Re: [Xen-devel] netback Oops then xenwatch stuck in D state

[Wei Liu <wei.liu2@xxxxxxxxxx>]

On Wed, 2013-02-13 at 02:51 +0000, Christopher S. Aker wrote:
> On 2/12/13 4:58 AM, Ian Campbell wrote:
> > Have you applied the XSA-39 fixes to this kernel?
> 
> Yes!  When I rebuilt with Wei's suggested patch for my original 
> netback/xenwatch problem I also brought us up to date with XSA patches.
> 
> We just hit the same (new) problem on another machine, and looking at 
> the BUG with more kernel output context gives a giant clue:
> 
> Feb 12 20:30:54: IPv6: ADDRCONF(NETDEV_UP): vif21.0: link is not ready
> Feb 12 20:30:54: device vif21.0 entered promiscuous mode
> Feb 12 20:30:56: xen-blkback:ring-ref 8, event-channel 31, protocol 2 
> (x86_32-abi)
> Feb 12 20:30:56: xen-blkback:ring-ref 9, event-channel 32, protocol 2 
> (x86_32-abi)
> Feb 12 20:30:56: IPv6: ADDRCONF(NETDEV_CHANGE): vif21.0: link becomes ready
> Feb 12 20:30:56: br0: port 5(vif21.0) entered forwarding state
> Feb 12 20:30:56: br0: port 5(vif21.0) entered forwarding state
> Feb 12 20:30:58: br0: port 5(vif21.0) entered forwarding state
> Feb 12 20:34:12: vif vif-21-0 vif21.0: Frag is bigger than frame.
> Feb 12 20:34:12: vif vif-21-0 vif21.0: fatal error; disabling device 
> <--------------
> Feb 12 20:34:12: BUG: unable to handle kernel NULL pointer dereference 
> at 00000000000008b8
> Feb 12 20:34:12: IP: [<ffffffff81011dda>] xen_spin_lock_flags+0x3a/0x80
> Feb 12 20:34:12: PGD 0
> Feb 12 20:34:12: Oops: 0002 [#1] SMP
> Feb 12 20:34:12: Modules linked in: ebt_comment ebt_arp ebt_set 
> ebt_limit ebt_ip6 ebt_ip ip_set_hash_net ip_set ebtable_nat xen_gntdev 
> bonding ebtable_filter e1000e
> Feb 12 20:34:12: CPU 3
> Feb 12 20:34:12: Pid: 1548, comm: netback/3 Not tainted 3.7.6-1-x86_64 
> #1 Supermicro X8DT6/X8DT6
> Feb 12 20:34:12: RIP: e030:[<ffffffff81011dda>]  [<ffffffff81011dda>] 
> xen_spin_lock_flags+0x3a/0x80
> Feb 12 20:34:12: RSP: e02b:ffff880083681b58  EFLAGS: 00010006
> Feb 12 20:34:12: RAX: 0000000000000400 RBX: 00000000000008b8 RCX: 
> 0000000000000663
> Feb 12 20:34:12: RDX: 0000000000000001 RSI: 0000000000000210 RDI: 
> 00000000000008b8
> Feb 12 20:34:12: RBP: ffff880083681b78 R08: 000000000000000d R09: 
> 0000000000000000
> Feb 12 20:34:12: R10: 0000000000000001 R11: 0000000000000001 R12: 
> 0000000000000001
> Feb 12 20:34:12: R13: 0000000000000200 R14: 0000000000000400 R15: 
> 0000000000000663
> Feb 12 20:34:12: FS:  00007f2bc1fb2700(0000) GS:ffff8801006c0000(0000) 
> knlGS:0000000000000000
> Feb 12 20:34:12: CS:  e033 DS: 0000 ES: 0000 CR0: 000000008005003b
> Feb 12 20:34:12: CR2: 00000000000008b8 CR3: 0000000001c0b000 CR4: 
> 0000000000002660
> Feb 12 20:34:12: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 
> 0000000000000000
> Feb 12 20:34:12: DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 
> 0000000000000400
> Feb 12 20:34:12: Process netback/3 (pid: 1548, threadinfo 
> ffff880083680000, task ffff8800837ec9c0)
> Feb 12 20:34:12: Stack:
> Feb 12 20:34:12: 0000000000000210 00000000000008b8 ffff880003baa700 
> ffff880003baa7d8
> Feb 12 20:34:12: ffff880083681b98 ffffffff817605da 0000000000000000 
> 00000000000008b8
> Feb 12 20:34:12: ffff880083681bd8 ffffffff8154446f ffff880003baa000 
> 0000000000000000

Fancy trying this *UNTESTED* patch a bit? I look through the code and
there seems to be an error.

Jan's advice is worth consideration, but I think we should fix this XSA
bug first.


Wei.

-----8<----
commit 6a06b51edd7124193322fd62244171eb099aff52
Author: Wei Liu <wei.liu2@xxxxxxxxxx>
Date:   Wed Feb 13 18:17:01 2013 +0000

    netback: fix netbk_count_requests
    
    Signed-off-by: Wei Liu <wei.liu2@xxxxxxxxxx>

diff --git a/drivers/net/xen-netback/netback.c 
b/drivers/net/xen-netback/netback.c
index 103294d..61aaeb0 100644
--- a/drivers/net/xen-netback/netback.c
+++ b/drivers/net/xen-netback/netback.c
@@ -927,7 +927,11 @@ static int netbk_count_requests(struct xenvif *vif,
                if (txp->size > first->size) {
                        netdev_err(vif->dev, "Frag is bigger than frame.\n");
                        netbk_fatal_tx_err(vif);
-                       return -frags;
+                       /* frag could be zero at this point if it
+                        * fails during first iteration, so we need to
+                        * set it to negative number if there is
+                        * error */
+                       return frags ? -frags : -1;
                }
 
                first->size -= txp->size;



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel