I've put new 9xeninst binaries on sources which should replace the
spurious "file does not exist" message with a more accurate (but still
not very helpful) "i/o error". I'm sorry I can't provide a better
diagnostic, but the xen virtual block device just returns BLKIF_RSP_OKAY
or BLKIF_RSP_ERROR.
I can't spot anything obvious. I'm not aware of anyone trying
Plan 9 on xen on NetBSD before, so maybe there's something different
in that environment.
> -rw-r--r-- 1 root wheel 1073741824 Sep 26 08:08 plan9.img
You could try 'chmod 666 plan9.img', just in case it's a permission thing.
> > When I specify the entire path to /bin/disk/fdisk, the "'./disk' file does
> > not exist" part doesn't happen.
> > /bin/disk/fdisk: diskread 68 at 0.446:
>
> Yes, this shows that %r in print(2) can be misleading if errstr
> hasn't been cleared from a previous error. The "file does not exist
> When I specify the entire path to /bin/disk/fdisk, the "'./disk' file does
> not exist" part doesn't happen.
> /bin/disk/fdisk: diskread 68 at 0.446:
Yes, this shows that %r in print(2) can be misleading if errstr
hasn't been cleared from a previous error. The "file does not exist"
message is
When running the plan9 installer in Xen, partdisk fails strangely:
init: starting /bin/rc
/bin/dossrv: serving #s/dos
dev 3 sector 18, write: 0, should be 4608
dev 3 sector 0, write: 0, should be 4608
dev 3 sector 18, write: 0, should be 4608
dev 3 sector 9, write: 0, should be 4608
dev 3 sector 4
