Subject: errors and bad144
To: John Vinopal <>
From: Steven Reiz <>
List: current-users
Date: 03/04/1994 11:43:36
>>>>> "JV" == John Vinopal <> writes:

JV> Is this quote (exerpted from the faq) true?  The term "bad sector
JV> map" is referring to the bad144 code?  I was under the impression
JV> that drives when low level formatted, wrote bad sector information
JV> in the sector headers and that bad144 (or DOS fats, etc) were only
JV> needed for those sectors which had errors in the sector header
JV> area.

We run a machine with two ESDI disks, both more than 1024 cylinders.
The bootblock code reports a bad bad144 table, but the kernel loads
the bad144 table correctly, so as long as your kernel isn't located
on mapped out blocks the 1024-cyl problem isn't an issue.

JV> What are the chances of upping bad144 sector size of 126?  I know
JV> its hard coded all over but 126 is pretty small and silly.

That's right, more so because ESDI disks frequently seem to have complete
bad tracks, at, say, 52 sectors/track...
We're running with a patch that makes the kernel interpret the bad144
table as mapping out tracks instead of blocks, allowing 126 bad tracks.

I also have a patched bad144 utility, though I'm not sure it's
100% reliable in moving around already-mapped-out blocks.

I'd be glad to supply the patches to anyone interested.