[ag-automation] Linux 2.6.31.12-rt21 hangs on AT91 based board

Martin Däumler mdae at cs.tu-chemnitz.de
Wed Sep 21 13:14:42 CEST 2011


Hello,

I want to use Linux 2.6.31.12-rt21 with a AT91SAM9G20 based board.
I followed the instructions described in [1] and set the kernel options
that are mentioned there (AEABI=y, OABI_COMPAT=n). Unfortunately,
Linux hangs after mounting the root file system:

#:rtc-pcf8563 0-0051: low voltage detected, date/time is not reliable.
#:rtc-pcf8563 0-0051: retrieved date/time is not valid.
#:rtc-pcf8563 0-0051: hctosys: invalid date/time
#:VFS: Mounted root (jffs2 filesystem) on device 31:3.
#:***************************************************************************** 

#:*                                                                           
*
#:*  REMINDER, the following debugging options are turned on in your 
.config: *
#:*                                                                           
*
#:*        
CONFIG_DEBUG_RT_MUTEXES                                            *
#:*        
CONFIG_LOCKDEP                                                     *
#:*                                                                           
*
#:*  they may increase runtime overhead and 
latencies.                        *
#:*                                                                           
*
#:***************************************************************************** 

#:Freeing init memory: 124K

Setting OABI_COMPAT=y leads to the same output. Setting AEABI=n leads to:

#:rtc-pcf8563 0-0051: low voltage detected, date/time is not reliable.
#:rtc-pcf8563 0-0051: retrieved date/time is not valid.
#:rtc-pcf8563 0-0051: hctosys: invalid date/time
#:VFS: Mounted root (jffs2 filesystem) on device 31:3.
#:***************************************************************************** 

#:*                                                                           
*
#:*  REMINDER, the following debugging options are turned on in your 
.config: *
#:*                                                                           
*
#:*        
CONFIG_DEBUG_RT_MUTEXES                                            *
#:*        
CONFIG_LOCKDEP                                                     *
#:*                                                                           
*
#:*  they may increase runtime overhead and 
latencies.                        *
#:*                                                                           
*
#:***************************************************************************** 

#:Freeing init memory: 132K
#:Kernel panic - not syncing: Attempted to kill init!
#:Backtrace:
#:[<c002dd60>] (dump_backtrace+0x0/0xfc) from [<c026a308>] 
(dump_stack+0x18/0x1c)
#: r6:00000001 r5:c1c1b020 r4:c0352edc
#:[<c026a2f0>] (dump_stack+0x0/0x1c) from [<c026a360>] (panic+0x54/0x12c)
#:[<c026a30c>] (panic+0x0/0x12c) from [<c003e2c4>] 
(find_new_reaper+0x74/0x90)
#: r3:c032d690 r2:c1c1fdc8 r1:00000000 r0:c02e0f22
#:[<c003e250>] (find_new_reaper+0x0/0x90) from [<c003f664>] 
(forget_original_parent+0x34/0x114)
#: r5:00000001 r4:c1c1b020
#:[<c003f630>] (forget_original_parent+0x0/0x114) from [<c003f75c>] 
(exit_notify+0x18/0x150)
#: r7:c1c1fee8 r6:00000001 r5:00000001 r4:c1c1b020
#:[<c003f744>] (exit_notify+0x0/0x150) from [<c003fa64>] 
(do_exit+0x1d0/0x2a8)
#: r5:00000004 r4:c1c1b020
#:[<c003f894>] (do_exit+0x0/0x2a8) from [<c003fbd4>] 
(do_group_exit+0x98/0xcc)
#:[<c003fb3c>] (do_group_exit+0x0/0xcc) from [<c004ac18>] 
(get_signal_to_deliver+0x268/0x29c)
#: r7:c1c1fee8 r6:c1c1e000 r5:c1d4d824 r4:0830009f
#:[<c004a9b0>] (get_signal_to_deliver+0x0/0x29c) from [<c002c6e0>] 
(do_signal+0xa4/0x1d8)
#:[<c002c63c>] (do_signal+0x0/0x1d8) from [<c002c844>] 
(do_notify_resume+0x30/0x34)
#: r8:00000000 r7:00900000 r6:00000000 r5:40026000 r4:00000000
#:[<c002c814>] (do_notify_resume+0x0/0x34) from [<c0029e4c>] 
(work_pending+0x1c/0x20)


I found a post on the RT-Preempt Mailing list that describes a similar 
problem [2]. I guess,
using Linux 2.6.33.7 does not solve the problem. So, there are 
Linux-RT-Preempt versions that
are known to work on AT91SAM9G20? Do you have a hint how to solve this 
problem?


With kind regards,
Martin Däumler


[1] 
http://www.at91.com/linux4sam/bin/view/Linux4SAM/BuildRootFAQ#Failed_to_mount_rootfs_system_ha
[2] http://www.spinics.net/lists/linux-rt-users/msg06366.html


More information about the ag-automation mailing list