[patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags bug

[patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags bug

[Ingo Molnar]

Subject: lock validator: fix ns83820.c irq-flags bug
From: Ingo Molnar <mingo@elte.hu>

Barry K. Nathan reported the following lockdep warning:

[  197.343948] BUG: warning at kernel/lockdep.c:1856/trace_hardirqs_on()
[  197.345928]  [<c010329b>] show_trace_log_lvl+0x5b/0x105
[  197.346359]  [<c0103896>] show_trace+0x1b/0x20
[  197.346759]  [<c01038ed>] dump_stack+0x1f/0x24
[  197.347159]  [<c012efa2>] trace_hardirqs_on+0xfb/0x185
[  197.348873]  [<c029b009>] _spin_unlock_irq+0x24/0x2d
[  197.350620]  [<e09034e8>] do_tx_done+0x171/0x179 [ns83820]
[  197.350895]  [<e090445c>] ns83820_irq+0x149/0x20b [ns83820]
[  197.351166]  [<c013b4b8>] handle_IRQ_event+0x1d/0x52
[  197.353216]  [<c013c6c2>] handle_level_irq+0x97/0xe1
[  197.355157]  [<c01048c3>] do_IRQ+0x8b/0xac
[  197.355612]  [<c0102d9d>] common_interrupt+0x25/0x2c

this is caused because the ns83820 driver re-enables irq flags
in hardirq context.

While legal in theory, in practice it should only be done if the
hardware is really old and has some very high overhead in its ISR.
(such as PIO IDE)

For modern hardware, running ISRs with irqs enabled is discouraged,
because 1) new hardware is fast enough to not cause latency problems
2) allowing the nesting of hardware interrupts only 'spreads out'
the handling of the current ISR, causing extra cachemisses that would
otherwise not happen. Furthermore, on architectures where ISRs share
the kernel stacks, enabling interrupts in ISRs introduces a much
higher kernel-stack-nesting and thus kernel-stack-overflow risk.
3) not managing irq-flags via the _irqsave / _irqrestore variants
is dangerous: it's easy to forget whether one function nests inside
another, and irq flags might be mismanaged.

In the few cases where re-enabling interrupts in an ISR is considered
useful (and unavoidable), it has to be taught to the lock validator
explicitly (because the lock validator needs the "no ISR ever enables
hardirqs" artificial simplification to keep the IRQ/softirq locking
dependencies manageable).

This teaching is done via the explicit use local_irq_enable_in_hardirq().
On a stock kernel this maps to local_irq_enable(). If the lock validator
is enabled then this does not enable interrupts.

Now, the analysis of drivers/net/ns83820.c's irq flags use: the
irq-enabling in irq context seems intentional, but i dont think it's
justified. Furthermore, the driver suffers from problem #3 above too,
in ns83820_tx_timeout() it disables irqs via local_irq_save(), but
then it calls do_tx_done() which does a spin_unlock_irq(),
re-enabling for a function that does not expect it! While currently
this bug seems harmless (only some debug printout seems to be
affected by it), it's nevertheless something to be fixed.

So this patch makes the ns83820 ISR irq-flags-safe, and cleans up
do_tx_done() use and locking to avoid the ns83820_tx_timeout() bug.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 drivers/net/ns83820.c |   23 ++++++++++++-----------
 1 file changed, 12 insertions(+), 11 deletions(-)

Index: v/drivers/net/ns83820.c
===================================================================
--- v.orig/drivers/net/ns83820.c
+++ v/drivers/net/ns83820.c
@@ -1013,8 +1013,6 @@ static void do_tx_done(struct net_device
 	struct ns83820 *dev = PRIV(ndev);
 	u32 cmdsts, tx_done_idx, *desc;
 
-	spin_lock_irq(&dev->tx_lock);
-
 	dprintk("do_tx_done(%p)\n", ndev);
 	tx_done_idx = dev->tx_done_idx;
 	desc = dev->tx_descs + (tx_done_idx * DESC_SIZE);
@@ -1070,7 +1068,6 @@ static void do_tx_done(struct net_device
 		netif_start_queue(ndev);
 		netif_wake_queue(ndev);
 	}
-	spin_unlock_irq(&dev->tx_lock);
 }
 
 static void ns83820_cleanup_tx(struct ns83820 *dev)
@@ -1308,6 +1305,8 @@ static irqreturn_t ns83820_irq(int foo, 
 static void ns83820_do_isr(struct net_device *ndev, u32 isr)
 {
 	struct ns83820 *dev = PRIV(ndev);
+	unsigned long flags;
+
 #ifdef DEBUG
 	if (isr & ~(ISR_PHY | ISR_RXDESC | ISR_RXEARLY | ISR_RXOK | ISR_RXERR | ISR_TXIDLE | ISR_TXOK | ISR_TXDESC))
 		Dprintk("odd isr? 0x%08x\n", isr);
@@ -1322,10 +1321,10 @@ static void ns83820_do_isr(struct net_de
 	if ((ISR_RXDESC | ISR_RXOK) & isr) {
 		prefetch(dev->rx_info.next_rx_desc);
 
-		spin_lock_irq(&dev->misc_lock);
+		spin_lock_irqsave(&dev->misc_lock, flags);
 		dev->IMR_cache &= ~(ISR_RXDESC | ISR_RXOK);
 		writel(dev->IMR_cache, dev->base + IMR);
-		spin_unlock_irq(&dev->misc_lock);
+		spin_unlock_irqrestore(&dev->misc_lock, flags);
 
 		tasklet_schedule(&dev->rx_tasklet);
 		//rx_irq(ndev);
@@ -1371,16 +1370,18 @@ static void ns83820_do_isr(struct net_de
 	 * work has accumulated
 	 */
 	if ((ISR_TXDESC | ISR_TXIDLE | ISR_TXOK | ISR_TXERR) & isr) {
+		spin_lock_irqsave(&dev->tx_lock, flags);
 		do_tx_done(ndev);
+		spin_unlock_irqrestore(&dev->tx_lock, flags);
 
 		/* Disable TxOk if there are no outstanding tx packets.
 		 */
 		if ((dev->tx_done_idx == dev->tx_free_idx) &&
 		    (dev->IMR_cache & ISR_TXOK)) {
-			spin_lock_irq(&dev->misc_lock);
+			spin_lock_irqsave(&dev->misc_lock, flags);
 			dev->IMR_cache &= ~ISR_TXOK;
 			writel(dev->IMR_cache, dev->base + IMR);
-			spin_unlock_irq(&dev->misc_lock);
+			spin_unlock_irqrestore(&dev->misc_lock, flags);
 		}
 	}
 
@@ -1391,10 +1392,10 @@ static void ns83820_do_isr(struct net_de
 	 * nature are expected, we must enable TxOk.
 	 */
 	if ((ISR_TXIDLE & isr) && (dev->tx_done_idx != dev->tx_free_idx)) {
-		spin_lock_irq(&dev->misc_lock);
+		spin_lock_irqsave(&dev->misc_lock, flags);
 		dev->IMR_cache |= ISR_TXOK;
 		writel(dev->IMR_cache, dev->base + IMR);
-		spin_unlock_irq(&dev->misc_lock);
+		spin_unlock_irqrestore(&dev->misc_lock, flags);
 	}
 
 	/* MIB interrupt: one of the statistics counters is about to overflow */
@@ -1456,7 +1457,7 @@ static void ns83820_tx_timeout(struct ne
         u32 tx_done_idx, *desc;
 	unsigned long flags;
 
-	local_irq_save(flags);
+	spin_lock_irqsave(&dev->tx_lock, flags);
 
 	tx_done_idx = dev->tx_done_idx;
 	desc = dev->tx_descs + (tx_done_idx * DESC_SIZE);
@@ -1483,7 +1484,7 @@ static void ns83820_tx_timeout(struct ne
 		ndev->name,
 		tx_done_idx, dev->tx_free_idx, le32_to_cpu(desc[DESC_CMDSTS]));
 
-	local_irq_restore(flags);
+	spin_unlock_irqrestore(&dev->tx_lock, flags);
 }
 
 static void ns83820_tx_watch(unsigned long data)

Re: [patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags bug

[Keith Owens]

Ingo Molnar (on Sun, 4 Jun 2006 10:30:17 +0200) wrote:
>2) allowing the nesting of hardware interrupts only 'spreads out'
>the handling of the current ISR, causing extra cachemisses that would
>otherwise not happen. Furthermore, on architectures where ISRs share
>the kernel stacks, enabling interrupts in ISRs introduces a much
>higher kernel-stack-nesting and thus kernel-stack-overflow risk.

It is worse than you think.  A third party network driver enabled
interrupts in its irq handler.  For reasons that are still not clear,
that allowed recursive interrupts from the same device.  Unexpected I
know, because the card's ISR should still have been masked, but the
stack trace said otherwise.  When multiple packets arrived for the same
driver it drove multiple levels of kernel functions to handle them and
completely blew the kernel stack, even though it was using a separate
IRQ stack.

Re: [patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags bug

[Ingo Molnar]

* Keith Owens <kaos@ocs.com.au> wrote:

> Ingo Molnar (on Sun, 4 Jun 2006 10:30:17 +0200) wrote:
> >2) allowing the nesting of hardware interrupts only 'spreads out'
> >the handling of the current ISR, causing extra cachemisses that would
> >otherwise not happen. Furthermore, on architectures where ISRs share
> >the kernel stacks, enabling interrupts in ISRs introduces a much
> >higher kernel-stack-nesting and thus kernel-stack-overflow risk.
> 
> It is worse than you think.  A third party network driver enabled 
> interrupts in its irq handler.  For reasons that are still not clear, 
> that allowed recursive interrupts from the same device.  Unexpected I 
> know, because the card's ISR should still have been masked, but the 
> stack trace said otherwise.  When multiple packets arrived for the 
> same driver it drove multiple levels of kernel functions to handle 
> them and completely blew the kernel stack, even though it was using a 
> separate IRQ stack.

ouch!

They seem to be quite rare in practice (Barry's was the only one so 
far), but when they happen they can be very nasty. So the lock validator 
creates some gentle pressure to get these fixed ;-) For the really 
really rare cases where the irq enabling is justified (mostly old, very 
slow hardware) it can be easily annotated for lockdep.

	Ingo

Re: [patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags bug part 2

[Barry K. Nathan]

On 6/4/06, Arjan van de Ven <arjan@linux.intel.com> wrote:
> just the preempt the next email from Barry; while fixing this one I
> looked at the usage of the locks more and found another patch needed...
[snip]

Nice try, but it didn't work. ~_^

I was about to reply to the previous ns83820 patch with my dmesg, when
I saw this one. I applied this patch too, and like the previous patch,
it reports an instance of illegal lock usage. My dmesg follows.
-- 
-Barry K. Nathan <barryn@pobox.com>

