cpu hotplug-unplug in parallel with suspend resume.. guest becomes unresponsive.. #16
Comments
|
@sam has fix for this issue, patch @ http://www.spinics.net/lists/kvm-ppc/msg12981.html .. |
|
------- Comment From bssrikanth@in.ibm.com 2017-12-04 04:54:15 EDT------- |
malcolmcrossley
pushed a commit
to NXPower/linux
that referenced
this issue
Mar 2, 2018
commit efdab99 upstream. syzkaller reported: WARNING: CPU: 0 PID: 12927 at arch/x86/kernel/traps.c:780 do_debug+0x222/0x250 CPU: 0 PID: 12927 Comm: syz-executor Tainted: G OE 4.15.0-rc2+ open-power-host-os#16 RIP: 0010:do_debug+0x222/0x250 Call Trace: <#DB> debug+0x3e/0x70 RIP: 0010:copy_user_enhanced_fast_string+0x10/0x20 </#DB> _copy_from_user+0x5b/0x90 SyS_timer_create+0x33/0x80 entry_SYSCALL_64_fastpath+0x23/0x9a The testcase sets a watchpoint (with perf_event_open) on a buffer that is passed to timer_create() as the struct sigevent argument. In timer_create(), copy_from_user()'s rep movsb triggers the BP. The testcase also sets the debug registers for the guest. However, KVM only restores host debug registers when the host has active watchpoints, which triggers a race condition when running the testcase with multiple threads. The guest's DR6.BS bit can escape to the host before another thread invokes timer_create(), and do_debug() complains. The fix is to respect do_debug()'s dr6 invariant when leaving KVM. Reported-by: Dmitry Vyukov <dvyukov@google.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Radim Krčmář <rkrcmar@redhat.com> Cc: David Hildenbrand <david@redhat.com> Cc: Dmitry Vyukov <dvyukov@google.com> Reviewed-by: David Hildenbrand <david@redhat.com> Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Radim Krčmář <rkrcmar@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
malcolmcrossley
pushed a commit
to NXPower/linux
that referenced
this issue
Mar 2, 2018
[ Upstream commit 5811767 ] According to LS1021A RM, the value of PAL can be set so that the start of the IP header in the receive data buffer is aligned to a 32-bit boundary. Normally, setting PAL = 2 provides minimal padding to ensure such alignment of the IP header. However every incoming packet's 8-byte time stamp will be inserted into the packet data buffer as padding alignment bytes when hardware time stamping is enabled. So we set the padding 8+2 here to avoid the flooded alignment faults: root@128:~# cat /proc/cpu/alignment User: 0 System: 17539 (inet_gro_receive+0x114/0x2c0) Skipped: 0 Half: 0 Word: 0 DWord: 0 Multi: 17539 User faults: 2 (fixup) Also shown when exception report enablement CPU: 0 PID: 161 Comm: irq/66-eth1_g0_ Not tainted 4.1.21-rt13-WR8.0.0.0_preempt-rt open-power-host-os#16 Hardware name: Freescale LS1021A [<8001b420>] (unwind_backtrace) from [<8001476c>] (show_stack+0x20/0x24) [<8001476c>] (show_stack) from [<807cfb48>] (dump_stack+0x94/0xac) [<807cfb48>] (dump_stack) from [<80025d70>] (do_alignment+0x720/0x958) [<80025d70>] (do_alignment) from [<80009224>] (do_DataAbort+0x40/0xbc) [<80009224>] (do_DataAbort) from [<80015398>] (__dabt_svc+0x38/0x60) Exception stack(0x86ad1cc0 to 0x86ad1d08) 1cc0: f9b3e080 86b3d072 2d78d287 00000000 866816c0 86b3d05e 86e785d0 00000000 1ce0: 00000011 0000000e 80840ab0 86ad1d3c 86ad1d08 86ad1d08 806d7fc0 806d806c 1d00: 40070013 ffffffff [<80015398>] (__dabt_svc) from [<806d806c>] (inet_gro_receive+0x114/0x2c0) [<806d806c>] (inet_gro_receive) from [<80660eec>] (dev_gro_receive+0x21c/0x3c0) [<80660eec>] (dev_gro_receive) from [<8066133c>] (napi_gro_receive+0x44/0x17c) [<8066133c>] (napi_gro_receive) from [<804f0538>] (gfar_clean_rx_ring+0x39c/0x7d4) [<804f0538>] (gfar_clean_rx_ring) from [<804f0bf4>] (gfar_poll_rx_sq+0x58/0xe0) [<804f0bf4>] (gfar_poll_rx_sq) from [<80660b10>] (net_rx_action+0x27c/0x43c) [<80660b10>] (net_rx_action) from [<80033638>] (do_current_softirqs+0x1e0/0x3dc) [<80033638>] (do_current_softirqs) from [<800338c4>] (__local_bh_enable+0x90/0xa8) [<800338c4>] (__local_bh_enable) from [<8008025c>] (irq_forced_thread_fn+0x70/0x84) [<8008025c>] (irq_forced_thread_fn) from [<800805e8>] (irq_thread+0x16c/0x244) [<800805e8>] (irq_thread) from [<8004e490>] (kthread+0xe8/0x104) [<8004e490>] (kthread) from [<8000fda8>] (ret_from_fork+0x14/0x2c) Signed-off-by: Zumeng Chen <zumeng.chen@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <alexander.levin@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I tried cpu hotplug-unplug in parallel with suspend resume.. guest becomes unresponsive..
i. On host run
terminal 1:
for i in {1..100};do sleep 2;virsh suspend srikanth_1710_Cdrom;sleep 2;virsh resume srikanth_1710_Cdrom;done
terminal 2:
for i in {1..100};do sleep 5;virsh setvcpus srikanth_1710_Cdrom 4 --live;sleep 5;virsh setvcpus srikanth_1710_Cdrom 2 --live;done
ii. After/while both above commands complete, will start seeing below messages on guest console:
[ 1209.368538] INFO: task jbd2/dm-0-8:1088 blocked for more than 120 seconds.
[ 1209.368633] Not tainted 4.12.0-11-generic #12-Ubuntu
[ 1209.368694] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
And the guest goes unresponsive..
I could'nt run sosreport on guest since it was unresponsive..
I tested same scenario as mentioned in previous comment with latest hostos guest.. I am hitting similar stall issues there as well...
I am attaching the traces of hostos here..
Guest is hung.. so could not get sosreports...
So these stall issues present on Ubuntu 17.10 [kernel: 4.13.0-11] and HostOS guest [kernel: 4.13.0-4.rel ]
In an internal comment I will post machine access details and guest details..
cde:info Mirrored with LTC bug #159338 </cde:info>
The text was updated successfully, but these errors were encountered: