On, more.

[ 9907.601105] Freezing user space processes ... 
[ 9909.280188] thinkpad_acpi: fan watchdog: enabling fan
[ 9927.620460] 
[ 9927.624458] Freezing of tasks failed after 20.01 seconds (1 tasks refusing to freeze, wq_busy=0):
[ 9927.630280] git             D 8179d659  5720 30852  30851 0x00800084
[ 9927.634513]  ef75ddb4 00000082 ef75dd38 8179d659 b770aecc ef75dd44 ca2f065f 00000902
[ 9927.638879]  8109173d 81c94900 81c94900 00000000 00000902 b770aa40 b7589520 ef75dd7c
[ 9927.643379]  00000000 00000000 b770aa40 00000001 001c3a40 ef75ddac 81088dca 22f7d887
[ 9927.647841] Call Trace:
[ 9927.652236]  [<8179d659>] ? sub_preempt_count.part.170+0x9/0x90
[ 9927.656673]  [<8109173d>] ? put_lock_stats.isra.20+0xd/0x30
[ 9927.661120]  [<81088dca>] ? ktime_get_ts+0xea/0x120
[ 9927.665564]  [<81797438>] io_schedule+0x78/0xb0
[ 9927.670088]  [<810fda8d>] sleep_on_page+0xd/0x20
[ 9927.674617]  [<81797def>] __wait_on_bit+0x4f/0x80
[ 9927.679177]  [<810fda80>] ? __lock_page+0xb0/0xb0
[ 9927.683825]  [<810fdc63>] wait_on_page_bit+0xb3/0xc0
[ 9927.688477]  [<8107bfa0>] ? autoremove_wake_function+0x50/0x50
[ 9927.693244]  [<811004a7>] __lock_page_or_retry+0x47/0x70
[ 9927.698017]  [<81100816>] filemap_fault+0x346/0x3b0
[ 9927.702872]  [<811004d0>] ? __lock_page_or_retry+0x70/0x70
[ 9927.707675]  [<8111d0ef>] __do_fault+0x4f/0x4d0
[ 9927.712476]  [<811004d0>] ? __lock_page_or_retry+0x70/0x70
[ 9927.717308]  [<81120497>] handle_pte_fault+0x87/0x240
[ 9927.722171]  [<81120818>] handle_mm_fault+0x138/0x260
[ 9927.726934]  [<8179d247>] do_page_fault+0x117/0x520
[ 9927.731709]  [<812940f0>] ? fuse_flush+0x40/0x140
[ 9927.736491]  [<8179aa2f>] ? error_code+0x5b/0x64
[ 9927.741285]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[ 9927.746005]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[ 9927.750679]  [<8179aa33>] error_code+0x5f/0x64
[ 9927.755260]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[ 9927.759879] 
[ 9927.762581] Restarting tasks ... done.

and one more (same?)

[10109.434374] Freezing user space processes ... 
[10111.840098] thinkpad_acpi: fan watchdog: enabling fan
[10129.450245] 
[10129.452542] Freezing of tasks failed after 20.01 seconds (1 tasks refusing to freeze, wq_busy=0):
[10129.455775] git             D 8179d659  6252 32330  32326 0x00800084
[10129.458196]  f1befdb4 00000082 f1befd38 8179d659 f0a219ac f1befd44 c9103491 00000931
[10129.460737]  8109173d 81c94900 81c94900 00000000 00000931 f0a21520 ee078000 f1befd7c
[10129.463286]  00000000 00000000 f0a21520 00000001 001cd784 f1befdac 81088dca 19b59094
[10129.465860] Call Trace:
[10129.468353]  [<8179d659>] ? sub_preempt_count.part.170+0x9/0x90
[10129.470899]  [<8109173d>] ? put_lock_stats.isra.20+0xd/0x30
[10129.473445]  [<81088dca>] ? ktime_get_ts+0xea/0x120
[10129.476013]  [<81797438>] io_schedule+0x78/0xb0
[10129.478597]  [<810fda8d>] sleep_on_page+0xd/0x20
[10129.481234]  [<81797def>] __wait_on_bit+0x4f/0x80
[10129.483864]  [<810fda80>] ? __lock_page+0xb0/0xb0
[10129.486513]  [<810fdc63>] wait_on_page_bit+0xb3/0xc0
[10129.489201]  [<8107bfa0>] ? autoremove_wake_function+0x50/0x50
[10129.491949]  [<811004a7>] __lock_page_or_retry+0x47/0x70
[10129.494698]  [<81100816>] filemap_fault+0x346/0x3b0
[10129.497425]  [<811004d0>] ? __lock_page_or_retry+0x70/0x70
[10129.500167]  [<8111d0ef>] __do_fault+0x4f/0x4d0
[10129.502908]  [<811004d0>] ? __lock_page_or_retry+0x70/0x70
[10129.505656]  [<81120497>] handle_pte_fault+0x87/0x240
[10129.508376]  [<81120818>] handle_mm_fault+0x138/0x260
[10129.511096]  [<8179d247>] do_page_fault+0x117/0x520
[10129.513814]  [<810813a1>] ? hrtimer_interrupt+0x151/0x280
[10129.516531]  [<8179d69d>] ? sub_preempt_count.part.170+0x4d/0x90
[10129.519244]  [<8179aa2f>] ? error_code+0x5b/0x64
[10129.521952]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10129.524607]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10129.527213]  [<8179aa33>] error_code+0x5f/0x64
[10129.529794]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10129.532430] 
[10129.535002] Restarting tasks ... done.

And one more (different again):

[10216.592699] Freezing user space processes ... 
[10234.880096] thinkpad_acpi: fan watchdog: enabling fan
[10236.610246] 
[10236.612850] Freezing of tasks failed after 20.01 seconds (1 tasks refusing to freeze, wq_busy=0):
[10236.616459] git             D 8179d659  6140  1900  32304 0x00800084
[10236.619227]  f0afbdb4 00000082 f0afbd38 8179d659 8c0eaecc f0afbd44 bc523cf1 0000094a
[10236.622089]  8109173d 81c94900 81c94900 00000000 0000094a 8c0eaa40 eca9aa40 f0afbd7c
[10236.624929]  00000000 00000000 8c0eaa40 00000001 001d2b34 f0afbdac 81088dca 23452c4c
[10236.627765] Call Trace:
[10236.630540]  [<8179d659>] ? sub_preempt_count.part.170+0x9/0x90
[10236.633302]  [<8109173d>] ? put_lock_stats.isra.20+0xd/0x30
[10236.636066]  [<81088dca>] ? ktime_get_ts+0xea/0x120
[10236.638805]  [<81797438>] io_schedule+0x78/0xb0
[10236.641580]  [<810fda8d>] sleep_on_page+0xd/0x20
[10236.644305]  [<81797def>] __wait_on_bit+0x4f/0x80
[10236.647015]  [<810fda80>] ? __lock_page+0xb0/0xb0
[10236.649709]  [<810fdc63>] wait_on_page_bit+0xb3/0xc0
[10236.652441]  [<8107bfa0>] ? autoremove_wake_function+0x50/0x50
[10236.655162]  [<811004a7>] __lock_page_or_retry+0x47/0x70
[10236.657892]  [<81100816>] filemap_fault+0x346/0x3b0
[10236.660633]  [<811004d0>] ? __lock_page_or_retry+0x70/0x70
[10236.663331]  [<8111d0ef>] __do_fault+0x4f/0x4d0
[10236.666019]  [<811004d0>] ? __lock_page_or_retry+0x70/0x70
[10236.668740]  [<81120497>] handle_pte_fault+0x87/0x240
[10236.671470]  [<81120818>] handle_mm_fault+0x138/0x260
[10236.674150]  [<8179d247>] do_page_fault+0x117/0x520
[10236.676819]  [<81154bdb>] ? sys_lstat64+0x2b/0x30
[10236.679470]  [<8179aa2f>] ? error_code+0x5b/0x64
[10236.682140]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10236.684760]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10236.687349]  [<8179aa33>] error_code+0x5f/0x64
[10236.689926]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10236.692501] 
[10236.694967] Restarting tasks ... done.


Interesingly, I have this problem only when STARTING compilation of kernel. Most problematic for some reason is git command (git describe --exact-match) and conf. When this commands are in progress, then i cannot suspend. But when they end, and normal compilation took place (still with pretty big disk usage), I have no problem suspending machine (10 success in 10 tries).

But when i put really big load on machine (make -j6), then also can reproduce this when compiling: 

[10345.074878] PM: Syncing filesystems ... done.
[10350.738837] Freezing user space processes ... 
[10370.754726] Freezing of tasks failed after 20.01 seconds (1 tasks refusing to freeze, wq_busy=0):
[10370.760811] ld              D 8179d659  6424  4532   4528 0x00800084
[10370.765250]  941cfdb4 00000082 941cfd38 8179d659 f38f6e2c 941cfd44 f6d76dfa 00000969
[10370.769809]  8109173d 81c94900 81c94900 00000000 00000969 f38f69a0 b7589520 941cfd7c
[10370.774510]  00000000 00000000 f38f69a0 00000001 001d93fa 941cfdac 81088dca 2ac4aabc
[10370.779148] Call Trace:
[10370.783733]  [<8179d659>] ? sub_preempt_count.part.170+0x9/0x90
[10370.788360]  [<8109173d>] ? put_lock_stats.isra.20+0xd/0x30
[10370.793045]  [<81088dca>] ? ktime_get_ts+0xea/0x120
[10370.797682]  [<81797438>] io_schedule+0x78/0xb0
[10370.802341]  [<810fda8d>] sleep_on_page+0xd/0x20
[10370.806943]  [<81797def>] __wait_on_bit+0x4f/0x80
[10370.811579]  [<810fda80>] ? __lock_page+0xb0/0xb0
[10370.816158]  [<810fdc63>] wait_on_page_bit+0xb3/0xc0
[10370.820774]  [<8107bfa0>] ? autoremove_wake_function+0x50/0x50
[10370.825366]  [<811004a7>] __lock_page_or_retry+0x47/0x70
[10370.829953]  [<81100816>] filemap_fault+0x346/0x3b0
[10370.834603]  [<811004d0>] ? __lock_page_or_retry+0x70/0x70
[10370.839221]  [<8111d0ef>] __do_fault+0x4f/0x4d0
[10370.843868]  [<811004d0>] ? __lock_page_or_retry+0x70/0x70
[10370.848482]  [<81120497>] handle_pte_fault+0x87/0x240
[10370.853124]  [<81120818>] handle_mm_fault+0x138/0x260
[10370.857713]  [<8179d247>] do_page_fault+0x117/0x520
[10370.862384]  [<8179aa2f>] ? error_code+0x5b/0x64
[10370.867012]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10370.871710]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10370.876288]  [<8179aa33>] error_code+0x5f/0x64
[10370.880902]  [<8179d130>] ? vmalloc_fault+0x190/0x190
[10370.885571] 
[10370.888072] Restarting tasks ... done.


In fact I could reproduce it also using cat *somebig_files* > /dev/null

[10561.423121] PM: Syncing filesystems ... done.
[10561.874139] Freezing user space processes ... 
[10571.360097] thinkpad_acpi: fan watchdog: enabling fan
[10581.890234] 
[10581.892674] Freezing of tasks failed after 20.01 seconds (1 tasks refusing to freeze, wq_busy=0):
[10581.895973] cat             D 8c0eaecc  6188  5057  14924 0x00800084
[10581.898497]  f1dd1d7c 00000086 8179d659 8c0eaecc f1dd1d08 8179d704 20ac0580 0000099b
[10581.901104]  f1dd1d28 81c94900 81c94900 00000000 0000099b 8c0eaa40 b758d480 00000000
[10581.903681]  00000000 8c0eaa40 00000001 8107bd68 001e38f0 f1dd1d74 81088dca 3404bc3c
[10581.906282] Call Trace:
[10581.908817]  [<8179d659>] ? sub_preempt_count.part.170+0x9/0x90
[10581.911419]  [<8179d704>] ? sub_preempt_count+0x24/0x50
[10581.913983]  [<8107bd68>] ? prepare_to_wait_exclusive+0x48/0x80
[10581.916562]  [<81088dca>] ? ktime_get_ts+0xea/0x120
[10581.919145]  [<81797438>] io_schedule+0x78/0xb0
[10581.921744]  [<810fdaad>] sleep_on_page_killable+0xd/0x40
[10581.924318]  [<81797ca7>] __wait_on_bit_lock+0x47/0xa0
[10581.926888]  [<810fdaa0>] ? sleep_on_page+0x20/0x20
[10581.929458]  [<810fd9c7>] __lock_page_killable+0xa7/0xb0
[10581.932062]  [<8107bfa0>] ? autoremove_wake_function+0x50/0x50
[10581.934645]  [<810ff2ef>] do_generic_file_read.constprop.33+0x2ef/0x4f0
[10581.937241]  [<811001a4>] generic_file_aio_read+0xd4/0x290
[10581.939836]  [<81291fd4>] fuse_file_aio_read+0xc4/0xf0
[10581.942475]  [<8114f00c>] do_sync_read+0xac/0xe0
[10581.945092]  [<812b474c>] ? security_file_permission+0x8c/0xa0
[10581.947727]  [<8114f34c>] ? rw_verify_area+0x6c/0x120
[10581.950377]  [<810942a2>] ? check_flags+0x12/0x20
[10581.953003]  [<8114f82c>] vfs_read+0x8c/0x160
[10581.955648]  [<8114ef60>] ? do_sync_write+0xe0/0xe0
[10581.958298]  [<8114f93d>] sys_read+0x3d/0x70
[10581.960960]  [<817a1098>] sysenter_do_call+0x12/0x38
[10581.963645] 
[10581.966249] Restarting tasks ... done.


I know some time ago, there was similar problem with sshfs. One could not suspend when sshfs was mounted. But i had no such problem on zfs-fuse previously. Also on sshfs it was that no file needed to be opened or operated on, it was sufficient to have sshfs mounted to prevent suspending. Here, one need to put considerable load to trigger some kind of race-condition.

AFAIK suspending logic is not sufficient. First kernel freezes all processes, but then it tries to perform something with filesystem, probably caches, but cannot do anything as process (zfs-fuse), is already freezed.