[Xenomai] system freeze on IMX6

[Xenomai] system freeze on IMX6

[TALBI Ahmed -EXT]

Hi all,


I'm new to this mailing list so please feel free to correct me if i'm not using it as it should.

We are facing a freeze on our platform based on a IMX6 soc (see backtrace dump bellow). Please, can anybody take a look to it and give us a feedback  or an idea on the root cause of this issue ?


thanks in advance


we use kernel 3.0.35 and xenomai 2.6.4.


cat /proc/cpuinfo:


Processor       : ARMv7 Processor rev 10 (v7l)
processor       : 0
BogoMIPS        : 1574.50
processor       : 1
BogoMIPS        : 1581.05
Features        : swp half thumb fastmult vfp edsp neon vfpv3
CPU implementer : 0x41
CPU architecture: 7
CPU variant     : 0x2
CPU part        : 0xc09
CPU revision    : 10

Hardware        : Phytec i.MX 6Quad phyFLEX Board
Revision        : ffffffea
Serial          : dbc7879c270f81d4


********* backtrace *****************

[ 8164.291368] xnheap: destroying shared heap 'private sem heap [4349]' with 8 bytes still in use.
[ 8164.465692] xnheap: destroying shared heap 'private sem heap [4352]' with 8 bytes still in use.
[ 8164.526577] Unable to handle kernel paging request at virtual address 78303d63
[ 8164.526766] Xenomai: stuck on nucleus lock 809c4d10
[ 8164.526772]    waiter = kernel/xenomai/nucleus/intr.c:114 (xnintr_clock_handler(), CPU #1)
[ 8164.526779]    owner  = kernel/xenomai/nucleus/shadow.c:2580 (do_taskexit_event(), CPU #0)
[ 8164.526792] Backtrace:
[ 8164.526827] [<8035981c>] (dump_backtrace+0x0/0x10c) from [<80359c40>] (show_stack+0x18/0x1c)
[ 8164.526840]  r6:00000000 r5:00000001 r4:809c4d10 r3:8004d4e8
[ 8164.526888] [<80359c28>] (show_stack+0x0/0x1c) from [<803f20e4>] (__xnlock_spin+0x12c/0x180)
[ 8164.526911] [<803f1fb8>] (__xnlock_spin+0x0/0x180) from [<803f159c>] (xnintr_clock_handler+0x430/0x498)
[ 8164.526938] [<803f116c>] (xnintr_clock_handler+0x0/0x498) from [<803d70f8>] (__ipipe_dispatch_wired_nochec)
[ 8164.526959] [<803d7024>] (__ipipe_dispatch_wired_nocheck+0x0/0x23c) from [<803d72bc>] (__ipipe_dispatch_wi)
[ 8164.526984] [<803d7260>] (__ipipe_dispatch_wired+0x0/0xa0) from [<803608a0>] (__ipipe_handle_irq+0xf0/0x29)
[ 8164.526997]  r7:fffffd9d r6:00000000 r5:809fa3a4 r4:0000001d
[ 8164.527032] [<803607b0>] (__ipipe_handle_irq+0x0/0x294) from [<80360b70>] (__ipipe_grab_irq+0x68/0x84)
[ 8164.527057] [<80360b08>] (__ipipe_grab_irq+0x0/0x84) from [<8035cb30>] (__ipipe_grab_localtimer+0x18/0x1c)
[ 8164.527069]  r6:0000001d r5:f2a00100 r4:ffffffff r3:80000013
[ 8164.527117] [<8035cb18>] (__ipipe_grab_localtimer+0x0/0x1c) from [<80355440>] (__irq_svc+0x40/0xd4)
[ 8164.527147] [<80367570>] (arch_idle+0x0/0x80) from [<80356718>] (__ipipe_halt_root+0x54/0x9c)
[ 8164.527165] [<803566c4>] (__ipipe_halt_root+0x0/0x9c) from [<80356788>] (default_idle+0x28/0x30)
[ 8164.527177]  r5:9ffaa000 r4:9ffaa000
[ 8164.527201] [<80356760>] (default_idle+0x0/0x30) from [<8035698c>] (cpu_idle+0x98/0x118)
[ 8164.527228] [<803568f4>] (cpu_idle+0x0/0x118) from [<8084a700>] (secondary_start_kernel+0x13c/0x160)
[ 8164.527241]  r8:1000406a r7:809e2f40 r6:10c03c7d r5:00000001 r4:809c6440
[ 8164.527270] r3:00000000
[ 8164.527292] [<8084a5c4>] (secondary_start_kernel+0x0/0x160) from [<1084a054>] (0x1084a054)
[ 8164.527303]  r5:00000015 r4:2ffac06a
[ 8164.719404] pgd = 80004000
[ 8164.722125] [78303d63] *pgd=00000000
[ 8164.725738] Internal error: Oops: 5 [#1] PREEMPT SMP
[ 8164.730716] Modules linked in: ksz8462_h rtpacket nomac rtmac rtnet xeno_can_flexcan xeno_can spartan
[ 8164.740153] CPU: 0    Not tainted  (3.0.43-tpcom_run2-PD13.2.4 #1)
[ 8164.746358] PC is at strnlen+0x18/0x60
[ 8164.750124] LR is at string.isra.2+0x34/0xd4
[ 8164.754409] pc : [<806b4384>]    lr : [<806b620c>]    psr: a0000093
[ 8164.754415] sp : 8f617b3c  ip : 8f617b50  fp : 8f617b4c
[ 8164.765917] r10: 8f617cdc  r9 : 8087d9a4  r8 : 00000000
[ 8164.771154] r7 : ffffffff  r6 : 78303d63  r5 : 809e48c0  r4 : 809e4545
[ 8164.777696] r3 : 00000000  r2 : 78303d63  r1 : ffffffff  r0 : 78303d63
[ 8164.784240] Flags: NzCv  IRQs off  FIQs on  Mode SVC_32  ISA ARM  Segment user
[ 8164.791479] Control: 10c53c7d  Table: 1f37404a  DAC: 00000015
[ 8164.797238] Process sh (pid: 4352, stack limit = 0x8f6162f0)
[ 8164.802912] Stack: (0x8f617b3c to 0x8f618000)
[ 8164.807283] 7b20:                                                                809e4545
[ 8164.815479] 7b40: 8f617b74 8f617b50 806b620c 806b4378 00000000 80932267 80932269 00000002
[ 8164.823674] 7b60: 809e48c0 809e4545 8f617bcc 8f617b78 806b71f4 806b61e4 ffffffff ffffffff
[ 8164.831870] 7b80: ffffffff ffffffff 8f617bcc 00000000 00000400 809e44c0 ff0a0004 ffffffff
[ 8164.840065] 7ba0: 8f617be4 00000400 809c217c 809e44c0 00000400 00000000 8f617cc4 80932214
[ 8164.848260] 7bc0: 8f617be4 8f617bd0 806b74cc 806b7030 8f617cc4 00000000 8f617c84 8f617be8
[ 8164.856455] 7be0: 8038d124 806b74c4 8b9f4f00 8bf0cfc8 8f617c14 8f617c00 803d7594 00000080
[ 8164.864649] 7c00: 8034d92c 8f616000 00000000 809e4408 809f4d20 8004a800 8f617c5c 8f617c28
[ 8164.872845] 7c20: 803d49e8 803d4054 00200200 00100100 00000000 00000000 809b5d68 809f4d20
[ 8164.881040] 7c40: 809f4d20 8004a800 8f617c8c 8f617c58 00000004 803d4054 803d7594 00000000
[ 8164.889235] 7c60: 60000093 00000000 00000001 80932214 80868810 00000034 8f617cbc 8f617c88
[ 8164.897429] 7c80: 8084d9fc 8038d030 00000000 8f617cc4 9603ccf0 9603ccf0 00000000 00000000
[ 8164.905625] 7ca0: 8004a800 80933fc8 80868810 00000034 8f617d24 8f617cd0 803f20d8 8084d824
[ 8164.913820] 7cc0: 80932214 9603ccf0 80933fc8 00000034 80868810 00000000 78303d63 37343232
[ 8164.922015] 7ce0: 66646164 3965362e 809c4d10 809b5d68 66646164 37343232 804d7bd8 9603cce0
[ 8164.930210] 7d00: ffff0f40 a0882600 00000000 809b5d68 809b5d68 00000001 8f617d7c 8f617d28
[ 8164.938405] 7d20: 804b153c 803f1fc4 8004a808 a0883170 0001fffe a0882f38 8004a800 8004a808
[ 8164.946600] 7d40: 809b5d68 80a0a8f8 bc3cfec0 000002f0 804d3654 809c4d10 00000000 a0882608
[ 8164.954795] 7d60: a0882600 00000003 809b5d68 80a0a8f8 8f617dc4 8f617d80 804b1884 804b0e70
[ 8164.962990] 7d80: 8050e5d0 803d7620 00000000 00000000 9fff2600 ffffffff 00000000 80a0d068
[ 8164.971185] 7da0: a0875a20 a0882608 00000008 80a0a8f0 00000001 80a0a8f8 8f617dec 8f617dc8
[ 8164.979380] 7dc0: 803f959c 804b16d4 a0875a30 8bc06808 a0882608 00000001 80a0a8e0 000000f5
[ 8164.987575] 7de0: 8f617e54 8f617df0 803ffd04 803f9554 80432bf4 804315fc 9fff2600 9fff88c0
[ 8164.995770] 7e00: 80a0a8f0 00000003 80851c78 9a7029c0 bc3cd840 000002f0 bc3cd95f 000002f0
[ 8165.003965] 7e20: 00000001 809b5d68 8f617e6c a0882608 80a0a8f8 809c4d10 80a0a8f0 80a0d770
[ 8165.012160] 7e40: 00000000 809b5d68 8f617ee4 8f617e58 80436fb0 803fed18 0000006f 8bc06808
[ 8165.020355] 7e60: 00000b30 809f4d20 8bc06808 00000b30 809f4d20 50534531 8f617ef4 8f617e88
[ 8165.028550] 7e80: 803d6524 803d4054 8004a800 95a19360 bc3ccc51 000002f0 8004a800 9a703860
[ 8165.036745] 7ea0: 00000000 00000011 00008000 00000000 8004a800 0000000f 8004a808 8bc06808
[ 8165.044940] 7ec0: 00000b30 809f4d20 809b5d68 80436de4 8bc06808 809c4680 8f617f54 8f617ee8
[ 8165.053135] 7ee0: 803d6610 80436df0 8004a800 8004a800 ffffbfff ffffffff 8004a800 95a19360
[ 8165.061330] 7f00: 809f4d20 00000010 00004000 00000000 8004a800 0000000e 8004a808 20000013
[ 8165.069525] 7f20: 00000000 809f4d20 8038ec58 95a19360 8f616000 00000001 00000000 9a703860
[ 8165.077720] 7f40: 0000006f 9ac3d540 8f617f74 8f617f58 803903d4 803d643c 8f616000 00000001
[ 8165.085915] 7f60: 8f617fb0 000000f8 8f617f94 8f617f78 803907f8 80390294 00000000 00084017
[ 8165.094110] 7f80: 00084008 000000f8 8f617fa4 8f617f98 80390884 803907b4 00000000 8f617fa8
[ 8165.102304] 7fa0: 80355ac0 80390878 00000000 00084017 00000000 2ab2c4c0 00000000 00000000
[ 8165.110500] 7fc0: 00000000 00084017 00084008 000000f8 ffffffff 00000000 2aae7000 00000000
[ 8165.118695] 7fe0: 00091dfc 7eb208d8 00091e10 2acfc340 60000010 00000000 00000000 00000000
[ 8165.126885] Backtrace:
[ 8165.129371] [<806b436c>] (strnlen+0x0/0x60) from [<806b620c>] (string.isra.2+0x34/0xd4)
[ 8165.137389]  r4:809e4545
[ 8165.139963] [<806b61d8>] (string.isra.2+0x0/0xd4) from [<806b71f4>] (vsnprintf+0x1d0/0x410)
[ 8165.148327]  r8:809e4545 r7:809e48c0 r6:00000002 r5:80932269 r4:80932267
[ 8165.154953] r3:00000000
[ 8165.157627] [<806b7024>] (vsnprintf+0x0/0x410) from [<806b74cc>] (vscnprintf+0x14/0x2c)
[ 8165.165657] [<806b74b8>] (vscnprintf+0x0/0x2c) from [<8038d124>] (vprintk+0x100/0x524)
[ 8165.173587]  r4:00000000 r3:8f617cc4
[ 8165.177228] [<8038d024>] (vprintk+0x0/0x524) from [<8084d9fc>] (printk+0x1e8/0x2a8)
[ 8165.184907] [<8084d814>] (printk+0x0/0x2a8) from [<803f20d8>] (__xnlock_spin+0x120/0x180)
[ 8165.193097]  r3:00000034 r2:80933fc8 r1:9603ccf0 r0:80932214
[ 8165.198864] [<803f1fb8>] (__xnlock_spin+0x0/0x180) from [<804b153c>] (thread_destroy+0x6d8/0x864)
[ 8165.207757] [<804b0e64>] (thread_destroy+0x0/0x864) from [<804b1884>] (thread_delete_hook+0x1bc/0x3cc)
[ 8165.217087] [<804b16c8>] (thread_delete_hook+0x0/0x3cc) from [<803f959c>] (xnpod_fire_callouts+0x54/0x98)
[ 8165.226678] [<803f9548>] (xnpod_fire_callouts+0x0/0x98) from [<803ffd04>] (xnpod_delete_thread+0xff8/0x1af)
[ 8165.236519]  r8:000000f5 r7:80a0a8e0 r6:00000001 r5:a0882608 r4:8bc06808
[ 8165.243145] r3:a0875a30
[ 8165.245825] [<803fed0c>] (xnpod_delete_thread+0x0/0x1afc) from [<80436fb0>] (taskexit_event+0x1cc/0x14c0)
[ 8165.255415] [<80436de4>] (taskexit_event+0x0/0x14c0) from [<803d6610>] (__ipipe_dispatch_event+0x1e0/0x484)
[ 8165.265183] [<803d6430>] (__ipipe_dispatch_event+0x0/0x484) from [<803903d4>] (do_exit+0x14c/0x3c8)
[ 8165.274251] [<80390288>] (do_exit+0x0/0x3c8) from [<803907f8>] (do_group_exit+0x50/0xc4)
[ 8165.282354]  r7:000000f8
[ 8165.284929] [<803907a8>] (do_group_exit+0x0/0xc4) from [<80390884>] (__wake_up_parent+0x0/0x28)
[ 8165.293640]  r7:000000f8 r6:00084008 r5:00084017 r4:00000000
[ 8165.299404] [<8039086c>] (sys_exit_group+0x0/0x18) from [<80355ac0>] (ret_fast_syscall+0x0/0x30)
[ 8165.308209] Code: e24cb004 e3510000 01a00001 089da810 (e5d0c000)
[ 8165.314345] ---[ end trace 1003734c85188a37 ]---
[ 8165.319144] Fixing recursive fault but reboot is needed!
[ 8165.324476] BUG: scheduling while atomic: sh/4352/0x00000003
[ 8165.330146] Modules linked in: ksz8462_h rtpacket nomac rtmac rtnet xeno_can_flexcan xeno_can spartan
[ 8165.339571] Backtrace:
[ 8165.342064] [<8035981c>] (dump_backtrace+0x0/0x10c) from [<8084d098>] (dump_stack+0x18/0x1c)
[ 8165.350518]  r6:8f616000 r5:95a19360 r4:00000000 r3:8004d4e8
[ 8165.356285] [<8084d080>] (dump_stack+0x0/0x1c) from [<8084d230>] (__schedule_bug+0x64/0x74)
[ 8165.364663] [<8084d1cc>] (__schedule_bug+0x0/0x74) from [<8085bce4>] (__schedule+0x574/0x6ec)
[ 8165.373202]  r4:8bf09ce0 r3:00000000
[ 8165.376839] [<8085b770>] (__schedule+0x0/0x6ec) from [<8085bfc0>] (schedule+0x50/0x68)
[ 8165.384779] [<8085bf70>] (schedule+0x0/0x68) from [<803905f0>] (do_exit+0x368/0x3c8)
[ 8165.392543] [<80390288>] (do_exit+0x0/0x3c8) from [<80359cfc>] (die+0xb8/0xdc)
[ 8165.399777]  r7:00000005
[ 8165.402350] [<80359c44>] (die+0x0/0xdc) from [<8084d0f8>] (__do_kernel_fault.part.2+0x5c/0x7c)
[ 8165.410975]  r7:8f617af0 r6:00000005 r5:78303d63 r4:00000000
[ 8165.416745] [<8084d09c>] (__do_kernel_fault.part.2+0x0/0x7c) from [<8036278c>] (do_page_fault+0x268/0x3a8)
[ 8165.426412]  r7:78303d63 r3:8f617af0
[ 8165.430049] [<80362524>] (do_page_fault+0x0/0x3a8) from [<80362b9c>] (do_translation_fault+0x20/0x24)
[ 8165.439292] [<80362b7c>] (do_translation_fault+0x0/0x24) from [<8034f4b8>] (do_DataAbort+0x3c/0x248)
[ 8165.448451] [<8034f47c>] (do_DataAbort+0x0/0x248) from [<803553d0>] (__dabt_svc+0x70/0xa0)
[ 8165.456948] Exception stack(0x8f617af0 to 0x8f617b38)
[ 8165.462099] 7ae0:                                     78303d63 ffffffff 78303d63 00000000
[ 8165.470297] 7b00: 809e4545 809e48c0 78303d63 ffffffff 00000000 8087d9a4 8f617cdc 8f617b4c
[ 8165.478493] 7b20: 8f617b50 8f617b3c 806b620c 806b4384 a0000093 ffffffff
[ 8165.485133] [<806b436c>] (strnlen+0x0/0x60) from [<806b620c>] (string.isra.2+0x34/0xd4)
[ 8165.493151]  r4:809e4545
[ 8165.495724] [<806b61d8>] (string.isra.2+0x0/0xd4) from [<806b71f4>] (vsnprintf+0x1d0/0x410)
[ 8165.504088]  r8:809e4545 r7:809e48c0 r6:00000002 r5:80932269 r4:80932267
[ 8165.510716] r3:00000000
[ 8165.513388] [<806b7024>] (vsnprintf+0x0/0x410) from [<806b74cc>] (vscnprintf+0x14/0x2c)
[ 8165.521416] [<806b74b8>] (vscnprintf+0x0/0x2c) from [<8038d124>] (vprintk+0x100/0x524)
[ 8165.529345]  r4:00000000 r3:8f617cc4
[ 8165.532985] [<8038d024>] (vprintk+0x0/0x524) from [<8084d9fc>] (printk+0x1e8/0x2a8)
[ 8165.540666] [<8084d814>] (printk+0x0/0x2a8) from [<803f20d8>] (__xnlock_spin+0x120/0x180)
[ 8165.548856]  r3:00000034 r2:80933fc8 r1:9603ccf0 r0:80932214
[ 8165.554709] [<803f1fb8>] (__xnlock_spin+0x0/0x180) from [<804b153c>] (thread_destroy+0x6d8/0x864)
[ 8165.563603] [<804b0e64>] (thread_destroy+0x0/0x864) from [<804b1884>] (thread_delete_hook+0x1bc/0x3cc)
[ 8165.572933] [<804b16c8>] (thread_delete_hook+0x0/0x3cc) from [<803f959c>] (xnpod_fire_callouts+0x54/0x98)
[ 8165.582605] [<803f9548>] (xnpod_fire_callouts+0x0/0x98) from [<803ffd04>] (xnpod_delete_thread+0xff8/0x1af)
[ 8165.592449]  r8:000000f5 r7:80a0a8e0 r6:00000001 r5:a0882608 r4:8bc06808
[ 8165.599077] r3:a0875a30
[ 8165.601754] [<803fed0c>] (xnpod_delete_thread+0x0/0x1afc) from [<80436fb0>] (taskexit_event+0x1cc/0x14c0)
[ 8165.611344] [<80436de4>] (taskexit_event+0x0/0x14c0) from [<803d6610>] (__ipipe_dispatch_event+0x1e0/0x484)
[ 8165.621109] [<803d6430>] (__ipipe_dispatch_event+0x0/0x484) from [<803903d4>] (do_exit+0x14c/0x3c8)
[ 8165.630176] [<80390288>] (do_exit+0x0/0x3c8) from [<803907f8>] (do_group_exit+0x50/0xc4)
[ 8165.638281]  r7:000000f8
[ 8165.640853] [<803907a8>] (do_group_exit+0x0/0xc4) from [<80390884>] (__wake_up_parent+0x0/0x28)
[ 8165.649565]  r7:000000f8 r6:00084008 r5:00084017 r4:00000000
[ 8165.655330] [<8039086c>] (sys_exit_group+0x0/0x18) from [<80355ac0>] (ret_fast_syscall+0x0/0x30)
[ 8178.084852] Xenomai: stuck on nucleus lock 809c4d10
[ 8178.084857]    waiter = kernel/xenomai/nucleus/intr.c:114 (xnintr_clock_handler(), CPU #1)
[ 8178.084865]    owner  = kernel/xenomai/nucleus/shadow.c:2580 (do_taskexit_event(), CPU #0)



________________________________
CONFIDENTIALITY : This e-mail and any attachments are confidential and may be privileged. If you are not a named recipient, please notify the sender immediately and do not disclose the contents to another person, use it for any purpose or store or copy the information in any medium.

Re: [Xenomai] system freeze on IMX6

[Gilles Chanteperdrix]

On Fri, May 13, 2016 at 01:25:31PM +0000, TALBI Ahmed -EXT wrote:
> Hi all,
> 
> 
> I'm new to this mailing list so please feel free to correct me if i'm not using it as it should.
> 
> We are facing a freeze on our platform based on a IMX6 soc (see backtrace dump bellow). Please, can anybody take a look to it and give us a feedback  or an idea on the root cause of this issue ?
> 
> 
> thanks in advance
> 
> 
> we use kernel 3.0.35 and xenomai 2.6.4.

Do you have the same problem with the 4.1 mainline kernel on Xenomai
3.0.2 ?

-- 
					    Gilles.
https://click-hack.org

Re: [Xenomai] system freeze on IMX6

[Rolf Wöhrmann]

Hi

not sure if this helps: 

I’m running fine Xenomai 3.0.1 with mainline 4.1 on a i.MX6Q phyCORE board.

Phytec has a new PD15.3.0 released which contains a mainline 4.1.18 only with some DTS additions and minor driver changes. Applying the ipipe-4.1.18arm 4 was without problems.
So, may it’s worth a try to upgrade.

Cheers
Rolf
 
 
> On 14 May 2016, at 14:30, Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org> wrote:
> 
> On Fri, May 13, 2016 at 01:25:31PM +0000, TALBI Ahmed -EXT wrote:
>> Hi all,
>> 
>> 
>> I'm new to this mailing list so please feel free to correct me if i'm not using it as it should.
>> 
>> We are facing a freeze on our platform based on a IMX6 soc (see backtrace dump bellow). Please, can anybody take a look to it and give us a feedback  or an idea on the root cause of this issue ?
>> 
>> 
>> thanks in advance
>> 
>> 
>> we use kernel 3.0.35 and xenomai 2.6.4.
> 
> Do you have the same problem with the 4.1 mainline kernel on Xenomai
> 3.0.2 ?
> 
> -- 
> 					    Gilles.
> https://click-hack.org
> 
> _______________________________________________
> Xenomai mailing list
> Xenomai@xenomai.org
> https://xenomai.org/mailman/listinfo/xenomai

Re: [Xenomai] system freeze on IMX6

[TALBI Ahmed -EXT]

Hi Rolf and Gilles,

Thanks for your responses and sorry for my delayed one (we had 3 days off ...)

Unfortunately we can not afford to upgrade to newer versions of kernel and xenomai. We are near to a major release and we have very short deadlines.

In the logs I attached I can see that xnlock  is locked in do_taskexit_event (by CPU0) and then in the same context we try to lock it a second time in xintr_clock_handler. May this be the issue ? or is xnlock a recursive lock ?

Cheers
Ahmed

________________________________________
De : Rolf Wöhrmann <info@temporubato.com>
Envoyé : samedi 14 mai 2016 15:01:14
À : Gilles Chanteperdrix; TALBI Ahmed -EXT; xenomai@xenomai.org
Objet : Re: [Xenomai] system freeze on IMX6

Hi

not sure if this helps:

I’m running fine Xenomai 3.0.1 with mainline 4.1 on a i.MX6Q phyCORE board.

Phytec has a new PD15.3.0 released which contains a mainline 4.1.18 only with some DTS additions and minor driver changes. Applying the ipipe-4.1.18arm 4 was without problems.
So, may it’s worth a try to upgrade.

