dfu-util: Invalid DFU suffix signature

[ali1234]

When flashing the latest firmware on linux I get this warning message:

dfu-util: Invalid DFU suffix signature
dfu-util: A valid DFU suffix will be required in a future dfu-util release!!!

I am not sure what this means but we should look into it to avoid it breaking at some point in the future.

I suspect it means there is some invalid header on the firmware.dfu.

[Gadgetoid]

The suffix is this portion here:

32blit-sdk/tools/dfu

Lines 117 to 122 in dd9d68d

	'bcdDevice' / Int16ul, # Firmware version, or 0xffff if ignored
	'idProduct' / Hex(Int16ul), # USB product ID or 0xffff to ignore
	'idVendor' / Hex(Int16ul), # USB vendor ID or 0xffff to ignore
	'bcdDFU' / Const(0x011A, Int16ul), # DFU specification number
	'ucDfuSignature' / Const(b'UFD'), # 0x44, 0x46 and 0x55 ie 'DFU' but reversed
	'bLength' / Const(16, Int8ul) # Length of the DFU suffix in bytes

Which, according to this datasheet: http://rc.fdr.hu/UM0391.pdf

● The bcdDevice field gives the firmware version contained in the file, or 0xFFFF if ignored.
● The idProduct and idVendor field give the Product ID and Vendor ID respectively of the device that the file is intended for, or 0xFFFF if the field is ignored.
● The bcdDFU field, fixed to 0x011A, gives the DFU specification number. This value differs from that specified in standard DFU rev1.1.
● The ucSignature field contains a fixed string of three unsigned characters (44h, 46h, 55h). In the file they appear in reverse order, allowing valid DFU files to be recognized.
● The bLength field, currently fixed to 16, gives the length of the DFU Suffix itself in bytes.
● The dwCRC (Cyclic Redundancy Check) field is the CRC calculated over the whole file except for the dwCRC data itself

And to meet the conditions for the CRC:

32blit-sdk/tools/dfu

Lines 125 to 130 in dd9d68d

	DFU = Struct(
	'fields' / RawCopy(DFU_body),
	'dwCRC' / Checksum(Int32ul, # CRC calculated over the whole file, except for itself
	lambda data: 0xffffffff ^ zlib.crc32(data),
	this.fields.data)
	)

This all looks right, insofar as there isn't much to get wrong.

Maybe - contrary to the comments at the top of the DFU tool - the suffix should be included in the file length:

32blit-sdk/tools/dfu

Line 55 in dd9d68d

# size += 16 # DFU Suffix Length

Edit: "DFU File Format" PDF attached here for posterity: DFU-File-Format-UM0391.pdf

[Gadgetoid]

P.S. What's your DFU Util version and flash command?

On Ubuntu I see a different flavour of wrong:

dfu-util -D firmware.dfu -d 0x483:0xdf11 -a 0
dfu-util 0.9

Copyright 2005-2009 Weston Schmidt, Harald Welte and OpenMoko Inc.
Copyright 2010-2016 Tormod Volden and Stefan Schmidt
This program is Free Software and has ABSOLUTELY NO WARRANTY
Please report bugs to http://sourceforge.net/p/dfu-util/tickets/

Opening DFU capable USB device...
ID 0483:df11
Run-time device DFU version 011a
Claiming USB DFU Interface...
Setting Alternate Setting #0 ...
Determining device status: state = dfuIDLE, status = 0
dfuIDLE, continuing
DFU mode device DFU version 011a
Device returned transfer size 1024
dfu-util: Error: File ID 0483:0000 does not match device (0483:df11 or 0483:df11)

IE: The DFU file has 0000 as the Product ID so it will refuse to flash.

[ali1234]

Okay I found the problem. I typed firmware.bin instead of firmware.dfu and apparently this actually works, but produces the warning because no suffix.

[ali1234]

0000 is probably because of this:

https://github.com/32blit/32blit-sdk/blob/master/tools/dfu#L190

[ali1234]

Over in the new DFU tool the default is still 0 for product ID:

https://github.com/32blit/32blit-tools/blob/master/src/ttblit/core/dfu.py#L121

[Gadgetoid]

That's definitely an oversight, should be df11. Me abandoning DFuSe in favour of Linux has not helped this.