Bug 111864 - [CI][DRMTIP] igt@gem_mmap_gtt@fault-concurrent - dmesg-warn - page allocation failure
Summary: [CI][DRMTIP] igt@gem_mmap_gtt@fault-concurrent - dmesg-warn - page allocation...
Status: RESOLVED MOVED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: DRI git
Hardware: Other All
: low minor
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2019-09-30 10:12 UTC by Lakshmi
Modified: 2019-11-29 19:36 UTC (History)
1 user (show)

See Also:
i915 platform: I915G
i915 features: GEM/Other

Attachments

Description Lakshmi 2019-09-30 10:12:53 UTC 
https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_379/fi-gdg-551/igt@gem_mmap_gtt@fault-concurrent.html
<6> [106.712677] [IGT] gem_mmap_gtt: executing
<6> [106.726615] [IGT] gem_mmap_gtt: starting subtest fault-concurrent
<7> [111.468711] [drm:intel_sdvo_detect [i915]] [CONNECTOR:48:DVI-D-1]
<7> [111.468842] [drm:intel_sdvo_debug_write [i915]] SDVOB: W: 0B                         (GET_ATTACHED_DISPLAYS)
<7> [111.476823] [drm:intel_sdvo_read_response [i915]] SDVOB: R: (Success) 00 00
<7> [111.476922] [drm:intel_sdvo_detect [i915]] SDVO response 0 0 [1]
<7> [121.708760] [drm:intel_sdvo_detect [i915]] [CONNECTOR:48:DVI-D-1]
<7> [121.708900] [drm:intel_sdvo_debug_write [i915]] SDVOB: W: 0B                         (GET_ATTACHED_DISPLAYS)
<7> [121.720377] [drm:intel_sdvo_read_response [i915]] SDVOB: R: (Success) 00 00
<7> [121.720478] [drm:intel_sdvo_detect [i915]] SDVO response 0 0 [1]
<4> [126.804555] gem_mmap_gtt: page allocation failure: order:0, mode:0x400d0(__GFP_IO|__GFP_FS|__GFP_COMP|__GFP_RECLAIMABLE), nodemask=(null)
<4> [126.804594] CPU: 0 PID: 1053 Comm: gem_mmap_gtt Tainted: G     U            5.3.0-g80fa0e042cdb-drmtip_379+ #1
<4> [126.804598] Hardware name: Dell Inc.                 OptiPlex GX280               /0G8310, BIOS A04 02/09/2005
<4> [126.804602] Call Trace:
<4> [126.804615]  dump_stack+0x67/0x9b
<4> [126.804624]  warn_alloc+0xfa/0x180
<4> [126.804643]  __alloc_pages_nodemask+0xd34/0x1110
<4> [126.804663]  ? get_partial_node.isra.30+0x1fe/0x3f0
<4> [126.804677]  new_slab+0x81b/0x1400
<4> [126.804691]  ___slab_alloc.constprop.34+0x2d3/0x380
<4> [126.804697]  ? xas_nomem+0x49/0x70
<4> [126.804704]  ? rcu_read_lock_sched_held+0x4d/0x80
<4> [126.804710]  ? kmem_cache_alloc+0x2bd/0x2f0
<4> [126.804722]  ? xas_nomem+0x49/0x70
<4> [126.804729]  ? __slab_alloc.isra.28.constprop.33+0x3d/0x70
<4> [126.804733]  __slab_alloc.isra.28.constprop.33+0x3d/0x70
<4> [126.804740]  ? xas_nomem+0x49/0x70
<4> [126.804746]  kmem_cache_alloc+0x28d/0x2f0
<4> [126.804755]  xas_nomem+0x49/0x70
<4> [126.804761]  add_to_swap_cache+0x2b2/0x370
<4> [126.804778]  __read_swap_cache_async+0x11d/0x240
<4> [126.804787]  swap_cluster_readahead+0x1ae/0x280
<4> [126.804794]  ? __lock_acquire+0x4ac/0x1ea0
