Re: [xen-unstable test] 152067: regressions - trouble: fail/pass/starved

[Jürgen Groß <jgross@xxxxxxxx>]

On 22.07.20 11:30, Roger Pau Monné wrote:
> On Wed, Jul 22, 2020 at 11:23:20AM +0200, Jürgen Groß wrote:
> > On 22.07.20 11:02, Roger Pau Monné wrote:
> > > On Wed, Jul 22, 2020 at 10:59:48AM +0200, Jürgen Groß wrote:
> > > > On 22.07.20 10:38, Roger Pau Monné wrote:
> > > > > On Wed, Jul 22, 2020 at 12:37:46AM +0000, osstest service owner wrote:
> > > > > > flight 152067 xen-unstable real [real]
> > > > > > http://logs.test-lab.xenproject.org/osstest/logs/152067/
> > > > > >
> > > > > > Regressions :-(
> > > > > >
> > > > > > Tests which did not succeed and are blocking,
> > > > > > including tests which could not be run:
> > > > > >     test-amd64-amd64-dom0pvh-xl-intel 18 guest-localmigrate/x10 fail REGR. vs. 
> > > > > > 152045
> > > > > Failure was caused by:
> > > > >
> > > > > Jul 21 16:20:58.985209 [  530.412043] libxl-save-help: page allocation failure: 
> > > > > order:4, mode:0x60c0c0(GFP_KERNEL|__GFP_COMP|__GFP_ZERO), nodemask=(null)
> > > > > Jul 21 16:21:00.378548 [  530.412261] libxl-save-help cpuset=/ mems_allowed=0
> > > > > Jul 21 16:21:00.378622 [  530.412318] CPU: 1 PID: 15485 Comm: libxl-save-help 
> > > > > Not tainted 4.19.80+ #1
> > > > > Jul 21 16:21:00.390740 [  530.412377] Hardware name: Dell Inc. PowerEdge 
> > > > > R420/0K29HN, BIOS 2.4.2 01/29/2015
> > > > > Jul 21 16:21:00.390810 [  530.412448] Call Trace:
> > > > > Jul 21 16:21:00.402721 [  530.412499]  dump_stack+0x72/0x8c
> > > > > Jul 21 16:21:00.402801 [  530.412541]  warn_alloc.cold.140+0x68/0xe8
> > > > > Jul 21 16:21:00.402841 [  530.412585]  __alloc_pages_slowpath+0xc73/0xcb0
> > > > > Jul 21 16:21:00.414737 [  530.412640]  ? __do_page_fault+0x249/0x4d0
> > > > > Jul 21 16:21:00.414786 [  530.412681]  __alloc_pages_nodemask+0x235/0x250
> > > > > Jul 21 16:21:00.426555 [  530.412734]  kmalloc_order+0x13/0x60
> > > > > Jul 21 16:21:00.426619 [  530.412774]  kmalloc_order_trace+0x18/0xa0
> > > > > Jul 21 16:21:00.426671 [  530.412816]  alloc_empty_pages.isra.15+0x24/0x60
> > > > > Jul 21 16:21:00.438447 [  530.412867]  
> > > > > privcmd_ioctl_mmap_batch.isra.18+0x303/0x320
> > > > > Jul 21 16:21:00.438507 [  530.412918]  ? vmacache_find+0xb0/0xb0
> > > > > Jul 21 16:21:00.450475 [  530.412957]  privcmd_ioctl+0x253/0xa9b
> > > > > Jul 21 16:21:00.450540 [  530.412996]  ? mmap_region+0x226/0x630
> > > > > Jul 21 16:21:00.450592 [  530.413043]  ? selinux_mmap_file+0xb0/0xb0
> > > > > Jul 21 16:21:00.462757 [  530.413084]  ? selinux_file_ioctl+0x15c/0x200
> > > > > Jul 21 16:21:00.462823 [  530.413136]  do_vfs_ioctl+0x9f/0x630
> > > > > Jul 21 16:21:00.474698 [  530.413177]  ksys_ioctl+0x5b/0x90
> > > > > Jul 21 16:21:00.474762 [  530.413224]  __x64_sys_ioctl+0x11/0x20
