Skip to content

L
linux
Switch branch/tag
  1. 28 Oct, 2010 14 commits
    • Krishna Kumar's avatar
      cxgb3: Fix panic in free_tx_desc() · b1424ed9
      Krishna Kumar authored 
      I got a few of these panics (on 2.6.36-rc7) when running high
number of netperf sessions:

BUG: unable to handle kernel paging request at 0000100000000000
IP: [<ffffffff813125f0>] skb_release_data+0xa0/0xd0
Oops: 0000 [#1] SMP
Pid: 2155, comm: vhost-2115 Not tainted 2.6.36-rc7-ORG #1 49Y6512     /System x3650 M2 -[7947AC1]-
RIP: 0010:[<ffffffff813125f0>]  [<ffffffff813125f0>] skb_release_data+0xa0/0xd0
RSP: 0018:ffff880001803738  EFLAGS: 00010206
RAX: ffff880179b0fc00 RBX: ffff880178b441c0 RCX: 0000000000000000
RSP: 0018:ffff880001803738  EFLAGS: 00010206
RAX: ffff880179b0fc00 RBX: ffff880178b441c0 RCX: 0000000000000000
RDX: ffff880179b0fd40 RSI: 0000000000000000 RDI: 0000100000000000
RBP: ffff880001803748 R08: 0000000000000001 R09: ffff88017f117000
R10: ffff88017b990608 R11: ffff88017f117090 R12: ffff880178b441c0
R13: ffff88017f117090 R14: 0000000000000000 R15: ffff880178b441c0
FS:  0000000000000000(0000) GS:ffff880001800000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000100000000000 CR3: 000000017ea64000 CR4: 00000000000026e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process vhost-2115 (pid: 2155, threadinfo ffff88017d872000, task ffff88017e954680)
Stack:
ffff880178b441c0 0000000000000007 ffff880001803768 ffffffff81312119
<0> 0000000000000000 0000000000000002 ffff880001803778 ffffffff813121f9
<0> ffff880001803818 ffffffffa012d14c ffffffffa02de076 ffff880001803700
Call Trace:
<IRQ>
[<ffffffff81312119>] __kfree_skb+0x19/0xa0
[<ffffffff813121f9>] kfree_skb+0x19/0x40
[<ffffffffa012d14c>] free_tx_desc+0x2fc/0x350 [cxgb3]
[<ffffffffa02de076>] ? vhost_poll_wakeup+0x16/0x20 [vhost_net]
[<ffffffffa01323db>] t3_eth_xmit+0x28b/0x380 [cxgb3]
[<ffffffff8131ce47>] dev_hard_start_xmit+0x377/0x5a0
[<ffffffff81335a4a>] sch_direct_xmit+0xfa/0x1d0
[<ffffffff8131d1a9>] dev_queue_xmit+0x139/0x450
[<ffffffff81326225>] neigh_resolve_output+0x125/0x340
[<ffffffff8135a77c>] ip_finish_output+0x14c/0x320
[<ffffffff8135a9fe>] ip_output+0xae/0xc0
[<ffffffff8135620f>] ip_forward_finish+0x3f/0x50
[<ffffffff8135641f>] ip_forward+0x1ff/0x400
[<ffffffff81354789>] ip_rcv_finish+0x119/0x3e0
[<ffffffff81354c7d>] ip_rcv+0x22d/0x300
[<ffffffff8131a95b>] __netif_receive_skb+0x29b/0x570
[<ffffffff8131ba70>] ? netif_receive_skb+0x0/0x80
[<ffffffff8131bae8>] netif_receive_skb+0x78/0x80
[<ffffffffa02a96d8>] br_handle_frame_finish+0x198/0x260 [bridge]
[<ffffffffa02aebc8>] br_nf_pre_routing_finish+0x238/0x380 [bridge]
[<ffffffff813424bc>] ? nf_hook_slow+0x6c/0x100
[<ffffffffa02ae990>] ? br_nf_pre_routing_finish+0x0/0x380 [bridge]
[<ffffffffa02afb08>] br_nf_pre_routing+0x698/0x7a0 [bridge]
[<ffffffff81342414>] nf_iterate+0x64/0xa0
[<ffffffffa02a9540>] ? br_handle_frame_finish+0x0/0x260 [bridge]
[<ffffffff813424bc>] nf_hook_slow+0x6c/0x100
[<ffffffffa02a9540>] ? br_handle_frame_finish+0x0/0x260 [bridge]
[<ffffffffa02a9931>] br_handle_frame+0x191/0x240 [bridge]
[<ffffffffa02a97a0>] ? br_handle_frame+0x0/0x240 [bridge]
[<ffffffff8131a863>] __netif_receive_skb+0x1a3/0x570
[<ffffffff812ef3f6>] ? dma_issue_pending_all+0x76/0xa0
[<ffffffff8131ad32>] process_backlog+0x102/0x200
[<ffffffff8131c2d0>] net_rx_action+0x100/0x220
[<ffffffff810548ef>] __do_softirq+0xaf/0x140
[<ffffffff8100bcdc>] call_softirq+0x1c/0x30
[<ffffffff8100dfc5>] ? do_softirq+0x65/0xa0
[<ffffffff8131c6b8>] netif_rx_ni+0x28/0x30
[<ffffffffa02c305d>] tun_sendmsg+0x2cd/0x4b0 [tun]
[<ffffffffa02e01af>] handle_tx+0x1df/0x340 [vhost_net]
[<ffffffffa02e0340>] handle_tx_kick+0x10/0x20 [vhost_net]
[<ffffffffa02de29b>] vhost_worker+0xbb/0x130 [vhost_net]
[<ffffffffa02de1e0>] ? vhost_worker+0x0/0x130 [vhost_net]
[<ffffffffa02de1e0>] ? vhost_worker+0x0/0x130 [vhost_net]
[<ffffffff81069686>] kthread+0x96/0xa0
[<ffffffff8100bbe4>] kernel_thread_helper+0x4/0x10
[<ffffffff810695f0>] ? kthread+0x0/0xa0
[<ffffffff8100bbe0>] ? kernel_thread_helper+0x0/0x10
Code: 8b 94 24 d0 00 00 00 49 8b 84 24 d8 00 00 00 48 8d 14 10 0f b7 0a 39 d9 7f d1 48 8b 7a 10 48 85 ff 74 20 48 c7 42 10 00 00 00 00 <48> 8b 1f e8 e8 fb ff ff 48 85 db 48 89 df 75 f0 49 8b 84 24 d8

