So its not clear as to how the API is being used wrong, but for very low rates, every other call is reporting an overflow
Running a simple rate testing loop (which works great on single channel up to very high rates), I get overflow messages every other call to sync_rx() and the timestamp is 0x1234432112344321 every other call as well.
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 1016
[NOTICE] md.timestamp = 0x3
[NOTICE] md.status = 0x30001
[NOTICE] numElems read per channel = 508
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x1234432112344321
[NOTICE] md.status = 0x30001
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x3fb
[NOTICE] md.status = 0x1
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x1234432112344321
[NOTICE] md.status = 0x30001
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x5f7
[NOTICE] md.status = 0x1
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x1234432112344321
[NOTICE] md.status = 0x30001
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x7f3
[NOTICE] md.status = 0x1
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x1234432112344321
[NOTICE] md.status = 0x30001
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x9ef
[NOTICE] md.status = 0x1
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0x1234432112344321
[NOTICE] md.status = 0x30001
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
[NOTICE] readStream called with samples per channel = 4096
[NOTICE] calling bladerf_sync_rx(num_samples = 8192)
[NOTICE] md.actual_count = 508
[NOTICE] md.timestamp = 0xbeb
[NOTICE] md.status = 0x1
[NOTICE] numElems read per channel = 254
[NOTICE] BLADERF_META_STATUS_OVERRUN
If I force the num_samples passed to bladerf_sync_rx() to be 508 (the value actual_count returns), then I don't get the overflows, but every other time is 0x1234432112344321
So its not clear as to how the API is being used wrong, but for very low rates, every other call is reporting an overflow
SoapySDRUtil --rate=1e6 --direction=RX --channels="0,1"