Ooops 805 2.6.31.12-rt20

Ooops 805 2.6.31.12-rt20

[Tim Sander]

Hi

I have just tested the newest RT Kernel. It gives me an oops:
FEC Ethernet Driver
kernel BUG at kernel/rtmutex.c:807!
Unable to handle kernel NULL pointer dereference at virtual address 00000000
pgd = c0004000
[00000000] *pgd=00000000
Internal error: Oops: 805 [#1] PREEMPT
Modules linked in:
CPU: 0    Not tainted  (2.6.31.12-rt20-00001-g36e2ecd #1)
PC is at __bug+0x1c/0x28
LR is at __bug+0x18/0x28
pc : [<c00231d0>]    lr : [<c00231cc>]    psr: 00000093
sp : c784bef0  ip : c784be30  fp : 00000002
r10: c79cfc24  r9 : c02e2074  r8 : c030e7f4
r7 : 40000013  r6 : c0188624  r5 : c784a000  r4 : 600e0000
r3 : 00000000  r2 : 00000001  r1 : c784a000  r0 : 0000002a
Flags: nzcv  IRQs off  FIQs on  Mode SVC_32  ISA ARM  Segment kernel
Control: 00c5387d  Table: 80004008  DAC: 00000017
Process irq/57-fec (pid: 115, stack limit = 0xc784a268)
Stack: (0xc784bef0 to 0xc784c000)
bee0:                                     c022ef38 c022f134 00000000 40000013
bf00: ffffffff c784bf10 c022f0d4 c0030704 c0031a3c c00317f4 c78c2840 c02de460
bf20: 00000073 00000001 00000000 c784bf38 c005b4fc 600e0000 c79cf800 c0188624
bf40: c79cfc24 c030e7f4 c02e2074 c030e7f8 00000002 c01887ac c030e704 c79cfb20
bf60: c79cf800 c79cfc24 c030e7f4 c0189144 c02e2030 c02e206c c790a760 c784a000
bf80: c02e2030 c790a788 c02e208c c0061110 00000000 00000032 c784bfbc c781de4c
bfa0: c790a760 c0060fdc 00000000 00000000 00000000 00000000 00000000 c004ad68
bfc0: 00000000 00000000 c784bfc8 c784bfc8 c784bfd0 c784bfd0 00000000 00000000
bfe0: c784bfe0 c784bfe0 00000000 00000000 00000000 c00208b4 cfee7cd4 8fcd69f9
[<c00231d0>] (__bug+0x1c/0x28) from [<c022f134>] (rt_spin_lock_slowlock+0x220/0x234)
[<c022f134>] (rt_spin_lock_slowlock+0x220/0x234) from [<c01887ac>] (mii_queue+0x20/0xb4)
[<c01887ac>] (mii_queue+0x20/0xb4) from [<c0189144>] (fec_enet_interrupt+0x84/0x104)
[<c0189144>] (fec_enet_interrupt+0x84/0x104) from [<c0061110>] (irq_thread+0x134/0x1c0)
[<c0061110>] (irq_thread+0x134/0x1c0) from [<c004ad68>] (kthread+0x7c/0x84)
[<c004ad68>] (kthread+0x7c/0x84) from [<c00208b4>] (kernel_thread_exit+0x0/0x8)
Code: e1a01000 e59f000c eb08299b e3a03000 (e5833000)
---[ end trace 7ce2cf1764dda59d ]---
exiting task "irq/57-fec" (115) is an active IRQ thread (irq 57)
note: irq/57-fec[115] exited with preempt_count 1
physmap platform flash device: 02000000 at a0000000
physmap-flash.0: Found 1 x16 devices at 0x0 in 16-bit bank


...
and later than
------------[ cut here ]------------
WARNING: at net/sched/sch_generic.c:250 dev_watchdog+0x288/0x2a8()
NETDEV WATCHDOG: eth0 (fec): transmit queue 0 timed out
Modules linked in:
[<c002492c>] (unwind_backtrace+0x0/0xe4) from [<c0035ea8>] (warn_slowpath_common+0x4c/0x80)
[<c0035ea8>] (warn_slowpath_common+0x4c/0x80) from [<c0035f18>] (warn_slowpath_fmt+0x28/0x38)
[<c0035f18>] (warn_slowpath_fmt+0x28/0x38) from [<c01c8e58>] (dev_watchdog+0x288/0x2a8)
[<c01c8e58>] (dev_watchdog+0x288/0x2a8) from [<c0040350>] (run_timer_softirq+0x160/0x280)
[<c0040350>] (run_timer_softirq+0x160/0x280) from [<c003b32c>] (ksoftirqd+0x11c/0x224)
[<c003b32c>] (ksoftirqd+0x11c/0x224) from [<c004ad68>] (kthread+0x7c/0x84)
[<c004ad68>] (kthread+0x7c/0x84) from [<c00208b4>] (kernel_thread_exit+0x0/0x8)
---[ end trace 7ce2cf1764dda59f ]---

it seems as if the FEC ethernet driver of the phycore pcm043 freescale imx35 is unhappy.
If there is an easy way to test this i would give it a try.

