[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1321732145.2885.163.camel@deadeye>
Date: Sat, 19 Nov 2011 19:49:05 +0000
From: Ben Hutchings <ben@...adent.org.uk>
To: Greg KH <gregkh@...e.de>
Cc: linux-kernel@...r.kernel.org, stable@...r.kernel.org,
greg@...ah.com, torvalds@...ux-foundation.org,
akpm@...ux-foundation.org, alan@...rguk.ukuu.org.uk,
linux-wireless@...r.kernel.org, b43-dev@...ts.infradead.org,
roman-vl@...a.ua,
Rafał Miłecki <zajec5@...il.com>
Subject: Re: [17/26] b43: refuse to load unsupported firmware
On Sat, 2011-11-19 at 11:05 -0800, Greg KH wrote:
> On Sat, Nov 19, 2011 at 05:22:10PM +0000, Ben Hutchings wrote:
> > On Fri, 2011-11-18 at 16:02 -0800, Greg KH wrote:
> > > 3.1-stable review patch. If anyone has any objections, please let me know.
> > >
> > > ------------------
> > >
> > > [This patch is supposed to be applied in 3.1 (and maybe older) branches only.]
> > >
> > > New kernels support newer firmware that users may try to incorrectly use
> > > with older kernels. Display error and explain the problem in such a case
> > >
> > > Signed-off-by: RafaÅ MiÅecki <zajec5@...il.com>
> > > Signed-off-by: Greg Kroah-Hartman <gregkh@...e.de>
> > >
> > > ---
> > > drivers/net/wireless/b43/main.c | 7 +++++++
> > > 1 file changed, 7 insertions(+)
> > >
> > > --- a/drivers/net/wireless/b43/main.c
> > > +++ b/drivers/net/wireless/b43/main.c
> > > @@ -2508,6 +2508,13 @@ static int b43_upload_microcode(struct b
> > > b43_print_fw_helptext(dev->wl, 1);
> > > err = -EOPNOTSUPP;
> > > goto error;
> > > + } else if (fwrev >= 598) {
> > > + b43err(dev->wl, "YOUR FIRMWARE IS TOO NEW. Support for "
> > > + "firmware 598 and up requires kernel 3.2 or newer. You "
> > > + "have to install older firmware or upgrade kernel.\n");
> > > + b43_print_fw_helptext(dev->wl, 1);
> > > + err = -EOPNOTSUPP;
> > > + goto error;
> > [...]
> >
> > Again, I strongly object to this way of handling the incompatibility.
> > The revision check is fine, but the solution must be to assign a
> > different filename for the newer revisions. The changes to 3.0/3.1
> > should be deferred until this is resolved.
>
> You can object, however the driver author and maintainer wants it done
> this way for now, so that's a valid solution that I have to accept. It
> solves real bugs that people have already hit when running the 3.0 and
> 3.1 kernels.
The revision check fixes a bug, yes. What I'm concerned about is the
cop-out error message: 'We didn't deal with versioning properly, so you
get to solve the problem.' It should be possible to reboot between
multiple installed kernel versions without modifying files.
> Propagating a userspace tool change for those systems
> would be much harder, if not impossible, don't you think? It would also
> require a much more intrusive change to the driver infrastructure to
> handle such a proposed change.
It is the new kernel version and firmware files that should be changed,
not the old ones.
There is another option which is to use the /lib/firmware/<kernel-ver>
directories to separate incompatible firmware versions. That is not
nearly as good as using a API/ABI number in the filename, but it would
work and the error message could suggest it.
Ben.
--
Ben Hutchings
The world is coming to an end. Please log off.
Download attachment "signature.asc" of type "application/pgp-signature" (829 bytes)
Powered by blists - more mailing lists