Re: Fw: epoll oops on 2.6.x

[Louay Gammo <lgammo@cs.uwaterloo.ca>]

[-- Attachment #1: Type: text/plain, Size: 4287 bytes --]

Davide Libenzi wrote:

>> including some of these oopses as attachements. I am not sure what 
>>the 'normal' bug reporting mechanism is
>>for IA-64 kernels.
>>    
>>
>
>I'm getting this from Randy, since I was not subscribed to linux-ia64. Now 
>I am. I took a quick look at it and so far the clue is completely missing. 
>Only one of the oops actually happen inside epoll, even if the other ones 
>comes typically from __wake_up_common() (others from vfs_read(), that is 
>quite unrelated), that let me think that it might be epoll related (epoll 
>has a wait queue item dropped in). Can you try a few experiments?
>
>1) Disable pre-emption (if enabled)
>2) Load the server thru loopback
>3) #define DEBUG_EPOLL 10
>  
>

I did #1 and #2 and ran my original experiment. I did not get a panic but
I got 0 replies from the server. So I undid the #2 and I got a kernel 
panic (included below).
So I am going to try to overload the server on loopback interface (I think
this is what you are saying) and I will let you know of the results.

I am enclosing the kernel config just in case that might clue anyone to 
a configuration problem.

Thanks,

Louay Gammo

tissimo:~# Unable to handle kernel paging request at virtual address 
000000000001003e
userver[1284]: Oops 8804682956800 [1]
                                     
Pid: 1284, CPU 0, comm:              userver
psr : 0000101008022018 ifs : 8000000000000003 ip  : 
[<a000000100080620>]    Not tainted
ip is at prepare_to_wait_exclusive+0x80/0xe0
unat: 0000000000000000 pfs : 0000000000000288 rsc : 0000000000000003
rnat: 0000000000000000 bsps: 0000000000000000 pr  : 0000000000158659
ldrs: 0000000000000000 ccv : 0000000000000000 fpsr: 0009804c0270033f
csd : 0000000000000000 ssd : 0000000000000000
b0  : a000000100530360 b6  : a000000100002d70 b7  : a000000100581980
f6  : 000000000000000000000 f7  : 000000000000000000000
f8  : 000000000000000000000 f9  : 000000000000000000000
f10 : 000000000000000000000 f11 : 000000000000000000000
r1  : a000000100990000 r2  : e0000040f5f5fd58 r3  : 0000000000000001
r8  : 0000000000000000 r9  : 000000000001003e r10 : e0000040f5f5fd88
r11 : a000000100722900 r12 : e0000040f5f5fd30 r13 : e0000040f5f58000
r14 : e000004043d26160 r15 : e0000040f5f5fd58 r16 : e0000040f5f58000
r17 : e000004043880d48 r18 : 0000001008026018 r19 : 0000000000000000
r20 : 000000003fffff00 r21 : e0000040f5f5fd58 r22 : ffffffffffff0020
r23 : a0000001008e9ab0 r24 : e0000040f5f5fd78 r25 : a0000001008ea198
r26 : e0000040f5f5fd70 r27 : e000004042206088 r28 : e000004042206090
r29 : e0000040422060a8 r30 : e0000040422060b0 r31 : e0000040422060b8
                                                                    
Call Trace:
 [<a000000100017ae0>] show_stack+0x80/0xa0
                                sp=e0000040f5f5f900 bsp=e0000040f5f59378
 [<a00000010003a330>] die+0x130/0x1a0
                                sp=e0000040f5f5fad0 bsp=e0000040f5f59340
 [<a000000100054cd0>] ia64_do_page_fault+0x350/0x920
                                sp=e0000040f5f5fad0 bsp=e0000040f5f592d8
 [<a0000001000118a0>] ia64_leave_kernel+0x0/0x260
                                sp=e0000040f5f5fb60 bsp=e0000040f5f592d8
 [<a000000100080620>] prepare_to_wait_exclusive+0x80/0xe0
                                sp=e0000040f5f5fd30 bsp=e0000040f5f592c0
 [<a000000100530360>] __lock_sock+0x100/0x1a0
                                sp=e0000040f5f5fd30 bsp=e0000040f5f59298
 [<a000000100530f70>] lock_sock+0x50/0xa0
                                sp=e0000040f5f5fd90 bsp=e0000040f5f59278
 [<a0000001005c7610>] inet_accept+0xb0/0x1e0
                                sp=e0000040f5f5fd90 bsp=e0000040f5f59248
 [<a00000010052bdf0>] sys_accept+0x170/0x300
                                sp=e0000040f5f5fda0 bsp=e0000040f5f591a0
 [<a000000100011720>] ia64_ret_from_syscall+0x0/0x20
                                sp=e0000040f5f5fe30 bsp=e0000040f5f591a0
 <0>Kernel panic: Aiee, killing interrupt handler!
In interrupt handler - not syncing

>(Your machine does not look too much in shape looking at INIT messages 
>though)
>
>
>
>- Davide
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-ia64" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>  
>


[-- Attachment #2: config.gz --]
[-- Type: application/x-gzip, Size: 4836 bytes --]