Re: [Xen-devel] [PATCH v2] xl: track child processes for the benefit of libxl

[Ian Jackson <Ian.Jackson@xxxxxxxxxxxxx>]

Ian Jackson writes ("[PATCH v2] xl: track child processes for the benefit of 
libxl"):
> Each time xl forks, it needs to record the pid, so that its exit
> status can be preserved if it happens that libxl's event loop reaped
> it.  Consequently we also have a new wrapper for waitpid, which in
> that case returns the previously-reaped status.

I have tested this on top of current tip ("x86/mem_sharing: Allow
paging-in pages to be replaced by shared ones") plus my event
debugging series and:

Loading new save file incoming migration stream (new xl fmt info 0x0/0x0/853)
 Savefile contains xl domain config
xc: Saving memory: iter 0 (last sent 0 skipped 0): 131072/131072  100%
xc: Saving memory: iter 1 (last sent 130814 skipped 258): 131072/131072  100%
xc: Saving memory: iter 2 (last sent 306 skipped 0): 131072/131072  100%      
xc: Saving memory: iter 3 (last sent 53 skipped 0): 131072/131072  100%  
xc: Saving memory: iter 4 (last sent 0 skipped 0): 131072/131072  100%  
xc: error: Max batch size exceeded (-18). Giving up.: Internal error
xc: error: Error when reading batch (90 = Message too long): Internal error
libxl: error: libxl_dom.c:430:libxl__domain_restore_common: restoring domain: 
Resource temporarily unavailable
libxl: error: libxl_create.c:596:do_domain_create: cannot (re-)build domain: -3
libxl: error: libxl.c:1043:libxl_domain_destroy: non-existant domain 123
migration target: Domain creation failed (code -3).
libxl: error: libxl_utils.c:364:libxl_read_exactly: file/stream truncated 
reading ready message from migration receiver stream
libxl: info: libxl_exec.c:108:libxl_report_child_exitstatus: migration target 
process [28059] exited with error status 3
Migration failed, resuming at sender.
root@bedbug:~/shadow/tools/libxl#

I don't think this is my fault.  Also after this happens the guest
crashes:

root@bedbug:~# xl console debian.guest.osstest
[   47.118283] ------------[ cut here ]------------
[   47.118283] kernel BUG at 
/tmp/buildd/linux-2.6-2.6.32/debian/build/source_i386_none/drivers/xen/events.c:839!
[   47.118283] invalid opcode: 0000 [#1] SMP 
[   47.118283] last sysfs file: /sys/devices/virtual/net/lo/operstate
[   47.118283] Modules linked in: evdev snd_pcm snd_timer snd soundcore 
snd_page_alloc pcspkr ext3 jbd mbcache xen_netfront xen_blkfront
[   47.118283] 
[   47.118283] Pid: 531, comm: kstop/0 Not tainted (2.6.32-5-686-bigmem #1) 
[   47.118283] EIP: 0061:[<c1196709>] EFLAGS: 00010082 CPU: 0
[   47.118283] EIP is at xen_irq_resume+0xd9/0x21f
[   47.118283] EAX: ffffffef EBX: 00000000 ECX: deadbeef EDX: de2d5f40
[   47.118283] ESI: 00000000 EDI: de2d5fb0 EBP: c14f29fc ESP: de2d5f10
[   47.118283]  DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0069
[   47.118283] Process kstop/0 (pid: 531, ti=de2d4000 task=df968000 
task.ti=de2d4000)
[   47.118283] Stack:
[   47.118283]  00000000 00000003 c143b474 c24184f0 00000000 00000000 00000002 
00000010
[   47.118283] <0> c1005744 deadbeef 00000003 de2d5fb0 00000000 00000000 
0000f3f3 deadbeef
[   47.118283] <0> 00000003 de2d5fb0 00000000 c1197494 df851f8c c14995ec 
00000003 de2d5fb0
[   47.118283] Call Trace:
[   47.118283]  [<c1005744>] ? xen_force_evtchn_callback+0xc/0x10
[   47.118283]  [<c1197494>] ? xen_suspend+0x99/0xb0
[   47.118283]  [<c106abd0>] ? stop_cpu+0x6d/0xab
[   47.118283]  [<c1047c37>] ? worker_thread+0x141/0x1bd
[   47.118283]  [<c106ab63>] ? stop_cpu+0x0/0xab
[   47.118283]  [<c104a97a>] ? autoremove_wake_function+0x0/0x2d
[   47.118283]  [<c1047af6>] ? worker_thread+0x0/0x1bd
[   47.118283]  [<c104a748>] ? kthread+0x61/0x66
[   47.118283]  [<c104a6e7>] ? kthread+0x0/0x66
[   47.118283]  [<c1008d87>] ? kernel_thread_helper+0x7/0x10
[   47.118283] Code: fe 0f b7 45 08 39 d8 74 04 0f 0b eb fe 8b 44 24 10 8d 54 
24 30 89 5c 24 30 89 44 24 34 b8 01 00 00 00 e8 f3 fa ff ff 85 c0 74 04 <0f> 0b 
eb fe 8b 4c 24 38 8d 7c 24 24 89 0c 24 89 34 8d e0 1e 3c 
[   47.118283] EIP: [<c1196709>] xen_irq_resume+0xd9/0x21f SS:ESP 0069:de2d5f10
[   47.118283] ---[ end trace 838e9608136663d9 ]---
[   47.118283] ------------[ cut here ]------------
[   47.118283] WARNING: at 
/tmp/buildd/linux-2.6-2.6.32/debian/build/source_i386_none/kernel/time/timekeeping.c:260
 ktime_get+0x1f/0xcb()
[   47.118283] Modules linked in: evdev snd_pcm snd_timer snd soundcore 
snd_page_alloc pcspkr ext3 jbd mbcache xen_netfront xen_blkfront
[   47.118283] Pid: 0, comm: swapper Tainted: G      D    2.6.32-5-686-bigmem #1
[   47.118283] Call Trace:
[   47.118283]  [<c1051b1c>] ? ktime_get+0x1f/0xcb
[   47.118283]  [<c1051b1c>] ? ktime_get+0x1f/0xcb
[   47.118283]  [<c1036ac5>] ? warn_slowpath_common+0x5e/0x8a
[   47.118283]  [<c1036afb>] ? warn_slowpath_null+0xa/0xc
[   47.118283]  [<c1051b1c>] ? ktime_get+0x1f/0xcb
[   47.118283]  [<c1055bc8>] ? tick_nohz_stop_sched_tick+0x5d/0x36e
[   47.118283]  [<c104d83d>] ? hrtimer_start_range_ns+0xf/0x13
[   47.118283]  [<c1055b66>] ? tick_nohz_restart_sched_tick+0x128/0x12d
[   47.118283]  [<c1007395>] ? cpu_idle+0x67/0xa2
[   47.118283]  [<c13dc80d>] ? start_kernel+0x318/0x31d
[   47.118283]  [<c13de2eb>] ? xen_start_kernel+0x46a/0x471
[   47.118283]  [<c1409043>] ? udp_init+0x30/0x8c
[   47.118283] ---[ end trace 838e9608136663da ]---

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel