Beware. I've discovered, quite by accident (by inheriting a test system that had previously been used for TX testing), that configuring a Clearview IP tunnel on a labeled system results in the following panic:
> ::status debugging crash dump vmcore.2 (32-bit) from whitestar3-1.East.Sun.COM operating system: 5.11 clearview-build:01/09/09 (i86pc) panic message: BAD TRAP: type=e (#pf Page fault) rp=ce74bb88 addr=14 occurred in module "ip" du e to a NULL pointer dereference dump content: kernel pages only > $c ip_bind_connected_v4+0x215(d3d24c00, ce74bc60, 4, c896b8d8, 0, 2339080a) ip_bind_v4+0x1eb(ce0ef958, c8049d40, d3d24c00, 0) iptun_bind+0x21f(d3d25f08, c5f26f00, d1b9b490, fe9a89ed) iptun_create+0x172(c5f26f00) iptun_ioc_create+0x21(d29831c0, 80472a0, 100003, d1b9b490, ce74bef8, d29831c0) drv_ioctl+0xf8() cdev_ioctl+0x31(2c80000, 454a0001, 80472a0, 100003, d1b9b490, ce74bef8) spec_ioctl+0x6b(c85fb300, 454a0001, 80472a0, 100003, d1b9b490, ce74bef8) fop_ioctl+0x49(c85fb300, 454a0001, 80472a0, 100003, d1b9b490, ce74bef8) ioctl+0x171() sys_call+0x1aa() I'll file a bug, but I thought you should be warning (particularly if you're sucking in the clearview bits into other project gates to do more comprehensive testing). -Seb
