there's another VFS lock that just popped up, hopefully the last one.
Fix below. (All this is still related to proc_subdir_lock, and the
original BKL bug it fixed.)
Ingo
-------------
&proc_inum_lock also nests within proc_subdir_lock, and &proc_inum_lock
is used in a softirq-unsafe manner. The lock validator noticed the
following scenario:
=====================================
[ BUG: lock inversion bug detected! ]
-------------------------------------
ifup-eth/2308 just changed the state of lock {proc_subdir_lock} at:
[<c0197083>] remove_proc_entry+0x33/0x1f0
but this lock took lock {proc_inum_lock} in the past, acquired at:
[<c0196fee>] free_proc_entry+0x2e/0x90
and interrupts could create an inverse lock dependency between them,
which could lead to deadlocks!
other info that might help in debugging this:
locks held by ifup-eth/2308:
[<c010432d>] show_trace+0xd/0x10
[<c0104347>] dump_stack+0x17/0x20
[<c0137a41>] check_no_lock_2_mask+0x131/0x180
[<c013852b>] mark_lock+0xfb/0x2a0
[<c0138b63>] debug_lock_chain+0x493/0xdc0
[<c01394cd>] debug_lock_chain_spin+0x3d/0x60
[<c026594d>] _raw_spin_lock+0x2d/0x90
[<c04d91a2>] _spin_lock_bh+0x12/0x20
[<c0197083>] remove_proc_entry+0x33/0x1f0
[<c01429e9>] unregister_handler_proc+0x19/0x20
[<c01421ab>] free_irq+0x7b/0xe0
[<c02f25b2>] floppy_release_irq_and_dma+0x1b2/0x210
[<c02f0aa7>] set_dor+0xc7/0x1b0
[<c02f3b21>] motor_off_callback+0x21/0x30
[<c01274d5>] run_timer_softirq+0xf5/0x1f0
[<c0122e27>] __do_softirq+0x97/0x130
[<c0105519>] do_softirq+0x69/0x100
=======================
the fix is to take proc_inum_lock in a softirq-safe manner.
Signed-off-by: Ingo Molnar <[email protected]>
----
fs/proc/generic.c | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)
Index: linux/fs/proc/generic.c
===================================================================
--- linux.orig/fs/proc/generic.c
+++ linux/fs/proc/generic.c
@@ -329,9 +329,9 @@ retry:
if (idr_pre_get(&proc_inum_idr, GFP_KERNEL) == 0)
return 0;
- spin_lock(&proc_inum_lock);
+ spin_lock_bh(&proc_inum_lock);
error = idr_get_new(&proc_inum_idr, NULL, &i);
- spin_unlock(&proc_inum_lock);
+ spin_unlock_bh(&proc_inum_lock);
if (error == -EAGAIN)
goto retry;
else if (error)
@@ -350,9 +350,9 @@ static void release_inode_number(unsigne
{
int id = (inum - PROC_DYNAMIC_FIRST) | ~MAX_ID_MASK;
- spin_lock(&proc_inum_lock);
+ spin_lock_bh(&proc_inum_lock);
idr_remove(&proc_inum_idr, id);
- spin_unlock(&proc_inum_lock);
+ spin_unlock_bh(&proc_inum_lock);
}
static void *proc_follow_link(struct dentry *dentry, struct nameidata *nd)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]
