7 Apr
2015
7 Apr
'15
11:42 a.m.
On 2015-04-07 19:21, Pete Turnbull wrote:
On 07/04/2015 16:25, Paul Koning wrote:
I worded it poorly. The driver provides the data to remap blocks, but
doesn't necessarily use a replacement when it gets an error. As I wrote
elsewhere, there are two mechanisms: the files called FILE.BAD being
analogous to the file BADBLK.SYS in RSX, and the bad block substitution
table in the home block which is used when you perform normal file
operations - and which only exists for devices such as RL01/02 and
RK06/07, and is stored in the driver's memory-resident tables. I don't
know in detail how substitution happens, but in RT-11, it does.
This sounds identical to how OS/8 does it. It essentially treats the RL
disks as have a few less blocks than they have. And then the actual
block numbers are contiguous, but any bad blocks are skipped over in the
device driver. So at the file system level, it appears as if the device
have no bad blocks.
Johnny
> On Apr 7, 2015, at 8:41 AM, Pete Turnbull
> <pete at dunnington.plus.com> wrote:
DEC does
it by having the driver software remap sectors.
No, DEC does not do that. At least not in RSX, as Johnny described,
nor in RSTS.