> I have fooled with this for some time. I am using the Debian GNU/Hurd > packages from the ftp site, as per recommendations. The kernal itself > seems to boot fine, but I get this error right before it starts the > exec server: > > hd12: bad access: block=28, count=2 > end_request: I/O error, dev 03:42, sector 28
That is an error from the disk driver. Does it really say `hd12'?? > However, it does go on: > > Hurd server boostrap: ext2fs.static[hd1s1] execext2fs.static: > ../../libdiskfs/boot-start.c:412: diskfs_execboot_fsys_startup: Unexpected > error: No such file or directory. That error means it failed to find /servers/exec (which should be an empty file, but must exist) in the root partition. Given your others problems, I think that is in fact because of the disk driver problem crashing the filesystem. > memory_object_data_request(0x0, 0x0, 0x2000, 0x1000, 0x1) failed, > 268435459 That is a message from the kernel saying that a pager (one of the filesystem's internal pagers) didn't respond properly. It is most likely just more fallout from the problem with the disk. > I have an ugly suspicion that these are linked. What exactly does this > mean, and does anyone have suggestions for fixing it? > > Also, while detecting my disks, the kernel claims that it has 'failed to > initalized WD-700 SCSI card!' (number maybe different, I only caught it on > the fly as it shot by). Is this autodetection, or should I disable it > somewhere? I get that too. I think it's autodetection of something I don't have. You could always try building a kernel with just the devices you need.
