IMX SDMA scheduling issue

saurabh206 — Wed, 24 Dec 2014 08:17:41 GMT

We are using Linux kernel 3.10.17.

We are seeing following crash log while trying to play sequence of audio file with CSI data capture.

Unable to handle kernel NULL pointer dereference at virtual address 000000a4

pgd = a84fc000

[000000a4] *pgd=384d4831, *pte=00000000, *ppte=00000000

Internal error: Oops: 17 [#1] PREEMPT SMP ARM

Modules linked in: tvin galcore

CPU: 0 PID: 118 Comm: demo Not tainted 3.10.17-1.0.0_ga #1

task: a84ac3c0 ti: a831c000 task.ti: a831c000

PC is at snd_pcm_period_elapsed+0x44/0xd8

LR is at _raw_spin_lock+0x10/0x4c

pc : [<802bc274>] lr : [<80306cd8>] psr: 60000193

sp : a831def8 ip : 00000020 fp : 00000100

r10: 00000000 r9 : 00404000 r8 : 00000000

r7 : 60000113 r6 : a81f14c0 r5 : a842a800 r4 : a81f1400

r3 : 00000000 r2 : 00007fe8 r1 : 00000000 r0 : 00000002

Flags: nZCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user

Control: 10c53c7d Table: 384fc04a DAC: 00000015

Process demo (pid: 118, stack limit = 0xa831c238)

Stack: (0xa831def8 to 0xa831e000)

dee0: a8401e40 a81a00f0

df00: 0000000c 00000001 00000003 801a336c 801a32dc a81a01c0 a81a01c4 804725c4

df20: 00000000 80029b00 00000001 8047809c a831c000 0000000a 7fb87f7c 80029240

df40: a80079c0 a8199400 a80079c0 00000018 00000006 00004b11 7eab5770 60000193

df60: 80472ef0 00000000 7eab5700 7eab5730 00000009 5fa69d78 7eab5770 800293b4

df80: a831c008 80029608 00000022 8000e1a8 f4000100 a831dfb0 8047e83c 80008448

dfa0: 76e3c1b8 80000010 ffffffff 8000d67c 00000001 00000000 00000000 00000000

dfc0: 5fa5b300 00000002 6dd86f88 7eab5700 7eab5730 00000009 5fa69d78 7eab5770

dfe0: 00000000 7eab56c0 76e3aeb4 76e3c1b8 80000010 ffffffff 3bf58811 3bf58c11

[<802bc274>] (snd_pcm_period_elapsed+0x44/0xd8) from [<801a336c>] (sdma_tasklet+0x90/0x134)

[<801a336c>] (sdma_tasklet+0x90/0x134) from [<80029b00>] (tasklet_action+0x98/0xfc)

[<80029b00>] (tasklet_action+0x98/0xfc) from [<80029240>] (__do_softirq+0xd4/0x1c0)

[<80029240>] (__do_softirq+0xd4/0x1c0) from [<800293b4>] (do_softirq+0x44/0x50)

[<800293b4>] (do_softirq+0x44/0x50) from [<80029608>] (irq_exit+0x74/0xc0)

[<80029608>] (irq_exit+0x74/0xc0) from [<8000e1a8>] (handle_IRQ+0x6c/0x90)

[<8000e1a8>] (handle_IRQ+0x6c/0x90) from [<80008448>] (gic_handle_irq+0x3c/0x60)

[<80008448>] (gic_handle_irq+0x3c/0x60) from [<8000d67c>] (__irq_usr+0x3c/0x60)

Exception stack(0xa831dfb0 to 0xa831dff8)

dfa0: 00000001 00000000 00000000 00000000

dfc0: 5fa5b300 00000002 6dd86f88 7eab5700 7eab5730 00000009 5fa69d78 7eab5770

dfe0: 00000000 7eab56c0 76e3aeb4 76e3c1b8 80000010 ffffffff

Code: e1a07000 e1a00006 eb012a95 e59430a8 (e59330a4)

---[ end trace bde07178062cfd02 ]---

Kernel panic - not syncing: Fatal exception in interrupt

CPU1: stopping

CPU: 1 PID: 0 Comm: swapper/1 Tainted: G D 3.10.17-1.0.0_ga #1

[<80012b28>] (unwind_backtrace+0x0/0xe0) from [<80010a48>] (show_stack+0x10/0x14)

[<80010a48>] (show_stack+0x10/0x14) from [<80011df8>] (handle_IPI+0xc0/0x118)

[<80011df8>] (handle_IPI+0xc0/0x118) from [<80008464>] (gic_handle_irq+0x58/0x60)

[<80008464>] (gic_handle_irq+0x58/0x60) from [<8000d4c0>] (__irq_svc+0x40/0x70)

Exception stack(0xa808df50 to 0xa808df98)

df40: a808df98 00000072 a8a80283 00000072

df60: 977b2776 00000072 00000001 80d15128 8048191c 80481968 804f16c4 00000000

df80: 00000009 a808df98 80051ad8 8026a624 600f0013 ffffffff

[<8000d4c0>] (__irq_svc+0x40/0x70) from [<8026a624>] (cpuidle_enter_state+0x50/0xd8)

[<8026a624>] (cpuidle_enter_state+0x50/0xd8) from [<8026a78c>] (cpuidle_idle_call+0xe0/0x144)

[<8026a78c>] (cpuidle_idle_call+0xe0/0x144) from [<8000e484>] (arch_cpu_idle+0x10/0x48)

[<8000e484>] (arch_cpu_idle+0x10/0x48) from [<8005144c>] (cpu_startup_entry+0xc4/0x124)

[<8005144c>] (cpu_startup_entry+0xc4/0x124) from [<102ff044>] (0x102ff044)

imx-ipuv3 2400000.ipu: IPU Warning - IPU_INT_STAT_10 = 0x00000001

Re: IMX SDMA scheduling issue

Yuri — Thu, 25 Dec 2014 06:28:40 GMT

I hope the enclosed recommendations how to debug kernel crashes help.

Have a great day,
Yuri

-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------

Re: IMX SDMA scheduling issue

saurabh206 — Mon, 29 Dec 2014 09:42:35 GMT

Hello Yuri

Can you provide more detail about http://permalink.gmane.org/gmane.linux.ports.arm.kernel/335436

"imx-sdma cyclic handling to report residue"

Do we have any update for this in 3.10.17 kernel?

Thanks

Saurabh

Re: IMX SDMA scheduling issue

Yuri — Fri, 23 Jan 2015 07:20:10 GMT

Alas. I do not have information about the imx-sdma cyclic handling.

~Yuri.

topic Re: IMX SDMA scheduling issue in i.MX Processors

IMX SDMA scheduling issue

Re: IMX SDMA scheduling issue

Re: IMX SDMA scheduling issue

Re: IMX SDMA scheduling issue