* Problems with hard irq? (inconsistent lock state)
@ 2006-09-29 11:45 Axel C. Voigt
2006-10-04 5:43 ` Jarek Poplawski
0 siblings, 1 reply; 7+ messages in thread
From: Axel C. Voigt @ 2006-09-29 11:45 UTC (permalink / raw)
To: linux-kernel
Hello all,
today I received the following stack backtrace using debian-2.6.18 with our communications driver accessing a ACM device. This happened when removing (powering off and/or on) the mobile phone (nokia 6630) at /dev/ttyACMx
Are someone able to get a hint for me?
--schnipp--
Sep 29 13:29:53 mcs70 kernel:
Sep 29 13:29:53 mcs70 kernel: =================================
Sep 29 13:29:53 mcs70 kernel: [ INFO: inconsistent lock state ]
Sep 29 13:29:53 mcs70 kernel: ---------------------------------
Sep 29 13:29:53 mcs70 kernel: inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
Sep 29 13:29:53 mcs70 kernel: startDV24/3864 [HC1[1]:SC0[0]:HE0:SE1] takes:
Sep 29 13:29:53 mcs70 kernel: (&acm->read_lock){++..}, at: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
Sep 29 13:29:53 mcs70 kernel: {hardirq-on-W} state was registered at:
Sep 29 13:29:53 mcs70 kernel: [<c01321f3>] lock_acquire+0x56/0x73
Sep 29 13:29:53 mcs70 kernel: [<c02ce42a>] _spin_lock+0x1a/0x25
Sep 29 13:29:53 mcs70 kernel: [<e08953a8>] acm_rx_tasklet+0x52/0x2be [cdc_acm]
Sep 29 13:29:53 mcs70 kernel: [<c011f868>] tasklet_action+0x6d/0xd7
Sep 29 13:29:53 mcs70 kernel: [<c011f526>] __do_softirq+0x79/0xf2
Sep 29 13:29:53 mcs70 kernel: [<c011f5dc>] do_softirq+0x3d/0x5c
Sep 29 13:29:53 mcs70 kernel: [<c011faa7>] ksoftirqd+0x77/0xc7
Sep 29 13:29:53 mcs70 kernel: [<c012bee6>] kthread+0x84/0xad
Sep 29 13:29:53 mcs70 kernel: [<c0100e95>] kernel_thread_helper+0x5/0xb
Sep 29 13:29:53 mcs70 kernel: irq event stamp: 38544
Sep 29 13:29:53 mcs70 kernel: hardirqs last enabled at (38543): [<c02ce51e>] _spin_unlock_irqrestore+0x3a/0x43
Sep 29 13:29:53 mcs70 kernel: hardirqs last disabled at (38544): [<c010332f>] common_interrupt+0x1b/0x2c
Sep 29 13:29:53 mcs70 kernel: softirqs last enabled at (38488): [<e08c43ee>] unix_release_sock+0x61/0x1e3 [unix]
Sep 29 13:29:53 mcs70 kernel: softirqs last disabled at (38486): [<c02ce3f1>] _write_lock_bh+0xb/0x2a
Sep 29 13:29:53 mcs70 kernel:
Sep 29 13:29:53 mcs70 kernel: other info that might help us debug this:
Sep 29 13:29:53 mcs70 kernel: no locks held by startDV24/3864.
