While attempting to install gentoo 1.4 from "live" CD, I get stopped at the part where one is meant to chroot into the new environtment. Everytime I attempt to execute chroot /mnt/gentoo /bin/bash, I only get the response "Invalid Instruction". Reproducible: Always Steps to Reproduce: 1. Boot Virtual PC with Gentoo live CD. 2. Follow install to step 6a. 3. Issue "chroot /mnt/gentoo /bin/bash" command. Actual Results: Message "Invalid argument" appears and the prompt returns. Expected Results: The system should change root and return to a bash shell prompt. I'm using Virtual PC 6.1 on MacOS 10.3.1 on a Titanium PowerBook G4 1 Ghz with 1 GB RAM. I realize this is an unusual "hardware" environment which you probably don't have access to for testing. If you're interested in debugging/fixing this issue with chroot, I can offer to test for you.
what livecd version and what stage ?
Yes, I discovered and am reporting the bug. I was using i686-1.4-20030911-cd1.iso this time. I had tried the Pentium3 version first, but the result was the same for both.
Oh, doesn't matter what stage.. you never get that far. Heh.. sorry about that first bit in the last comment. I'm accustomed to the email sent from bugzilla.mozilla.org and thought the bottom part was you asking a question. I'll try to pay more attention.
we either need to find a way to support these emulators, or take make a official policy of no support
let's support just the ones we could be sure that we can fix or make the upstream dev fix (eg bochs)
i've got the exactly the same problem but on a real PC (VIA EPIA Mainboard and C3-CPU). I tried to install gentoo the first time (by liveCD) and a co-worker had switch off the machine while bootstrap.sh had been working. After that I retried the installation => "Invalid Instruction" I restarted at step 0 by deleting the partition-table but I still get the same bug. It's the same CD it's the same stage and the harddisk should be empty Tomorrow I'll try it with an new harddisk - maybe this will work
You should try blank stages, i.e. those that are compiled for 386 - especially for that via-cpu
we don't currently have the hardware to test this, so I am closing it until we do.