• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 99 subscribers
  • Views 595 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

loadmodules_sd.sh error

NINGMENG CHEN
Intellectual 895 points


The uImage-dm365 is copied from PSP_02_10_00_10.tar.gz.

The filesys is copied from the target of mvl_5_0_0801921_demo_sys_setuplinux.bin

The loadmodules_sd.sh is copied from dvsdk_2_10_00_11/dvsdk_demos_2_10_00_11/dm365.

When I loadmoudels, the error is as following:

cmemk: disagrees about version of symbol struct_module

edmak: disagrees about version of symbol struct_module

insmod: error inserting 'cmemk.ko': -1 Invalid module format

insmod: error inserting 'edmak.ko': -1 Invalid module format

irqk: disagrees about version of symbol struct_module

dm365mmap: disagrees about version of symbol struct_module

insmod: error inserting 'irqk.ko': -1 Invalid module format

insmod: error inserting 'dm365mmap.ko': -1 Invalid module format

 

How to solve it?

  • 0
    Intellectual 895 points

    cmemk.ko edmak.ko irqk.ko dm365mmap.ko is from dvsdk_2_10_00_11/kernel_binaries/dm365.

     

  • 0 in reply to NINGMENG CHEN
    TI__Mastermind 37340 points

    "disagrees about version of symbol" errors are normally seen as a result of trying to load incompatible software components; the three components to keep in mind are u-boot (bootloader), uImage (Linux kernel) and the Linux File System (includes cmemk.ko edmak.ko irqk.ko dm365mmap.ko ).  How are you loading each of these components (from flash, via TFTP, NFS....).  Make sure all three of these components come from the same DVSDK release.  loading an older version of one of these components perhaps residing in flash memory, along with a newer file-system can lead to the errors you are seeing.

  • 0 in reply to Juan Gonzales
    Intellectual 895 points

    Juan Gonzales said:
    How are you loading each of these components (from flash, via TFTP, NFS....).

    I load them from NFS, I copy these filse from dvsdk_dm365_setuplinux_2_10_00_11( cmemk.ko edmak.ko irqk.ko dm365mmap.ko) and psp_02_10_00_10(uImage u-boot).

  • 0 in reply to NINGMENG CHEN
    Intellectual 895 points

    Would you please give me a tar which include all these files which have been verified?

    I not sure psp_02_10_00_10 is compatible with dvsdk_dm365_setuplinux_2_10_00_11,though I get them at the same time from the Chinese FAE.

  • 0 in reply to NINGMENG CHEN
    TI__Mastermind 37340 points

    My understanding is that the software is still in Early Adoptor stage and is available via FAEs; personally, I do not know where to ge this software yet.  I would ask your local FAE for assistance on this.

  • 0 in reply to Juan Gonzales
    Intellectual 895 points

    Thanks
    I rebuild the ko. Now the ko is loaded successfully as follow:

    root@192.168.1.161:~/decode# ./loadmodules_sd.sh
    ioremap_nocache(0x85400000, 46137344)=0xd5000000
    allocated heap buffer 0xd5000000 of size 0x3bf000
    cmem initialized 15 pools between 0x85400000 and 0x88000000
    irqk initialized
    Module                  Size  Used by
    dm365mmap               5336  0
    irqk                    8328  0
    edmak                   8680  0
    cmemk                  26276  0


    But when I execute the decode demo, it has error as follow:




    root@192.168.1.161:~/decode# ./decode -v davincieffect_ntsc.264 -y 1
    Decode demo started.
    davincifb davincifb.0: check_fb_var: calling convert_fb_var_to_osd, var->bits_pe
    r_pixel: 4
    davincifb davincifb.0: convert_fb_var_to_osd: bits per pixel: 4
    davincifb davincifb.0: davincifb_set_par: calling convert_fb_info_to_osd bpp: 4
    davincifb davincifb.0: convert_fb_info_to_osd: var->bits_per_pixel: 4
    davincifb davincifb.0: convert_fb_var_to_osd: bits per pixel: 4
    davincifb davincifb.0: check_fb_var: calling convert_fb_var_to_osd, var->bits_pe
    r_pixel: 4
    davincifb davincifb.0: convert_fb_var_to_osd: bits per pixel: 4
    davincifb davincifb.0: davincifb_set_par: calling convert_fb_info_to_osd bpp: 4
    davincifb davincifb.0: convert_fb_info_to_osd: var->bits_per_pixel: 4
    davincifb davincifb.0: convert_fb_var_to_osd: bits per pixel: 4
    DavinciDisplay DavinciDisplay.1: Before finishing with S_FMT:
    layer.pix_fmt.bytesperline = 736
    layer.pix_fmt.width = 720
    layer.pix_fmt.height = 480
    layer.pix_fmt.sizeimage =529920
    DavinciDisplay DavinciDisplay.1: pixfmt->width = 720
    layer->layer_info.config.line_

    length = 736
    CMEMK Error: Failed to find a pool which fits 1548288
    CMEM Error: getPool: Failed to get a pool fitting a size 1548288
    Failed to allocate memory.
    Error: Failed to create BufTab for display pipe


    I add the 1548288 pools number in the loadmodules_sd.sh, it's still wrong. The memory in the bootargs is mem=70M.

    root@192.168.1.161:~/decode# ./decode -v davincieffect_pal.mpeg4 -y 1
    Decode demo started.
    davincifb davincifb.0: check_fb_var: calling convert_fb_var_to_osd, var->bits_pe
    r_pixel: 4
    davincifb davincifb.0: convert_fb_var_to_osd: bits per pixel: 4
    davincifb davincifb.0: davincifb_set_par: calling convert_fb_info_to_osd bpp: 4
    davincifb davincifb.0: convert_fb_info_to_osd: var->bits_per_pixel: 4
    davincifb davincifb.0: convert_fb_var_to_osd: bits per pixel: 4
    davincifb davincifb.0: check_fb_var: calling convert_fb_var_to_osd, var->bits_pe
    r_pixel: 4
    davincifb davincifb.0: convert_fb_var_to_osd: bits per pixel: 4
    davincifb davincifb.0: davincifb_set_par: calling convert_fb_info_to_osd bpp: 4
    davincifb davincifb.0: convert_fb_info_to_osd: var->bits_per_pixel: 4
    davincifb davincifb.0: convert_fb_var_to_osd: bits per pixel: 4
    DavinciDisplay DavinciDisplay.1: Before finishing with S_FMT:
    layer.pix_fmt.bytesperline = 736
    layer.pix_fmt.width = 720
    layer.pix_fmt.height = 480
    layer.pix_fmt.sizeimage =529920
    DavinciDisplay DavinciDisplay.1: pixfmt->width = 720
    layer->layer_info.config.line_length = 736
    VICP2: IMX0 Granted
    VICP2: MJCP Granted
    Error: Failed to decode video buffer
  • 0 in reply to NINGMENG CHEN
    TI__Mastermind 41665 points

    The current early adopter DVSDK has two loadmodules script files one for SD and one for HD, you may want to try running the loadmodules_hd.sh instead as it looks like the demo application you are trying to run is attempting to allocate buffers of a size that is allocated in a larger number in the HD version of the script (SD allocates 8x1548288 whereas the HD allocates 22x1548288 for the version I have). Please keep in mind that this is early adopter software and is still under development so this may all change within the next month or so as we work toward an initial beta release.

  • 0 in reply to Bernie Thompson TI
    Intellectual 895 points

    Yes,when I run the loadmodules_hd.sh, it's ok now. Thanks.

    Lemon Chan
    Beijing AVSolution Technology Co.,Ltd