Sep 29 13:29:53 mcs70 kernel:
Sep 29 13:29:53 mcs70 kernel: stack backtrace:
Sep 29 13:29:53 mcs70 kernel: [<c010376e>] show_trace+0x16/0x18
Sep 29 13:29:53 mcs70 kernel: [<c010383c>] dump_stack+0x19/0x1b
Sep 29 13:29:53 mcs70 kernel: [<c0130ad8>] print_usage_bug+0x1e1/0x1eb
Sep 29 13:29:53 mcs70 kernel: [<c0130b86>] mark_lock+0xa4/0x4d9
Sep 29 13:29:53 mcs70 kernel: [<c0131823>] __lock_acquire+0x41a/0x841
Sep 29 13:29:53 mcs70 kernel: [<c01321f3>] lock_acquire+0x56/0x73
Sep 29 13:29:53 mcs70 kernel: [<c02ce42a>] _spin_lock+0x1a/0x25
Sep 29 13:29:53 mcs70 kernel: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
Sep 29 13:29:53 mcs70 kernel: [<e0831e6a>] usb_hcd_giveback_urb+0x2b/0x5b [usbcore]
Sep 29 13:29:53 mcs70 kernel: [<e08ab055>] ehci_urb_done+0x7c/0x8b [ehci_hcd]
Sep 29 13:29:53 mcs70 kernel: [<e08ab0ed>] qh_completions+0x89/0x29b [ehci_hcd]
Sep 29 13:29:53 mcs70 kernel: [<e08abd9c>] scan_async+0x7f/0x11f [ehci_hcd]
Sep 29 13:29:53 mcs70 kernel: [<e08adf88>] ehci_work+0x38/0xa0 [ehci_hcd]
Sep 29 13:29:53 mcs70 kernel: [<e08ae441>] ehci_irq+0x145/0x15c [ehci_hcd]
Sep 29 13:29:53 mcs70 kernel: [<e0831ec5>] usb_hcd_irq+0x2b/0x58 [usbcore]
Sep 29 13:29:53 mcs70 kernel: [<c013c08e>] handle_IRQ_event+0x18/0x4a
Sep 29 13:29:53 mcs70 kernel: [<c013c152>] __do_IRQ+0x92/0xf0
Sep 29 13:29:53 mcs70 kernel: [<c0104b26>] do_IRQ+0x4e/0x5f
Sep 29 13:29:53 mcs70 kernel: [<c0103339>] common_interrupt+0x25/0x2c
Sep 29 13:29:53 mcs70 DV24[3925]: DV24: close: dev->openCnt--
Sep 29 13:29:53 mcs70 kernel: [<c012c144>] add_wait_queue+0x30/0x35
Sep 29 13:29:53 mcs70 kernel: [<c01694fc>] __pollwait+0x47/0x4e
Sep 29 13:29:53 mcs70 kernel: [<c0164260>] pipe_poll+0x26/0x91
Sep 29 13:29:53 mcs70 kernel: [<c0169794>] do_select+0x1bf/0x385
Sep 29 13:29:53 mcs70 kernel: [<c0169b6c>] core_sys_select+0x212/0x307
Sep 29 13:29:53 mcs70 kernel: [<c0169cf1>] sys_select+0x90/0x13f
Sep 29 13:29:53 mcs70 kernel: [<c0102897>] syscall_call+0x7/0xb
Sep 29 13:29:53 mcs70 DV24[3925]: DV24: close: /dev/mobile1: openCnt: 0
Sep 29 13:29:53 mcs70 DV24[3925]: DV24: executing tcsetattr(dev->fd: 6, TCSADRAIN, &dev->oldtio)
Sep 29 13:29:53 mcs70 DV24[3925]: DV24: executing close(dev->fd): 6
Sep 29 13:29:53 mcs70 DV24[3925]: DV24: close: write(gControl.toAdm, &cmd, sizeof(cmd)
Sep 29 13:29:53 mcs70 DV24[3925]: DV24: close: gControl.openCnt--
Sep 29 13:29:53 mcs70 DV24[3925]: DV24: close: OPEN_CNT_ALL_CLOSE == gControl.openCnt
Sep 29 13:29:53 mcs70 DV24[3925]: DV24: close: returning CMS_OK
--schnapp--
Mit freundlichen Grüssen/ with kind regards,
Axel C. Voigt
--
Qosmotec Software Solutions GmbH QQQQ tel: +49.241.879 75 13
Schloss-Rahe-Strasse 15 Q _\ Q mob: +49.163.879 75 13
52072 Aachen Q \ \Q fax: +49.241.879 75 15
Germany QQ\_\ http://www.qosmotec.com
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Problems with hard irq? (inconsistent lock state)
2006-09-29 11:45 Problems with hard irq? (inconsistent lock state) Axel C. Voigt
@ 2006-10-04 5:43 ` Jarek Poplawski
2006-10-04 5:43 ` Greg KH
0 siblings, 1 reply; 7+ messages in thread
From: Jarek Poplawski @ 2006-10-04 5:43 UTC (permalink / raw)
To: Axel C. Voigt; +Cc: linux-kernel, David Kubicek, Greg Kroah-Hartman
On 29-09-2006 13:45, Axel C. Voigt wrote:
>
> Hello all,
>
> today I received the following stack backtrace using debian-2.6.18 with our communications driver accessing a ACM device. This happened when removing (powering off and/or on) the mobile phone (nokia 6630) at /dev/ttyACMx
>
> Are someone able to get a hint for me?
