Re: [linux-dvb] [BUG] flexcop lockdep
I've been using Antti's fix for a while, it seems to be working fine.
Any chance of getting it pushed into the main repository?
Cheers,
Tim.
Removing of unnecessary spinlock
Based on the discussion on linux-dvb it seems that ISR:s are already
serialized by the linux kernel and thus drivers won't have to specify
their own facilities for ISR serialization.
Signed-off-by: Antti Seppälä [EMAIL PROTECTED]
--
Antti Seppälä
--- linux/drivers/media/dvb/b2c2/flexcop-pci.c~ 2007-04-07 09:44:50.0
+0300
+++ linux/drivers/media/dvb/b2c2/flexcop-pci.c 2007-04-07 09:43:27.0
+0300
@@ -59,8 +59,6 @@
u32 last_dma1_cur_pos; /* position of the pointer last time the
timer/packet irq occured */
int count;
- spinlock_t irq_lock;
-
unsigned long last_irq;
#if LINUX_VERSION_CODE KERNEL_VERSION(2,6,20)
@@ -143,12 +141,9 @@
{
struct flexcop_pci *fc_pci = dev_id;
struct flexcop_device *fc = fc_pci-fc_dev;
- unsigned long flags;
flexcop_ibi_value v;
irqreturn_t ret = IRQ_HANDLED;
- spin_lock_irqsave(fc_pci-irq_lock,flags);
-
v = fc-read_ibi_reg(fc,irq_20c);
/* errors */
@@ -211,8 +206,6 @@
ret = IRQ_NONE;
}
- spin_unlock_irqrestore(fc_pci-irq_lock,flags);
-
return ret;
}
@@ -310,7 +303,6 @@
}
pci_set_drvdata(fc_pci-pdev, fc_pci);
- spin_lock_init(fc_pci-irq_lock);
if ((ret = request_irq(fc_pci-pdev-irq, flexcop_pci_isr,
IRQF_SHARED, DRIVER_NAME, fc_pci)) != 0)
goto err_pci_iounmap;
___
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
___
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
mp3geek wrote: safe to use this directly on the 2.6.20.x source? The fix is valid but the patch I posted won't apply to 2.6.20. Use this one instead. -- Antti Seppälä --- linux-2.6.20/drivers/media/dvb/b2c2/flexcop-pci.c~ 2007-02-04 20:44:54.0 +0200 +++ linux-2.6.20/drivers/media/dvb/b2c2/flexcop-pci.c 2007-04-04 22:10:25.0 +0300 @@ -59,8 +59,6 @@ u32 last_dma1_cur_pos; /* position of the pointer last time the timer/packet irq occured */ int count; - spinlock_t irq_lock; - unsigned long last_irq; struct delayed_work irq_check_work; @@ -130,8 +128,6 @@ flexcop_ibi_value v; irqreturn_t ret = IRQ_HANDLED; - spin_lock_irq(fc_pci-irq_lock); - v = fc-read_ibi_reg(fc,irq_20c); /* errors */ @@ -194,8 +190,6 @@ ret = IRQ_NONE; } - spin_unlock_irq(fc_pci-irq_lock); - return ret; } @@ -298,8 +292,6 @@ IRQF_SHARED, DRIVER_NAME, fc_pci)) != 0) goto err_pci_iounmap; - spin_lock_init(fc_pci-irq_lock); - fc_pci-init_state |= FC_PCI_INIT; return ret; ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Antti Seppälä wrote: Removing of unnecessary spinlock Based on the discussion on linux-dvb it seems that ISR:s are already serialized by the linux kernel and thus drivers won't have to specify their own facilities for ISR serialization. There is a good summary of concurrency and ISRs on page 268 Linux Device Drivers 3rd Edition: http://lwn.net/images/pdf/LDD3/ch10.pdf Thanks for the debugging! Cheers, Tim. ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Tim Small wrote: There is a good summary of concurrency and ISRs on page 268 Linux Device Drivers 3rd Edition: http://lwn.net/images/pdf/LDD3/ch10.pdf Sorry, I should also have referenced this: http://lwn.net/Kernel/LDD3/ Cheers, Tim. ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Patrick Boettcher wrote:
On Fri, 6 Apr 2007, Antti Seppälä wrote:
Patrick Boettcher wrote:
Someone outthere with a SMP-system and a flexcop who could see if e.g.
dvbscan or dvbtraffic is killing the system/flexcop when removing the
irq_lock?
