10 Mar
2016
10 Mar
'16
6:07 p.m.
On Mar 10, 2016, at 8:51 PM, Don North <north at
alum.mit.edu> wrote:
I was wondering too. But then again, RT11 can do partial sector reads (and even writes).
They go straight from application through the kernel to the driver. (A partial block
write requires the device, or the driver, to zero-fill to the end of the block -- and yes,
some applications require that and will break if it isn't done right, as I found out
debugging the RC11 driver.)
If memory serves, what you have is part 1 of the boot (1 sector, block 0), part 2 (2
sectors, block 2), then something is looking through the directory (sectors 6 and up) and
then the read from block 134 is presumably a file being read (perhaps a program) and the
stuff at 85 and 86 is ???
It might be informative to dump the directory from the tape and analyze it to see what
files those later reads refer to.
paul
...
info: boot unit=0 blk=0x0000 cnt=0x0200
info: read unit=0 sw=0x00 mod=0x00 blk=0x0002 cnt=0x0800
info: read unit=0 sw=0x00 mod=0x00 blk=0x0006 cnt=0x0400
info: read unit=0 sw=0x00 mod=0x00 blk=0x0008 cnt=0x0400
info: read unit=0 sw=0x00 mod=0x00 blk=0x000A cnt=0x0400
info: read unit=0 sw=0x00 mod=0x00 blk=0x000C cnt=0x0400
info: read unit=0 sw=0x00 mod=0x00 blk=0x0134 cnt=0x0A00
info: read unit=0 sw=0x00 mod=0x00 blk=0x0085 cnt=0x0132
info: read unit=0 sw=0x00 mod=0x00 blk=0x0086 cnt=0x3CC8
At this point, the RUN LED goes off and the system halts. It looks like the boot loader
is loading and running the second-level loader on the tape, but something happens after
that. Not sure where to go from here.
These last two read commands look awful suspicious to me. A single logical device block
is 512B (or 0x200) so the count
parameter is 0x200 for one block, 0x400 for two consecutive blocks, etc. 0xA00 is five
consecutive blocks.
However, 0x0132 (306. bytes) is a partial block read, and 0x3CC8 (15560.) is a read of
30.+ blocks. So I would speculate
that the code went south prior to these two read commands. Just a guess of course as I
have never done this particular
task. But I would guess that the image you are using has been corrupted in the process.