>
> --schnipp--
> Sep 29 13:29:53 mcs70 kernel:
> Sep 29 13:29:53 mcs70 kernel: =================================
> Sep 29 13:29:53 mcs70 kernel: [ INFO: inconsistent lock state ]
> Sep 29 13:29:53 mcs70 kernel: ---------------------------------
> Sep 29 13:29:53 mcs70 kernel: inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
> Sep 29 13:29:53 mcs70 kernel: startDV24/3864 [HC1[1]:SC0[0]:HE0:SE1] takes:
> Sep 29 13:29:53 mcs70 kernel: (&acm->read_lock){++..}, at: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
> Sep 29 13:29:53 mcs70 kernel: {hardirq-on-W} state was registered at:
> Sep 29 13:29:53 mcs70 kernel: [<c01321f3>] lock_acquire+0x56/0x73
> Sep 29 13:29:53 mcs70 kernel: [<c02ce42a>] _spin_lock+0x1a/0x25
> Sep 29 13:29:53 mcs70 kernel: [<e08953a8>] acm_rx_tasklet+0x52/0x2be [cdc_acm]
> Sep 29 13:29:53 mcs70 kernel: [<c011f868>] tasklet_action+0x6d/0xd7
> Sep 29 13:29:53 mcs70 kernel: [<c011f526>] __do_softirq+0x79/0xf2
...
> Sep 29 13:29:53 mcs70 kernel: stack backtrace:
> Sep 29 13:29:53 mcs70 kernel: [<c010376e>] show_trace+0x16/0x18
> Sep 29 13:29:53 mcs70 kernel: [<c010383c>] dump_stack+0x19/0x1b
> Sep 29 13:29:53 mcs70 kernel: [<c0130ad8>] print_usage_bug+0x1e1/0x1eb
> Sep 29 13:29:53 mcs70 kernel: [<c0130b86>] mark_lock+0xa4/0x4d9
> Sep 29 13:29:53 mcs70 kernel: [<c0131823>] __lock_acquire+0x41a/0x841
> Sep 29 13:29:53 mcs70 kernel: [<c01321f3>] lock_acquire+0x56/0x73
> Sep 29 13:29:53 mcs70 kernel: [<c02ce42a>] _spin_lock+0x1a/0x25
> Sep 29 13:29:53 mcs70 kernel: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
...
> Sep 29 13:29:53 mcs70 kernel: [<c0104b26>] do_IRQ+0x4e/0x5f
> Sep 29 13:29:53 mcs70 kernel: [<c0103339>] common_interrupt+0x25/0x2c
...
It looks in drivers/usb/class/cdc-acm.c acm_rx_tasklet could be preempted
with acm->read_lock by acm_read_bulk which uses the same lock from hardirq
context.
So probably spin_lock_irqsave is needed.
> Mit freundlichen Grüssen/ with kind regards,
>
> Axel C. Voigt
Serdecznie pozdrawiam/ cheers,
Jarek P.
