linux-flexiantxendom0-natty.git
13 years agoALSA: hda - Handle missing NID 0x1b on ALC259 codec
Takashi Iwai [Fri, 30 Jul 2010 08:51:10 +0000 (10:51 +0200)]
ALSA: hda - Handle missing NID 0x1b on ALC259 codec

BugLink: http://bugs.launchpad.net/bugs/582199
BugLink: http://bugs.launchpad.net/bugs/586418
BugLink: http://bugs.launchpad.net/bugs/588031

Since ALC259/269 use the same parser of ALC268, the pin 0x1b was ignored
as an invalid widget.  Just add this NID to handle properly.
This will add the missing mixer controls for some devices.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
(cherry picked from commit 5d4abf93ea3192cc666430225a29a4978c97c57d in
git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6.git)

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: blkfront: default to sd devices
John Johansen [Wed, 28 Jul 2010 18:44:04 +0000 (11:44 -0700)]
UBUNTU: SAUCE: blkfront: default to sd devices

EC2 defaults to "/dev/sda" rather than "/dev/xvd". Once they finally
switch then this patch should be dropped.

Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: fix pv-ops for legacy Xen
John Johansen [Tue, 27 Jul 2010 13:06:07 +0000 (06:06 -0700)]
UBUNTU: SAUCE: fix pv-ops for legacy Xen

Import fix_xen_guest_on_old_EC2.patch from fedora 14

Legacy hypervisors (RHEL 5.0 and RHEL 5.1) do not handle guest writes to
cr4 gracefully. If a guest attempts to write a bit of cr4 that is
unsupported, then the HV is so offended it crashes the domain. While
later guest kernels (such as RHEL6) don't assume the HV supports all
features, they do expect nicer responses. That assumption introduced
code that probes whether or not xsave is supported early in the boot. So
now when attempting to boot a RHEL6 guest on RHEL5.0 or RHEL5.1 an early
crash will occur.

This patch is quite obviously an undesirable hack. The real fix for this
problem should be in the HV, and is, in later HVs. However, to support
running on old HVs, RHEL6 can take this small change. No impact will
occur for running on any RHEL HV (not even RHEL 5.5 supports xsave).
There is only potential for guest performance loss on upstream Xen.

All this by way of explanation for why is this patch not going upstream.

Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: AppArmor: Allow dfa backward compatibility with broken userspace
John Johansen [Fri, 23 Jul 2010 12:53:56 +0000 (05:53 -0700)]
UBUNTU: SAUCE: AppArmor: Allow dfa backward compatibility with broken userspace

Allow broken Lucid userspace tools to load policy, on Maverick kernel.
The fix for http://launchpad.net/bugs/581525 blocks Lucid tools from
loading policy, this provides compatibility with Lucid tools without
reintroducing the bug.

The apparmor_parser when compiling policy could generate invalid dfas
that did not have sufficient padding to avoid invalid references, when
used by the kernel.  The kernels check to verify the next/check table
size was broken meaning invalid dfas were being created by userspace
and not caught.

To remain compatible with old tools that are not fixed, pad the loaded
dfas next/check table.  The dfa's themselves are valid except for the
high padding for potentially invalid transitions (high bounds error),
which have a maximimum is 256 entries.  So just allocate an extra null filled
256 entries for the next/check tables.  This will guarentee all bounds
are good and invalid transitions go to the null (0) state.

Signed-off-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: vt -- allow grub to request automatic vt_handoff
Andy Whitcroft [Thu, 29 Jul 2010 15:36:31 +0000 (16:36 +0100)]
UBUNTU: SAUCE: vt -- allow grub to request automatic vt_handoff

Grub may be able to select a graphics mode and paint a splash screen
for us.  If so it needs to be able to tell us it has done so.  Add
support for detecting a new graphics mode selected bit in the
screen_info passed over at boot.  Use this to automatically enable
vt_handoff mode.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: vt -- maintain bootloader screen mode and content until vt switch
Andy Whitcroft [Thu, 29 Jul 2010 15:36:30 +0000 (16:36 +0100)]
UBUNTU: SAUCE: vt -- maintain bootloader screen mode and content until vt switch

Introduce a new VT mode KD_TRANSPARENT which endevours to leave the current
content of the framebuffer untouched.  This allows the bootloader to insert
a graphical splash and have the kernel maintain it until the OS splash
can take over.  When we finally switch away (either through programs like
plymouth or manually) the content is lost and the VT reverts to text mode.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: add tracing for user initiated readahead requests
Andy Whitcroft [Thu, 29 Jul 2010 08:46:41 +0000 (09:46 +0100)]
UBUNTU: SAUCE: add tracing for user initiated readahead requests