> > > > > Jul 21 16:21:00.474813 [  530.413264]  do_syscall_64+0x57/0x130
> > > > > Jul 21 16:21:00.486480 [  530.413305]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
> > > > > Jul 21 16:21:00.486548 [  530.413357] RIP: 0033:0x7f4f7ecde427
> > > > > Jul 21 16:21:00.486600 [  530.413395] Code: 00 00 90 48 8b 05 69 aa 0c 00 64 c7 00 26 
> > > > > 00 00 00 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 b8 10 00 00 00 0f 05 
> > > > > <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 39 aa 0c 00 f7 d8 64 89 01 48
> > > > > Jul 21 16:21:00.510766 [  530.413556] RSP: 002b:00007ffc1ef6eb38 EFLAGS: 
> > > > > 00000213 ORIG_RAX: 0000000000000010
> > > > > Jul 21 16:21:00.522758 [  530.413629] RAX: ffffffffffffffda RBX: 
> > > > > 0000000000000000 RCX: 00007f4f7ecde427
> > > > > Jul 21 16:21:00.534632 [  530.413699] RDX: 00007ffc1ef6eb90 RSI: 
> > > > > 0000000000205004 RDI: 0000000000000007
> > > > > Jul 21 16:21:00.534702 [  530.413810] RBP: 00007ffc1ef6ebe0 R08: 
> > > > > 0000000000000007 R09: 0000000000000000
> > > > > Jul 21 16:21:00.547013 [  530.413881] R10: 0000000000000001 R11: 
> > > > > 0000000000000213 R12: 000055d754136200
> > > > > Jul 21 16:21:00.558751 [  530.413951] R13: 00007ffc1ef6f340 R14: 
> > > > > 0000000000000000 R15: 0000000000000000
> > > > > Jul 21 16:21:00.558846 [  530.414079] Mem-Info:
> > > > > Jul 21 16:21:00.558928 [  530.414123] active_anon:1724 inactive_anon:3931 
> > > > > isolated_anon:0
> > > > > Jul 21 16:21:00.570481 [  530.414123]  active_file:7862 inactive_file:86530 
> > > > > isolated_file:0
> > > > > Jul 21 16:21:00.582599 [  530.414123]  unevictable:0 dirty:18 writeback:0 
> > > > > unstable:0
> > > > > Jul 21 16:21:00.582668 [  530.414123]  slab_reclaimable:4704 
> > > > > slab_unreclaimable:4036
> > > > > Jul 21 16:21:00.594782 [  530.414123]  mapped:3461 shmem:124 pagetables:372 
> > > > > bounce:0
> > > > > Jul 21 16:21:00.594849 [  530.414123]  free:1863 free_pcp:16 free_cma:0
> > > > > Jul 21 16:21:00.606733 [  530.414579] Node 0 active_anon:6896kB 
> > > > > inactive_anon:15724kB active_file:31448kB inactive_file:346120kB 
> > > > > unevictable:0kB isolated(anon):0kB isolated(file):0kB mapped:13844kB dirty:72kB 
> > > > > writeback:0kB shmem:496kB writeback_tmp:0kB unstable:0kB all_unreclaimable? no
> > > > > Jul 21 16:21:00.630626 [  530.414870] DMA free:1816kB min:92kB low:112kB 
> > > > > high:132kB active_anon:0kB inactive_anon:0kB active_file:76kB 
> > > > > inactive_file:9988kB unevictable:0kB writepending:0kB present:15980kB 
> > > > > managed:14328kB mlocked:0kB kernel_stack:0kB pagetables:0kB bounce:0kB 
> > > > > free_pcp:0kB local_pcp:0kB free_cma:0kB
> > > > > Jul 21 16:21:00.658448 [  530.415329] lowmem_reserve[]: 0 431 431 431
