Re: 3.7-rc7: BUG: MAX_STACK_TRACE_ENTRIES too low!

[Li Zhong]

On Tue, 2012-11-27 at 19:22 -0800, Christian Kujau wrote:
> On Tue, 27 Nov 2012 at 19:06, Christian Kujau wrote:
> > the same thing[0] happened again in 3.7-rc7, after ~20h uptime:
> 
> I found the following on patchwork, but this seems to deal with powerpc64 
> only, while this PowerBook G4 of mine is powerpc32:
> 
>   http://patchwork.ozlabs.org/patch/193414/
> 
> It looks related, but then again, I fail to parse assember...

Hi Christian, 

Would you please help to try the following fix? I don't have a powerpc32
machine for test...

Thanks, Zhong

==============================================
diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S
index 9499385..cadebfd 100644
--- a/arch/powerpc/kernel/entry_32.S
+++ b/arch/powerpc/kernel/entry_32.S
@@ -439,6 +439,8 @@ ret_from_fork:
 ret_from_kernel_thread:
 	REST_NVGPRS(r1)
 	bl	schedule_tail
+	li	r3,0
+	stw	r3,0(r1)
 	mtlr	r14
 	mr	r3,r15
 	PPC440EP_ERR42
==============================================

> 
> Christian.
> 
> > [40007.339487] [sched_delayed] sched: RT throttling activated
> > [69731.388717] BUG: MAX_STACK_TRACE_ENTRIES too low!
> > [69731.390371] turning off the locking correctness validator.
> > [69731.391942] Call Trace:
> > [69731.393525] [c9a61c10] [c0009064] show_stack+0x70/0x1bc (unreliable)
> > [69731.395152] [c9a61c50] [c0077460] save_trace+0xfc/0x114
> > [69731.396735] [c9a61c60] [c007be20] __lock_acquire+0x1568/0x19b8
> > [69731.398296] [c9a61d00] [c007c2c0] lock_acquire+0x50/0x70
> > [69731.399857] [c9a61d20] [c0550e28] _raw_spin_lock_irq+0x5c/0x78
> > [69731.401419] [c9a61d40] [c054fb58] __schedule+0xd8/0x534
> > [69731.402972] [c9a61da0] [c0550094] _cond_resched+0x50/0x68
> > [69731.404527] [c9a61db0] [c0479908] dst_gc_task+0xbc/0x258
> > [69731.406070] [c9a61e40] [c004eeb8] process_one_work+0x1f4/0x49c
> > [69731.407585] [c9a61e80] [c004f644] worker_thread+0x14c/0x400
> > [69731.409075] [c9a61eb0] [c0057634] kthread+0xbc/0xc0
> > [69731.410521] [c9a61f40] [c0011ad4] ret_from_kernel_thread+0x5c/0x64
> > [...repeated 54 times...]
> > 
> > Anyone knows what this is about?
> > 
> > Thanks,
> > Christian.
> > 
> > [0] http://lkml.indiana.edu/hypermail/linux/kernel/1211.0/03025.html


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/