Track pages which undergo readahead and for each record which were
actually consumed, via either read or faulted into a map.  This allows
userspace readahead applications (such as ureadahead) to track which
pages in core at the end of a boot are actually required and generate an
optimal readahead pack.  It also allows pack adjustment and optimisation
in parallel with readahead, allowing the pack to evolve to be accurate
as userspace paths change.  The status of the pages are reported back via
the mincore() call using a newly allocated bit.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) Modularize vesafb -- fix initialisation
Andy Whitcroft [Thu, 29 Jul 2010 08:42:21 +0000 (09:42 +0100)]
UBUNTU: SAUCE: (no-up) Modularize vesafb -- fix initialisation

When this patch was rolled forward, likely between Dapper and Hardy
a chunk of initialisation was lost.  Pull this back in so we actually
have an vesafb_info structure initialised.  Else we may well panic when
we rmmod vesafb.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Acked-by: Brad Figg <brad.figg@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: OMAP: DSS2: enable hsclk in dsi_pll_init for OMAP36XX
Lee Jones [Wed, 21 Jul 2010 09:32:26 +0000 (11:32 +0200)]
UBUNTU: SAUCE: OMAP: DSS2: enable hsclk in dsi_pll_init for OMAP36XX

BugLink: http://bugs.launchpad.net/bugs/608095

This helps provide the required setup to enable USB Ethernet (usb0) and
USB host on the XM Beagleboard (A rev).  This will be submitted upstream
by Steve Sakoman.

Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Lee Jones <lee.jones@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: OMAP: DSS2: check for both cpu type and revision, rather than just...
Lee Jones [Wed, 21 Jul 2010 09:30:33 +0000 (11:30 +0200)]
UBUNTU: SAUCE: OMAP: DSS2: check for both cpu type and revision, rather than just revision

BugLink: http://bugs.launchpad.net/bugs/608095

This helps provide the required setup to enable USB Ethernet (usb0) and
USB host on the XM Beagleboard (A rev).  This will be submitted upstream
by Steve Sakoman.

Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Lee Jones <lee.jones@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: ARM: OMAP: Add macros for comparing silicon revision
Lee Jones [Wed, 21 Jul 2010 09:24:40 +0000 (11:24 +0200)]
UBUNTU: SAUCE: ARM: OMAP: Add macros for comparing silicon revision

BugLink: http://bugs.launchpad.net/bugs/608095

Adapted from arago project patch by Sanjeev Premi <premi@ti.com>

This helps provide the required setup to enable USB Ethernet (usb0) and
USB host on the XM Beagleboard (A rev).  This will be submitted upstream
by Steve Sakoman.

Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Lee Jones <lee.jones@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: Stop ARM boards crashing when CUPS is loaded
Lee Jones [Fri, 16 Jul 2010 08:37:31 +0000 (09:37 +0100)]
UBUNTU: Stop ARM boards crashing when CUPS is loaded

BugLink: http://bugs.launchpad.net/bugs/601226

When CUPS loads, it tries to load several drivers it may need. When
one of these drivers, specifically parport_pc is loaded, it attempts
to write to address space normally reserved for ISA transactions.
On OMAP based systems, this causes a segmentation fault.

Signed-off-by: Lee Jones <lee.jones@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: AUFS -- update to b37c575759dc4535ccc03241c584ad5fe69e3b25
Andy Whitcroft [Thu, 15 Jul 2010 16:24:41 +0000 (17:24 +0100)]
UBUNTU: ubuntu: AUFS -- update to b37c575759dc4535ccc03241c584ad5fe69e3b25

    commit b37c575759dc4535ccc03241c584ad5fe69e3b25
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Sun Jul 4 21:06:02 2010 +0900

        aufs: minor update abput the doubling donations

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 799db4b1d59ea0ffc999889bc6985397333e9a13
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Fri Jul 2 00:37:56 2010 +0900

        aufs: compat_ioctl, implement the operations

        (A commit in a series of supporting 32bit emulation under 64bit kernel.
        While every commit is git-bisect-able, you shoule read all commits in
        the series since a single commit may have less meaning.)

        Implement f_op->compat_ioctl().

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 765378b55abcabfe3344e5fcf2eabd6a1d52abc0
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Fri Jul 2 00:35:17 2010 +0900

        aufs: compat_ioctl, make a room for a pointer

        (A commit in a series of supporting 32bit emulation under 64bit kernel.
        While every commit is git-bisect-able, you shoule read all commits in
        the series since a single commit may have less meaning.)