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Problems with hard irq? (inconsistent lock state)
2006-10-04 5:43 ` Jarek Poplawski
@ 2006-10-04 5:43 ` Greg KH
2006-10-04 6:45 ` Jarek Poplawski
0 siblings, 1 reply; 7+ messages in thread
From: Greg KH @ 2006-10-04 5:43 UTC (permalink / raw)
To: Jarek Poplawski, Axel C. Voigt, linux-kernel, David Kubicek
On Wed, Oct 04, 2006 at 07:43:08AM +0200, Jarek Poplawski wrote:
> On 29-09-2006 13:45, Axel C. Voigt wrote:
> >
> > Hello all,
> >
> > today I received the following stack backtrace using debian-2.6.18 with our communications driver accessing a ACM device. This happened when removing (powering off and/or on) the mobile phone (nokia 6630) at /dev/ttyACMx
> >
> > Are someone able to get a hint for me?
> >
> > --schnipp--
> > Sep 29 13:29:53 mcs70 kernel:
> > Sep 29 13:29:53 mcs70 kernel: =================================
> > Sep 29 13:29:53 mcs70 kernel: [ INFO: inconsistent lock state ]
> > Sep 29 13:29:53 mcs70 kernel: ---------------------------------
> > Sep 29 13:29:53 mcs70 kernel: inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
> > Sep 29 13:29:53 mcs70 kernel: startDV24/3864 [HC1[1]:SC0[0]:HE0:SE1] takes:
> > Sep 29 13:29:53 mcs70 kernel: (&acm->read_lock){++..}, at: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
> > Sep 29 13:29:53 mcs70 kernel: {hardirq-on-W} state was registered at:
> > Sep 29 13:29:53 mcs70 kernel: [<c01321f3>] lock_acquire+0x56/0x73
> > Sep 29 13:29:53 mcs70 kernel: [<c02ce42a>] _spin_lock+0x1a/0x25
> > Sep 29 13:29:53 mcs70 kernel: [<e08953a8>] acm_rx_tasklet+0x52/0x2be [cdc_acm]
> > Sep 29 13:29:53 mcs70 kernel: [<c011f868>] tasklet_action+0x6d/0xd7
> > Sep 29 13:29:53 mcs70 kernel: [<c011f526>] __do_softirq+0x79/0xf2
> ...
> > Sep 29 13:29:53 mcs70 kernel: stack backtrace:
> > Sep 29 13:29:53 mcs70 kernel: [<c010376e>] show_trace+0x16/0x18
> > Sep 29 13:29:53 mcs70 kernel: [<c010383c>] dump_stack+0x19/0x1b
> > Sep 29 13:29:53 mcs70 kernel: [<c0130ad8>] print_usage_bug+0x1e1/0x1eb
> > Sep 29 13:29:53 mcs70 kernel: [<c0130b86>] mark_lock+0xa4/0x4d9
> > Sep 29 13:29:53 mcs70 kernel: [<c0131823>] __lock_acquire+0x41a/0x841
> > Sep 29 13:29:53 mcs70 kernel: [<c01321f3>] lock_acquire+0x56/0x73
> > Sep 29 13:29:53 mcs70 kernel: [<c02ce42a>] _spin_lock+0x1a/0x25
> > Sep 29 13:29:53 mcs70 kernel: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
> ...
> > Sep 29 13:29:53 mcs70 kernel: [<c0104b26>] do_IRQ+0x4e/0x5f
> > Sep 29 13:29:53 mcs70 kernel: [<c0103339>] common_interrupt+0x25/0x2c
> ...
>
> It looks in drivers/usb/class/cdc-acm.c acm_rx_tasklet could be preempted
> with acm->read_lock by acm_read_bulk which uses the same lock from hardirq
> context.
>
> So probably spin_lock_irqsave is needed.
Yup. Care to send a patch?
thanks,
greg k-h
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Problems with hard irq? (inconsistent lock state)
2006-10-04 5:43 ` Greg KH
@ 2006-10-04 6:45 ` Jarek Poplawski
2006-10-05 18:05 ` Greg KH
0 siblings, 1 reply; 7+ messages in thread
From: Jarek Poplawski @ 2006-10-04 6:45 UTC (permalink / raw)
To: Greg KH; +Cc: Axel C. Voigt, linux-kernel, David Kubicek
On Tue, Oct 03, 2006 at 10:43:09PM -0700, Greg KH wrote:
> On Wed, Oct 04, 2006 at 07:43:08AM +0200, Jarek Poplawski wrote:
> > On 29-09-2006 13:45, Axel C. Voigt wrote:
...