Patrick.
I've now been torturing my SMP box for over three hours with various dvb
-related activities (dvbtraffic, dvbscan, vdr) without the irq_lock in flexcop
driver. So far I haven't noticed any problems in operation whatsoever.
Maybe the spinlock really should go?
I vote for it, too.
If someone will have a problem with the flexcop we will see during
development phase of the kernel after integration.
regards,
Patrick.
I've created a patch for removing the spinlock from latest v4l-dvb -tree:
Removing of unnecessary spinlock
Based on the discussion on linux-dvb it seems that ISR:s are already
serialized by the linux kernel and thus drivers won't have to specify
their own facilities for ISR serialization.
Signed-off-by: Antti Seppälä [EMAIL PROTECTED]
--
Antti Seppälä
--- linux/drivers/media/dvb/b2c2/flexcop-pci.c~ 2007-04-07 09:44:50.0 +0300
+++ linux/drivers/media/dvb/b2c2/flexcop-pci.c 2007-04-07 09:43:27.0 +0300
@@ -59,8 +59,6 @@
u32 last_dma1_cur_pos; /* position of the pointer last time the timer/packet irq occured */
int count;
- spinlock_t irq_lock;
-
unsigned long last_irq;
#if LINUX_VERSION_CODE KERNEL_VERSION(2,6,20)
@@ -143,12 +141,9 @@
{
struct flexcop_pci *fc_pci = dev_id;
struct flexcop_device *fc = fc_pci-fc_dev;
- unsigned long flags;
flexcop_ibi_value v;
irqreturn_t ret = IRQ_HANDLED;
- spin_lock_irqsave(fc_pci-irq_lock,flags);
-
v = fc-read_ibi_reg(fc,irq_20c);
/* errors */
@@ -211,8 +206,6 @@
ret = IRQ_NONE;
}
- spin_unlock_irqrestore(fc_pci-irq_lock,flags);
-
return ret;
}
@@ -310,7 +303,6 @@
}
pci_set_drvdata(fc_pci-pdev, fc_pci);
- spin_lock_init(fc_pci-irq_lock);
if ((ret = request_irq(fc_pci-pdev-irq, flexcop_pci_isr,
IRQF_SHARED, DRIVER_NAME, fc_pci)) != 0)
goto err_pci_iounmap;
___
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
safe to use this directly on the 2.6.20.x source? I've created a patch for removing the spinlock from latest v4l-dvb -tree: Removing of unnecessary spinlock Based on the discussion on linux-dvb it seems that ISR:s are already serialized by the linux kernel and thus drivers won't have to specify their own facilities for ISR serialization. Signed-off-by: Antti Seppälä [EMAIL PROTECTED] -- Antti Seppälä ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
On Thu, 5 Apr 2007, Trent Piepho wrote: On Thu, 5 Apr 2007, Patrick Boettcher wrote: On Thu, 5 Apr 2007, Borgi2008 wrote: Am Mittwoch, den 04.04.2007, 23:29 +0300 schrieb Antti Sepp?l?: Actually, looking at the code I cannot figure out why there has to be a spinlock in the first place. The lock is only taken in the interrupt handler which already runs in atomic context so there is no use in making the handler protected by a spinlock. Am I missing something here? I think the spinlock is unnecessary and should be removed entirely. Even on SMP systems? ISRs are only atomic on one CPU. I thought ISRs were normally atomic even on SMP systems. When an interrupt occurs, that interrupt is disabled until the ISR returns. Except in fast handlers (which flexcop is not) all interrupts aren't disabled, and so an ISR can be interrupted by a _different_ ISR, and a different ISR could be running at the same time on another CPU. But an ISR doesn't have to worry about two copies of itself running at the same time. At least, that's how I think it works. I don't see how a spin lock that is only used from the ISR could be useful, assuming the ISR is only called via an interrupt. There is no reason you couldn't call the isr function from some system call handler, but that would be an unusual thing to do. Normally an ISR does need a spinlock, to access data shared by the non-interrupt part of the driver. I wonder if there is a bug in flexcop in that nothing else uses irq_lock? You can only take a spinlock at IRQ-level and not at normal kernel level. So you cannot protect from shared data access without delegating the work to a lower level and there to use mutexes to protect. As there is only the isr which is at IRQ-level you might be right, it can be removed. Someone outthere with a SMP-system and a flexcop who could see if e.g. dvbscan or dvbtraffic is killing the system/flexcop when removing the irq_lock? Patrick. ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Patrick Boettcher wrote: Someone outthere with a SMP-system and a flexcop who could see if e.g. dvbscan or dvbtraffic is killing the system/flexcop when removing the irq_lock? Patrick. I've now been torturing my SMP box for over three hours with various dvb -related activities (dvbtraffic, dvbscan, vdr) without the irq_lock in flexcop driver. So far I haven't noticed any problems in operation whatsoever. Maybe the spinlock really should go? -- Antti Seppälä ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
On Fri, 6 Apr 2007, Antti Seppälä wrote: Patrick Boettcher wrote: Someone outthere with a SMP-system and a flexcop who could see if e.g. dvbscan or dvbtraffic is killing the system/flexcop when removing the irq_lock? Patrick. I've now been torturing my SMP box for over three hours with various dvb -related activities (dvbtraffic, dvbscan, vdr) without the irq_lock in flexcop driver. So far I haven't noticed any problems in operation whatsoever. Maybe the spinlock really should go? I vote for it, too. If someone will have a problem with the flexcop we will see during development phase of the kernel after integration. regards, Patrick. ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
On Fri, 6 Apr 2007, Patrick Boettcher wrote:
On Thu, 5 Apr 2007, Trent Piepho wrote:
On Thu, 5 Apr 2007, Patrick Boettcher wrote:
On Thu, 5 Apr 2007, Borgi2008 wrote:
Am Mittwoch, den 04.04.2007, 23:29 +0300 schrieb Antti Sepp?l?:
Actually, looking at the code I cannot figure out why there has to be
a
spinlock in the first place.
The lock is only taken in the interrupt handler which already runs in
atomic context so there is no use in making the handler protected by a
spinlock. Am I missing something here?
I think the spinlock is unnecessary and should be removed entirely.
Even on SMP systems? ISRs are only atomic on one CPU.
I thought ISRs were normally atomic even on SMP systems. When an interrupt
occurs, that interrupt is disabled until the ISR returns. Except in fast
handlers (which flexcop is not) all interrupts aren't disabled, and so an
ISR can be interrupted by a _different_ ISR, and a different ISR could be
running at the same time on another CPU. But an ISR doesn't have to worry
about two copies of itself running at the same time.
At least, that's how I think it works.
I don't see how a spin lock that is only used from the ISR could be useful,
assuming the ISR is only called via an interrupt. There is no reason you
couldn't call the isr function from some system call handler, but that
would be an unusual thing to do.
Normally an ISR does need a spinlock, to access data shared by the
non-interrupt part of the driver. I wonder if there is a bug in flexcop in
that nothing else uses irq_lock?
You can only take a spinlock at IRQ-level and not at normal kernel level.
Sure you can. There is a ton of code that does this. It would be pretty
much impossible to create SMP safe ISRs if there was no locking mechanism
that worked across interrupt-level to kernel-level.
Look in Documentation/cli-sti-removal.txt:
irq_handler (...)
{
unsigned long flags;
spin_lock_irqsave(driver_lock, flags);
driver_data.finish = 1;
driver_data.new_work = 0;
spin_unlock_irqrestore(driver_lock, flags);
}
ioctl_func (...)
{
...
spin_lock_irq(driver_lock);
...
driver_data.finish = 0;
driver_data.new_work = 2;
...
spin_unlock_irq(driver_lock);
...
