supplierdeeply

eos M10 port

  • 148 Replies
  • 29383 Views
*

Offline axman

  • ***
  • 138
  • Publish
    Re: eos M10 port
    « Reply #140 on: 28 / May / 2017, 18:07:50 »
    Advertisements
    Tested my m10 using your provided build
    Build 4806? It matters when evaluating romlogs.

    Yes, m10-110d-1.5.0-4806-full.zip from topic=12758.msg130850.

    USB hotplug crash logs attached.

    *

    Offline axman

    • ***
    • 138
  • Publish
    Re: eos M10 port
    « Reply #141 on: 28 / May / 2017, 18:09:27 »
    Here's the chdkptp cli I/O error ROMLOGs.

    *

    Offline axman

    • ***
    • 138
  • Publish
    Re: eos M10 port
    « Reply #142 on: 28 / May / 2017, 19:53:04 »
    You need to check for

    "WARNING! Incorrect dissassembly is likely"

    in the output when you rebuild stubs.

    Edit:
    I confirmed that 3.0.5-rc2 does not fix this issue. It was fixed in the "next" branch, but that branch has changed in other incompatible ways. So, you need to build from source with the attached patch.
    @reyalp, thanks for advice.  Capstone is my doom.

    Is there a better thread for discussing capstone?  Or sig finding using capstone in various firmwares?

    *

    Offline reyalp

    • ******
    • 10700
  • Publish
    Re: eos M10 port
    « Reply #143 on: 28 / May / 2017, 20:06:12 »
    Is there a better thread for discussing capstone?  Or sig finding using capstone in various firmwares?
    There's some information in http://chdk.wikia.com/wiki/Capdis_Disassembly_Tool

    Discussion should probably go in the development thread: https://chdk.setepontos.com/index.php?topic=12714.0
    Don't forget what the H stands for.


    *

    Offline srsa_4c

    • ******
    • 3453
  • Publish
    Re: eos M10 port
    « Reply #144 on: 28 / May / 2017, 20:51:29 »
    USB hotplug crash logs attached.
    This one seems displaybusyonscreen related when called from SpyTask (others have also reported crashes related to that fw function). A pointer used by Canon code appears corrupted.
    Here's the chdkptp cli I/O error ROMLOGs.
    Failure to allocate UI memory - also Canon GUI related, but in a fw task. UI memory resides in the malloc heap and has its own suballocator. This might be due to corrupted memory inside the UI memory region.

    Something bad is going on, but I'm not sure if displaybusyonscreen is the cause or something else. Should probably check tasks' stack usage next...

    *

    Offline reyalp

    • ******
    • 10700
  • Publish
    Re: eos M10 port
    « Reply #145 on: 28 / May / 2017, 21:20:01 »
    Previous displaybusyonescreen crash discussion https://chdk.setepontos.com/index.php?topic=12788.msg132368#msg132368

    I also noted this seemed to be associated with play/rec switches.
    Don't forget what the H stands for.

    *

    Offline axman

    • ***
    • 138
  • Publish
    Re: eos M10 port
    « Reply #146 on: 29 / May / 2017, 23:49:24 »
    Reading other posts, not specifically related, but caught my eye - the 'bad' SD card vs 'good' issue.. Fwiw, my M10's SD card is     SanDisk Ultra II sdhc c4 8Gb

    I've used it a lot with chdk.  Lots of partitions, BOOTDISK and SCRIPTDISK written to it..

    I will test M10 again, same failing tests, but with *brand new* card.

    san disk Extreme Plus sdhc-I c10 16Gb 

    Try to rule out a garbage sd card.  Will advise if similar results, will capture romlogs if any.

    *

    Offline srsa_4c

    • ******
    • 3453
  • Publish
    Re: eos M10 port
    « Reply #147 on: 16 / July / 2017, 16:27:21 »
    I have uploaded new builds to the usual place.

    The only M10 related change is that the new builds no longer use the DisplayBusyOnScreen Canon function to erase the screen.
    The result is that faint blinks of "Busy" are no longer visible when using the CHDK user interface.
    This is actually in test phase, so reports about any new problems are welcome.


    *

    Offline srsa_4c

    • ******
    • 3453
  • Publish
    Re: eos M10 port
    « Reply #148 on: 26 / July / 2017, 15:55:58 »
    Builds are updated again. The new screen refresh method should now work better in modes that have Canon animations (focus rectangle, animated icon).
    Note that this does not solve the excessive blinking of CHDK overlay that can be experienced in several recording modes.

    The PTP crash reported by axman might be gone. Or at least, I could not provoke it when I dared to try again (once).

     

    Related Topics