> > > Sep 29 13:29:53 mcs70 kernel: =================================
> > > Sep 29 13:29:53 mcs70 kernel: [ INFO: inconsistent lock state ]
> > > Sep 29 13:29:53 mcs70 kernel: ---------------------------------
> > > Sep 29 13:29:53 mcs70 kernel: inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
> > > Sep 29 13:29:53 mcs70 kernel: startDV24/3864 [HC1[1]:SC0[0]:HE0:SE1] takes:
> > > Sep 29 13:29:53 mcs70 kernel: (&acm->read_lock){++..}, at: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
> > > Sep 29 13:29:53 mcs70 kernel: {hardirq-on-W} state was registered at:
...
> > It looks in drivers/usb/class/cdc-acm.c acm_rx_tasklet could be preempted
> > with acm->read_lock by acm_read_bulk which uses the same lock from hardirq
> > context.
> >
> > So probably spin_lock_irqsave is needed.
>
> Yup. Care to send a patch?
If it could help?
Jarek P.
diff -Nurp linux-2.6-18-git20-/drivers/usb/class/cdc-acm.c linux-2.6-18-git20/drivers/usb/class/cdc-acm.c
--- linux-2.6-18-git20-/drivers/usb/class/cdc-acm.c 2006-10-04 07:59:46.000000000 +0200
+++ linux-2.6-18-git20/drivers/usb/class/cdc-acm.c 2006-10-04 08:16:03.000000000 +0200
@@ -325,7 +325,7 @@ static void acm_rx_tasklet(unsigned long
struct acm_rb *buf;
struct tty_struct *tty = acm->tty;
struct acm_ru *rcv;
- //unsigned long flags;
+ unsigned long flags;
int i = 0;
dbg("Entering acm_rx_tasklet");
@@ -333,15 +333,15 @@ static void acm_rx_tasklet(unsigned long
return;
next_buffer:
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
if (list_empty(&acm->filled_read_bufs)) {
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
goto urbs;
}
buf = list_entry(acm->filled_read_bufs.next,
struct acm_rb, list);
list_del(&buf->list);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
dbg("acm_rx_tasklet: procesing buf 0x%p, size = %d", buf, buf->size);
@@ -356,29 +356,29 @@ next_buffer:
memmove(buf->base, buf->base + i, buf->size - i);
buf->size -= i;
spin_unlock(&acm->throttle_lock);
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
list_add(&buf->list, &acm->filled_read_bufs);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
return;
}
spin_unlock(&acm->throttle_lock);
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
list_add(&buf->list, &acm->spare_read_bufs);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
goto next_buffer;
urbs:
while (!list_empty(&acm->spare_read_bufs)) {
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
if (list_empty(&acm->spare_read_urbs)) {
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
return;
}
rcv = list_entry(acm->spare_read_urbs.next,
struct acm_ru, list);
list_del(&rcv->list);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
buf = list_entry(acm->spare_read_bufs.next,
struct acm_rb, list);
@@ -400,9 +400,9 @@ urbs:
free-urbs-pool and resubmited ASAP */
if (usb_submit_urb(rcv->urb, GFP_ATOMIC) < 0) {
list_add(&buf->list, &acm->spare_read_bufs);
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
list_add(&rcv->list, &acm->spare_read_urbs);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
return;
}
}
^ permalink raw reply [flat|nested] 7+ messages in thread
* RE: Problems with hard irq? (inconsistent lock state)
@ 2006-10-04 20:58 Axel C. Voigt
2006-10-05 8:39 ` Jarek Poplawski
0 siblings, 1 reply; 7+ messages in thread
From: Axel C. Voigt @ 2006-10-04 20:58 UTC (permalink / raw)
To: Jarek Poplawski, Greg KH; +Cc: linux-kernel, David Kubicek
Hello,
problem seems solved, testing extensively does not show the described behavior any more. What exactly does the patch using _irqsave functions instead of their counterparts? Is this worth to make public patch or just a quick hack to solve this particular problem?
