[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH] Fix when installing on pationless but partionable medium

From: Robert Millan
Subject: Re: [PATCH] Fix when installing on pationless but partionable medium
Date: Wed, 22 Jul 2009 19:16:12 +0200
User-agent: Mutt/1.5.18 (2008-05-17)

On Sun, Jul 19, 2009 at 11:07:41AM +0100, Colin Watson wrote:
> On Sat, Jul 18, 2009 at 09:17:41PM +0200, Robert Millan wrote:
> > On Sat, Jul 18, 2009 at 09:01:38PM +0200, Vladimir 'phcoder' Serbinenko 
> > wrote:
> > > On Sat, Jul 18, 2009 at 8:45 PM, Robert Millan<address@hidden> wrote:
> > > > I might be missing something about this check, but GRUB doesn't require 
> > > > that
> > > > the boot flag is present.  Therefore, its non presence doesn't imply 
> > > > this is
> > > > not a real msdos label.
> > > 
> > > He refers to boot flag as a byte in msdos structure which can only be
> > > 0x00 (not set) or 0x80 (set)
> > 
> > Yes.  GRUB's boot.img doesn't do anything with it AFAICT.
> That's as may be, sure; but checking that that byte is one of the two
> permitted values in all four partitions happens to be a good sanity
> check for whether it's really an MS-DOS label or in fact something else.

As a bootloader, most of the decisions GRUB takes have a critical effect.  We
can't make GRUB take those decisions based on heuristic.  If we determine
that something ought to be done, and we do it, we're taking responsibility for
it.  Bootloader-related breakage tends to be severe, and we can't afford to
be blamed for it just because "our heuristic got it wrong".

Going down this path, if our heuristic is not good enough, we'll have to
improve it.  We'd make it more complex and check for more things.  But no
matter what we do, it can never be completely reliable.

Similarly, and as I said before (in this thread or another, I'm not sure), the
situation in which someone using a _sane_ MBR label with a sane filesystem in
it is our primary use case, and the ability of GRUB to install in it must not
depend on heuristic.

Robert Millan

  The DRM opt-in fallacy: "Your data belongs to us. We will decide when (and
  how) you may access your data; but nobody's threatening your freedom: we
  still allow you to remove your data and not access it at all."

reply via email to

[Prev in Thread] Current Thread [Next in Thread]