Patch below fixes the panic. cxgb4 and cxgb4vf already have this fix.
Signed-off-by: default avatarKrishna Kumar <krkumar2@in.ibm.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      b1424ed9
    • Nishanth Aravamudan's avatar
      cxgb3: fix crash due to manipulating queues before registration · 69dcfc8a
      Nishanth Aravamudan authored 
      Along the same lines as "cxgb4: fix crash due to manipulating queues
before registration" (8f6d9f40), before
commit "net: allocate tx queues in register_netdevice"
netif_tx_stop_all_queues and related functions could be used between
device allocation and registration but now only after registration.
cxgb4 has such a call before registration and crashes now.  Move it
after register_netdev.
Signed-off-by: default avatarNishanth Aravamudan <nacc@us.ibm.com>
Cc: eric.dumazet@gmail.com
Cc: sonnyrao@us.ibm.com
Cc: Divy Le Ray <divy@chelsio.com>
Cc: Dimitris Michailidis <dm@chelsio.com>
Cc: netdev@vger.kernel.org
Cc: linux-kernel@vger.kernel.org
Tested-by: default avatarNishanth Aravamudan <nacc@us.ibm.com>
Acked-by: default avatarDivy Le Ray <divy@chelsio.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      69dcfc8a
    • Pavel Emelyanov's avatar
      8390: Don't oops on starting dev queue · b7126d8c
      Pavel Emelyanov authored 
      The __NS8390_init tries to start the device queue before the
device is registered. This results in an oops (snipped):