Cheers
Rolf


> On 14 May 2016, at 14:30, Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org> wrote:
>
> On Fri, May 13, 2016 at 01:25:31PM +0000, TALBI Ahmed -EXT wrote:
>> Hi all,
>>
>>
>> I'm new to this mailing list so please feel free to correct me if i'm not using it as it should.
>>
>> We are facing a freeze on our platform based on a IMX6 soc (see backtrace dump bellow). Please, can anybody take a look to it and give us a feedback  or an idea on the root cause of this issue ?
>>
>>
>> thanks in advance
>>
>>
>> we use kernel 3.0.35 and xenomai 2.6.4.
>
> Do you have the same problem with the 4.1 mainline kernel on Xenomai
> 3.0.2 ?
>
> --
>                                           Gilles.
> https://click-hack.org
>
> _______________________________________________
> Xenomai mailing list
> Xenomai@xenomai.org
> https://xenomai.org/mailman/listinfo/xenomai


________________________________
CONFIDENTIALITY : This e-mail and any attachments are confidential and may be privileged. If you are not a named recipient, please notify the sender immediately and do not disclose the contents to another person, use it for any purpose or store or copy the information in any medium.

Re: [Xenomai] system freeze on IMX6

[Gilles Chanteperdrix]

On Tue, May 17, 2016 at 09:11:09AM +0000, TALBI Ahmed -EXT wrote:
> Hi Rolf and Gilles,
> 
> Thanks for your responses and sorry for my delayed one (we had 3 days off ...)
> 
> Unfortunately we can not afford to upgrade to newer versions of kernel and xenomai. We are near to a major release and we have very short deadlines.

We are not asking you to upgrade. We are asking you to test Xenomai
3.0.2 with Linux 4.1.18 so as to know whether the issue you have has
already been fixed. On our side, we do not have much time either,
and certainly no time to waste on an issue that was already fixed.

-- 
					    Gilles.
https://click-hack.org

Re: [Xenomai] system freeze on IMX6

[Gilles Chanteperdrix]

On Tue, May 17, 2016 at 09:11:09AM +0000, TALBI Ahmed -EXT wrote:
> In the logs I attached I can see that xnlock  is locked in do_taskexit_event (by CPU0) and then in the same context we try to lock it a second time in xintr_clock_handler. May this be the issue ? or is xnlock a recursive lock ?

You are misreading the logs. xnintr_clock_handler tries to get the
lock on one cpu, while do_taskexit_event got it on another cpu. The
root cause of the problem is a fault when xnlock_spin is trying to
print some message. Probably because it has spin for too long. You
can try to disable spinlock debugging, but I am afraid you will get
a lockup instead.

-- 
					    Gilles.
https://click-hack.org

Re: [Xenomai] system freeze on IMX6

[TALBI Ahmed -EXT]

Ok, got it Gilles...

I will give a try for both proposals. The issue is hard to reproduce (may be 2 or 3 days under stress...).
Back to you as soon as I have news

Ahmed

________________________________________
De : Gilles Chanteperdrix <gilles.chanteperdrix@xenomai.org>
Envoyé : mardi 17 mai 2016 11:41:27
À : TALBI Ahmed -EXT
Cc : Rolf Wöhrmann; xenomai@xenomai.org
Objet : Re: [Xenomai] system freeze on IMX6

On Tue, May 17, 2016 at 09:11:09AM +0000, TALBI Ahmed -EXT wrote:
> In the logs I attached I can see that xnlock  is locked in do_taskexit_event (by CPU0) and then in the same context we try to lock it a second time in xintr_clock_handler. May this be the issue ? or is xnlock a recursive lock ?

You are misreading the logs. xnintr_clock_handler tries to get the
lock on one cpu, while do_taskexit_event got it on another cpu. The
root cause of the problem is a fault when xnlock_spin is trying to
print some message. Probably because it has spin for too long. You
can try to disable spinlock debugging, but I am afraid you will get
a lockup instead.

--
                                            Gilles.
https://click-hack.org

________________________________
CONFIDENTIALITY : This e-mail and any attachments are confidential and may be privileged. If you are not a named recipient, please notify the sender immediately and do not disclose the contents to another person, use it for any purpose or store or copy the information in any medium.