> > > > > Jul 21 16:21:00.658513 [  530.415404] DMA32 free:5512kB min:2608kB low:3260kB 
> > > > > high:3912kB active_anon:6896kB inactive_anon:15724kB active_file:31372kB 
> > > > > inactive_file:336132kB unevictable:0kB writepending:72kB present:508300kB 
> > > > > managed:451760kB mlocked:0kB kernel_stack:2848kB pagetables:1488kB bounce:0kB 
> > > > > free_pcp:184kB local_pcp:0kB free_cma:0kB
> > > > > Jul 21 16:21:00.694702 [  530.415742] lowmem_reserve[]: 0 0 0 0
> > > > > Jul 21 16:21:00.694778 [  530.415806] DMA: 8*4kB (UM) 3*8kB (UM) 4*16kB (UM) 
> > > > > 3*32kB (M) 5*64kB (UM) 2*128kB (UM) 4*256kB (UM) 0*512kB 0*1024kB 0*2048kB 
> > > > > 0*4096kB = 1816kB
> > > > > Jul 21 16:21:00.706798 [  530.416015] DMA32: 4*4kB (UH) 459*8kB (MH) 2*16kB (H) 
> > > > > 6*32kB (H) 5*64kB (H) 4*128kB (H) 3*256kB (H) 0*512kB 0*1024kB 0*2048kB 
> > > > > 0*4096kB = 5512kB
> > > > > Jul 21 16:21:00.718789 [  530.416287] Node 0 hugepages_total=0 hugepages_free=0 
> > > > > hugepages_surp=0 hugepages_size=1048576kB
> > > > > Jul 21 16:21:00.730785 [  530.416413] Node 0 hugepages_total=0 hugepages_free=0 
> > > > > hugepages_surp=0 hugepages_size=2048kB
> > > > > Jul 21 16:21:00.742847 [  530.416538] 94608 total pagecache pages
> > > > > Jul 21 16:21:00.742881 [  530.416598] 79 pages in swap cache
> > > > > Jul 21 16:21:00.754859 [  530.416670] Swap cache stats: add 702, delete 623, 
> > > > > find 948/1025
> > > > > Jul 21 16:21:00.754924 [  530.416759] Free swap  = 1947124kB
> > > > > Jul 21 16:21:00.766880 [  530.416822] Total swap = 1949692kB
> > > > > Jul 21 16:21:00.766960 [  530.416924] 131070 pages RAM
> > > > > Jul 21 16:21:00.767021 [  530.416988] 0 pages HighMem/MovableOnly
> > > > > Jul 21 16:21:00.778697 [  530.417051] 14548 pages reserved
> > > > >
> > > > > AFAICT from the kernel config used for the test [0]
> > > > > CONFIG_XEN_BALLOON_MEMORY_HOTPLUG is enabled, so I'm not sure where
> > > > > the memory exhaustion is coming from. Maybe 512M is too low for a PVH
> > > > > dom0, even when using hotplug balloon memory?
> > > > I don't see how CONFIG_XEN_BALLOON_MEMORY_HOTPLUG would help here, as it
> > > > will be used for real memory hotplug only. Well, you _can_ use it for
> > > > mapping of foreign pages, but you'd have to:
> > > >
> > > > echo 1 > /proc/sys/xen/balloon/hotplug_unpopulated
> > > Uh, I've completely missed the point then. I assume there's some
> > > reason for not doing it by default then? (using empty hotplug ranges
> > > to map foreign memory)
> > This dates back to 2015. See commit 1cf6a6c82918c9aad.
> >
> > I guess we could initialize xen_hotplug_unpopulated with 1 for PVH
> > dom0.
> Would you like me to enabled it in osstest first and then we can see
> about enabling it by default?
Yes, good idea.


Juergen