[gentoo-user] Re: Flashing hardware via WINE ?

[Kai Krakow]

Am Sun, 19 Mar 2017 11:38:56 +0100
schrieb tuxic@××××××.de:

> On 03/19 10:27, Neil Bothwick wrote:
> > On Sun, 19 Mar 2017 09:09:51 +0100, tuxic@××××××.de wrote:
> >
> > > I have a smart NiMH-charger with serial port (normally used to
> > > dump chargeing curves to the PC).
> > > The chargers firmware can bei flashed with a flashtool provided
> > > by the vendor. The communication is via serial port. I have
> > > a PCI=>serial.ports-card installed in my PC.
> > >
> > > The command
> > >
> > >     file <flashing tool>
> > >
> > > results in this information
> > >
> > >     PE32 executable (GUI) Intel 80386, for MS Windows
> > >
> > > . As Linux user by heart I have no Windows.
> >
> > Have you tried contacting the manufacturer to see if flashing from
> > Linux is possible?
> >
> >
> > --
> > Neil Bothwick
> >
> > I am MODERATOR of BORG. Follow the rules or be assimilated.
>
> Hi Neil,
>
> yes...no chance.
> Linux is insecure...you know.
> It makes it possible to spy the firmware and decrypt it on
> the way to the charger.
>
> Windows is much more secyre.
>
> Earth is flat and there was no landing on the moon.
>
> I believe in Santa Claus.

And then there is Wireshark in Windows, with libpcap for USB... It can
easily decode and record all USB frames. I successfully used it some
time ago to debug a custom made USB board and found bugs in the
firmware, which I had access to that way. In the end, there was a bug
in the interrupt handler: a race condition overwriting the USB
communication buffer too early and thus sending back wrong data in the
USB HID frames back to the PC (part old data, part new data). The bug
was uncovered by our implementation using a highly optimized threaded
implementation to use the maximum bandwidth the device could provide.
It wasn't possible to convince the firmware programmer about the bug
(which was really expensive for our customer) and the project was
eventually stopped. Strangely, they later provided a firmware without
this failing behavior, much later. Project still paused. But the
preliminary implementation now runs and show very high time resolution
of measurement data (it's an electrochemical spectrometer, just if
you're curious).

Just that in Linux tools are available more easily doesn't mean that
Windows is more "secure". ;-)

Choose your partners wisely.

-- 
Regards,
Kai

Replies to list-only preferred.