[    0.000000] Linux version 2.6.17-rc5-mm3 (barryn@nserv) (gcc
version 4.1.1) #1 Sun Jun 4 02:03:43 PDT 2006
[    0.000000] BIOS-provided physical RAM map:
[    0.000000] sanitize start
[    0.000000] sanitize end
[    0.000000] copy_e820_map() start: 0000000000000000 size:
000000000009fc00 end: 000000000009fc00 type: 1
[    0.000000] copy_e820_map() type is E820_RAM
[    0.000000] add_memory_region(0000000000000000, 000000000009fc00, 1)
[    0.000000] copy_e820_map() start: 000000000009fc00 size:
0000000000000400 end: 00000000000a0000 type: 2
[    0.000000] add_memory_region(000000000009fc00, 0000000000000400, 2)
[    0.000000] copy_e820_map() start: 00000000000f0000 size:
0000000000010000 end: 0000000000100000 type: 2
[    0.000000] add_memory_region(00000000000f0000, 0000000000010000, 2)
[    0.000000] copy_e820_map() start: 0000000000100000 size:
000000001fef0000 end: 000000001fff0000 type: 1
[    0.000000] copy_e820_map() type is E820_RAM
[    0.000000] add_memory_region(0000000000100000, 000000001fef0000, 1)
[    0.000000] copy_e820_map() start: 000000001fff0000 size:
0000000000008000 end: 000000001fff8000 type: 3
[    0.000000] add_memory_region(000000001fff0000, 0000000000008000, 3)
[    0.000000] copy_e820_map() start: 000000001fff8000 size:
0000000000008000 end: 0000000020000000 type: 4
[    0.000000] add_memory_region(000000001fff8000, 0000000000008000, 4)
[    0.000000] copy_e820_map() start: 00000000ffff0000 size:
0000000000010000 end: 0000000100000000 type: 2
[    0.000000] add_memory_region(00000000ffff0000, 0000000000010000, 2)
[    0.000000]  BIOS-e820: 0000000000000000 - 000000000009fc00 (usable)
[    0.000000]  BIOS-e820: 000000000009fc00 - 00000000000a0000 (reserved)
[    0.000000]  BIOS-e820: 00000000000f0000 - 0000000000100000 (reserved)
[    0.000000]  BIOS-e820: 0000000000100000 - 000000001fff0000 (usable)
[    0.000000]  BIOS-e820: 000000001fff0000 - 000000001fff8000 (ACPI data)
[    0.000000]  BIOS-e820: 000000001fff8000 - 0000000020000000 (ACPI NVS)
[    0.000000]  BIOS-e820: 00000000ffff0000 - 0000000100000000 (reserved)
[    0.000000] 511MB LOWMEM available.
[    0.000000] On node 0 totalpages: 131056
[    0.000000]   DMA zone: 4096 pages, LIFO batch:0
[    0.000000]   Normal zone: 126960 pages, LIFO batch:31
[    0.000000] DMI 2.3 present.
[    0.000000] ACPI: RSDP (v000 AMI
) @ 0x000fa8f0
[    0.000000] ACPI: RSDT (v001 AMIINT          0x00000010 MSFT
0x00000097) @ 0x1fff0000
[    0.000000] ACPI: FADT (v001 AMIINT          0x00000010 MSFT
0x00000097) @ 0x1fff0030
[    0.000000] ACPI: DSDT (v001 AMD75X IRONGATE 0x00001000 MSFT
0x0100000b) @ 0x00000000
[    0.000000] Allocating PCI resources starting at 30000000 (gap:
20000000:dfff0000)
[    0.000000] Detected 805.687 MHz processor.
[   73.586916] Built 1 zonelists
[   73.586923] Kernel command line: BOOT_IMAGE=bzimage lapic
root=/dev/sdc1 vga=6
[   73.587430] Local APIC disabled by BIOS -- reenabling.
[   73.587438] Found and enabled local APIC!
[   73.587447] mapped APIC to ffffd000 (fee00000)
[   73.587456] Enabling fast FPU save and restore... done.
[   73.587465] Initializing CPU#0
[   73.587609] CPU 0 irqstacks, hard=c03f3000 soft=c03f2000
[   73.587619] PID hash table entries: 2048 (order: 11, 8192 bytes)
[   73.591125] Console: colour VGA+ 80x60
[   73.593702] Lock dependency validator: Copyright (c) 2006 Red Hat,
Inc., Ingo Molnar
[   73.593851] ... MAX_LOCKDEP_SUBTYPES:    8
[   73.593943] ... MAX_LOCK_DEPTH:          30
[   73.594035] ... MAX_LOCKDEP_KEYS:        2048
[   73.594124] ... TYPEHASH_SIZE:           1024
[   73.594215] ... MAX_LOCKDEP_ENTRIES:     8192
[   73.594311] ... MAX_LOCKDEP_CHAINS:      8192
[   73.594406] ... CHAINHASH_SIZE:          4096
[   73.594496]  memory used by lock dependency info: 696 kB
[   73.594592]  per task-struct memory footprint: 1080 bytes
[   73.594687] ------------------------
[   73.594782] | Locking API testsuite:
[   73.594876] ----------------------------------------------------------------------------
[   73.595029]                                  | spin |wlock |rlock
|mutex | wsem | rsem |
[   73.595181]
--------------------------------------------------------------------------
[   73.595357]                      A-A deadlock:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.597758]                  A-B-B-A deadlock:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.600104]              A-B-B-C-C-A deadlock:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.602480]              A-B-C-A-B-C deadlock:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.604863]          A-B-B-C-C-D-D-A deadlock:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.607310]          A-B-C-D-B-D-D-A deadlock:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.609796]          A-B-C-D-B-C-D-A deadlock:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.612246]                     double unlock:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.614529]                  bad unlock order:  ok  |  ok  |  ok  |
 ok  |  ok  |  ok  |
[   73.616838]
--------------------------------------------------------------------------
[   73.616984]               recursive read-lock:             |  ok  |
