Subject: Re: problem installing NetBSD-1.1 on hp425t
To: Jason Thorpe <thorpej@nas.nasa.gov>
From: Don Lewis <gdonl@gv.ssi1.com>
List: port-hp300
Date: 12/04/1995 14:36:46
On Dec 3,  4:07pm, Jason Thorpe wrote:
} Subject: Re: problem installing NetBSD-1.1 on hp425t
} On Sun, 3 Dec 1995 01:07:10 -0800 
}  gdonl@gv.ssi1.com (Don Lewis) wrote:
} 
}  > Even a 4K/1K filesystem isn't stable.  I ran into problems when trying
}  > to build a new kernel.  In one case, "ld -r -o fpsp.o" failed due to an
}  > I/O error.  Re-running make got past this OK, but I got frequent panics:
}  > 
}  > panic: ffs_valloc: dup alloc
}  > panic: ifree: freeing free inode
}  > panic: blkfree: freeing free block
}  > 
}  > All these problems were on the root filesystem and went away when I made
}  > /tmp an mfs filesystem.  After I did this, I got another worrysome
}  > sounding error message
}  > 
}  > swap_pager_clean: clean of page ff8f20000 failed
}  > 
}  > but the kernel build succeeded.
} 
} I can't help but wonder if your disk is bad or if the SCSI chain is flakey 
} somehow.  The I/O error, filesystem problems, and swap problems could all 
} be attributed to something like that.

It appears to have been a bum disk, but why didn't the kernel report
any SCSI errors?  The only manifestations were error returns in userland
and silent data corruption on the disk that eventually resulted in
panics.  If the disk was just corrupting data on it's own while reporting
that it was happy, then open() shouldn't fail with EIO.  When I ran
dump, it reported some unreadable blocks (the block numbers looked OK),
but the kernel didn't complain, and I could run dd on on the raw disk
without error.