[regression] external HDD in USB3 enclosure cannot be dynamically removed (Re: Linux 3.7.5)

Sarah Sharp sarah.a.sharp at linux.intel.com
Wed Feb 13 22:12:50 UTC 2013


On Wed, Feb 13, 2013 at 01:31:29PM -0800, Greg KH wrote:
> On Wed, Feb 13, 2013 at 01:08:46PM -0800, Sarah Sharp wrote:
> > On Wed, Feb 13, 2013 at 09:04:13PM +0100, Matthias Schniedermeyer wrote:
> > > On 13.02.2013 11:33, Sarah Sharp wrote:
> > > > On Wed, Feb 13, 2013 at 06:16:56PM +0100, Matthias Schniedermeyer wrote:
> > > > > On 13.02.2013 09:28, Holger Hoffstätte wrote:
> > > > > > On 12.02.2013 21:42, Sarah Sharp wrote:
> > > > > > > [..]
> > > > > > > There was a further set of patches queued for 3.9 to deal with connected
> > > > > > > devices going to the Inactive state, but they looked like they were too
> > > > > > > big for stable:
> > > > > > > 
> > > > > > > d3b9d7a USB: Fix connected device switch to Inactive state.
> > > > > > > a24a607 USB: Rip out recursive call on warm port reset.
> > > > > > > 2d4fa94 USB: Prepare for refactoring by adding extra udev checks.
> > > > > > > 0fe51aa USB: Don't use EHCI port sempahore for USB 3.0 hubs.
> > > > 
> > > > Holger and Matthias, can you double check that cherry picking just those
> > > > four patches on top of 3.7.7 or 3.8 works as well?
> > > 
> > > 3.7.7 + those 4 patches workes for me.
> > > 
> > > As i'm not very firm in git i generated separate diffs for each commit 
> > > and applying them in the order 2d4fa94 0fe51aa d3b9d7a a24a607 worked 
> > > without hunks dropped.
> > > 
> > > I have attached `git diff` against vanilla 3.7.7, so you can check that 
> > > i got it right.
> > 
> > Yep, that diff looks fine compared to the git diff of those four patches.
> > 
> > Greg,
> > 
> > How do you want to handle this?  The above four patches should go into
> > 3.8 and stable, but they're not currently in Linus' tree and it's
> > probably too late in the cycle to merge them this week.  Should we just
> > wait until 3.9 is out and put the patches into the stable trees then?
> 
> > My email shows that the bad commit
> > f7965c0846d74b270e246c1470ca955d5078eb07 has been added to the 3.2, 3.4,
> > and 3.7 stable trees, as well as Canonical's 3.7 stable tree.  I'm also
> > fine with just reverting that commit from 3.8 and stable.
> 
> It's probably easier at this point in time to just revert that commit,
> leave the stable kernels alone, and then everything will be fixed in
> 3.9.  When the commits go into Linus's tree, we can backport everything
> to the stable releases (including 3.8.1) at that point.
> 
> As long as someone remembers to send the needed information to
> stable at vger.kernel.org, I know I'll forget :)
> 
> Sound ok?

Yep, sounds fine.  I'll attempt to remember to send email to stable once
3.9 is out.

Sarah Sharp




More information about the kernel-team mailing list