            |  ok  |
[   73.617956]
--------------------------------------------------------------------------
[   73.618106]                 non-nested unlock:  ok  |  ok  |  ok  |  ok  |
[   73.619665]   ------------------------------------------------------------
[   73.619757]      hard-irqs-on + irq-safe-A/12:  ok  |  ok  |  ok  |
[   73.620943]      soft-irqs-on + irq-safe-A/12:  ok  |  ok  |  ok  |
[   73.622135]      hard-irqs-on + irq-safe-A/21:  ok  |  ok  |  ok  |
[   73.623315]      soft-irqs-on + irq-safe-A/21:  ok  |  ok  |  ok  |
[   73.624505]        sirq-safe-A => hirqs-on/12:  ok  |  ok  |  ok  |
[   73.625696]        sirq-safe-A => hirqs-on/21:  ok  |  ok  |  ok  |
[   73.626879]          hard-safe-A + irqs-on/12:  ok  |  ok  |  ok  |
[   73.628085]          soft-safe-A + irqs-on/12:  ok  |  ok  |  ok  |
[   73.629273]          hard-safe-A + irqs-on/21:  ok  |  ok  |  ok  |
[   73.630454]          soft-safe-A + irqs-on/21:  ok  |  ok  |  ok  |
[   73.631635]     hard-safe-A + unsafe-B #1/123:  ok  |  ok  |  ok  |
[   73.632844]     soft-safe-A + unsafe-B #1/123:  ok  |  ok  |  ok  |
[   73.634051]     hard-safe-A + unsafe-B #1/132:  ok  |  ok  |  ok  |
[   73.635251]     soft-safe-A + unsafe-B #1/132:  ok  |  ok  |  ok  |
[   73.636451]     hard-safe-A + unsafe-B #1/213:  ok  |  ok  |  ok  |
[   73.637679]     soft-safe-A + unsafe-B #1/213:  ok  |  ok  |  ok  |
[   73.638890]     hard-safe-A + unsafe-B #1/231:  ok  |  ok  |  ok  |
[   73.640092]     soft-safe-A + unsafe-B #1/231:  ok  |  ok  |  ok  |
[   73.641297]     hard-safe-A + unsafe-B #1/312:  ok  |  ok  |  ok  |
[   73.642484]     soft-safe-A + unsafe-B #1/312:  ok  |  ok  |  ok  |
[   73.643679]     hard-safe-A + unsafe-B #1/321:  ok  |  ok  |  ok  |
[   73.644888]     soft-safe-A + unsafe-B #1/321:  ok  |  ok  |  ok  |
[   73.646088]     hard-safe-A + unsafe-B #2/123:  ok  |  ok  |  ok  |
[   73.647296]     soft-safe-A + unsafe-B #2/123:  ok  |  ok  |  ok  |
[   73.648535]     hard-safe-A + unsafe-B #2/132:  ok  |  ok  |  ok  |
[   73.649752]     soft-safe-A + unsafe-B #2/132:  ok  |  ok  |  ok  |
[   73.650965]     hard-safe-A + unsafe-B #2/213:  ok  |  ok  |  ok  |
[   73.652178]     soft-safe-A + unsafe-B #2/213:  ok  |  ok  |  ok  |
[   73.653382]     hard-safe-A + unsafe-B #2/231:  ok  |  ok  |  ok  |
[   73.654585]     soft-safe-A + unsafe-B #2/231:  ok  |  ok  |  ok  |
[   73.655794]     hard-safe-A + unsafe-B #2/312:  ok  |  ok  |  ok  |
[   73.656999]     soft-safe-A + unsafe-B #2/312:  ok  |  ok  |  ok  |
[   73.658237]     hard-safe-A + unsafe-B #2/321:  ok  |  ok  |  ok  |
[   73.659450]     soft-safe-A + unsafe-B #2/321:  ok  |  ok  |  ok  |
[   73.660660]       hard-irq lock-inversion/123:  ok  |  ok  |  ok  |
[   73.661870]       soft-irq lock-inversion/123:  ok  |  ok  |  ok  |
[   73.663088]       hard-irq lock-inversion/132:  ok  |  ok  |  ok  |
[   73.664300]       soft-irq lock-inversion/132:  ok  |  ok  |  ok  |
[   73.665519]       hard-irq lock-inversion/213:  ok  |  ok  |  ok  |
[   73.666732]       soft-irq lock-inversion/213:  ok  |  ok  |  ok  |
[   73.667974]       hard-irq lock-inversion/231:  ok  |  ok  |  ok  |
[   73.669179]       soft-irq lock-inversion/231:  ok  |  ok  |  ok  |
[   73.670386]       hard-irq lock-inversion/312:  ok  |  ok  |  ok  |
[   73.671606]       soft-irq lock-inversion/312:  ok  |  ok  |  ok  |
[   73.672830]       hard-irq lock-inversion/321:  ok  |  ok  |  ok  |
[   73.674036]       soft-irq lock-inversion/321:  ok  |  ok  |  ok  |
[   73.675254]       hard-irq read-recursion/123:  ok  |
[   73.675748]       soft-irq read-recursion/123:  ok  |
[   73.676246]       hard-irq read-recursion/132:  ok  |
[   73.676740]       soft-irq read-recursion/132:  ok  |
[   73.677234]       hard-irq read-recursion/213:  ok  |
[   73.677747]       soft-irq read-recursion/213:  ok  |
[   73.678240]       hard-irq read-recursion/231:  ok  |
[   73.678736]       soft-irq read-recursion/231:  ok  |
[   73.679229]       hard-irq read-recursion/312:  ok  |
[   73.679725]       soft-irq read-recursion/312:  ok  |
[   73.680217]       hard-irq read-recursion/321:  ok  |
[   73.680710]       soft-irq read-recursion/321:  ok  |
[   73.681207] -------------------------------------------------------
[   73.681306] Good, all 210 testcases passed! |
[   73.681399] ---------------------------------
[   73.682349] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes)
[   73.683329] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes)
[   73.731760] Memory: 514628k/524224k available (1572k kernel code,
9028k reserved, 909k data, 160k init, 0k highmem)
[   73.731928] Checking if this processor honours the WP bit even in
supervisor mode... Ok.
[   73.877388] Calibrating delay using timer specific routine..
1613.94 BogoMIPS (lpj=8069734)
[   73.877840] Mount-cache hash table entries: 512
[   73.878796] CPU: After generic identify, caps: 0183fbff c1c3fbff
00000000 00000000 00000000 00000000 00000000
[   73.878828] CPU: After vendor identify, caps: 0183fbff c1c3fbff
00000000 00000000 00000000 00000000 00000000
[   73.878860] CPU: L1 I Cache: 64K (64 bytes/line), D cache 64K (64 bytes/line)
