Fw: "partnew" Command Writes Wrong Ending Cylinder in MPT

[Steve Burtchin]

----- Original Message ----- 
From: "Steve Burtchin" <address@hidden>
To: "adrian15" <address@hidden>
Sent: Thursday, January 25, 2007 4:14 PM
Subject: Re: "partnew" Command Writes Wrong Ending Cylinder in MPT


>
> ----- Original Message ----- 
> From: "adrian15" <address@hidden>
> To: "sburtchin" <address@hidden>
> Sent: Sunday, January 21, 2007 7:07 AM
> Subject: Re: "partnew" Command Writes Wrong Ending Cylinder in MPT
>
>
> > Did you receive this email?
> Yes.
>
> >
> > adrian15
> >
> > sburtchin escribió:
> >
> > > I would like to propose the following code change to allow blanking of
a
> > > partition table entry
> > > with a command like "partnew (hd0,3) 0x00 0 0".  New/Changed code is
in
> > > bold.
> > >
> > >    /* Convert a LBA address to a CHS address in the INT 13 format.  */
> > >    auto void lba_to_chs (int lba, int *cl, int *ch, int *dh);
> > >    void lba_to_chs (int lba, int *cl, int *ch, int *dh)
> > >      {
> > >        int cylinder, head, sector;
> > >
> > >        if (lba <= 0)
> > >        {
> > >          *cl = 0;
> > >          *ch = 0;
> > >          *dh = 0;
> > >         }
> > >         else
> > >         {
> > >          sector = lba % buf_geom.sectors + 1;
> > >          head = (lba / buf_geom.sectors) % buf_geom.heads;
> > >          cylinder = lba / (buf_geom.sectors * buf_geom.heads);
> > >
> > >          if (cylinder >= buf_geom.cylinders)
> > >           cylinder = buf_geom.cylinders - 1;
> > >
> > >          *cl = sector | ((cylinder & 0x300) >> 2);
> > >          *ch = cylinder & 0xFF;
> > >          *dh = head;
> > >        }
> > >      }
> >
> > Hi sburtchin,
> >
> > Sorry about the late answer I've been very busy with Super Grub Disk and
> > university lately.
> >
> > Currently grub legacy source code is frozen. Developers only work on
> > grub2 which promises to be smarter and better than grub legacy.
> >
> > I am however working on a kind of grub legacy fork named Super Grub
Disk.
> >
> > You have proposed us a source code change. Have you tried it yourself or
> > were not you able to build a grub floppy or cdrom with the changes?
> I did'nt know how.  I will have to read up on that.
>
> > I can build a SGD cdrom with the changes if you want to and you can try
> > yourself to see if it works ok or not.
> That would be ideal!
>
> > About the bug do you think that if we write:
> >
> >        if (cylinder >= buf_geom.cylinders)
> >          cylinder = buf_geom.cylinders - 1;
> >
> > like this:
> > // cylinders correction
> > buf_geom.cylinders+=2;
> > if (cylinder >= buf_geom.cylinders)
> >          cylinder = buf_geom.cylinders - 1;
> >
> > we will fix the bug?
> >
> > adrian15
> >
> That would definitely provide a workaround to the bug on MY COMPUTER  (for
> the "partnew" command) provided "buf_geom.cylinders" is not used anywhere
> else (I don't think that it is, but its value may be).  "Fix" is a
stronger
> word implying that "buf_geom.cylinders" would be assigned the correct
value
> to begin with.  I still have'nt figured out how that gets assigned, or if
> its value gets used anywhere else that matters.  The data in the
"buf_geom"
> structure seems to get passed around quite a bit (eg. the "geometry"
> function reports 1021 cylinders also [here its the value of
> "geom.cylinders"], if that matters?).
>
> I suspect the source of the bug was a 'dirty' fix to some earlier bug.  A
> safer approach might be:
>
>
> int cylinder, head, sector, bufgeomcylinders;
>
>  bufgeomcylinders = buf_geom.cylinders
>  bufgeomcylinders+=2;
>  if (cylinder >= bufgeomcylinders)
>           cylinder = bufgeomcylinders - 1;
>
>
> leaving the "buf_geom" structure unchanged, and avoiding the potential of
> awakening that earlier bug.
>
> There are far too many global identifiers in GRUB legacy, adding greatly
to
> the risk that one bug fix will cause disastrous consequences somewhere
else.
> I am hoping that GRUB2 development will place a strong emphasis on
> identifier scope so that future bug fixes won't need a 'Band-Aid' solution
> or discussions about what is safe.  The other BIG advantage of keeping
> things local is that novices like me could contribute with a much quicker
> learning curve.  I hope I'm not being too critical, but the more well
> structured the code, the more people like me who want to contribute, can.
>