usb devices not recognized automatically when plugged in a hub

I tried to recompile the kernel and remove the ubuntu modification from drivers/usb/core/hub.c (see attached patch), and the usb hubs are working back. I guess the bug is somewhere in this modification.

How I did :

apt-get source linux-image-2.6.24-21-generic
patch -p1 -R < hub.patch
cd linux-2.6.24.2/
cp /boot/config-2.6.24-21-generic .config
make deb-pkg
dpkg -i ../linux-2.6.24.3_2.6.24.3_i386.deb
mkinitramfs -o /boot/initrd.img-2.6.24.3
edit /boot/grub/menu.lst

reboot on this kernel.

I had a look in hub.c and I propose this patch which works for me. I'm not quite sure if it has side effects since it's the first time I'm looking in the usbcore code.

I noticed that when I'm plugging a memory stick in my hub, the hub makes a soft reset first. I guess it can be one of this "side effects".

I just tried to backport the hub_reset() and persistent_device() functions from hub.c in 2.6.26, but it didn't help. (So the patch I attach to this comment DOESN'T WORK).

I guess there are also modifications in the ehci driver but I'm not fit enought to look in it.

I searched throught the diff file and through the changelogs of the vanilla kernel, but I can't find from which revision comes this patch between -19 and -21.

Does someone knows why this patch was integrated in -21 ?
Was the code in -19 buggy ? If not, it should be quite usefull to roll back to the code of -19 for hub.c.

This bug is quite annoying for my users who can't use their intern usb-hub anymore. (which means that they have to boot under -19 to work).

Just to say that it still exists in -22 (but it's quite normal because it was just a security update).

Per a decision made by the Ubuntu Kernel Team, bugs will longer be assigned to the ubuntu-kernel-team in Launchpad as part of the bug triage process. The ubuntu-kernel-team is being unassigned from this bug report. Refer to https://wiki.ubuntu.com/KernelTeamBugPolicies for more information. Thanks.

Still exists in 2.6.24-23.

Does this still happen on the released version of Jaunty ?

Same problem with kubuntu 8.04.2 (linux 2.6.24-24)

when I plug the external usb hub:
[ 182.788947] usb 5-4: new high speed USB device using ehci_hcd and address 6
[ 182.921415] usb 5-4: configuration #1 chosen from 1 choice
[ 182.922560] hub 5-4:1.0: USB hub found
[ 182.929209] hub 5-4:1.0: 4 ports detected

and then all usb devices stop working, I mean the usb device plugged in the hub and the usb device directly plugged in the other free usb pc ports.
Also if i unplugged the usb hub the kernel doesn't care, usb stops working at all.

I tried all the combination of pc/usb hub/kernel and the only that fails is kubuntu 8.04.2
 - Same hardware on an old gentoo 2007 works fine.
 - Same usb hub connected on a eepc901 with jaunty works fine
 - Same usb hub connected to a different hardware on kubuntu 8.0.4 fails.

Can you confirm this issue exists with the Karmic Alpha 9.10 release. Images for testing are available at http://cdimage.ubuntu.com/daily-live/current/ . Please let us know your results.

If the issue remains while still running Karmic, please run the following command which will automatically gather and attach updated debug information:

apport-collect -p linux <bug #>

Thanks in advance.

I was working with the workaround "don't update the kernel with any version > 2.6.24-19" for a while.

I just retested with 2.6.24-24 and 2.6.24-25, and this issue seems to be fixed.
With the 2.6.31 from 9.10 it also works fine.

I checked the linux_2.6.24-25.63.diff.gz (got with apt-get source linux-image-2.6.24-24-generic) and I searched for hub.c

I was able to find some stuff including that :

--- linux-2.6.24.orig/drivers/usb/core/hub.c
+++ linux-2.6.24/drivers/usb/core/hub.c
@@ -327,6 +327,9 @@
[...]
+static void hub_restart(struct usb_hub *hub, int type)
+{
+ struct usb_device *hdev = hub->hdev;
+ int port1;
+
+ /* Check each of the children to see if they require
+ * USB-PERSIST handling or disconnection. Also check
+ * each unoccupied port to make sure it is still disabled.
+ */
[...]

I guess that's part of the change which solved this issue.

I guess it comes from this entry of the changelog

linux (2.6.24-24.60) hardy-proposed; urgency=low
[...]
  [Upstream Kernel Changes]

  * USB: EHCI: fix remote-wakeup regression
    - LP: #406419

 -- Stefan Bader <email address hidden> Tue, 18 Aug 2009 18:25:47 +0200

So I guess this bug was a duplicated to https://bugs.launchpad.net/ubuntu/+source/linux/+bug/406419

So far I was able to test (3 computers, 2 hubs), this bug is for me fixed.

If you are still having issues it must be another similar issue and needs a new bug