Mit freundlichen Grüssen/ with kind regards,
Axel C. Voigt
--
Business Development
Qosmotec Software Solutions GmbH QQQQ tel: +49.241.879 75 13
Schloss-Rahe-Strasse 15 Q _\ Q mob: +49.163.879 75 13
52072 Aachen Q \ \Q fax: +49.241.879 75 15
Germany QQ\_\ http://www.qosmotec.com
-----Original Message-----
From: Jarek Poplawski [mailto:jarkao2@o2.pl]
Sent: Wednesday, October 04, 2006 8:46 AM
To: Greg KH
Cc: Axel C. Voigt; linux-kernel@vger.kernel.org; David Kubicek
Subject: Re: Problems with hard irq? (inconsistent lock state)
On Tue, Oct 03, 2006 at 10:43:09PM -0700, Greg KH wrote:
> On Wed, Oct 04, 2006 at 07:43:08AM +0200, Jarek Poplawski wrote:
> > On 29-09-2006 13:45, Axel C. Voigt wrote:
...
> > > Sep 29 13:29:53 mcs70 kernel: =================================
> > > Sep 29 13:29:53 mcs70 kernel: [ INFO: inconsistent lock state ]
> > > Sep 29 13:29:53 mcs70 kernel: ---------------------------------
> > > Sep 29 13:29:53 mcs70 kernel: inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
> > > Sep 29 13:29:53 mcs70 kernel: startDV24/3864 [HC1[1]:SC0[0]:HE0:SE1] takes:
> > > Sep 29 13:29:53 mcs70 kernel: (&acm->read_lock){++..}, at:
> > > [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm] Sep 29 13:29:53 mcs70 kernel: {hardirq-on-W} state was registered at:
...
> > It looks in drivers/usb/class/cdc-acm.c acm_rx_tasklet could be
> > preempted with acm->read_lock by acm_read_bulk which uses the same
> > lock from hardirq context.
> >
> > So probably spin_lock_irqsave is needed.
>
> Yup. Care to send a patch?
If it could help?
Jarek P.
diff -Nurp linux-2.6-18-git20-/drivers/usb/class/cdc-acm.c linux-2.6-18-git20/drivers/usb/class/cdc-acm.c
--- linux-2.6-18-git20-/drivers/usb/class/cdc-acm.c 2006-10-04 07:59:46.000000000 +0200
+++ linux-2.6-18-git20/drivers/usb/class/cdc-acm.c 2006-10-04 08:16:03.000000000 +0200
@@ -325,7 +325,7 @@ static void acm_rx_tasklet(unsigned long
struct acm_rb *buf;
struct tty_struct *tty = acm->tty;
struct acm_ru *rcv;
- //unsigned long flags;
+ unsigned long flags;
int i = 0;
dbg("Entering acm_rx_tasklet");
@@ -333,15 +333,15 @@ static void acm_rx_tasklet(unsigned long
return;
next_buffer:
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
if (list_empty(&acm->filled_read_bufs)) {
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
goto urbs;
}
buf = list_entry(acm->filled_read_bufs.next,
struct acm_rb, list);
list_del(&buf->list);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
dbg("acm_rx_tasklet: procesing buf 0x%p, size = %d", buf, buf->size);
@@ -356,29 +356,29 @@ next_buffer:
memmove(buf->base, buf->base + i, buf->size - i);
buf->size -= i;
spin_unlock(&acm->throttle_lock);
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
list_add(&buf->list, &acm->filled_read_bufs);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
return;
}
spin_unlock(&acm->throttle_lock);
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
list_add(&buf->list, &acm->spare_read_bufs);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
goto next_buffer;
urbs:
while (!list_empty(&acm->spare_read_bufs)) {
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
if (list_empty(&acm->spare_read_urbs)) {
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
return;
}
rcv = list_entry(acm->spare_read_urbs.next,
struct acm_ru, list);
