Re: linux-next: Tree for July 30

[Rafael J. Wysocki]

On Thursday, 31 of July 2008, Dmitry Torokhov wrote:
> On Thu, Jul 31, 2008 at 08:48:56PM +0200, Rafael J. Wysocki wrote:
> > On Thursday, 31 of July 2008, Dmitry Torokhov wrote:
> > > On Thu, Jul 31, 2008 at 10:44:37AM -0700, Andrew Morton wrote:
> > > > On Thu, 31 Jul 2008 11:56:48 -0400 Dmitry Torokhov <dmitry.torokhov@xxxxxxxxx> wrote:
> > > > 
> > > > > On Thu, Jul 31, 2008 at 05:36:16PM +0200, Bartlomiej Zolnierkiewicz wrote:
> > > > > > On Thu, Jul 31, 2008 at 4:07 PM, Dmitry Torokhov
> > > > > > <dmitry.torokhov@xxxxxxxxx> wrote:
> > > > > > > On Wed, Jul 30, 2008 at 11:10:29PM -0700, Andrew Morton wrote:
> > > > > > >> On Wed, 30 Jul 2008 17:06:35 +1000 Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx> wrote:
> > > > > > >>
> > > > > > >> > I have created today's linux-next tree at
> > > > > > >> > git://git.kernel.org/pub/scm/linux/kernel/git/sfr/linux-next.git
> > > > > > >>
> > > > > > >> The X server broke on my FC8 t61p thinkpad.  Mainline is OK.
> > > > > > >>
> > > > > > >> Various information is at http://userweb.kernel.org/~akpm/mo/
> > > > > > >>
> > > > > > >> I'm suspecting the input layer - my synaptics device seems to have
> > > > > > >> disappeared?  See http://userweb.kernel.org/~akpm/mo/Xorg-log-diff.txt
> > > > > > >>
> > > > > > >
> > > > > > > I think this patch should help with Synaptics:
> > > > > > 
> > > > > > Which unfortunately doesn't help all people running with older synaptics
> > > > > > user-space after commit 0571c5d20aca71c735222132b02aebddf593045c
> > > > > > ("Input: expand keycode space").
> > > > > > 
> > > > > > Can't this be solved without breaking Xorg on newer kernels running
> > > > > > older synaptics?
> > > > > > 
> > > > > 
> > > > > No. The X driver is broken. It tells kernel to use buffer bugger than
> > > > > allocated and gets its stack smashed. Tslib has also soma funkiness
> > > > > in the ioctl handling as well... *shrug*
> > > > > 
> > > > > We have a couple months to get distros updated...
> > > > > 
> > > > 
> > > > aaarrrrgggggghhh.  I don't think this is practical.  This means that
> > > > (for example) FC5 machines (of which I happen to have one) are dead. 
> > > > And lots of other older-distro-based systems.
> > > > 
> > > > Is there some userspace workaround which doesn't require an X server
> > > > update?
> > > > 
> > > > Surely it must be possible to make the kernel contiue to support these
> > > > servers?
> > > > 
> > > 
> > > Andrew,
> > > 
> > > It is not like we broke ABI here. The progam (synaptics driver) had a
> > > grave bug. Older kernels happened to paper over the bug because they
> > > did not fill the whole buffer that was advertised as available. Now
> > > that we have more data to report the bug bit us. What do you want me
> > > to do?
> > > 
> > > Synaptics driver is a small package and takes 2 minutes to recompile.
> > > You don't have to update entire X server with it (in fact I don't think
> > > it is even part of X distribution because it is GPL).
> > 
> > Well, we're not supposed to break user space that we used to work with, even
> > if it is known to be buggy.
> 
> No, I am sorry. We are not supposed to break userspace ABI, but that
> is it. Can you vouch that 2.6.25 did not break a single userspace
> program out there?

It probably did break some, but that doesn't imply it's acceptable.

> >  Many people use the older user space on their
> > test systems which are not practical to upgrade.
> > 
> 
> I don't understand this - it is expected that everyone jumps and
> upgrades their kernels with ease but updating broken userspace
> bits is super-hard... Plus, in this case the fixed driver will
> happily work with old kernels.

Please look at it from a different angle.

Kernel testers usually don't expect the kernel to break their X servers,
whatever the reason.  So, if they are not warned to expect the breakage, they
will search for the reason of it, as Andrew has just done, losing their time.
To prevent such losses of time from happening, it's better not to break user
space.

Still, if you really think there's no other way to go (like deferring the change
in question until everyone can be safely assumed to have upgraded their user
space already), please do something to keep people informed.

Thanks,
Rafael
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/