[   73.878966] CPU: L2 Cache: 512K (64 bytes/line)
[   73.879060] CPU: After all inits, caps: 0183fbff c1c3fbff 00000000
00000420 00000000 00000000 00000000
[   73.879089] Intel machine check architecture supported.
[   73.879186] Intel machine check reporting enabled on CPU#0.
[   73.879289] Compat vDSO mapped to ffffe000.
[   73.879399] CPU: AMD Athlon(tm) Processor stepping 01
[   73.879567] Checking 'hlt' instruction... OK.
[   73.917472] SMP alternatives: switching to UP code
[   73.917563] Freeing SMP alternatives: 0k freed
[   73.926938] ACPI: setting ELCR to 0010 (from 0e00)
[   73.927753] lockdep: disabled NMI watchdog.
[   74.053321] NET: Registered protocol family 16
[   74.053729] ACPI: bus type pci registered
[   74.065724] PCI: PCI BIOS revision 2.10 entry at 0xfdaf1, last bus=1
[   74.065832] Setting up standard PCI resources
[   74.068330] spurious 8259A interrupt: IRQ7.
[   74.081547] ACPI: Subsystem revision 20060310
[   74.097165] ACPI: Interpreter enabled
[   74.097320] ACPI: Using PIC for interrupt routing
[   74.100331] ACPI: PCI Root Bridge [PCI0] (0000:00)
[   74.100466] PCI: Probing PCI hardware (bus 00)
[   74.110098] Boot video device is 0000:01:05.0
[   74.110458] ACPI: PCI Interrupt Routing Table [\_SB_.PCI0._PRT]
[   74.121205] ACPI: Power Resource [URP1] (off)
[   74.121717] ACPI: Power Resource [URP2] (off)
[   74.122187] ACPI: Power Resource [FDDP] (off)
[   74.122657] ACPI: Power Resource [LPTP] (off)
[   74.124439] ACPI: PCI Interrupt Link [LNKA] (IRQs 3 4 5 6 7 9 *10
11 12 14 15)
[   74.125981] ACPI: PCI Interrupt Link [LNKB] (IRQs 3 4 5 6 7 9 10
*11 12 14 15)
[   74.127521] ACPI: PCI Interrupt Link [LNKC] (IRQs 3 4 5 6 7 *9 10
11 12 14 15)
[   74.129033] ACPI: PCI Interrupt Link [LNKD] (IRQs 3 4 5 6 7 9 10 11
12 14 15) *0, disabled.
[   74.130089] Linux Plug and Play Support v0.97 (c) Adam Belay
[   74.130299] pnp: PnP ACPI init
[   74.136579] pnp: PnP ACPI: found 8 devices
[   74.136688] PnPBIOS: Disabled by ACPI PNP
[   74.137816] SCSI subsystem initialized
[   74.138209] PCI: Using ACPI for IRQ routing
[   74.138301] PCI: If a device doesn't work, try "pci=routeirq".  If
it helps, post a report
[   74.149421] PCI: Bridge: 0000:00:01.0
[   74.149534]   IO window: 9000-9fff
[   74.149658]   MEM window: e7d00000-efdfffff
[   74.149775]   PREFETCH window: e3b00000-e3bfffff
[   74.150036] NET: Registered protocol family 2
[   74.247378] IP route cache hash table entries: 4096 (order: 2, 16384 bytes)
[   74.248173] TCP established hash table entries: 16384 (order: 8,
1376256 bytes)
[   74.256402] TCP bind hash table entries: 8192 (order: 7, 720896 bytes)
[   74.260910] TCP: Hash tables configured (established 16384 bind 8192)
[   74.261061] TCP reno registered
[   74.263479] Machine check exception polling timer started.
[   74.269336] Initializing Cryptographic API
[   74.269474] io scheduler noop registered
[   74.269644] io scheduler cfq registered (default)
[   74.270194] ACPI: Power Button (FF) [PWRF]
[   74.270306] ACPI: Sleep Button (FF) [SLPF]
[   74.270430] ACPI: Power Button (CM) [PWRB]
[   74.271006] isapnp: Scanning for PnP cards...
[   74.628027] isapnp: No Plug & Play device found
[   74.645301] Floppy drive(s): fd0 is 1.44M
[   74.658905] FDC 0 is a post-1991 82077
[   74.661870] libata version 1.30 loaded.
[   74.662184] pata_pdc2027x 0000:00:0a.0: version 0.74
[   74.663009] ACPI: PCI Interrupt Link [LNKB] enabled at IRQ 11
[   74.663121] PCI: setting IRQ 11 as level-triggered
[   74.663131] ACPI: PCI Interrupt 0000:00:0a.0[A] -> Link [LNKB] ->
GSI 11 (level, low) -> IRQ 11
[   74.763521] pata_pdc2027x 0000:00:0a.0: PLL input clock 16813 kHz
[   74.793952] ata1: PATA max UDMA/133 cmd 0xE08097C0 ctl 0xE0809FDA
bmdma 0xE0809000 irq 11
[   74.794290] ata2: PATA max UDMA/133 cmd 0xE08095C0 ctl 0xE0809DDA
bmdma 0xE0809008 irq 11
[   74.947356] ata1.00: cfg 49:2f00 82:7c6b 83:7f09 84:4003 85:7c69
86:3e01 87:4003 88:407f
[   74.947372] ata1.00: ATA-7, max UDMA/133, 490234752 sectors: LBA48
[   74.948337] ata1.00: configured for UDMA/133
[   74.948440] scsi0 : pata_pdc2027x
[   75.107315] ata2.00: cfg 49:2f00 82:346b 83:7f01 84:4003 85:3469
86:3c01 87:4003 88:203f
[   75.107328] ata2.00: ATA-6, max UDMA/100, 625142448 sectors: LBA48
[   75.108429] ata2.00: configured for UDMA/100
[   75.108531] scsi1 : pata_pdc2027x
[   75.109758]   Vendor: ATA       Model: Maxtor 6Y250P0    Rev: YAR4
[   75.110792]   Type:   Direct-Access                      ANSI SCSI
revision: 05
[   75.112186]   Vendor: ATA       Model: WDC WD3200JB-00K  Rev: 08.0
[   75.113248]   Type:   Direct-Access                      ANSI SCSI
revision: 05
[   75.114516] sata_sil 0000:00:0d.0: version 1.0
[   75.115265] ACPI: PCI Interrupt Link [LNKA] enabled at IRQ 10
[   75.115374] PCI: setting IRQ 10 as level-triggered
[   75.115384] ACPI: PCI Interrupt 0000:00:0d.0[A] -> Link [LNKA] ->
GSI 10 (level, low) -> IRQ 10
[   75.115653] sata_sil 0000:00:0d.0: Applying R_ERR on DMA activate
FIS errata fix
[   75.116016] ata3: SATA max UDMA/100 cmd 0xE080E480 ctl 0xE080E48A
bmdma 0xE080E400 irq 10
[   75.116357] ata4: SATA max UDMA/100 cmd 0xE080E4C0 ctl 0xE080E4CA
