[TUHS] 211bsd: kernel panic after a 'here document' in tcsh

Walter F.J. Mueller w.f.j.mueller at retro11.de
Mon Jun 26 02:25:25 AEST 2017


Hi,

two remarks on the issues around FPSIM and tcsh:

I of course wondered by a line like

    mov     $4..,r0

is accepted by 'as', I naively expected that this should cause an error.
I didn't locate the 211bsd 'as' manual, so checked 7th Edition manuals,
which can be found under

   https://wolfram.schneider.org/bsd/7thEdManVol2/

The assembler manual, see
   https://wolfram.schneider.org/bsd/7thEdManVol2/assembler/assembler.pdf

states

    6.1  Expression  operators
         The operators are:
            (blank)     when there  is  no  operand  between  operands,
                        the  effect  is  exactly  the  same  as  if  a
                        ‘+’ had  appeared.

So the lexer sees two tokens

   $4.    --> number
   .      --> symbol for location counter

and, because the default operator is '+', interprets this as

    mov     $4. + . , r0

which ends up being a number in the 160000 to 177777 range.

So 'as' is not to blame, works as designed.

Noel Chippa wrote:
 > I'm fairly amazed that apparently nobody has run across one of these 4 before!
 > (Or, at least, not bothered to report it.)
 > I wonder how long that bug has been in the code?

The answer is: this bug was in 211bsd all the time.
Steven Schultz told me that that they simply didn't have a way to
test FPSIM because all machines had FPP, and the only way of testing
would have been to physically remove the FP11 from a 11/70.


		With best regards,   Walter


More information about the TUHS mailing list