b44 unable to enable network on 64bit

[Stefan Bader]

Release: Quantal (alpha2)
Architecture: amd64

Starting to file this against the kernel, though I am not sure. On a Dell1521 with a Broadcom BCM4401-B0 NIC, the automatic network setup fails and the log shows repeatedly the following message:

Jun 27 10:37:11 kernel: [ 49.858545] b44 ssb1:0: eth0: powering down PHY

(this continues even after network detection is reported to have failed). I get the same result on both amd64 server and alternate images. The i386 versions don't cause that problem and network comes up.

[Brad Figg (brad-figg)]

This bug is missing log files that will aid in diagnosing the problem. From a terminal window please run:

apport-collect 1018456

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

[Stefan Bader (smb)]

This really seems to be a regression in the kernel. Installing the current 3.5.0-2.2 Quantal kernel onto a Precise base (where the network does work), produces the same type of messages. It works at least a little bit, though. So it looks like the link is up but it drops often. Even a simple ping gets about 50% loss.

[Stefan Bader (smb)]

3.4.0-5.11: good
3.5.0-1.1: bad

So the regression seems to have been introduced by 3.5-rc1. I will look further into that delta.

[Stefan Bader (smb)]

Bisection complete. Bad commit:

commit a1c7fff7e18f59e684e07b0f9a770561cd39f395
Author: Eric Dumazet <email address hidden>
Date: Thu May 17 07:34:16 2012 +0000

    net: netdev_alloc_skb() use build_skb()

    netdev_alloc_skb() is used by networks driver in their RX path to
    allocate an skb to receive an incoming frame.

    With recent skb->head_frag infrastructure, it makes sense to change
    netdev_alloc_skb() to use build_skb() and a frag allocator.

    This permits a zero copy splice(socket->pipe), and better GRO or TCP
    coalescing.

    Signed-off-by: Eric Dumazet <email address hidden>
    Signed-off-by: David S. Miller <email address hidden>

Though it does not seem to be simple revert to test this against current master. It seems there should be a way for the driver to enforce the old style somehow...

[Launchpad Janitor (janitor)]

This bug was fixed in the package linux - 3.5.0-4.4

---------------
linux (3.5.0-4.4) quantal-proposed; urgency=low

  [ Andy Whitcroft ]

  * [Packaging] getabis should be extracting all packages
    - LP: #1021174
  * [Config] getabis -- series uses linux-image-extra
    - LP: #1021174
  * rebase to v3.5-rc6

  [ Bryan Wu ]

  * [Config] built-in CONFIG_MICREL_PHY as other PHY drivers for all
    flavours
  * [Config] sync CONFIG_MOUSE_PS2_ config for all flavours

  [ Leann Ogasawara ]

  * [Config] Enable CONFIG_RT2800USB_RT35XX and CONFIG_RT2800USB_RT53XX
    - LP: #1019561

  [ Paolo Pisati ]

  * [Config] SND_OMAP_SOC, SND_OMAP_SOC_MCBSP and SND_OMAP_SOC_OMAP3_BEAGLE =y
    - LP: #1019321

  [ Stefan Bader ]

  * SAUCE: (pre-up) net: dont use __netdev_alloc_skb for bounce buffer
    - LP: #1018456
  * (config) Disable ACPI_PROCFS_POWER

  [ Tim Gardner ]

  * [Config] CONFIG_ACPI_BGRT=y
  * Extract firmware module info during getabi
    - LP: #1021174

  [ Upstream Kernel Changes ]

  * rebase to v3.5-rc6
 -- Leann Ogasawara <email address hidden> Mon, 09 Jul 2012 08:50:20 -0700

[Adam Conrad (adconrad)]

The verification of this Stable Release Update has completed successfully and the package has now been released to -updates. Subsequently, the Ubuntu Stable Release Updates Team is being unsubscribed and will not receive messages about this bug report. In the event that you encounter a regression using the package from -updates please report a new bug using ubuntu-bug and tag the bug report regression-update so we can easily find any regresssions.