bmdma 0xE080E408 irq 10
[   75.116738] ata5: SATA max UDMA/100 cmd 0xE080E680 ctl 0xE080E68A
bmdma 0xE080E600 irq 10
[   75.117051] ata6: SATA max UDMA/100 cmd 0xE080E6C0 ctl 0xE080E6CA
bmdma 0xE080E608 irq 10
[   75.486123] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[   75.487064] ata3.00: cfg 49:2f00 82:7c6b 83:7f09 84:4003 85:7c69
86:3e01 87:4003 88:207f
[   75.487078] ata3.00: ATA-7, max UDMA/133, 585940320 sectors: LBA48
[   75.487187] ata3.00: applying bridge limits
[   75.488298] ata3.00: configured for UDMA/100
[   75.488395] scsi2 : sata_sil
[   75.855823] ata4: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[   75.857125] ata4.00: cfg 49:2f00 82:346b 83:7d01 84:4023 85:3469
86:3c01 87:4023 88:203f
[   75.857137] ata4.00: ATA-7, max UDMA/100, 781422768 sectors: LBA48
[   75.857235] ata4.00: applying bridge limits
[   75.858607] ata4.00: configured for UDMA/100
[   75.858710] scsi3 : sata_sil
[   76.225523] ata5: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[   76.226496] ata5.00: cfg 49:2f00 82:346b 83:7f01 84:4003 85:3469
86:3e01 87:4003 88:203f
[   76.226509] ata5.00: ATA-6, max UDMA/100, 625142448 sectors: LBA48
[   76.226605] ata5.00: applying bridge limits
[   76.227685] ata5.00: configured for UDMA/100
[   76.227783] scsi4 : sata_sil
[   76.435337] ata6: SATA link down (SStatus 0 SControl 310)
[   76.435450] scsi5 : sata_sil
[   76.436397]   Vendor: ATA       Model: Maxtor 4A300J0    Rev: RAMB
[   76.437451]   Type:   Direct-Access                      ANSI SCSI
revision: 05
[   76.438841]   Vendor: ATA       Model: ST3400832A        Rev: 3.01
[   76.439911]   Type:   Direct-Access                      ANSI SCSI
revision: 05
[   76.441340]   Vendor: ATA       Model: WDC WD3200JB-00K  Rev: 08.0
[   76.442397]   Type:   Direct-Access                      ANSI SCSI
revision: 05
[   76.444277] SCSI device sda: 490234752 512-byte hdwr sectors (251000 MB)
[   76.444481] sda: Write Protect is off
[   76.444572] sda: Mode Sense: 00 3a 00 00
[   76.444744] SCSI device sda: drive cache: write back
[   76.445381] SCSI device sda: 490234752 512-byte hdwr sectors (251000 MB)
[   76.445594] sda: Write Protect is off
[   76.445688] sda: Mode Sense: 00 3a 00 00
[   76.445885] SCSI device sda: drive cache: write back
[   76.445984]  sda: unknown partition table
[   76.448576] sd 0:0:0:0: Attached scsi disk sda
[   76.449210] SCSI device sdb: 625142448 512-byte hdwr sectors (320073 MB)
[   76.449391] sdb: Write Protect is off
[   76.449478] sdb: Mode Sense: 00 3a 00 00
[   76.449645] SCSI device sdb: drive cache: write back
[   76.450141] SCSI device sdb: 625142448 512-byte hdwr sectors (320073 MB)
[   76.450332] sdb: Write Protect is off
[   76.450420] sdb: Mode Sense: 00 3a 00 00
[   76.450608] SCSI device sdb: drive cache: write back
[   76.450706]  sdb: unknown partition table
[   76.453773] sd 1:0:0:0: Attached scsi disk sdb
[   76.454378] SCSI device sdc: 585940320 512-byte hdwr sectors (300001 MB)
[   76.454567] sdc: Write Protect is off
[   76.454670] sdc: Mode Sense: 00 3a 00 00
[   76.454837] SCSI device sdc: drive cache: write back
[   76.455366] SCSI device sdc: 585940320 512-byte hdwr sectors (300001 MB)
[   76.455564] sdc: Write Protect is off
[   76.455670] sdc: Mode Sense: 00 3a 00 00
[   76.455864] SCSI device sdc: drive cache: write back
[   76.455961]  sdc: sdc1
[   76.466882] sd 2:0:0:0: Attached scsi disk sdc
[   76.467443] SCSI device sdd: 781422768 512-byte hdwr sectors (400088 MB)
[   76.467636] sdd: Write Protect is off
[   76.467725] sdd: Mode Sense: 00 3a 00 00
[   76.467892] SCSI device sdd: drive cache: write back
[   76.468408] SCSI device sdd: 781422768 512-byte hdwr sectors (400088 MB)
[   76.468616] sdd: Write Protect is off
[   76.468716] sdd: Mode Sense: 00 3a 00 00
[   76.468906] SCSI device sdd: drive cache: write back
[   76.469014]  sdd: unknown partition table
[   76.485819] sd 3:0:0:0: Attached scsi disk sdd
[   76.486391] SCSI device sde: 625142448 512-byte hdwr sectors (320073 MB)
[   76.486581] sde: Write Protect is off
[   76.486681] sde: Mode Sense: 00 3a 00 00
[   76.486848] SCSI device sde: drive cache: write back
[   76.487359] SCSI device sde: 625142448 512-byte hdwr sectors (320073 MB)
[   76.487560] sde: Write Protect is off
[   76.487664] sde: Mode Sense: 00 3a 00 00
[   76.487853] SCSI device sde: drive cache: write back
[   76.487960]  sde: sde1 sde2
[   76.508717] sd 4:0:0:0: Attached scsi disk sde
[   76.511901] PNP: PS/2 Controller [PNP0303:PS2K,PNP0f03:PS2M] at
0x60,0x64 irq 1,12
[   76.512803] serio: i8042 AUX port at 0x60,0x64 irq 12
[   76.512945] serio: i8042 KBD port at 0x60,0x64 irq 1
[   76.514756] mice: PS/2 mouse device common for all mice
[   76.515161] NET: Registered protocol family 1
[   76.515349] Using IPI Shortcut mode
[   76.516174] ACPI: (supports S0 S1 S4 S5)
[   76.517145] Freeing unused kernel memory: 160k freed
[   76.517418] Write protecting the kernel read-only data: 312k
[   76.525255] Time: tsc clocksource has been installed.
[   76.530890] ReiserFS: sdc1: found reiserfs format "3.6" with standard journal
[   76.550312] input: AT Translated Set 2 keyboard as /class/input/input0
