Bug 60729 - [linux] 3.10.x causes floppy hang when trying to umount
Summary: [linux] 3.10.x causes floppy hang when trying to umount
Status: NEW
Alias: None
Product: IO/Storage
Classification: Unclassified
Component: Other (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: io_other
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-08-11 18:07 UTC by bgc1954
Modified: 2016-03-21 18:15 UTC (History)
4 users (show)

See Also:
Kernel Version: 3.9
Subsystem:
Regression: No
Bisected commit-id:


Attachments

Description bgc1954 2013-08-11 18:07:54 UTC
Upgrading from linux-3.9.9-1 to linux-3.10.2-1 and any 3.10.x varieties after cause a hang from the kernel when trying to umount the floppy.  This is in Arch linux i686.  I thought this might get fixed in later versions but last tried 3.10.5-1 and still the same.  Downgrading back to 3.9.9-1 and floppy is fine.

dmesg output after hang:
[ 42.609185] EXT4-fs (fd0): mounting ext2 file system using the ext4 subsystem
[ 42.609820] EXT4-fs (fd0): warning: mounting unchecked fs, running e2fsck is recommended
[ 43.220225] EXT4-fs (fd0): mounted filesystem without journal. Opts: (null)
[ 53.483095] EXT4-fs (fd0): mounting ext2 file system using the ext4 subsystem
[ 53.484243] EXT4-fs (fd0): warning: mounting unchecked fs, running e2fsck is recommended
[ 54.094030] EXT4-fs (fd0): mounted filesystem without journal. Opts: (null)
[ 240.253352] INFO: task umount:696 blocked for more than 120 seconds.
[ 240.253361] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 240.253366] umount D f6d12b00 0 696 1 0x00000004
[ 240.253376] f511fdc8 00000086 c016e997 f6d12b00 00000000 00000001 f6735c00 c0754b00
[ 240.253387] c06641a0 f511fd84 c0754b00 f6d12b00 f1876680 f511fd90 c018791f 00003988
[ 240.253398] 00000246 f5c16920 00000246 f511fdc0 c01bd9f3 eba39ec8 ffffffff 273e68c8
[ 240.253408] Call Trace:
[ 240.253424] [<c016e997>] ? set_next_entity+0xb7/0xe0
[ 240.253432] [<c018791f>] ? ktime_get_ts+0x3f/0x100
[ 240.253441] [<c01bd9f3>] ? delayacct_end+0x93/0xa0
[ 240.253448] [<c018791f>] ? ktime_get_ts+0x3f/0x100
[ 240.253455] [<c01f1f00>] ? filemap_fdatawait+0x50/0x50
[ 240.253462] [<c0505933>] schedule+0x23/0x60
[ 240.253468] [<c0505b78>] io_schedule+0x78/0xc0
[ 240.253475] [<c01f1f0d>] sleep_on_page+0xd/0x20
[ 240.253483] [<c0503931>] __wait_on_bit+0x51/0x70
[ 240.253489] [<c01f1f00>] ? filemap_fdatawait+0x50/0x50
[ 240.253496] [<c01f1d16>] wait_on_page_bit+0x86/0x90
[ 240.253504] [<c015b7c0>] ? autoremove_wake_function+0x40/0x40
[ 240.253511] [<c01f1e38>] filemap_fdatawait_range+0xc8/0x140
[ 240.253518] [<c01f1ef4>] filemap_fdatawait+0x44/0x50
[ 240.253526] [<c0265649>] sync_inodes_sb+0x109/0x170
[ 240.253542] [<c02699a4>] sync_filesystem+0x54/0x90
[ 240.253549] [<c0243c1b>] generic_shutdown_super+0x2b/0xc0
[ 240.253556] [<c0243eb2>] kill_block_super+0x22/0x70
[ 240.253563] [<c02441b3>] deactivate_locked_super+0x33/0x50
[ 240.253570] [<c024463a>] deactivate_super+0x3a/0x60
[ 240.253577] [<c025bcd7>] mntput_no_expire+0xb7/0x110
[ 240.253583] [<c025cbca>] SyS_umount+0x9a/0x330
[ 240.253591] [<c050d0cd>] sysenter_do_call+0x12/0x28