        In order to make it compatible, make a room for a pointer and always
        handle it as 64bit size.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit abd339c757ea095f1affdd648d0e9e598213a790
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Fri Jul 2 00:29:00 2010 +0900

        aufs: compat_ioctl, remove verifying the size of ptr

        (A commit in a series of supporting 32bit emulation under 64bit kernel.
        While every commit is git-bisect-able, you shoule read all commits in
        the series since a single commit may have less meaning.)

        Remove verifying the size of ptr which is meaningless.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 3c979528184058b184608e6e3086c4c59b0a6c86
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Fri Jul 2 00:04:33 2010 +0900

        aufs: follow 2.6.28, new flag LOOKUP_EXCL

        NFS replaces the internal test for LOOKUP_CREATE by a new flag
        LOOKUP_EXCL.
        Aufs has to prohibit this internal test in order to know whether the
        file exists or not.

        Reported-by: "Ian Stakenvicius, Aerobiology Research" <ian@aerobiology.ca>
Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 396a2f097a6b278fc2a9e9da83b87255f60075fb
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Tue Jun 29 14:32:26 2010 +0900

        aufs: bugfix, separate the workqueue for preprocessing mmap

        variation of common AB-BA deadlock problem.

        ProcessA:
        - aufs_mmap
          + pre-process with using a workqueue
          + wait until return from the workqueue

        Workqueue task for ProcessA:
        - acquire aufs rwsem

        Processb
        - lookup or readdir in aufs
          + acquire aufs rwsem
          + assign a new inode number
          + write the value to the XINO file using a workqueue
          + wait until return from the workqueue

        Since the workqueue handles the request one by one, both of processA and
        B waits forever.

        This bug was introduced by the commit
        d986fa5 2010-03-08
         aufs: bugfix, another approach to keep the lock order of mmap_sem
        which is the last added workqueue task.
        And this is the only one task which acquires such lock in workqueue.
        To fix it, introduce another workqueue which is for preprocessing mmap only.
        This commit will make the approach more ugly, I don't have another option.

        Reported-by: Oliver Welter <mail@oliwel.de>
Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 7c7f493d58745e45160ccebfbfb5e6244dbd0b52
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Fri Jun 25 10:52:59 2010 +0900

        aufs: tiny, debug print [if]_version

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit ade8662cb1184703ebdd7c7d07cb42d2056a605b
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Sat Jun 19 16:38:15 2010 +0900

        aufs: follow the changes in 2.6.35, lockdep for sb->s_vfs_rename_mutex

        lockdep_set_class() is applied to sb->s_vfs_rename_mutex, and
        lockdep_off/on() in aufs become unnecessary.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit d405a78a328658600f1928309bfbd2ded6136c59
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Sat Jun 19 03:36:21 2010 +0900

        aufs: tiny, remove unused lockdep_off/on()

        In linux-2.6.31, lockdep_set_class() was applited to them.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit a4862273fc684dd24a9e0b4c29f89aaf53723afc
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Thu Jun 17 23:39:13 2010 +0900

        aufs: several GIT servers

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 135ac88ed89e4780dc71bef119b803d4c594ea07
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Wed Jun 16 18:11:40 2010 +0900

        aufs: possible bugfix, sbinfo lock in deleting inode, lockdep

        (A commit in a series of introducing pid map/tree and making sure to
        acquire sbinfo lock in deleting inode. While every commit is
        git-bisect-able, you shoule read all commits in the series since a
        single commit may have less meaning.)

        A debugging feature in linux kernel, lockdep, warns
        "inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-R} usage."
        about sb->s_umount and aufs sbinfo lock.
        This is bogus or "false positive" since {RECLAIM_FS-ON-W} state was
        registered at allcating the root inode at mounting. This is definitly no
        RECLAIM state. It may be a limitation of s_umount in lockdep.
        Let's simply make it quiet.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 6ef8250091897e1cf7e8b77a7cfc61c7ffe58b58
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Wed Jun 16 18:03:48 2010 +0900

        aufs: possible bugfix, sbinfo lock in deleting inode, core

        (A commit in a series of introducing pid map/tree and making sure to
        acquire sbinfo lock in deleting inode. While every commit is
        git-bisect-able, you shoule read all commits in the series since a
        single commit may have less meaning.)

        s_umount rwsem in struct super_block prevents a race condition among
        umount, remount and kswapd. It is good.
        But if an inode is going to be deleted by other than kswapd, aufs may be
        doing another operation which requires a lock for sbinfo. In this case,
        si_noflush_read_trylock() in au_iinfo_fin() will not acquire the
        lock. Before au_iinfo_fin() completes or during its operations, another
        operation may release the lock and remount or a branch management
        process may start. Here if the branch management process changes the
        union members, then xino management in au_iinfo_fin() will not work
        correctly.