[   76.725051] logips2pp: Detected unknown logitech mouse model 0
[   77.245394] input: PS/2 Logitech Mouse as /class/input/input1
[   94.958327] ReiserFS: sdc1: using ordered data mode
[   94.991667] ReiserFS: sdc1: journal params: device sdc1, size 8192,
journal first block 18, max trans len 1024, max batch 900, max commit
age 30, max trans age 30
[   94.994994] ReiserFS: sdc1: checking transaction log (sdc1)
[   95.080519] ReiserFS: sdc1: Using r5 hash to sort names
[  102.790993] Real Time Clock Driver v1.12ac
[  105.362817] Linux Tulip driver version 1.1.13-NAPI (December 15, 2004)
[  105.364136] ACPI: PCI Interrupt Link [LNKC] enabled at IRQ 9
[  105.364252] PCI: setting IRQ 9 as level-triggered
[  105.364263] ACPI: PCI Interrupt 0000:00:0b.0[A] -> Link [LNKC] ->
GSI 9 (level, low) -> IRQ 9
[  105.380719] tulip0:  EEPROM default media type Autosense.
[  105.380828] tulip0:  Index #0 - Media MII (#11) described by a
21142 MII PHY (3) block.
[  105.385259] tulip0:  MII transceiver #1 config 1000 status 782d
advertising 01e1.
[  105.390474] eth0: Digital DS21143 Tulip rev 65 at 0001d000,
00:C0:F0:4A:0C:46, IRQ 9.
[  105.412046] pcnet32.c:v1.32 18.Mar.2006 tsbogend@alpha.franken.de
[  105.515961] NET: Registered protocol family 5
[  105.613710] ns83820.c: National Semiconductor DP83820 10/100/1000 driver.
[  105.614319] ACPI: PCI Interrupt 0000:00:09.0[A] -> Link [LNKA] ->
GSI 10 (level, low) -> IRQ 10
[  105.614625] eth1: ns83820.c: 0x22c: 00000000, subsystem: 0000:0000
[  105.641785] eth1: ns83820 v0.23: DP83820 v1.2: 00:4f:4a:00:13:5a
io=0xefffb000 irq=10 f=sg
[  106.289402] Loading Reiser4. See www.namesys.com for a description
of Reiser4.
[  139.047608] JFS: nTxBlock = 4023, nTxLock = 32185
[  157.171109] Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
[  158.611757] eth1: link now 1000 mbps, full duplex and up.
[  161.526960] eth0: Setting full-duplex based on MII#1 link partner
capability of cde1.
[  168.051671]
[  168.051676] ============================
[  168.055000] [ BUG: illegal lock usage! ]
[  168.055087] ----------------------------
[  168.055176] illegal {softirq-on-W} -> {in-softirq-W} usage.
[  168.055273] idle/0 [HC0[0]:SC1[1]:HE0:SE0] takes:
[  168.055368]  (&dev->rx_info.lock){-+..}, at: [<e090a98f>]
rx_irq+0x24/0x18f [ns83820]
[  168.055664] {softirq-on-W} state was registered at:
[  168.055755]   [<c012d27d>] trace_hardirqs_on+0x142/0x16d
[  168.055972]   [<c028815a>] _spin_unlock_irq+0x21/0x25
[  168.056166]   [<e090b78e>] ns83820_open+0x25b/0x381 [ns83820]
[  168.056357]   [<c0242c60>] dev_open+0x2d/0x64
[  168.056543]   [<c024171e>] dev_change_flags+0x51/0xf1
[  168.056728]   [<c0274e42>] devinet_ioctl+0x238/0x556
[  168.056915]   [<c02753d3>] inet_ioctl+0x73/0x91
[  168.057095]   [<c023929f>] sock_ioctl+0x190/0x1b6
[  168.057279]   [<c0164701>] do_ioctl+0x19/0x4f
[  168.057466]   [<c0164940>] vfs_ioctl+0x209/0x220
[  168.057648]   [<c0164983>] sys_ioctl+0x2c/0x59
[  168.057827]   [<c0288283>] syscall_call+0x7/0xb
[  168.058009] irq event stamp: 575395
[  168.058098] hardirqs last  enabled at (575394): [<c011a64d>]
tasklet_action+0x20/0x71
[  168.058299] hardirqs last disabled at (575395): [<c0287ffd>]
_spin_lock_irqsave+0xf/0x2f
[  168.058491] softirqs last  enabled at (575386): [<c011aa6a>]
__do_softirq+0x97/0x9c
[  168.058680] softirqs last disabled at (575391): [<c01045e3>]
do_softirq+0x45/0xb6
[  168.058873]
[  168.058875] other info that might help us debug this:
[  168.059056] no locks held by idle/0.
[  168.059149]
[  168.059151] stack backtrace:
[  168.059732]  [<c010311a>] show_trace_log_lvl+0x54/0xfd
[  168.059889]  [<c0103709>] show_trace+0xd/0x10
[  168.060031]  [<c0103750>] dump_stack+0x19/0x1b
[  168.060173]  [<c012b89a>] print_usage_bug+0x1a4/0x1ae
[  168.060512]  [<c012be3d>] mark_lock+0x16b/0x502
[  168.060844]  [<c012cae0>] __lockdep_acquire+0x3cb/0xa26
[  168.061181]  [<c012d56f>] lockdep_acquire+0x69/0x82
[  168.061515]  [<c028800e>] _spin_lock_irqsave+0x20/0x2f
[  168.061880]  [<e090a98f>] rx_irq+0x24/0x18f [ns83820]
[  168.062026]  [<e090acd9>] rx_action+0x19/0x5a [ns83820]
[  168.062164]  [<c011a672>] tasklet_action+0x45/0x71
[  168.062422]  [<c011aa19>] __do_softirq+0x46/0x9c
[  168.062681]  [<c01045e3>] do_softirq+0x45/0xb6
[  168.062838]  [<c011aa9f>] irq_exit+0x30/0x32
[  168.063095]  [<c01046d9>] do_IRQ+0x85/0x90
[  168.063246]  [<c0102c45>] common_interrupt+0x25/0x2c
[  170.236049] NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state
recovery directory
[  170.236840] NFSD: starting 90-second grace period
[  172.087631] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.088785] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.091643] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.092365] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.095425] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.096145] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.099023] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.099742] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.102564] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO
[  172.103278] program smartd is using a deprecated SCSI ioctl, please
convert it to SG_IO

[patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags part 3

[Arjan van de Ven]

On Sun, 2006-06-04 at 02:53 -0700, Barry K. Nathan wrote:
> On 6/4/06, Arjan van de Ven <arjan@linux.intel.com> wrote:
> > just the preempt the next email from Barry; while fixing this one I
> > looked at the usage of the locks more and found another patch needed...
> [snip]
> 
> Nice try, but it didn't work. ~_^
> 
> I was about to reply to the previous ns83820 patch with my dmesg, when
> I saw this one. I applied this patch too, and like the previous patch,
> it reports an instance of illegal lock usage. My dmesg follows.
> -- 


ok this is a real driver deadlock:

The ns83820 driver enabled interrupts (by unlocking the misc_lock with
_irq) while still holding the rx_info.lock, which is required to be irq
safe since it's used in the ISR like this:
                writel(1, dev->base + IER);
                spin_unlock_irq(&dev->misc_lock);
                kick_rx(ndev);
                spin_unlock_irq(&dev->rx_info.lock);

This is can cause a deadlock if an irq was pending at the first
spin_unlock_irq already, or if one would hit during kick_rx(). 
Simply remove the first _irq solves this

Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>


---
 drivers/net/ns83820.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Index: linux-2.6.17-rc5-mm3/drivers/net/ns83820.c
===================================================================
--- linux-2.6.17-rc5-mm3.orig/drivers/net/ns83820.c
+++ linux-2.6.17-rc5-mm3/drivers/net/ns83820.c
@@ -808,7 +808,7 @@ static int ns83820_setup_rx(struct net_d
 
 		writel(dev->IMR_cache, dev->base + IMR);
 		writel(1, dev->base + IER);
-		spin_unlock_irq(&dev->misc_lock);
+		spin_unlock(&dev->misc_lock);
 
 		kick_rx(ndev);
 

Re: [patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags part 3

[Ingo Molnar]

* Arjan van de Ven <arjan@linux.intel.com> wrote:

> On Sun, 2006-06-04 at 02:53 -0700, Barry K. Nathan wrote:
> > On 6/4/06, Arjan van de Ven <arjan@linux.intel.com> wrote:
> > > just the preempt the next email from Barry; while fixing this one I
> > > looked at the usage of the locks more and found another patch needed...
> > [snip]
> > 
> > Nice try, but it didn't work. ~_^
> > 
> > I was about to reply to the previous ns83820 patch with my dmesg, when
> > I saw this one. I applied this patch too, and like the previous patch,
> > it reports an instance of illegal lock usage. My dmesg follows.
> > -- 
> 
> ok this is a real driver deadlock:

preexisting bug, right? So this fix should go into 2.6.16/17 too, 
correct?

	Ingo

Re: [patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags part 3

[Arjan van de Ven]

On Sun, 2006-06-04 at 12:11 +0200, Ingo Molnar wrote:
> * Arjan van de Ven <arjan@linux.intel.com> wrote:
> 
> > On Sun, 2006-06-04 at 02:53 -0700, Barry K. Nathan wrote:
> > > On 6/4/06, Arjan van de Ven <arjan@linux.intel.com> wrote:
> > > > just the preempt the next email from Barry; while fixing this one I
> > > > looked at the usage of the locks more and found another patch needed...
> > > [snip]
> > > 
> > > Nice try, but it didn't work. ~_^
> > > 
> > > I was about to reply to the previous ns83820 patch with my dmesg, when
> > > I saw this one. I applied this patch too, and like the previous patch,
> > > it reports an instance of illegal lock usage. My dmesg follows.
> > > -- 
> > 
> > ok this is a real driver deadlock:
> 
> preexisting bug, right? So this fix should go into 2.6.16/17 too, 
> correct?

yes; this is real driver deadlock that has been there for quite some
time

Re: [patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags part 3

[Barry K. Nathan]

On 6/4/06, Arjan van de Ven <arjan@linux.intel.com> wrote:
> ok this is a real driver deadlock:
[snip]

With this third patch added, it boots cleanly, with no lockdep
messages. (And, each time, I've been checking things by sshing into
the box via the ns83820 NIC, so I can also confirm that the card works
with these patches.)
-- 
-Barry K. Nathan <barryn@pobox.com>

Re: [patch, -rc5-mm3] lock validator: fix ns83820.c irq-flags part 3

[Ingo Molnar]

* Barry K. Nathan <barryn@pobox.com> wrote:

> On 6/4/06, Arjan van de Ven <arjan@linux.intel.com> wrote:
> >ok this is a real driver deadlock:
> [snip]
> 
> With this third patch added, it boots cleanly, with no lockdep 
> messages. (And, each time, I've been checking things by sshing into 
> the box via the ns83820 NIC, so I can also confirm that the card works 
> with these patches.)

great and thanks for all the testing!

	Ingo