}
___
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Am Mittwoch, den 04.04.2007, 23:29 +0300 schrieb Antti Seppälä: Borgi2008 wrote: Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst Actually, looking at the code I cannot figure out why there has to be a spinlock in the first place. The lock is only taken in the interrupt handler which already runs in atomic context so there is no use in making the handler protected by a spinlock. Am I missing something here? I think the spinlock is unnecessary and should be removed entirely. -- Antti Seppälä Hello, it could be that the spinlock can be completely removed. I only saw that the spinlock is called before initialization. :-) So I fixed because this crashed my system. I would like to see that the bug is fixed official so i have not to patch my kernels anymore. Best greets Hendrik Borghorst ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
On Thu, 5 Apr 2007, Borgi2008 wrote: Am Mittwoch, den 04.04.2007, 23:29 +0300 schrieb Antti Seppälä: Borgi2008 wrote: Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst Actually, looking at the code I cannot figure out why there has to be a spinlock in the first place. The lock is only taken in the interrupt handler which already runs in atomic context so there is no use in making the handler protected by a spinlock. Am I missing something here? I think the spinlock is unnecessary and should be removed entirely. Even on SMP systems? ISRs are only atomic on one CPU. Patrick.___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Patrick Boettcher wrote: On Thu, 5 Apr 2007, Borgi2008 wrote: Am Mittwoch, den 04.04.2007, 23:29 +0300 schrieb Antti Seppälä: Borgi2008 wrote: Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst Actually, looking at the code I cannot figure out why there has to be a spinlock in the first place. The lock is only taken in the interrupt handler which already runs in atomic context so there is no use in making the handler protected by a spinlock. Am I missing something here? I think the spinlock is unnecessary and should be removed entirely. Even on SMP systems? ISRs are only atomic on one CPU. Redhat has a bugzilla ticket open about this issue. Patrick, please take a look at the patch in bugzilla: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=234900 Regards, Michael Krufky ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Michael Krufky wrote: Patrick Boettcher wrote: On Thu, 5 Apr 2007, Borgi2008 wrote: Am Mittwoch, den 04.04.2007, 23:29 +0300 schrieb Antti Seppälä: Borgi2008 wrote: Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst Actually, looking at the code I cannot figure out why there has to be a spinlock in the first place. The lock is only taken in the interrupt handler which already runs in atomic context so there is no use in making the handler protected by a spinlock. Am I missing something here? I think the spinlock is unnecessary and should be removed entirely. Even on SMP systems? ISRs are only atomic on one CPU. Redhat has a bugzilla ticket open about this issue. Patrick, please take a look at the patch in bugzilla: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=234900 Actually, the bugzilla patch is also from Hendrik Borghorst ... Sorry about that... Nothing in the bugzilla report that hasnt already been said in this thread. -- Michael Krufky ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Em Qua, 2007-04-04 às 15:28 +0200, P. van Gaans escreveu: I've got a Technisat Airstar USB that doesn't lock, so I'd like to try the patch whatever it's for. How do I apply it? I've just commited Hendrik's patch at v4l-dvb tree. So, you can just do: hg clone http://linuxtv.org/hg/v4l-dvb or cd v4l-dvb;hg pull (if you have already a copy of the tree) Borgi2008 wrote: Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst Hendrik, You've forgot to sign your patch. This is not really required for such very simple