<4> [126.804811]  ? shmem_swapin+0x7b/0xa0
<4> [126.804815]  shmem_swapin+0x7b/0xa0
<4> [126.804828]  ? find_get_entry+0x1b5/0x2f0
<4> [126.804851]  shmem_swapin_page+0x568/0x770
<4> [126.804868]  shmem_getpage_gfp.isra.8+0x2c2/0x850
<4> [126.804889]  shmem_read_mapping_page_gfp+0x3e/0x70
<4> [126.805056]  shmem_get_pages+0x203/0x680 [i915]
<4> [126.805159]  ? __i915_gem_object_get_pages+0x18/0xb0 [i915]
<4> [126.805259]  ____i915_gem_object_get_pages+0x1d/0xa0 [i915]
<4> [126.805350]  __i915_gem_object_get_pages+0x59/0xb0 [i915]
<4> [126.805441]  i915_gem_fault+0x27e/0x970 [i915]
<4> [126.805448]  ? _raw_spin_unlock_irq+0x2f/0x50
<4> [126.805455]  ? finish_task_switch+0x96/0x250
<4> [126.805472]  __do_fault+0x2c/0xb0
<4> [126.805479]  __handle_mm_fault+0x98f/0xfa0
<4> [126.805498]  handle_mm_fault+0x155/0x360
<4> [126.805508]  __do_page_fault+0x249/0x4f0
<4> [126.805520]  page_fault+0x34/0x40
<4> [126.805526] RIP: 0033:0x564f9677e60d
<4> [126.805532] Code: 44 89 00 83 fa 20 74 2a 8b 47 08 01 d0 89 c1 c1 f9 1f c1 e9 1b 01 c8 83 e0 1f 29 c8 f6 c2 01 48 98 48 8b 04 c6 75 d6 83 c2 01 <44> 8b 00 83 fa 20 75 d6 31 c0 c3 0f 1f 84 00 00 00 00 00 48 63 35
<4> [126.805536] RSP: 002b:00007fef13c36cb8 EFLAGS: 00010202
<4> [126.805541] RAX: 00007fef29c4a000 RBX: 0000000000000000 RCX: 0000000000000000
<4> [126.805545] RDX: 0000000000000001 RSI: 00007ffc4fdcfbc0 RDI: 00007ffc4fdcfe70
<4> [126.805548] RBP: 0000000000000000 R08: 0000000000000000 R09: 00007fef13c37700
<4> [126.805552] R10: 00007fef13c379d0 R11: 0000000000000202 R12: 00007fef13c36d80
<4> [126.805556] R13: 0000000000000000 R14: 00007ffc4fdcfe70 R15: 00007ffc4fdcf750
<4> [126.805646] Mem-Info:
<4> [126.805655] active_anon:23166 inactive_anon:99074 isolated_anon:0
 active_file:24424 inactive_file:20478 isolated_file:0
 unevictable:15582 dirty:36 writeback:1 unstable:0
 slab_reclaimable:10156 slab_unreclaimable:10815
 mapped:24807 shmem:36356 pagetables:1413 bounce:0
 free:32152 free_pcp:62 free_cma:0
<4> [126.805662] Node 0 active_anon:92664kB inactive_anon:396296kB active_file:97696kB inactive_file:81912kB unevictable:62328kB isolated(anon):0kB isolated(file):0kB mapped:99228kB dirty:144kB writeback:4kB shmem:145424kB shmem_thp: 0kB shmem_pmdmapped: 0kB anon_thp: 38912kB writeback_tmp:0kB unstable:0kB all_unreclaimable? no
<4> [126.805669] DMA free:4436kB min:740kB low:924kB high:1108kB active_anon:8kB inactive_anon:11228kB active_file:0kB inactive_file:0kB unevictable:144kB writepending:0kB present:15996kB managed:15908kB mlocked:0kB kernel_stack:0kB pagetables:40kB bounce:0kB free_pcp:0kB local_pcp:0kB free_cma:0kB
<4> [126.805673] lowmem_reserve[]: 0 926 926 926
<4> [126.805683] DMA32 free:124172kB min:124184kB low:135260kB high:146336kB active_anon:92656kB inactive_anon:385064kB active_file:97696kB inactive_file:81912kB unevictable:62172kB writepending:144kB present:1022488kB managed:952424kB mlocked:0kB kernel_stack:2348kB pagetables:5612kB bounce:0kB free_pcp:248kB local_pcp:248kB free_cma:0kB
