This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

why it happens to my AP execution?



My program is running in DM365, sometimes it happens the following...

Bad pte = 2e0b77c5, process = ???, vm_flags = 875, vaddr = 40001000

Bad pte = 2e0b77c5, process = ???, vm_flags = 875, vaddr = 40003000

Bad pte = 00c90075, process = ???, vm_flags = 875, vaddr = 40006000

Bad pte = 770046ed, process = ???, vm_flags = 875, vaddr = 40008000

...

swap_free: Bad swap file entry 18000380
swap_free: Bad swap file entry 600bb706
swap_free: Bad swap file entry 0004ac0c
swap_free: Bad swap file entry 600bb706
swap_free: Bad swap file entry 0004ac0c
swap_free: Bad swap file entry 08000100
swap_free: Bad swap file entry 28001980
swap_free: Bad swap file entry 68091d05
swap_free: Bad swap file entry 68091d05

Could someone tell me why those happen?

It's annoying me.

 

  • I have not see this before, can you provide more details on your setup (hardware EVM used, software and version used, conditions under which this occurs)?  For example, this does only occurr when using a particular file-system (e.g. JFFS2)....

  • It happens both in NFS and ramdisk way.
    I am using APPRO DM365 development kit.
    ( filesystem is from montavista/pro/devkit/arm/v5t_le/images/ramdisk.gz )
    ( cross compiler version is 4.2.0 )
    ( boot argument is with mem=128M )

    The condition is as following:
    (1)connect to the embeded system by telnet.
    (2)scan memory if the program exists
    (3)kill the program if it exists
    (4)execution

    After many times doing so , the situation happens.

    I change many ways to terminate program such as
    kill -9 PID
    kill -INT PID
    killall -INT PNAME

    All will happens crash... :(

    ===
    By the way , it goes with AEW enable. (If disable AEW thread , it never happens)
    But my code is the same with DM355.
    In DM355 environment, it never happens...

  • doing a little research, it appears this is a widely spread issue with Linux.  See http://answers.google.com/answers/threadview?id=314578.  This article suggest a potential fix near the bottom; personally, I have not see this myself.