-----BEGIN PGP SIGNED MESSAGE-----
Alex Karpovic wrote:
>> That in mind, what's wrong with bpatch? I've used it for binary patching, it
>> works just fine that that (if my first assumption is totally off-base). You
>> download from the device, change any required data, and (if the device allows
>> writes) write it back to the device. Of course, not all devices allow
> It's a bit uncomfortable. Just for example: I need to search some
> signatures, which could be anywhere in 640Gb disk, and make some
> changes around them. And I don't have spare 640+ Gb to copy whole disk
> to. And even if I would have enough space, it is painfully slow to
> move 640Gb twice just to make ten minutes editing.
That's an unusual requirement, but folks ought to listen here, because
optimizing such a problem, it's an interesting challenge. There is NO
established tool which will do such an outre' task well, just because it's so
unusual) I won't probe into your reasons, although a request so very odd
usually means that there's some misunderstanding at the back of it.
Anyhow, if I were given this task, I really think that the problem is in
localizing the area you need to change, not in changing it. I'd use whatever
language you feel comfortable with, then using that language (either directly,
or by piping dd or nc to help out) so that you could do a global search for your
target. Your search could trivially do extra things, like uniquely identifiying
the target area, even dumping surrounding blocks into work file, so you could
follow up with bpatch to actually change things.
Don't expect such a thing to go quickly ... however, this is one of those tasks
that can be made to operate significantly quicker, if you choose an efficient
language and (easily as important) choose a good search/comparison algorithm.
Actually, this sort of thing mgiht well have been given as homework to an
undergrad, a very good learning opportunity indeed. Lot's of room for
optimization of all kinds, and that task is big enough to really show obvious
Done wrong, with tools bent into shape, this task is really too large to be
reasonably contemplated. Unless you have a few extra months to use waiting for
results, and you'd have to keep your mitts off the disk in the meanwhile. Just
not a good idea to take that approach.
> email@example.com mailing list
> To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (FreeBSD)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
-----END PGP SIGNATURE-----
firstname.lastname@example.org mailing list
To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"