[Bug 569610] Re: "Kernel unaligned access at TPC" causing network/system to become slow and/or unresponsive

[Luke J Militello]

I did some searching and found these two entries in some kernel change
logs.  I have no idea if it helps any but I figured it couldn't hurt to
ask...

 ~ ChangeLog-2.6.28 ~

 11696	commit 33cf71cee14743185305c61625c4544885055733
 11697	Author: Petr Tesarik <ptesarik at suse.cz>
 11698	Date:   Fri Nov 21 16:42:58 2008 -0800
 11699	
 11700	    tcp: Do not use TSO/GSO when there is urgent data
 11701	    
 11702	    This patch fixes http://bugzilla.kernel.org/show_bug.cgi?id=12014
 11703	    
 11704	    Since most (if not all) implementations of TSO and even the in-kernel
 11705	    software GSO do not update the urgent pointer when splitting a large
 11706	    segment, it is necessary to turn off TSO/GSO for all outgoing traffic
 11707	    with the URG pointer set.
 11708	    
 11709	    Looking at tcp_current_mss (and the preceding comment) I even think
 11710	    this was the original intention. However, this approach is insufficient,
 11711	    because TSO/GSO is turned off only for newly created frames, not for
 11712	    frames which were already pending at the arrival of a message with
 11713	    MSG_OOB set. These frames were created when TSO/GSO was enabled,
 11714	    so they may be large, and they will have the urgent pointer set
 11715	    in tcp_transmit_skb().
 11716	    
 11717	    With this patch, such large packets will be fragmented again before
 11718	    going to the transmit routine.
 11719	    
 11720	    As a side note, at least the following NICs are known to screw up
 11721	    the urgent pointer in the TCP header when doing TSO:
 11722	    
 11723	    	Intel 82566MM (PCI ID 8086:1049)
 11724	    	Intel 82566DC (PCI ID 8086:104b)
 11725	    	Intel 82541GI (PCI ID 8086:1076)
 11726	    	Broadcom NetXtreme II BCM5708 (PCI ID 14e4:164c)
 11727	    
 11728	    Signed-off-by: Petr Tesarik <ptesarik at suse.cz>
 11729	    Signed-off-by: David S. Miller <davem at davemloft.net>

 ~ ChangeLog-2.6.29 ~

168533	commit ef711cf1d156428d4c2911b8c86c6ce90519dc45
168534	Author: Eric Dumazet <dada1 at cosmosbay.com>
168535	Date:   Fri Nov 14 00:53:54 2008 -0800
168536	
168537	    net: speedup dst_release()
168538	    
168539	    During tbench/oprofile sessions, I found that dst_release() was in third position.
168540	    
168541	    CPU: Core 2, speed 2999.68 MHz (estimated)
168542	    Counted CPU_CLK_UNHALTED events (Clock cycles when not halted) with a unit mask of 0x00 (Unhalted core cycles) count 100000
168543	    samples  %        symbol name
168544	    483726    9.0185  __copy_user_zeroing_intel
168545	    191466    3.5697  __copy_user_intel
168546	    185475    3.4580  dst_release
168547	    175114    3.2648  ip_queue_xmit
168548	    153447    2.8608  tcp_sendmsg
168549	    108775    2.0280  tcp_recvmsg
168550	    102659    1.9140  sysenter_past_esp
168551	    101450    1.8914  tcp_current_mss
168552	    95067     1.7724  __copy_from_user_ll
168553	    86531     1.6133  tcp_transmit_skb
168554	    
168555	    Of course, all CPUS fight on the dst_entry associated with 127.0.0.1
168556	    
168557	    Instead of first checking the refcount value, then decrement it,
168558	    we use atomic_dec_return() to help CPU to make the right memory transaction
168559	    (ie getting the cache line in exclusive mode)
168560	    
168561	    dst_release() is now at the fifth position, and tbench a litle bit faster ;)
168562	    
168563	    CPU: Core 2, speed 3000.1 MHz (estimated)
168564	    Counted CPU_CLK_UNHALTED events (Clock cycles when not halted) with a unit mask of 0x00 (Unhalted core cycles) count 100000
168565	    samples  %        symbol name
168566	    647107    8.8072  __copy_user_zeroing_intel
168567	    258840    3.5229  ip_queue_xmit
168568	    258302    3.5155  __copy_user_intel
168569	    209629    2.8531  tcp_sendmsg
168570	    165632    2.2543  dst_release
168571	    149232    2.0311  tcp_current_mss
168572	    147821    2.0119  tcp_recvmsg
168573	    137893    1.8767  sysenter_past_esp
168574	    127473    1.7349  __copy_from_user_ll
168575	    121308    1.6510  ip_finish_output
168576	    118510    1.6129  tcp_transmit_skb
168577	    109295    1.4875  tcp_v4_rcv
168578	    
168579	    Signed-off-by: Eric Dumazet <dada1 at cosmosbay.com>
168580	    Signed-off-by: David S. Miller <davem at davemloft.net>

** Bug watch added: Linux Kernel Bug Tracker #12014
   http://bugzilla.kernel.org/show_bug.cgi?id=12014

-- 
"Kernel unaligned access at TPC" causing network/system to become slow and/or unresponsive
https://bugs.launchpad.net/bugs/569610
You received this bug notification because you are a member of Kernel
Bugs, which is subscribed to linux in ubuntu.