/drivers/macintosh/windfarm_smu_controls.c
@@ -173,6 +173,7 @@ static struct smu_fan_control *smu_fan_create(struct device_node *node,
 
 	fct->fan_type = pwm_fan;
 	fct->ctrl.type = pwm_fan ? WF_CONTROL_PWM_FAN : WF_CONTROL_RPM_FAN;
+	sysfs_attr_init(&fct->ctrl.attr);
 
 	/* We use the name & location here the same way we do for SMU sensors,
 	 * see the comment in windfarm_smu_sensors.c. The locations are a bit