fixes, but it would be nice if you can provide your SOB for me to add on it, before I submit it to mainstream. Cheers, Mauro ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Patrick Boettcher wrote: On Thu, 5 Apr 2007, Borgi2008 wrote: Am Mittwoch, den 04.04.2007, 23:29 +0300 schrieb Antti Seppälä: Borgi2008 wrote: Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst Actually, looking at the code I cannot figure out why there has to be a spinlock in the first place. The lock is only taken in the interrupt handler which already runs in atomic context so there is no use in making the handler protected by a spinlock. Am I missing something here? I think the spinlock is unnecessary and should be removed entirely. Even on SMP systems? ISRs are only atomic on one CPU. Patrick. Apparently I've used to thinking too much in the UP world. It seems that flexcop interrupts are not acked with a special register write so an interrupt can then occur even while it is being processed on another CPU.(?) In that case the patch from Hendrik is correct. -- Antti Seppälä ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
On Thu, 5 Apr 2007, Patrick Boettcher wrote: On Thu, 5 Apr 2007, Borgi2008 wrote: Am Mittwoch, den 04.04.2007, 23:29 +0300 schrieb Antti Sepp?l?: Actually, looking at the code I cannot figure out why there has to be a spinlock in the first place. The lock is only taken in the interrupt handler which already runs in atomic context so there is no use in making the handler protected by a spinlock. Am I missing something here? I think the spinlock is unnecessary and should be removed entirely. Even on SMP systems? ISRs are only atomic on one CPU. I thought ISRs were normally atomic even on SMP systems. When an interrupt occurs, that interrupt is disabled until the ISR returns. Except in fast handlers (which flexcop is not) all interrupts aren't disabled, and so an ISR can be interrupted by a _different_ ISR, and a different ISR could be running at the same time on another CPU. But an ISR doesn't have to worry about two copies of itself running at the same time. At least, that's how I think it works. I don't see how a spin lock that is only used from the ISR could be useful, assuming the ISR is only called via an interrupt. There is no reason you couldn't call the isr function from some system call handler, but that would be an unusual thing to do. Normally an ISR does need a spinlock, to access data shared by the non-interrupt part of the driver. I wonder if there is a bug in flexcop in that nothing else uses irq_lock? ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst flexcop-lockdep.patch Description: application/mbox ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
I've got a Technisat Airstar USB that doesn't lock, so I'd like to try the patch whatever it's for. How do I apply it? Borgi2008 wrote: Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Borgi2008 wrote: Hello, i've created a bugfixes. Hope it could helps you. Hendrik Borghorst Actually, looking at the code I cannot figure out why there has to be a spinlock in the first place. The lock is only taken in the interrupt handler which already runs in atomic context so there is no use in making the handler protected by a spinlock. Am I missing something here? I think the spinlock is unnecessary and should be removed entirely. -- Antti Seppälä ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
