On Mon, 2013-01-14 at 14:12 -0500, Chris Perl wrote: +AD4- On Mon, Jan 14, 2013 at 03:25:50PM +-0000, Myklebust, Trond wrote: +AD4- +AD4- On Mon, 2013-01-14 at 10:09 -0500, Chris Perl wrote: +AD4- +AD4- +AD4- On Fri, Jan 11, 2013 at 11:19:44AM -0500, Chris Perl wrote: +AD4- +AD4- +AD4- +AD4- On Thu, Jan 10, 2013 at 09:30:58PM +-0000, Myklebust, Trond wrote: +AD4- +AD4- +AD4- +AD4- +AD4- On Wed, 2013-01-09 at 15:52 -0500, Trond Myklebust wrote: +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- On Wed, 2013-01-09 at 12:55 -0500, Chris Perl wrote: +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- Hrm. I guess I'm in over my head here. Apologoies if I'm just asking +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- silly bumbling questions. You can start ignoring me at any time. :) +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- I stared at the code for a while and more and now see why what I +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- outlined is not possible. Thanks for helping to clarify+ACE- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- I decided to pull your git repo and compile with HEAD at +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- 87ed50036b866db2ec2ba16b2a7aec4a2b0b7c39 (linux-next as of this +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- morning). Using this kernel, I can no longer induce any hangs. +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- Interestingly, I tried recompiling the CentOS 6.3 kernel with +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- both the original patch (v4) and the last patch you sent about fixing +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- priority queues. With both of those in place, I still run into a +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- problem. +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- echo 0 +AD4- /proc/sys/sunrpc/rpc+AF8-debug after the hang shows (I left in the +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- previous additional prints and added printing of the tasks pointer +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- itself): +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +ADw-6+AD4-client: ffff88082896c200, xprt: ffff880829011000, snd+AF8-task: ffff880829a1aac0 +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +ADw-6+AD4-client: ffff8808282b5600, xprt: ffff880829011000, snd+AF8-task: ffff880829a1aac0 +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +ADw-6+AD4---task-- -pid- flgs status -client- --rqstp- -timeout ---ops-- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +ADw-6+AD4-ffff88082a463180 22007 0080 -11 ffff8808282b5600 (null) 0 ffffffffa027b7a0 nfsv3 ACCESS a:call+AF8-reserveresult q:xprt+AF8-sending +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +ADw-6+AD4-client: ffff88082838cc00, xprt: ffff88082b7c5800, snd+AF8-task: (null) +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +ADw-6+AD4-client: ffff8808283db400, xprt: ffff88082b7c5800, snd+AF8-task: (null) +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +ADw-6+AD4-client: ffff8808283db200, xprt: ffff880829011000, snd+AF8-task: ffff880829a1aac0 +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- Any thoughts about other patches that might affect this? +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- Hmm... The only one that springs to mind is this one (see attachment) +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- and then the 'connect' fixes that you helped us with previously. +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- Never mind. I suspect that the main reason why RHEL-6.3 is still +AD4- +AD4- +AD4- +AD4- +AD4- vulnerable is that it lacks commit +AD4- +AD4- +AD4- +AD4- +AD4- 961a828df64979d2a9faeeeee043391670a193b9 (SUNRPC: Fix potential races in +AD4- +AD4- +AD4- +AD4- +AD4- xprt+AF8-lock+AF8-write+AF8-next()). +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- Great, thanks+ACE- I've add this on top of the others and am now testing. +AD4- +AD4- +AD4- +AD4- I'll let you know how it goes. +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- With all 4 patches in place, I am no longer able to hang my CentOS 6.3 +AD4- +AD4- +AD4- system. I have not tested all the various combinations of the 4 +AD4- +AD4- +AD4- patches, but can definitely confirm that without either of: +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- 961a828df64979d2a9faeeeee043391670a193b9 SUNRPC: Fix potential races in xprt+AF8-lock+AF8-write+AF8-next() +AD4- +AD4- +AD4- 87ed50036b866db2ec2ba16b2a7aec4a2b0b7c39 SUNRPC: Ensure we release the socket write lock if the rpc+AF8-task exits early +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- I can hang the system using the test program I sent in my first email. +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- I'll follow up with Red Hat and ask that they include all 4 patches for +AD4- +AD4- +AD4- 6.4 (some of the earlier ones they may already have). +AD4- +AD4- +AD4- +AD4- +AD4- +AD4- Thanks so much for all the help+ACE- +AD4- +AD4- +AD4- +AD4- Likewise. Thank you for all the work you did in debugging and testing. +AD4- +AD4- It is much appreciated. +AD4- +AD4- Oh, I had just one other question about the final version of the patch +AD4- you created. In xprt+AF8-release, is it safe to access xprt-+AD4-snd+AF8-task +AD4- without the transport+AF8-lock because of the rcu+AF8-lock()/rcu+AF8-unlock() stuff, +AD4- or because all architectures can atomically read pointer sized values +AD4- that are aligned in certain ways, or something else? It is safe to access the xprt structure because of the rcu+AF8-lock(). As for xprt-+AD4-snd+AF8-task: no other processes are allowed to change the value of xprt-+AD4-snd+AF8-task once it has been assigned to +ACI-our+ACI- rpc+AF8-task, so if we just want to know the answer to the question +ACI-am I holding the socket lock+ACI-, then it is safe to read the value of xprt-+AD4-snd+AF8-task outside the transport lock. Cheers Trond -- Trond Myklebust Linux NFS client maintainer NetApp Trond.Myklebust+AEA-netapp.com www.netapp.com -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
