Thanks for the report. Can you provide your DiagReport output and any info on whether other pcap functions are working for you? MODE_CAPT is the default, so unless the descriptor was in MODE_STAT previously, it shouldn't make a difference whether that function is called. (This is an implementation detail and should not be taken to mean that pcap_setmode will never need to be called in other versions, releases, or forks. I mentioned it because it may help us with debugging only.)

Which exactly other pcap functions are you interesting in? There are several at initialization process (I'm using that over pcap++, they have related issue for this seladb/PcapPlusPlus#680 which I came from) like pcap_create, pcap_set_timeout, pcap_set_buffer_size, pcap_activate, pcap_datalink - they are ok.
Wireshark capture is working fine on the same setup.
Diag report is attached.
DiagReport-20211102-183847.txt

This is a regression in Npcap 1.50 and 1.55 due to the change in f82685f which assumed incorrectly that each MODE_* constant was a set bit in the mode parameter. MODE_CAPT is actually defined as 0, so pcap_setmode(p, 0) is rejected. Since this call is effectively a no-op for capture handles that have not previously been set to MODE_STAT, it did not affect the software we were testing with. We will investigate better tests to avoid similar regressions in the future.

The handler for BIOCSMODE IoControl has been rewritten for the next release, and this bug was not reproduced. The next release of Npcap will not be affected. As a workaround, software can avoid calling pcap_setmode with values other than MODE_STAT.

Fixed in Npcap 1.60.