[Bug 939240] Re: [public] armadaxp kernel slow to umount
Build Version/Date: Linux armadaxp1 3.2.0-1600-armadaxp #3-Ubuntu SMP Fri Mar 16 19:14:51 UTC 2012 armv7l armv7l armv7l GNU/Linux Release: 12.04 Environment used for testing: armadaxp1 # mount | tail -1 /dev/sda6 on /mnt type ext2 (rw) # cp /boot/* /mnt/ ; ls -la /mnt/ ; time umount /mnt ... real0m1.509s user0m0.000s sys 0m0.110s ** Changed in: eilt Status: New => Fix Committed ** Changed in: eilt Status: Fix Committed => Fix Released ** Changed in: linux-armadaxp (Ubuntu) Status: Fix Committed => Fix Released -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/939240 Title: [public] armadaxp kernel slow to umount To manage notifications about this bug go to: https://bugs.launchpad.net/eilt/+bug/939240/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 939240] Re: [public] armadaxp kernel slow to umount
This appears to be fixed with the 3.2 kernel. Marking as fix-committed. Will change once the 3.2 kernel hits the pool. ** Changed in: linux-armadaxp (Ubuntu) Status: New => Fix Committed -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/939240 Title: [public] armadaxp kernel slow to umount To manage notifications about this bug go to: https://bugs.launchpad.net/eilt/+bug/939240/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 939240] Re: [public] armadaxp kernel slow to umount
** Changed in: eilt Importance: Undecided => Low -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/939240 Title: [public] armadaxp kernel slow to umount To manage notifications about this bug go to: https://bugs.launchpad.net/eilt/+bug/939240/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 939240] Re: [public] armadaxp kernel slow to umount
Tested Precise userspace with 3.0.0.-12-generic-pae i686 kernel as sudo against an ext2 (/boot) umount and it was faster than the same umount using the 3.2 kernel (0.079s vs. 0.089s), all in a VM. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/939240 Title: [public] armadaxp kernel slow to umount To manage notifications about this bug go to: https://bugs.launchpad.net/eilt/+bug/939240/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 939240] Re: [public] armadaxp kernel slow to umount
To successfully reproduce this, mahmoh and I discovered that you seem to have to run it from a root shell - running it under sudo produces reasonable (i.e. immediate) umount times. I therefore suspected some obscure capability thing, but I don't see any such differences in the kernel code. Further, I've found that this only seems to be a problem when there's one runnable process on the system. If you start another program, the umount will immediately exit. Profiling turned up no cpu obvious cpu intensive kernel functions. I took a few samples of a hung umount using sysrq+t, and found that we seem to always be stuck w/ the following backtrace: [ 5093.862717] umount D c0557538 0 11174 2118 0x [ 5093.862729] Backtrace: [ 5093.862741] [] (__schedule+0x0/0x5bc) from [] (schedule+0x50/0x68) [ 5093.862756] [] (schedule+0x0/0x68) from [] (schedule_timeout+0x1d4/0x254) [ 5093.862770] [] (schedule_timeout+0x0/0x254) from [] (wait_for_common+0xc4/0x17c) [ 5093.862785] [] (wait_for_common+0x0/0x17c) from [] (wait_for_completion+0x18/0x1c) [ 5093.862802] [] (wait_for_completion+0x0/0x1c) from [] (_rcu_barrier.isra.31+0x98/0xb8) [ 5093.862817] [] (_rcu_barrier.isra.31+0x0/0xb8) from [] (rcu_barrier_sched+0x18/0x1c) [ 5093.862827] r6:c06eb748 r5:c06fb8c4 r4:ed68c000 r3:c01397b4 [ 5093.862847] [] (rcu_barrier_sched+0x0/0x1c) from [] (rcu_barrier+0x10/0x14) [ 5093.862864] [] (rcu_barrier+0x0/0x14) from [] (deactivate_locked_super+0x54/0x68) [ 5093.862879] [] (deactivate_locked_super+0x0/0x68) from [] (deactivate_super+0x68/0x70) [ 5093.862888] r5:ed68c000 r4:ed68c000 [ 5093.862904] [] (deactivate_super+0x0/0x70) from [] (mntput_no_expire+0xc8/0x11c) [ 5093.862913] r4:ed6c6780 r3:db28e34c [ 5093.862927] [] (mntput_no_expire+0x0/0x11c) from [] (sys_umount+0x68/0xc8) [ 5093.862936] r7:0034 r6:00efe8b0 r5: r4: [ 5093.862955] [] (sys_umount+0x0/0xc8) from [] (ret_fast_syscall+0x0/0x30) [ 5093.862964] r5:0001b320 r4:00efe8b0 Reading up on RCU, it looks like we are waiting for all rcu callbacks to complete on all CPUs before moving forward - but I'm not sure how to identify why that has not occurred. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/939240 Title: [public] armadaxp kernel slow to umount To manage notifications about this bug go to: https://bugs.launchpad.net/eilt/+bug/939240/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 939240] Re: [public] armadaxp kernel slow to umount
** Summary changed: - armadaxp kernel slow to unmount + [public] armadaxp kernel slow to umount -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/939240 Title: [public] armadaxp kernel slow to umount To manage notifications about this bug go to: https://bugs.launchpad.net/eilt/+bug/939240/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs