Because in_addr is stored in network byte order, this function could
extract the octets in MSB-to-LSB order or in LSB-to-MSB order. This
didn't matter in the case of resolved names, because the same order was
used when generating the octet array and later when matching against it.
But the function parse_ipv4_ranges, which handles literal IPv4
addresses, always uses octet[0] as the MSB, so comparisons failed in
later matching.
As it was, the code worked on little-endian architectures but didn't on
big-endian.
The patch that moves "config.h" to the top of scanner.l (and hence
scanner.c) has to happen before the files are pregenerated, otherwise
there is an error on AIX:
In file included from scanner.c:2792:
/usr/include/unistd.h:171: error: conflicting types for 'lseek64'
/usr/include/unistd.h:169: error: previous declaration of 'lseek64' was here
Solaris 11 uses BPF packet capture rather than DLPI, which requires
different handling in many cases. The new preprocessor symbol tells when
this is the case; it is additional granularity on top of SOLARIS.
Setting --min-parallelism without also setting --max-parallelism would
force the max parallelism to the same value. So, for example,
--min-parallelism=1 would also limit the max parallelism to 1. This
patch, based on one by Chris Woodbury, allows the max parallelism to
rise above this minimum, up to the maximum defined by each scan phase.
The Nsock select engine stores readiness information in a mask instead
of doing an FD_ISSET against the descriptor every time. This is a good
idea, but it removed a special case in the pre-engines code: in case
PCAP_CAN_DO_SELECT is not defined, we can't do FD_ISSET on the
descriptor because it is -1. Here we check for that occurrence and set
EV_READ in the mask so that a non-blocking pcap read can happen every
iteration.