list_del(&rcv->list);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
buf = list_entry(acm->spare_read_bufs.next,
struct acm_rb, list);
@@ -400,9 +400,9 @@ urbs:
free-urbs-pool and resubmited ASAP */
if (usb_submit_urb(rcv->urb, GFP_ATOMIC) < 0) {
list_add(&buf->list, &acm->spare_read_bufs);
- spin_lock(&acm->read_lock);
+ spin_lock_irqsave(&acm->read_lock, flags);
list_add(&rcv->list, &acm->spare_read_urbs);
- spin_unlock(&acm->read_lock);
+ spin_unlock_irqrestore(&acm->read_lock, flags);
return;
}
}
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Problems with hard irq? (inconsistent lock state)
2006-10-04 20:58 Axel C. Voigt
@ 2006-10-05 8:39 ` Jarek Poplawski
0 siblings, 0 replies; 7+ messages in thread
From: Jarek Poplawski @ 2006-10-05 8:39 UTC (permalink / raw)
To: Axel C. Voigt; +Cc: Greg KH, linux-kernel
On Wed, Oct 04, 2006 at 10:58:48PM +0200, Axel C. Voigt wrote:
> Hello,
>
> problem seems solved, testing extensively does not show the described behavior any more. What exactly does the patch using _irqsave functions instead of their counterparts? Is this worth to make public patch or just a quick hack to solve this particular problem?
It is a real problem but you have to have some "luck" to hit this.
Wish you luck,
Jarek P.
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Problems with hard irq? (inconsistent lock state)
2006-10-04 6:45 ` Jarek Poplawski
@ 2006-10-05 18:05 ` Greg KH
0 siblings, 0 replies; 7+ messages in thread
From: Greg KH @ 2006-10-05 18:05 UTC (permalink / raw)
To: Jarek Poplawski, Axel C. Voigt, linux-kernel, David Kubicek
On Wed, Oct 04, 2006 at 08:45:43AM +0200, Jarek Poplawski wrote:
> On Tue, Oct 03, 2006 at 10:43:09PM -0700, Greg KH wrote:
> > On Wed, Oct 04, 2006 at 07:43:08AM +0200, Jarek Poplawski wrote:
> > > On 29-09-2006 13:45, Axel C. Voigt wrote:
> ...
> > > > Sep 29 13:29:53 mcs70 kernel: =================================
> > > > Sep 29 13:29:53 mcs70 kernel: [ INFO: inconsistent lock state ]
> > > > Sep 29 13:29:53 mcs70 kernel: ---------------------------------
> > > > Sep 29 13:29:53 mcs70 kernel: inconsistent {hardirq-on-W} -> {in-hardirq-W} usage.
> > > > Sep 29 13:29:53 mcs70 kernel: startDV24/3864 [HC1[1]:SC0[0]:HE0:SE1] takes:
> > > > Sep 29 13:29:53 mcs70 kernel: (&acm->read_lock){++..}, at: [<e08952d8>] acm_read_bulk+0x60/0xde [cdc_acm]
> > > > Sep 29 13:29:53 mcs70 kernel: {hardirq-on-W} state was registered at:
> ...
> > > It looks in drivers/usb/class/cdc-acm.c acm_rx_tasklet could be preempted
> > > with acm->read_lock by acm_read_bulk which uses the same lock from hardirq
> > > context.
> > >
> > > So probably spin_lock_irqsave is needed.
> >
> > Yup. Care to send a patch?
>
> If it could help?
Great, care to resend this to me with a signed-off-by: line?
thanks,
greg k-h
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2006-10-05 18:48 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-09-29 11:45 Problems with hard irq? (inconsistent lock state) Axel C. Voigt
2006-10-04 5:43 ` Jarek Poplawski
2006-10-04 5:43 ` Greg KH
2006-10-04 6:45 ` Jarek Poplawski
2006-10-05 18:05 ` Greg KH
-- strict thread matches above, loose matches on Subject: below --
2006-10-04 20:58 Axel C. Voigt
2006-10-05 8:39 ` Jarek Poplawski
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox