[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [microblaze-uclinux] BUG: scheduling while atomic: (Kernel 2.6)

More info:
BUG: scheduling while atomic: more/0x7200546f/101

Stack:
 239afd20 00000001 00000018 239afefc 00000ff9 22001cb4 22155cc0 238698f0
 7200546f 00000065 00989680 00000000 239ae000 00000008 00000074 0000000a
 00000001 00000018 239afefc 00000ff9 2397fec8 00000001 00000000 00000000
Call Trace:
[<22001cb4>] ret_from_intr+0x1c/0x40

I can get this with thttpd as well with:
while [ 1 ]; do curl http://[ipaddress] > /dev/null; done

yields

BUG: scheduling while atomic: thttpd/0xfff90000/46

Stack:
 2389ba30 00000001 00000016 00000000 00000005 22001cb4 22155cc0 23869b50
 fff90000 0000002e 00989680 00000019 2389a000 00000008 2389ffa8 00000009
 00000001 00000016 00000000 00000005 2389c0f4 2389c018 23894fac 00000000
Call Trace:
[<22001cb4>] ret_from_intr+0x1c/0x40
[<22000000>] _start+0x0/0x50
[<220ea494>] skb_release_data+0x88/0xf4
[<220ea494>] skb_release_data+0x88/0xf4
[<220ea260>] kfree_skbmem+0x84/0xf0
[<220ea1e8>] kfree_skbmem+0xc/0xf0
[<2211b0a0>] tcp_cong_avoid+0x14/0x38
[<220ea1e8>] kfree_skbmem+0xc/0xf0
[<220094fc>] __call_console_drivers+0x64/0x88
[<2200ed70>] __do_softirq+0x28/0x3c
[<220095b8>] _call_console_drivers+0x98/0xb0
[<22026488>] handle_IRQ_event+0x5c/0xc8
[<220aeb80>] vscnprintf+0xc/0x38
[<220ae8c0>] vsnprintf+0x37c/0x630
[<22009904>] release_console_sem+0x110/0x2b0
[<22009c98>] vprintk+0x1d0/0x3b8
[<22009b30>] vprintk+0x68/0x3b8
[<22001728>] do_IRQ+0x40/0x54
[<22009ea0>] printk+0x20/0x34
[<220aec78>] sprintf+0x28/0x3c
[<22001cb4>] ret_from_intr+0x1c/0x40
[<220055f4>] scheduler_tick+0x18/0xa8
Reverting to the 2.4 kernel doesn't produce these messages, but I think that's just because 
the checking isn't in there.  It still barfs, but silently.

This is on a microblaze spartan3E 1600.

Any ideas?

thanks,
jim

Jim Van Vorst wrote:

I get this a lot as well:

# more /proc/partitions
<< /proc/partitions >>
major minor  #blocks  name

 31     0        768 mtdblock0
 31     1        256 mtdblock1
 31     2        256 mtdblock2
 31     3        128 mtdblock3
 31     4       4096 mtdblock4
 31     5       1024 mtdblock5
 31     6   BUG: scheduling while atomic: sh/0x00000008/47

Stack:
 23effc08 2200ed08 23e9140c 00000000 00000000 22001c54 22161ca0 23c9d8ec
 00000008 0000002f 01c9c380 0000000e 23efe000 00000008 00000000 00000000
 23efffbd 23e9140c 00000000 00000000 2397ff1c 00000000 00000000 23e91c54
Call Trace: [<2200ed08>] [<22001c54>] [<2215daa4>] [<2200ed08>] [<220016b4>] 
   9216 mtdblock6
 31     7       1424 mtdblock7

I then tried enabling CONFIG_PREEMPT:

# more /proc/partitions
<< /proc/partitions >>
maBUG: scheduling while atomic: more/0x7200546f/54

Stack:
 238efd34 00000000 2397ff1c 00000000 22001758 22001dd4 2217dcb0 238818f0
 7200546f 00000036 221b9bf0 00000000 00000008 238ee670 238ee474 0000001b
 00000000 238eff78 00000000 2397ff1c 00000000 00000000 00000000 238efe24
Call Trace: [<22001758>] [<22001dd4>]
jor minor  #blocks  name

 31     0        768 mtdblock0
 31     1        BUG: scheduling while atomic: more/0x7200546f/54

Stack:
 238efda0 00000000 2397ff1c 00000000 22001758 22001dd4 2217dcb0 238818f0
 7200546f 00000036 00989680 00000000 00000008 00000020 00000005 00000001
 0000000f 238eff78 00000000 2397ff1c 00000000 00000000 00000000 238efe90
Call Trace: [<22001758>] [<22001dd4>]
256 mtdblock1
 31     2        256 mtdblock2
 31     3        128 mtdblock3
 31     4       4096 mtdblock4
 31     5       1024 mtdblock5
 31     6       9216 mtdblock6
 31     7       1424 mtdblock7

Woohoo!  Crashes in more instead of sh...

I don't get these all the time but it's very easy to replicate.
Sometimes my board keeps going, sometimes it freezes.
It's not just more or cat.  Basically any command will do it.
I don't remember getting these with the 2.4 kernel.

John Williams wrote:

Hi Chun Yeow,
Thanks for the report - we'll look into it and let you know via the list. 

Thanks again,

John

Yeoh Chun Yeow wrote:

Dear all,
I compile the 2.6 kernel in the petalogix distribution. My file system is in romfs format.
I have found "BUG: scheduling while atomic: sh/0x00000008/43" in my kernel execution. 

BUG: scheduling while atomic: sh/0x00000008/43

Stack:
27efeac4 27ece5cc 245d1d20 00000001 0000005b 24001ed4 2418a0e4 244ea1c4 00000008 0000002b 0ba0f600 27efea30 00000008 245d15e0 245d15e0 2459e000 27ece5cc 245d1d20 00000001 0000005b 27effc30 00000000 00000000 27efebb4 
Call Trace:
[<24001ed4>] ret_from_intr+0x1c/0x40


After I enable in CONFIG_PREEMPT, thing get better.

But, now when I doing the ping, the following bugs occur:
bytes from 199.20.6.174 <http://199.20.6.174>: icmp_BUG: scheduling while atomic: ping/0x00000002/62 

Stack:
2465fc14 27cb28b0 000000c0 00000000 27d1e120 241a7f9c 241ab0a8 27d1b1d8 00000002 0000003e 2465e000 000000a2 240166c0 7fffffff 00000000 2465fc8c 2465fce0 2465fd18 00000000 00000000 27cb28b0 000000c0 00000000 2411b36c 
Call Trace:

What is the reason behind that? Any ideas?

Please advice. Thanks

regards,
chun yeow



___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/ 



___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/ 



___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/