Using hg from last night, with 2.6.21-rc4, I get: Mar 22 08:34:46 192.168.222.2 b2c2-flexcop: B2C2 FlexcopII/II(b)/III digital TV receiver chip loaded successfully Mar 22 08:34:52 192.168.222.2 flexcop-pci: will use the HW PID filter. Mar 22 08:34:52 192.168.222.2 flexcop-pci: card revision 2 Mar 22 08:34:52 192.168.222.2 ACPI: PCI Interrupt :04:09.0[A] - Mar 22 08:34:52 192.168.222.2 GSI 21 (level, low) - IRQ 21 Mar 22 08:34:52 192.168.222.2 INFO: trying to register non-static key. Mar 22 08:34:52 192.168.222.2 the code is fine but needs lockdep annotation. Mar 22 08:34:52 192.168.222.2 turning off the locking correctness validator. Mar 22 08:34:52 192.168.222.2 Mar 22 08:34:52 192.168.222.2 Call Trace: Mar 22 08:34:52 192.168.222.2 [8029d49d] __lock_acquire+0x162/0xbe0 Mar 22 08:34:52 192.168.222.2 [8027839e] flush_kernel_map+0x0/0x82 Mar 22 08:34:52 192.168.222.2 [884ed1a1] :b2c2_flexcop_pci:flexcop_pci_isr+0x2e/0x2fe Mar 22 08:34:52 192.168.222.2 [8029df97] lock_acquire+0x7c/0xa0 Mar 22 08:34:52 192.168.222.2 [884ed1a1] :b2c2_flexcop_pci:flexcop_pci_isr+0x2e/0x2fe Mar 22 08:34:52 192.168.222.2 [884ed173] :b2c2_flexcop_pci:flexcop_pci_isr+0x0/0x2fe Mar 22 08:34:52 192.168.222.2 [802628fd] _spin_lock_irq+0x2b/0x38 Mar 22 08:34:52 192.168.222.2 [884ed1a1] :b2c2_flexcop_pci:flexcop_pci_isr+0x2e/0x2fe Mar 22 08:34:52 192.168.222.2 [8020a5c6] kmem_cache_alloc+0x101/0x111 Mar 22 08:34:52 192.168.222.2 [884ed173] :b2c2_flexcop_pci:flexcop_pci_isr+0x0/0x2fe Mar 22 08:34:52 192.168.222.2 [802b3561] request_irq+0xce/0x11f Mar 22 08:34:52 192.168.222.2 [884ed5f6] :b2c2_flexcop_pci:flexcop_pci_probe+0x185/0x2d4 Mar 22 08:34:52 192.168.222.2 [80340278] pci_device_probe+0xd6/0x13d Mar 22 08:34:52 192.168.222.2 [8039992d] really_probe+0xce/0x162 Mar 22 08:34:52 192.168.222.2 [80399a6f] driver_probe_device+0xae/0xba Mar 22 08:34:52 192.168.222.2 [80399bac] __driver_attach+0x89/0xc9 Mar 22 08:34:52 192.168.222.2 [80399b23] __driver_attach+0x0/0xc9 Mar 22 08:34:52 192.168.222.2 [80398d47] bus_for_each_dev+0x49/0x7a Mar 22 08:34:52 192.168.222.2 [8039976b] driver_attach+0x1c/0x1e Mar 22 08:34:52 192.168.222.2 [803990c3] bus_add_driver+0x78/0x19c Mar 22 08:34:52 192.168.222.2 [80399e22] driver_register+0x85/0x89 Mar 22 08:34:52 192.168.222.2 [80340493] __pci_register_driver+0x95/0xd1 Mar 22 08:34:52 192.168.222.2 [8806701e] :b2c2_flexcop_pci:flexcop_pci_module_init+0x1e/0x20 Mar 22 08:34:52 192.168.222.2 [802a42d7] sys_init_module+0x15ca/0x1723 Mar 22 08:34:52 192.168.222.2 [8033b11c] pci_bus_read_config_byte+0x0/0x76 Mar 22 08:34:52 192.168.222.2 [8026215e] trace_hardirqs_on_thunk+0x35/0x37 Mar 22 08:34:52 192.168.222.2 [8025b11e] system_call+0x7e/0x83 Mar 22 08:34:52 192.168.222.2 The machine then reboots on watchdog. Thanks, Tim. Nico Sabbi wrote: Hi, using a very recent hg pull I can see in the logs the following message (tuning and receiving work fine) BUG: at kernel/lockdep.c:1860 trace_hardirqs_on() [c010318c] show_trace_log_lvl+0x19/0x2e [c0103285] show_trace+0x12/0x14 [c010329b] dump_stack+0x14/0x16 [c012c573] trace_hardirqs_on+0xc2/0x141 [c02b2e91] _spin_unlock_irq+0x22/0x26 [e0a75433] flexcop_pci_isr+0x294/0x29e [b2c2_flexcop_pci] [c013b557] handle_IRQ_event+0x18/0x44 [c013ca39] handle_level_irq+0x93/0xe2 [c0104176] do_IRQ+0x95/0xaf [c0102e4a] common_interrupt+0x2e/0x34 [c01019e0] cpu_idle+0x41/0x59 [c010033a] rest_init+0x1e/0x20 [c0386700] start_kernel+0x309/0x30b [] 0x0 === ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
Re: [linux-dvb] [BUG] flexcop lockdep
BTW, the non-static key message is from linux-2.6.21-rc4/kernel/lockdep.c. Tim. ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
[linux-dvb] [BUG] flexcop lockdep
Hi, using a very recent hg pull I can see in the logs the following message (tuning and receiving work fine) BUG: at kernel/lockdep.c:1860 trace_hardirqs_on() [c010318c] show_trace_log_lvl+0x19/0x2e [c0103285] show_trace+0x12/0x14 [c010329b] dump_stack+0x14/0x16 [c012c573] trace_hardirqs_on+0xc2/0x141 [c02b2e91] _spin_unlock_irq+0x22/0x26 [e0a75433] flexcop_pci_isr+0x294/0x29e [b2c2_flexcop_pci] [c013b557] handle_IRQ_event+0x18/0x44 [c013ca39] handle_level_irq+0x93/0xe2 [c0104176] do_IRQ+0x95/0xaf [c0102e4a] common_interrupt+0x2e/0x34 [c01019e0] cpu_idle+0x41/0x59 [c010033a] rest_init+0x1e/0x20 [c0386700] start_kernel+0x309/0x30b [] 0x0 === ___ linux-dvb mailing list linux-dvb@linuxtv.org http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
