sdm: Magic header check fails

Some scat users have pointed out to me that the magic header check often fails:

Line 276 in 6e3fd14

print('Skipping packet as magic does not match')

Gathering some info, this table came up:

Model	Modem	ICD	Log Type	Magic header check	Magic2 header
A30 A307F	S327	5.32	usb dm	KO	0x0172
A40 A405F	S327	5.32	usb dm	KO	0x0173
S6 G925F	S333	4.91	usb dm	KO	0x0172
S9 G960F	S360	5.33	usb dm	KO	0x0175
S9 G965F	S360	5.33	cplog	KO	0x0000
Note 9 N960B	S360	5.33	usb dm	OK	0x0174
Note 10 N976B	S5100	6.10	usb dm	KO	0x0172
S20 G981B	S5123	6.07	usb dm	KO	0x0172
S21 G998B	S5123A	6.15	silent log	KO	0x0177
S21 G996B	S5123A	6.22	silent log	KO	0x0183
S22 S908B	S5133	7.00	silent log	KO	0x0184
Pixel 7a	S5300	7.01	vendor log	KO	0x0188
Pixel 7	S5300	7.01	vendor log	OK	0x0186
Pixel 6a	S5123	//	vendor log	KO	//
Pixel 6a	S5123	6.22	vendor log	OK	0x0186
Pixel 6	S5123	6.22	triggered manually from adb	KO	0x017c

Note that only one sample has been tested for each row and it looks like different logs of the same (or similar) device may get different results.

What do you think about turning that skip into a warning?
In all these cases disabling that check made the parser do its job (except some issues with timestamps similar to #73)

Thanks for the extensive research. I also found that even on the same device the second "magic" check changes on the same device with baseband/system firmware update. As I was not sure on how to deal with it, I think I can change the check to just debug information instead of adding all possible values..

I have updated the table adding the magic2 header.
Looks like some kind of versioning? Older logs have lower values

Then it will be better to skip the check, as this value is not used in parsing anyway.

Disabled check through 034248c.

I did further research, that magic is a timestamp 🙃

length: ac 01
magic: 39 7f
timestamp1: b3 80 8a 85 72 01 = 1591378346163 = Friday 5 June 2020 17:32:26.163
seqnr: 05 29
direction: a0
group: 22
command: 52 
timestamp2: 3c a6 fa 1d = 502965820 = 503s (since modem boot? since log start?)
payload: ...