[Karmic] SRU: Update to stable kernel

Stefan Bader stefan.bader at canonical.com
Wed Nov 4 15:33:27 UTC 2009

Leann Ogasawara wrote:
> On Wed, 2009-10-28 at 17:19 +0000, Stefan Bader wrote:
>> Leann Ogasawara wrote:
>>> Hi All,
>>> I realize that Karmic isn't quite out the door just yet, but I wanted to
>>> get our first Karmic SRU in order.  Note, updating to will also
>>> result in an ABI Bump.
>>> https://bugs.launchpad.net/ubuntu/karmic/+source/linux/+bug/461556
>>> I've pushed these to a stable branch in my tree for review:
>>> git://kernel.ubuntu.com/ogasawara/ubuntu-karmic stable
>>> SRU Justification/Impact: The upstream process for stable tree updates
>>> is quite similar in scope to the Ubuntu SRU process, e.g., each patch
>>> has to demonstrably fix a bug, and each patch is vetted by upstream by
>>> originating either directly from Linus' tree or in a minimally
>>> backported form of that patch. The upstream stable patch set is
>>> now available. It should be included in the Ubuntu kernel as well.
>>> http://kernel.org/pub/linux/kernel/v2.6/ChangeLog-
>>> http://kernel.org/pub/linux/kernel/v2.6/patch-
>>> The Upstream Stable Kernel brings in the following 47 patches:
>>> * SCSI: Retry ADD_TO_MLQUEUE return value for EH commands
>>> * SCSI: Fix protection scsi_data_buffer leak
>>> * SCSI: sg: Free data buffers after calling blk_rq_unmap_user
>>> * ARM: pxa: workaround errata #37 by not using half turbo switching
>>> * tracing/filters: Fix memory leak when setting a filter
>>> * x86/paravirt: Use normal calling sequences for irq enable/disable
>>> * USB: ftdi_sio: remove tty->low_latency
>>> * USB: ftdi_sio: remove unused rx_byte counter
>>> * USB: ftdi_sio: clean up read completion handler
>>> * USB: ftdi_sio: re-implement read processing

A rather large change but reading the description it seems to address
quite a few problems. And we know usb serial has not been in its best
state in 2.6.31. Seems one of the things to require an ABI change.

>>> * USB: pl2303: fix error characters not being reported to ldisc
>>> * USB: digi_acceleport: Fix broken unthrottle.
>>> * USB: serial: don't call release without attach
>>> * USB: option: Toshiba G450 device id
>>> * USB: ipaq: fix oops when device is plugged in
>>> * USB: cp210x: Add support for the DW700 UART
>>> * USB: Fix throttling in generic usbserial driver
>>> * USB: storage: When a device returns no sense data, call it a Hardware Error
>>> * arm, cris, mips, sparc, powerpc, um, xtensa: fix build with bash 4.0

One of the bigger ones, but as it changes the build environment and not the
code itself it would be expected to cause build failures if wrong.

>>> * intel-iommu: Cope with broken HP DC7900 BIOS
>>> * futex: Detect mismatched requeue targets
>>> * futex: Fix wakeup race by setting TASK_INTERRUPTIBLE before queue_me()
>>> * tpm-fixup-pcrs-sysfs-file-update
>>> * TPM: fix pcrread
>>> * Bluetooth: Disconnect HIDRAW devices on disconnect
>>> * Bluetooth: Add extra device reference counting for connections
>>> * Bluetooth: Let HIDP grab the device reference for connections
>>> * connector: Keep the skb in cn_callback_data
>>> * connector: Provide the sender's credentials to the callback

Another one that causes the ABI to bump. All changes to functions to
pass additional information. It does not say much about the reasons but
additionally passing some caller's credentials sounds a bit like security.

>>> * connector: Removed the destruct_data callback since it is always kfree_skb()
>>> * dm/connector: Only process connector packages from privileged processes
>>> * dst/connector: Disallow unpliviged users to configure dst
>>> * pohmelfs/connector: Disallow unpliviged users to configure pohmelfs
>>> * uvesafb/connector: Disallow unpliviged users to send netlink packets
>>> * e1000e: swap max hw supported frame size between 82574 and 82583
>>> * MAINTAINERS: Fix Riku Voipio's address
>>> * macintosh: Don't assume i2c device probing always succeeds
>>> * i2c: Hide probe errors caused by ACPI resource conflicts
>>> * ALSA: Don't assume i2c device probing always succeeds
>>> * bsdacct: switch credentials for writing to the accounting file
>>> * sysfs: Allow sysfs_notify_dirent to be called from interrupt context.
>>> * Staging: rt2860sta: prevent a panic when disabling when associated
>>> * usb-storage: Workaround devices with bogus sense size
>>> * iwlwifi: incorrect method used for finding valid OTP blocks
>>> * mac80211: fix vlan and optimise RX
>>> * tty: Make flush_to_ldisc() locking more robust
>>> * Linux
>>> Fixes LP Bug 404626:
>>> Staging: rt2860sta: prevent a panic when disabling when associated
>>> Fixes LP Bug 446146:
>>> usb-storage: Workaround devices with bogus sense size
>>> Fixes LP Bug 445572:
>>> e1000e: swap max hw supported frame size between 82574 and 82583
> Fixes LP Bug 400652:
> USB: storage: When a device returns no sense data, call it a Hardware
> Error
>> As we seem to have gathered at least one thing that we might want to get out
>> of the door slightly quicker, we probably need to go for those first and then
>> add the stable update on top of that. Problem with the stable updates in general
>> is that we get them in bulk, so we need a longer timer in proposed for them.
>> But hopefully everything you got now just can be rebased onto the new top.
> I have rebased the stable patch set onto the new head and have
> repushed the patches to the stable branch in my repo for review:
> git://kernel.ubuntu.com/ogasawara/ubuntu-karmic stable

When looking through those I did not catch anything bad. I had a
closer look at the three which were bigger than the usual and added
some comments above.

Acked-by: Stefan Bader <stefan.bader at canonical>

> Stefan, I will get in touch with you tomorrow about uploading a
> pre-proposed test package.

I tweaked the headers a bit (mainly adding a buglink to the common tracking
bug for all of those) and sent the result to my PPA[1]. For the final take I'd
also add Ack lines to them.

I would suggest that at least us in the kernel team, that run karmic add this
to their apt lines. So we might find out ourselves when we break important
things. This maybe needs a bit more rubbing in on UDS as I am not sure how
well mails like this are read. :)

One more thing, I updated the maint-modify-patch script, as it failed in some
cases, and added it to the maint-tools repo[2].


> Thanks,
> Leann

[1] https://launchpad.net/~stefan-bader-canonical/+archive/karmic
[2] git://kernel.ubuntu.com/smb/maint-tools.git

More information about the kernel-team mailing list