pixicore: UEFI boot faild with '"ready": no such image

[teutonet]

Hi. When booting a Supermicro Server in UEFI mode, the boot stops after transferring boot script, kernel and initrd with:

http://172.17.0.2:80/_/booting?mac=0c%3AC4%3A7a%3Ad2%3A89.. Input/Output error (http://ipxe.org/1d0c6598)
"ready": no such image

[danderson]

It's possible those errors are not relevant - those commands are both optional, so even if iPXE prints an error, it's actually executing the rest of the boot script. I suspect that in reality, the boot is starting, and it's crashing in kernel init.

What kernel/initrd are you trying to boot?

[teutonet]

We are trying to boot core-os current as described in pixiecore Readme.md 'Using Pixiecore in static mode'. So you can be right that it is an issue with booting core-os in UEFI mode. Thank you for your support.

[danderson]

I'll need more information in that case:

• What machine are you trying to boot? (type of CPU, motherboard, brand/model name if it's an off the shelf server)
• Please give me the exact commandline you're using to run Pixiecore.
• Please run sudo pixiecore debug tcpdump <interface> issue13.pcap in a separate terminal, then try to boot the machine again with Pixiecore. This will capture the network traffic seen by Pixiecore, so I can see exactly what the machine is trying to do. Please attach the pcap file to this bug, or email it to me at dave natulte.net.

The only time I've seen a boot failure like this in the past, was because the kernel CPU architecture/mode (i386, x86-64, ...) was not exactly the same as the architecture/mode that UEFI was running. The information above will let me verify that everything should be compatible... Or maybe I'll see something else in the pcap...

[teutonet]

hw.txt

issue13.pcap.zip

For hardware information see lshw -short output in attached hw.txt.

[panchoh]

In case it helps, we had the same issue with our HP environment, which we worked around by booting in legacy mode instead of EFI mode.

[danderson]

I'm sorry it's taken so long for me to update this bug.

Looking at the pcap, I don't see anything obviously wrong. The server is advertising that it's an x86_64 UEFI machine, and the boot process happens normally (at least for the pieces that are visible in the pcap).

I don't really have a solution for this :(. Is there any way you can give me temporary access to the hardware in question (so I can run my own pixiecore and packet captures, and use the IPMI console to iterate on debugging) ? It sucks, but that's the only way I can think of to make progress :(.

@panchoh Same question for you, I'd love to debug the EFI issue on your HP machines, but I don't have expensive HP servers myself, so it's ~impossible to hack until I find the solution if I don't have access.

[panchoh]

I'm sorry it's taken so long for me to update this bug.

I think my "sorry" dwarfs yours by a couple orders of magnitude :-P

@panchoh Same question for you, I'd love to debug the EFI issue on your HP machines, but I don't have expensive HP servers myself, so it's ~impossible to hack until I find the solution if I don't have access.

I'm sorry to report that we lost access to that nifty HP hardware years ago, so sadly I couldn't (and can't) accomodate you. Sorry it took me so long to reply.

Thanks a lot for your help and for pixiecore!

Cheers.