Re: [colorforth] Booting from HDD

["John Comeau" <jc@xxxxxxxxxxxx>]

Thanks for digging that up, Jason! If you have a compatible GNU build
environment (on Cygwin), you could try moving that "showprogress" call
forward to locate the problem more precisely, but I'm afraid I was
wrong about USB pendrive emulation. As you and someone else indicated,
it must be emulating a hard drive rather than a floppy, and so the
BIOS calls are very likely where it's failing.

On Mon, Mar 17, 2008 at 7:05 AM, Jason Kemp <jason.kemp@xxxxxxxxxxxxxxxx> wrote:
> Good morning all,
>
>  This is from the color.lst
>
>   92 008c E85F00           call whereami ;# set EBP to where we relocated
>   93 008f E89E00           showprogress
>   94 0092 66E80D01         data32 call protected_mode
>   94      0000
>   95                  .code32
>   96 0098 E84F0100         call a20 ;# set A20 gate to enable access to
>  addresses with 1xxxxx
>   96      00
>
>  So it does look as if it is fine up until "call protected_mode" then the
>  problem is somewhere after that point...
>  But please don't spend more time trying to work this one out for me--it
>  works with the USB floppy so I'm happy now!  Of course, if anyone wants
>  to go further then I'm happy to help.
>
>  And anyway, unless floppy emulation can be forced in BIOS, wouldn't the
>  INT 13 read/write routines have to be changed to work on drive 0 instead
>  of first floppy?  Hey, I might try anyway, but these things are never
>  *that* simple are they?
>
>  Thank you,
>  Jason
>
>  Nick Maroudas said the following on 17/03/2008 09:16:
>
>
> > Nick here,
>  >
>  > Thanks Ray, for your boot.asm printout (below).  It
>  > shows a boot process that has already set up video
>  > 4FO2/4117 and copied the 512 byte boot sector before
>  > relocating to "at0:" to copy the remaining code via
>  > floppy then jumping to "start2" via the floppy call
>  > "stop" motor.  For comparison, here is JC's boot.asm.
>  > Unlike yours (or Howerd's original CFDOS4 floppy) this
>  > CD version does not set up video before relocating, and
>  > seems to rely on MSDOS or ElTorito (non-emulation
>  > mode) to have copied all of the code before setting up
>  > the Graphic Output Display ("god"). See lines 98, 99 in
>  > JC's version (below). This is a long way from the
>  > simplicity of Chuck's original boot.asm (which one can
>  > consult on JC's site under "as/", together with a very
>  > helpful commentary by John himself). As Albert van der
>  > Horst has just remarked, "CF is simple but this is the
>  > other side of the coin".  I find that Howerd's
>  > CFDOS4.blk (simply DeeDee'd onto a floppy) boots a new
>  > Pentium D with no OS apart from BIOS, but John's CFDOS4
>  > CD does not.
>  >
>  > IMHO Jason is on the right track: floppy emulation; just
>  > dd a reliable floppy image onto the flash stick
>  > (after all, CF is meant to be self booting). I could
>  > expatiate if anyone is interested.
>  >
>  > Caritas,
>  >
>  > NickM
>  >
>  >
>  >
>  >
>
>
>  ---------------------------------------------------------------------
>
>
> To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
>  For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
>  Main web page - http://www.colorforth.com
>
>



-- 
John Comeau <jc@xxxxxxxxxxxx> http://jc.unternet.net/
"A place for everything, and everything
all over the place"

---------------------------------------------------------------------
To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
Main web page - http://www.colorforth.com