diff options
Diffstat (limited to 'pcap-bpf.c')
| -rw-r--r-- | pcap-bpf.c | 18 |
1 files changed, 17 insertions, 1 deletions
@@ -2758,7 +2758,11 @@ get_if_flags(const char *name, bpf_u_int32 *flags, char *errbuf) strncpy(req.ifm_name, name, sizeof(req.ifm_name)); if (ioctl(sock, SIOCGIFMEDIA, &req) < 0) { if (errno == EOPNOTSUPP || errno == EINVAL || errno == ENOTTY || - errno == ENODEV || errno == EPERM) { + errno == ENODEV || errno == EPERM +#ifdef EPWROFF + || errno == EPWROFF +#endif + ) { /* * Not supported, so we can't provide any * additional information. Assume that @@ -2773,6 +2777,18 @@ get_if_flags(const char *name, bpf_u_int32 *flags, char *errbuf) * So, just as we do for some ethtool ioctls * on Linux, which makes the same mistake, we * also treat EPERM as meaning "not supported". + * + * And it appears that Apple's llw0 device, which + * appears to be part of the Skywalk subsystem: + * + * http://newosxbook.com/bonus/vol1ch16.html + * + * can sometimes return EPWROFF ("Device power + * is off") for that ioctl, so we treat *that* + * as another indication that we can't get a + * connection status. (If it *isn't* "powered + * off", it's reported as a wireless device, + * complete with an active/inactive state.) */ *flags |= PCAP_IF_CONNECTION_STATUS_NOT_APPLICABLE; close(sock); |