[TiLP-devel] Re: [TiLP-users] direct USB link support {Scanned for spam & virus}

[Romain <ro...@li...>]

Hi Benjamin,

I have updated your doc about FLASH transfers.

roms.

Benjamin Moody a écrit :
> Hi,
>
>> in the request to send packet header, there is a '09 01' (noted as ??).
>
> I now suspect the number 00 00 00 09 is the size of the variable to be
> sent.  I don't know about the 1.
>
>> I have already seen this 2-bytes number sequence in the D-BUS 'Version'
>> packet.
>
> Perhaps that is the current language?  09 01 is English (defined by the
> SDK; I think it's also a Windows language ID.)
>
> Initial thoughts on the new logs:
>
> It looks like app transfer works exactly like variable transfer, with the
> data type 0xF0070024.  Surprisingly, it also looks like you can pad the
> end of the app with zeroes.
>
> It doesn't appear that anything actually got sent in "send_cert.pkt", but
> it looks like that will be normal variable transfer as well.
>
> Parameter 0x31 is... interesting.  It appears to contain:
>
> - Certain fields from the calculator's certificate area.  Specifically,
> the calculator ID, calculator public key, ID validation bytes, date stamp
> and signature.  Except that the calculator public key is zeroed out.
>
> - The current OS header.
>
> - Current app headers.
>
> - An RSA signature (02 0D block) of something.
>
> - The last 768 bytes of the certificate area.
>
> About the OS sending protocol:
>
> It seems it is a little more complicated than I thought.  First, the 0002
> packet contains the complete length of the data to be transferred
> (including the address, page and flag bytes at the start of each packet.)
>
> Second, the header and signature should be set to the actual addresses
> (page 7A, address 4000 and 4100 respectively) rather than to address zero
> as in the DBUS protocol.  The flag byte should still be set to 80.
>
> Third, it seems you must start by sending the OS header as an 0004 packet.
>  This packet is identical to an 0005 packet, with address 4000, page 7A,
> and flag 80, and in fact it seems you must send the same data again as the
> first 0005 packet.  The 0004 packet will be acknowledged by a second 0003.
>
> Don't worry about the buffer size request in the middle of the transfer.
> That's probably the point when the OS transfers control to the boot code.
>
> Benjamin
>
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting
> language
> that extends applications into web and mobile media. Attend the live
> webcast
> and join the prime developer group breaking into this new coding
> territory!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
> _______________________________________________
> TiLP-users mailing list
> TiL...@li...
> https://lists.sourceforge.net/lists/listinfo/tilp-users
>


-- 
Romain Liévin :		<ro...@li...>
Web site :		http://www.lievin.net
"Linux, y'a moins bien mais c'est plus cher !"