POLA dictates that VLAN_MTU be enabled by default.

In particular, disabling it was likely to break configurations
involving ng_vlan(4) since the latter couldn't control
the parent's VLAN_MTU in the way vlan(4) did.

Pointed out by:	ru
This commit is contained in:
yar 2004-05-27 14:36:58 +00:00
parent 1606045a14
commit 4e00954200

View File

@ -595,6 +595,12 @@ fxp_attach(device_t dev)
/* turn on the extended TxCB feature */
sc->flags |= FXP_FLAG_EXT_TXCB;
/* enable reception of long frames for VLAN */
sc->flags |= FXP_FLAG_LONG_PKT_EN;
} else {
/* a hack to get long VLAN frames on a 82557 */
sc->flags |= FXP_FLAG_SAVE_BAD;
}
/*
@ -820,7 +826,7 @@ fxp_attach(device_t dev)
*/
ifp->if_data.ifi_hdrlen = sizeof(struct ether_vlan_header);
ifp->if_capabilities |= IFCAP_VLAN_MTU;
/* this driver lets vlan(4) control the bit in if_capenable via ioctl */
ifp->if_capenable |= IFCAP_VLAN_MTU; /* the hw bits already set */
/*
* Let the system queue as many packets as we have available