+++ This bug was initially created as a clone of Bug #835019 +++ If you try to mount an empty partition, the mount command hangs using 100% of CPU and being unkillable. Stack trace points to a problem in the isofs driver. Here is a simple reproducer: size=2048 guestfish -x -v <<EOF sparse test1.img $size run mount /dev/sda / EOF Any size between 2048 and 203775 (== 199*1024 - 1) will give the same problem. Sizes under 2048 hit bug 835019. Sizes over 203775 cause the mount to return an error. Stack trace of mount, capturing using sysrq + t: [ 7.761017] mount R running task 0 116 115 0x00000000 [ 7.761017] ffff88001c8c9ac8 0000000000000082 ffff88001c899700 ffff88001c8c9fd8 [ 7.761017] ffff88001c8c9fd8 ffff88001c8c9fd8 ffff88001ee14500 ffff88001c899700 [ 7.761017] 0000000000000000 ffff88001c8c8000 0000000000000030 ffff88001dc24560 [ 7.761017] Call Trace: [ 7.761017] [<ffffffff8108671a>] __cond_resched+0x2a/0x40 [ 7.761017] [<ffffffff815ef820>] _cond_resched+0x30/0x40 [ 7.761017] [<ffffffff8111d2eb>] find_lock_page+0x3b/0x80 [ 7.761017] [<ffffffff8111d9df>] find_or_create_page+0x3f/0xb0 [ 7.761017] [<ffffffff811acf12>] __getblk+0xf2/0x2a0 [ 7.761017] [<ffffffff811ad113>] __bread+0x13/0xb0 [ 7.761017] [<ffffffff81247b30>] isofs_fill_super+0x240/0xf40 [ 7.761017] [<ffffffff8118055b>] mount_bdev+0x1cb/0x210 [ 7.761017] [<ffffffff812478f0>] ? isofs_iget+0x8a0/0x8a0 [ 7.761017] [<ffffffff8116b611>] ? __kmalloc_track_caller+0x51/0x180 [ 7.761017] [<ffffffff81246bd5>] isofs_mount+0x15/0x20 [ 7.761017] [<ffffffff81181063>] mount_fs+0x43/0x1b0 [ 7.761017] [<ffffffff8113de80>] ? __alloc_percpu+0x10/0x20 [ 7.761017] [<ffffffff81199bc7>] vfs_kern_mount+0x67/0xf0 [ 7.761017] [<ffffffff8119a6e4>] do_kern_mount+0x54/0x110 [ 7.761017] [<ffffffff8119bf4a>] do_mount+0x26a/0x840 [ 7.761017] [<ffffffff8113832b>] ? strndup_user+0x5b/0x80 [ 7.761017] [<ffffffff8119c65d>] sys_mount+0x8d/0xe0 [ 7.761017] [<ffffffff815f8ae9>] system_call_fastpath+0x16/0x1b
Thanks to Jeff Moyer who suggested the following patch: https://lkml.org/lkml/2012/6/25/306 which fixes this bug.
We'll fix these with one bug. *** This bug has been marked as a duplicate of bug 835019 ***