<4> [126.805687] lowmem_reserve[]: 0 0 0 0
<4> [126.805693] DMA: 1*4kB (U) 4*8kB (UME) 3*16kB (UM) 2*32kB (ME) 3*64kB (UME) 2*128kB (UM) 3*256kB (UME) 2*512kB (ME) 2*1024kB (UM) 0*2048kB 0*4096kB = 4436kB
<4> [126.805714] DMA32: 4043*4kB (M) 1850*8kB (ME) 843*16kB (UME) 461*32kB (MEH) 203*64kB (MH) 96*128kB (UME) 73*256kB (UM) 39*512kB (MEH) 1*1024kB (H) 0*2048kB 0*4096kB = 124172kB
<6> [126.805737] Node 0 hugepages_total=0 hugepages_free=0 hugepages_surp=0 hugepages_size=2048kB
<4> [126.805740] 157562 total pagecache pages
<4> [126.805746] 76295 pages in swap cache
<4> [126.805750] Swap cache stats: add 83832, delete 7537, find 997/1086
<4> [126.805754] Free swap  = 1767420kB
<4> [126.805757] Total swap = 2097148kB
<4> [126.805760] 259621 pages RAM
<4> [126.805763] 0 pages HighMem/MovableOnly
<4> [126.805766] 17538 pages reserved
<4> [126.805805] SLUB: Unable to allocate memory on node -1, gfp=0xc0(__GFP_IO|__GFP_FS)
<4> [126.805810]   cache: radix_tree_node, object size: 576, buffer size: 912, default order: 2, min order: 0
<4> [126.805815]   node 0: slabs: 377, objs: 5863, free: 15
<7> [131.948728] [drm:intel_sdvo_detect [i915]] [CONNECTOR:48:DVI-D-1]
<7> [131.948858] [drm:intel_sdvo_debug_write [i915]] SDVOB: W: 0B                         (GET_ATTACHED_DISPLAYS)
<7> [131.955276] [drm:intel_sdvo_read_response [i915]] SDVOB: R: (Success) 00 00
<7> [131.955387] [drm:intel_sdvo_detect [i915]] SDVO response 0 0 [1]
<6> [132.339627] gem_mmap_gtt (1036) used greatest stack depth: 11928 bytes left
<6> [134.118710] [IGT] gem_mmap_gtt: exiting, ret=0
<5> [134.133339] Setting dangerous option reset - tainting kernel
<6> [134.319924] Console: switching to colour frame buffer device 128x48
Comment 1 CI Bug Log 2019-09-30 10:14:38 UTC 
The CI Bug Log issue associated to this bug has been updated.

### New filters associated

* GDG: igt@gem_mmap_gtt@fault-concurrent - dmesg-warn - page allocation failure
  - https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_379/fi-gdg-551/igt@gem_mmap_gtt@fault-concurrent.html
Comment 2 Chris Wilson 2019-09-30 10:23:45 UTC 
Same old swap allocation failure warning. The only question here is that fault-concurrent isn't meant to be hitting swap.

It uses 32 x 16MiB objects, an estimated 512 MiB which seems perhaps a little excessive, but not really. The object needs to be large enough to hit partial mmaps, but ideally we would have a mixture of object sizes to avoid them all hitting the same code paths.
Comment 3 Martin Peres 2019-11-29 19:36:15 UTC 
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/intel/issues/467.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.