[    2.865493] BUG: unable to handle kernel NULL pointer dereference at 0000000000000010
[    2.866106] IP: [<ffffffffa000602a>] netif_start_queue+0xb/0x12 [8390]
[    2.881267] Call Trace:
[    2.881437]  [<ffffffffa000624d>] __NS8390_init+0x102/0x15a [8390]
[    2.881999]  [<ffffffffa00062ae>] NS8390_init+0x9/0xb [8390]
[    2.882237]  [<ffffffffa000d820>] ne2k_pci_init_one+0x297/0x354 [ne2k_pci]
[    2.882955]  [<ffffffff811c7a0e>] local_pci_probe+0x12/0x16
[    2.883308]  [<ffffffff811c85ad>] pci_device_probe+0xc3/0xef
[    2.884049]  [<ffffffff8129218d>] driver_probe_device+0xbe/0x14b
[    2.884937]  [<ffffffff81292260>] __driver_attach+0x46/0x62
[    2.885170]  [<ffffffff81291788>] bus_for_each_dev+0x49/0x78
[    2.885781]  [<ffffffff81291fbb>] driver_attach+0x1c/0x1e
[    2.886089]  [<ffffffff812912ab>] bus_add_driver+0xba/0x227
[    2.886330]  [<ffffffff8129259a>] driver_register+0x9e/0x115
[    2.886933]  [<ffffffff811c8815>] __pci_register_driver+0x50/0xac
[    2.887785]  [<ffffffffa001102c>] ne2k_pci_init+0x2c/0x2e [ne2k_pci]
[    2.888093]  [<ffffffff81000212>] do_one_initcall+0x7c/0x130
[    2.888693]  [<ffffffff8106d74f>] sys_init_module+0x99/0x1da
[    2.888946]  [<ffffffff81002a2b>] system_call_fastpath+0x16/0x1b

This happens because the netif_start_queue sets respective bit on the dev->_tx
array which is not yet allocated.

As far as I understand the code removing the netif_start_queue from __NS8390_init
is OK, since queue will be started later on device open. Plz, correct me if I'm wrong.

Found in the Dave's current tree, so he's in Cc.
Signed-off-by: default avatarPavel Emelyanov <xemul@openvz.org>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      b7126d8c
    • Gerrit Renker's avatar
      dccp ccid-2: Stop polling · 1c0e0a05
      Gerrit Renker authored 
      This updates CCID-2 to use the CCID dequeuing mechanism, converting from
previous continuous-polling to a now event-driven mechanism.
Signed-off-by: default avatarGerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      1c0e0a05
    • Gerrit Renker's avatar
      dccp: Refine the wait-for-ccid mechanism · b1fcf55e
      Gerrit Renker authored 
      This extends the existing wait-for-ccid routine so that it may be used with
different types of CCID, addressing the following problems:

 1) The queue-drain mechanism only works with rate-based CCIDs. If CCID-2 for
    example has a full TX queue and becomes network-limited just as the
    application wants to close, then waiting for CCID-2 to become unblocked
    could lead to an indefinite  delay (i.e., application "hangs").
 2) Since each TX CCID in turn uses a feedback mechanism, there may be changes
    in its sending policy while the queue is being drained. This can lead to
    further delays during which the application will not be able to terminate.
 3) The minimum wait time for CCID-3/4 can be expected to be the queue length
    times the current inter-packet delay. For example if tx_qlen=100 and a delay
    of 15 ms is used for each packet, then the application would have to wait
    for a minimum of 1.5 seconds before being allowed to exit.
 4) There is no way for the user/application to control this behaviour. It would
    be good to use the timeout argument of dccp_close() as an upper bound. Then
    the maximum time that an application is willing to wait for its CCIDs to can
    be set via the SO_LINGER option.