Steps to reproduce:

upgrade from linux-3.9.9-1 to any 3.10.x version and floppy will mount once using terminal. Try to umount floppy from terminal and it appears to be ok. Mount floppy again and floppy light activates and terminal doesn't return to prompt. Open another terminal and check dmesg and you get the kernel hang message above.
Comment 1 bgc1954 2013-09-01 17:05:10 UTC
I just tried to upgrade to the latest Arch Linux kernel 3.10.10-1 and now the floppy is operating normally.  No hangs now so if someone did something, thanks, otherwise I guess it sorted itself out.
Comment 2 bgc1954 2013-09-03 16:48:17 UTC
I'm not sure what happened this morning but now I'm encountering the same problems although now it doesn't seem to mount correctly even once.  The floppy light goes on and off and the terminal won't return to prompt.  You get the same dmesg output as before.  Maybe I didn't reboot before trying.  I didn't notice this yesterday since I only use the floppy to backup gnucash data which I didn't use yesterday.  Now I'm back to kernel 3.9.9-1 with no problems.
Comment 3 bgc1954 2013-09-23 18:08:46 UTC
I originally put a bug report on Arch Linux flyspray #36485 and they told me to report upstream which is what I did.  Someone asked if anything changed with kernel 3.11.1 so I tried installing and I rebooted.  Floppy mount, umount seemed fine until I shut down my machine and xtarted it up the next day.  Then I started receiving the same hang with the same dmesg output.  This seems to be exactly the same as comment 1 and 2.  It seems to be ok for one reboot and then something happens after the next startup or reboot.
Comment 4 spooober 2014-01-15 04:48:02 UTC
I just want to comment that I seem to be having the same problem. I'm using kernel 3.12.0 and when I umount a floppy umount hangs.  I'm going to try installing a newer kernel and maybe compiling floppy support as a module if that's alowed

my dmesg output:
[   75.463035] EXT4-fs (fd0): warning: mounting unchecked fs, running e2fsck is recommended
[   75.709692] EXT4-fs (fd0): mounted filesystem without journal. Opts: (null)
[  480.236033] INFO: task umount:3555 blocked for more than 120 seconds.
[  480.236037]       Tainted: P           O 3.12.0 #2
[  480.236038] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  480.236039] umount          D ffff8802356c46c8     0  3555   3406 0x00000000
[  480.236042]  ffff8802356c43c0 0000000000000082 0000000000000002 ffff8802369dab20
[  480.236045]  ffff88023e92ffd8 ffff88023e92ffd8 ffff88023e92ffd8 ffff8802356c43c0
[  480.236047]  0000000000000001 ffff8802356c43c0 ffff88023fd931c0 ffff88023e92fd00
[  480.236049] Call Trace:
[  480.236055]  [<ffffffff810ef9d0>] ? __lock_page+0x70/0x70
[  480.236058]  [<ffffffff814b4ec8>] ? io_schedule+0x88/0xd0
[  480.236060]  [<ffffffff810ef9d9>] ? sleep_on_page+0x9/0x10
[  480.236062]  [<ffffffff814b27c2>] ? __wait_on_bit+0x52/0x80
[  480.236065]  [<ffffffff810f00fb>] ? find_get_pages_tag+0xcb/0x180
[  480.236067]  [<ffffffff810efaf3>] ? wait_on_page_bit+0x73/0x80
[  480.236069]  [<ffffffff810628c0>] ? wake_atomic_t_function+0x30/0x30
[  480.236071]  [<ffffffff810efbd8>] ? filemap_fdatawait_range+0xd8/0x150
[  480.236074]  [<ffffffff814b4750>] ? wait_for_completion+0xb0/0x120
[  480.236076]  [<ffffffff810706c0>] ? try_to_wake_up+0x270/0x270
[  480.236079]  [<ffffffff81164389>] ? sync_inodes_sb+0x139/0x1a0
[  480.236082]  [<ffffffff8116a483>] ? sync_filesystem+0x53/0xa0
[  480.236084]  [<ffffffff8113f1c4>] ? generic_shutdown_super+0x34/0x100
[  480.236086]  [<ffffffff8113f2b8>] ? kill_block_super+0x28/0x80
[  480.236089]  [<ffffffff8113fb43>] ? deactivate_locked_super+0x43/0x70
[  480.236091]  [<ffffffff8115b651>] ? SyS_umount+0xa1/0x3c0
[  480.236093]  [<ffffffff814b6a26>] ? system_call_fastpath+0x1a/0x1f

Note You need to log in before you can comment on or make changes to this bug.