WARNING: at kernel/lockdep.c:2658 check_flags()
Marcin Slusarz
Today I've got this (while i was upgrading my gentoo box):
WARNING: at kernel/lockdep.c:2658 check_flags()
Pid: 21680, comm: conftest Not tainted 2.6.24-rc6 #63
Call Trace:
[<ffffffff80253457>] check_flags+0x1c7/0x1d0
[<ffffffff80257217>] lock_acquire+0x57/0xc0
[<ffffffff8024d5c0>] __atomic_notifier_call_chain+0x60/0xd0
[<ffffffff8024d641>] atomic_notifier_call_chain+0x11/0x20
[<ffffffff8024d67e>] notify_die+0x2e/0x30
[<ffffffff8020da0a>] do_divide_error+0x5a/0xa0
[<ffffffff80522bdd>] trace_hardirqs_on_thunk+0x35/0x3a
[<ffffffff80255b89>] trace_hardirqs_on+0xd9/0x180
[<ffffffff80522bdd>] trace_hardirqs_on_thunk+0x35/0x3a
[<ffffffff80523c2d>] error_exit+0x0/0xa9
possible reason: unannotated irqs-off.
irq event stamp: 4693
hardirqs last enabled at (4693): [<ffffffff80522bdd>] trace_hardirqs_on_thunk+0x35/0x3a
hardirqs last disabled at (4692): [<ffffffff80522c17>] trace_hardirqs_off_thunk+0x35/0x37
softirqs last enabled at (3546): [<ffffffff80238343>] __do_softirq+0xb3/0xd0
softirqs last disabled at (3521): [<ffffffff8020c97c>] call_softirq+0x1c/0x30
This is 11ee29577cc1637f94d903a6ea322cf1ed8ce1b3:
Merge git://git.kernel.org/pub/scm/linux/kernel/git/bart/ide-2.6
from 25 Dec + some patches for udf (but I didn't mount it) and simple
patch for slab: http://lkml.org/lkml/2007/12/27/110
Marcin
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Peter Zijlstra
On Sat, 2007-12-29 at 18:06 +0100, Marcin Slusarz wrote:
> Hi
> Today I've got this (while i was upgrading my gentoo box):
>
> WARNING: at kernel/lockdep.c:2658 check_flags()
> Pid: 21680, comm: conftest Not tainted 2.6.24-rc6 #63
>
> Call Trace:
> [<ffffffff80253457>] check_flags+0x1c7/0x1d0
> [<ffffffff80257217>] lock_acquire+0x57/0xc0
> [<ffffffff8024d5c0>] __atomic_notifier_call_chain+0x60/0xd0
> [<ffffffff8024d641>] atomic_notifier_call_chain+0x11/0x20
> [<ffffffff8024d67e>] notify_die+0x2e/0x30
> [<ffffffff8020da0a>] do_divide_error+0x5a/0xa0
> [<ffffffff80522bdd>] trace_hardirqs_on_thunk+0x35/0x3a
> [<ffffffff80255b89>] trace_hardirqs_on+0xd9/0x180
> [<ffffffff80522bdd>] trace_hardirqs_on_thunk+0x35/0x3a
> [<ffffffff80523c2d>] error_exit+0x0/0xa9
>
> possible reason: unannotated irqs-off.
> irq event stamp: 4693
> hardirqs last enabled at (4693): [<ffffffff80522bdd>] trace_hardirqs_on_thunk+0x35/0x3a
> hardirqs last disabled at (4692): [<ffffffff80522c17>] trace_hardirqs_off_thunk+0x35/0x37
> softirqs last enabled at (3546): [<ffffffff80238343>] __do_softirq+0xb3/0xd0
> softirqs last disabled at (3521): [<ffffffff8020c97c>] call_softirq+0x1c/0x30
more early fixups for notify_die()..
Signed-off-by: Peter Zijlstra <a.p.zi...@chello.nl>
---
diff --git a/arch/x86/kernel/traps_32.c b/arch/x86/kernel/traps_32.c
index 8fc3a27..3a10cf6 100644
--- a/arch/x86/kernel/traps_32.c
+++ b/arch/x86/kernel/traps_32.c
@@ -543,6 +543,7 @@ void do_##name(struct pt_regs * regs, long error_code) \
info.si_errno = 0; \
info.si_code = sicode; \
info.si_addr = (void __user *)siaddr; \
+ trace_hardirqs_fixup(); \
if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
== NOTIFY_STOP) \
return; \
diff --git a/arch/x86/kernel/traps_64.c b/arch/x86/kernel/traps_64.c
index 778667a..5e07af1 100644
--- a/arch/x86/kernel/traps_64.c
+++ b/arch/x86/kernel/traps_64.c
@@ -635,6 +635,7 @@ asmlinkage void do_##name(struct pt_regs * regs, long error_code) \
info.si_errno = 0; \
info.si_code = sicode; \
info.si_addr = (void __user *)siaddr; \
+ trace_hardirqs_fixup(); \
if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) \
== NOTIFY_STOP) \
return; \
Ingo Molnar
* Peter Zijlstra <a.p.zi...@chello.nl> wrote:
> > softirqs last disabled at (3521): [<ffffffff8020c97c>]
> > call_softirq+0x1c/0x30
>
> more early fixups for notify_die()..
thanks - i've picked this up into sched.git, for more testing.
Ingo