On Wed, Feb 6, 2013 at 10:33 PM, Benjamin Huntsman <[email protected]> wrote: > Network boot from Intel E1000 > Copyright (C) 2003-2008 VMware, Inc. > Copyright (C) 1997-2000 Intel Corporation > > CLIENT MAC ADDR: 00 0C 29 D2 AC AC GUID: 564DA193-9D84-E902-D0E8-F20CCBD2ACAC > CLIENT IP: 10.0.0.133 MASK: 255.255.255.0 DHCP IP: 10.0.0.101 > GATEWAY IP: 10.0.0.1 > Protected-mode bootstrap... > ELCR: 0E00 > cpuidentify: cpuidax 0x306a9 cpuiddx 0xfabfbff > apm ax=f000 cx=f000 dx=40 di=ffff ebx=5770 esi=-1 > Boot devices: fd0 > boot from: > > > I'd think that at this point, I should just be able to type > "ether0!/amd64/9k8cpu", but alas that just gets me another "boot from:" > prompt. One gotcha that I can think of is that my TFTP server is not the > same as my DHCP server. Is there a way to specify the TFTP server? Or am I > missing something really stupidly simple like ppxeload not supporting the > E1000 that VMware Fusion emulates, or that my ppxeload binary may need to be > recompiled? > > Many thanks in advance! > > -Ben >
If ether0 isn't listed under "Boot devices", then the bootloader wasn't able to detect it, so you can't load your kernel over ethernet. Give it a shot with a different network card or try one of the other methods suggested in this thread. If you compile an ELF kernel it's also theoretically multiboot-compatible, so you may be able to boot using something like pxelinux instead of ppxeload. You can boot an amd64 kernel with GRUB, but I had to change one line of 6l's ELF header construction code. john