These problems are addressed by giving the CCID a grace period of up to the
`timeout' value.

The wait-for-ccid function is, as before, used when the application
 (a) has read all the data in its receive buffer and
 (b) if SO_LINGER was set with a non-zero linger time, or
 (c) the socket is either in the OPEN (active close) or in the PASSIVE_CLOSEREQ
     state (client application closes after receiving CloseReq).

In addition, there is a catch-all case of __skb_queue_purge() after waiting for
the CCID. This is necessary since the write queue may still have data when
 (a) the host has been passively-closed,
 (b) abnormal termination (unread data, zero linger time),
 (c) wait-for-ccid could not finish within the given time limit.
Signed-off-by: default avatarGerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      b1fcf55e
    • Gerrit Renker's avatar
      dccp: Extend CCID packet dequeueing interface · dc841e30
      Gerrit Renker authored 
      This extends the packet dequeuing interface of dccp_write_xmit() to allow
 1. CCIDs to take care of timing when the next packet may be sent;
 2. delayed sending (as before, with an inter-packet gap up to 65.535 seconds).

The main purpose is to take CCID-2 out of its polling mode (when it is network-
limited, it tries every millisecond to send, without interruption).

The mode of operation for (2) is as follows:
 * new packet is enqueued via dccp_sendmsg() => dccp_write_xmit(),
 * ccid_hc_tx_send_packet() detects that it may not send (e.g. window full),
 * it signals this condition via `CCID_PACKET_WILL_DEQUEUE_LATER',
 * dccp_write_xmit() returns without further action;
 * after some time the wait-condition for CCID becomes true,
 * that CCID schedules the tasklet,
 * tasklet function calls ccid_hc_tx_send_packet() via dccp_write_xmit(),
 * since the wait-condition is now true, ccid_hc_tx_packet() returns "send now",
 * packet is sent, and possibly more (since dccp_write_xmit() loops).

Code reuse: the taskled function calls dccp_write_xmit(), the timer function
            reduces to a wrapper around the same code.
Signed-off-by: default avatarGerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      dc841e30
    • Gerrit Renker's avatar
      dccp: Return-value convention of hc_tx_send_packet() · fe84f414
      Gerrit Renker authored 
      This patch reorganises the return value convention of the CCID TX sending
function, to permit more flexible schemes, as required by subsequent patches.

Currently the convention is
 * values < 0     mean error,
 * a value == 0   means "send now", and
 * a value x > 0  means "send in x milliseconds".

The patch provides symbolic constants and a function to interpret return values.

In addition, it caps the maximum positive return value to 0xFFFF milliseconds,
corresponding to 65.535 seconds.  This is possible since in CCID-3/4 the
maximum possible inter-packet gap is fixed at t_mbi = 64 sec.
Signed-off-by: default avatarGerrit Renker <gerrit@erg.abdn.ac.uk>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      fe84f414
    • Emil Tantilov's avatar
      igbvf: fix panic on load · de7fe787
      Emil Tantilov authored 
      Introduced by commit:e6484930
net: allocate tx queues in register_netdevice
Signed-off-by: default avatarEmil Tantilov <emil.s.tantilov@intel.com>
Acked-by: default avatarGreg Rose <greg.v.rose@intel.com>
Tested-by: default avatarJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: default avatarJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      de7fe787
    • Emil Tantilov's avatar
      ixgb: call pci_disable_device in ixgb_remove · ec43a81c
      Emil Tantilov authored 
      ixgb fails to work after reload on recent kernels:

rmmod ixgb (dev->current_state = PCI_UNKNOWN)
modprobe ixgb (pci_enable_device will bail leaving current_state to PCI_UNKNOWN)
ifup eth0
do_IRQ: 2.82 No irq handler for vector (irq -1)

