Jump of few hundreds ms in timestamp value with an acquisition at 1MHz. Run included 10000 samples.

  2555 0.82614179563085943236  0.00000000002539062500 seq: 2555 timestamp: 0.828697 raw: 00000000:062c9d1b:00000743, debug: 00009fb8 25.39062499988631316228
  2556 0.82614179573437507287  0.00000000010351562500 seq: 2556 timestamp: 0.828698 raw: 00000000:062c9d98:00000778, debug: 00009fc8 103.51562499988631316228
  2557 1.69460681297656257094  0.86846501724218738705 seq: 2557 timestamp: 1.697164 raw: 00000001:0531bc04:000009f4, debug: 00009fd8 868465017242.18737792968750000000
  2558 1.69460681300000004512  0.00000000002343750000 seq: 2558 timestamp: 1.697165 raw: 00000001:0531bc81:00000a00, debug: 00009fe8 23.43749999988631316228
  2559 1.69460681302539062365  0.00000000002539062500 seq: 2559 timestamp: 1.697166 raw: 00000001:0531bcfe:00000a0d, debug: 00009ff8 25.39062499988631316228

Probably related to #7 (closed)

The acquisition was done only for 10000 samples, so the buffer overflow should not happen.

If there was a buffer overflow, I would expect the sequence ID to be non contiguous.