Best regards
Tim

Re: Ooops 805 2.6.31.12-rt20

[Uwe Kleine-König]

Hi,

> it seems as if the FEC ethernet driver of the phycore pcm043 freescale imx35 is unhappy.
> If there is an easy way to test this i would give it a try.
This is already fixed in mainline.  (And you were Cc:d for the two
most relevant patches!)

	$ git lg v2.6.31.12-rt20..linus/master -- drivers/net/fec.c
	78abcb1 net: fix section mismatch in fec.c
	66bc4a6 Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/gerg/m68knommu
	4980296 m68knommu: remove ColdFire direct interrupt register access
	81538e7 fec: don't enable irqs in hard irq context
	84177a2 fec: fix recursive locking of mii_lock
	aa11d95 Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
	da81203 Merge branch 'master' of master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
	6ed1065 net: use NETDEV_TX_OK instead of 0 in ndo_start_xmit() functions

84177a2 and 81538e7 are the important ones for you, I didn't check the
others.

Uwe

-- 
Pengutronix e.K.                              | Uwe Kleine-König            |
Industrial Linux Solutions                    | http://www.pengutronix.de/  |
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: Ooops 805 2.6.31.12-rt20

[Tim Sander]

Hi 
> This is already fixed in mainline.  (And you were Cc:d for the two
> most relevant patches!)
Oh, then sorry for the noise. I really forgot it seems as if last year is like long ago...

> 84177a2 and 81538e7 are the important ones for you, I didn't check the
Ok thanks.

Btw. is there a up to date preempt rt git tree. I looked on git.kernel.org but i didn't 
find any recent preemt-rt linux tree? I mean one where the mainline 
fixes and the -rt patches are in one tree?

Best regards
Tim 

Btw heres a little typo fix (colon missing). Probably the wrong mailing list, but the right people 8-)

diff --git a/arch/arm/plat-mxc/include/mach/iomux-mx35.h b/arch/arm/plat-mxc/include/mach/iomux-mx35.h
index 00b0ac1..c88d407 100644
--- a/arch/arm/plat-mxc/include/mach/iomux-mx35.h
+++ b/arch/arm/plat-mxc/include/mach/iomux-mx35.h
@@ -671,7 +671,7 @@
 #define MX35_PAD_LD8__SDMA_SDMA_DEBUG_PC_8                     IOMUX_PAD(0x634, 0x1d0, 6, 0x0,   0, NO_PAD_CTRL)

 #define MX35_PAD_LD9__IPU_DISPB_DAT_9                          IOMUX_PAD(0x638, 0x1d4, 0, 0x0,   0, NO_PAD_CTRL)
-#define MX35_PAD_LD9__GPIO2_9                                  IOMUX_PAD(0x638, 0x1d4, 5, 0x8e4  0, NO_PAD_CTRL)
+#define MX35_PAD_LD9__GPIO2_9                                  IOMUX_PAD(0x638, 0x1d4, 5, 0x8e4, 0, NO_PAD_CTRL)
 #define MX35_PAD_LD9__SDMA_SDMA_DEBUG_PC_9                     IOMUX_PAD(0x638, 0x1d4, 6, 0x0,   0, NO_PAD_CTRL)

 #define MX35_PAD_LD10__IPU_DISPB_DAT_10                                IOMUX_PAD(0x63c, 0x1d8, 0, 0x0,   0, NO_PAD_CTRL)

Re: Ooops 805 2.6.31.12-rt20

[Uwe Kleine-König]

Hello Tim,

On Thu, Jan 28, 2010 at 08:41:30PM +0100, Tim Sander wrote:
> Hi 
> > This is already fixed in mainline.  (And you were Cc:d for the two
> > most relevant patches!)
> Oh, then sorry for the noise. I really forgot it seems as if last year is like long ago...
> 
> > 84177a2 and 81538e7 are the important ones for you, I didn't check the
> Ok thanks.
> 
> Btw. is there a up to date preempt rt git tree. I looked on git.kernel.org but i didn't 
> find any recent preemt-rt linux tree? I mean one where the mainline 
> fixes and the -rt patches are in one tree?
It's not trivial to merge Linus' tree and the rt tree.  In the mail
announcing 2.6.31.12-rt20 Thomas says to start working at 33-rcX-rt.  I
assume you have to wait for that.

> Btw heres a little typo fix (colon missing). Probably the wrong
> mailing list, but the right people 8-)
The right people are reading linux-arm-kernel ML, not this one.  I sent
a patch including a commit log to that list.

Best regards
Uwe

-- 
Pengutronix e.K.                              | Uwe Kleine-König            |
Industrial Linux Solutions                    | http://www.pengutronix.de/  |
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html