The issue was exposed by commit fcd097f3
PCI: MSI: Remove unsafe and unnecessary hardware access

which avoids HW writes for power states != PCI_D0

CC: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: default avatarEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: default avatarJeff Pieper <jeffrey.e.pieper@intel.com>
Signed-off-by: default avatarJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      ec43a81c
    • John Fastabend's avatar
      ixgbe: DCB, fix TX hang occurring in stress condition with PFC · 9806307a
      John Fastabend authored 
      The DCB credits refill quantum _must_ be greater than half the max
packet size. This is needed to guarantee that TX DMA operations
are not attempted during a pause state. Additionally, the min IFG
must be set correctly for DCB mode. If a DMA operation is
requested unexpectedly during the pause state the HW data
store may be corrupted leading to a DMA hang.  The DMA hang
requires a reset to correct. This fixes the HW configuration
to avoid this condition.
Signed-off-by: default avatarJohn Fastabend <john.r.fastabend@intel.com>
Tested-by: default avatarRoss Brattain <ross.b.brattain@intel.com>
Signed-off-by: default avatarJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      9806307a
    • Carolyn Wyborny's avatar
      e1000e: Add check for reset flags before displaying reset message · affa9dfb
      Carolyn Wyborny authored 
      Some parts need to execute resets during normal operation.  This flag
check ensures that those parts reset without needlessly alarming the
user.  Other unexpected resets by other parts will dump debug info
and message the reset action to the user, as originally intended.
Signed-off-by: default avatarCarolyn Wyborny <carolyn.wyborny@intel.com>
Acked-by: default avatarBruce Allan <bruce.w.allan@intel.com>
Tested-by: default avatarEmil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: default avatarJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      affa9dfb
    • Carolyn Wyborny's avatar
      e1000e: reset PHY after errors detected · ff10e13c
      Carolyn Wyborny authored 
      Some errors can be induced in the PHY via environmental testing
(specifically extreme temperature changes and electro static
discharge testing), and in the case of the PHY hanging due to
this input, this detects the problem and resets to continue.
This issue only applies to 82574 silicon.
Signed-off-by: default avatarJesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: default avatarCarolyn Wyborny <carolyn.wyborny@intel.com>
Tested-by: default avatarEmil Tantilov <emil.s.tantilov@intel.com>
Signed-off-by: default avatarJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      ff10e13c
    • David S. Miller's avatar
      pch_gbe: Select MII. · 116c1ea0
      David S. Miller authored 
      Reported-by: default avatarSteven Rostedt <rostedt@goodmis.org>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      116c1ea0
    • Jesse Gross's avatar
      igb: Fix unused variable warning. · c1758012
      Jesse Gross authored 
      Commit eab6d18d "vlan: Don't check for vlan group before
vlan_tx_tag_present" removed the need for the adapter variable
in igb_xmit_frame_ring_adv().  This removes the variable as well
to avoid the compiler warning.
Reported-by: default avatarStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: default avatarJesse Gross <jesse@nicira.com>
Signed-off-by: default avatarJeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      c1758012
  3. 27 Oct, 2010 25 commits
  5. 26 Oct, 2010 1 commit
    • Glenn Wurster's avatar
      IPv6: Temp addresses are immediately deleted. · 7a876b0e
      Glenn Wurster authored 
      There is a bug in the interaction between ipv6_create_tempaddr and
addrconf_verify.  Because ipv6_create_tempaddr uses the cstamp and tstamp
from the public address in creating a private address, if we have not
received a router advertisement in a while, tstamp + temp_valid_lft might be
< now.  If this happens, the new address is created inside
ipv6_create_tempaddr, then the loop within addrconf_verify starts again and
the address is immediately deleted.  We are left with no temporary addresses
on the interface, and no more will be created until the public IP address is
updated.  To avoid this, set the expiry time to be the minimum of the time
left on the public address or the config option PLUS the current age of the
public interface.
Signed-off-by: default avatarGlenn Wurster <gwurster@scs.carleton.ca>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      7a876b0e
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7