        In order to fix this potential problem, there another bad approch is
        introduced which uses a bitmap to mark the pid which acquired the sbinfo
        lock. In au_iinfo_fin(), if the bit is set, then the function will not
        try acquiring the lock.
        To support the pid larger than PID_MAX_DEFAULT, sbinfo prepares a radix
        tree too.

        With this commit, we can remove si_noflush_read_trylock(). But it will
        be necessary in aufs2-31 branch, so leave it now.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit eb6254f4b0080f01b3356a40bdd62d7130bf1951
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Wed Jun 16 17:51:00 2010 +0900

        aufs: possible bugfix, sbinfo lock in deleting inode, use si_pid

        (A commit in a series of introducing pid map/tree and making sure to
        acquire sbinfo lock in deleting inode. While every commit is
        git-bisect-able, you shoule read all commits in the series since a
        single commit may have less meaning.)

        Use si_pid which was declared and implmented by last commit.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 89d59298f045a8a7e2c5a667dfc3ee3b3fe8745a
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Wed Jun 16 17:43:12 2010 +0900

        aufs: possible bugfix, sbinfo lock in deleting inode, si_pid functions

        (A commit in a series of introducing pid map/tree and making sure to
        acquire sbinfo lock in deleting inode. While every commit is
        git-bisect-able, you shoule read all commits in the series since a
        single commit may have less meaning.)

        Declare and implement si_pid functions which are not used yet.
        The pid from 1 to PID_MAX_DEFAULT are marked in a new bitmap in
        sbinfo. The larger pids will go to a new radix tree in sbinfo.
        These marks will be referenced by au_iinfo_fin() in succeeding commit.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 40d554b0f6a6bc588fd88235f39b59f30ec27114
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Wed Jun 16 17:30:03 2010 +0900

        aufs: possible bugfix, sbinfo lock in deleting inode, __si_ lock

        (A commit in a series of introducing pid map/tree and making sure to
        acquire sbinfo lock in deleting inode. While every commit is
        git-bisect-able, you shoule read all commits in the series since a
        single commit may have less meaning.)

        Rename si_noflush_... lock macros to __si_..., and create new inlined
        functions si_noflush_....

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit f2f51daf576fa90ad7209288ed8b19c8151ff92a
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Wed Jun 9 17:03:00 2010 +0900

        aufs: tiny, simplify the locks in au_do_flush()

        The read-lock for dinfo is unnecessary.
        Also the write-lock for iinfo should be a read-lock.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit bdc941c464fd4181f7ab76fbb992b6ab150ac188
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Tue Jun 8 23:28:03 2010 +0900

        aufs: minor optimization au_iinfo_fin()

        - extract a part of au_iinfo_fin(), create a new function
          au_xino_delete_inode(), and remove au_iinfo_write0().
        - simplify au_xino_write0() and rename to au_xib_clear_bit().
        - convert the type of au_xigen_inc() into void.
        - stop testing 'xino' option in au_xigen_inc().

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 7e7b86c68cf5d3638ceb19a0a1d7a254d8501bde
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Tue Jun 8 00:08:00 2010 +0900

        aufs: tiny, remove an unnecessary variable

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit ea824b7a71cd905fdadcb0998ab2005fcb3786e9
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Tue Jun 8 00:06:55 2010 +0900

        aufs: minor optimization, pass the 'verbose' flag

        Stop testing the 'verbose' flag in all test_dentry_busy(),
        test_inode_busy() and au_br_del() functions. Instead pass the tested
        result from au_br_del() to test_dentry_busy() and test_inode_busy().

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit cabc97fe740d2c7664988e76f218318aecd5ec18
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Mon Jun 7 14:45:15 2010 +0900

        aufs: tiny, test task flags instead of mm

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 575bf8c6484fd12a0719a7ebcf6c0d2fd0af31f3
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Fri Jun 4 14:55:42 2010 +0900

        aufs: follow linux-2.6.35, simple_setsize()

        Replace vmtruncate() by a new function simple_setsize().

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit c503a51d10c9c7767f1eba8e47de545d31ab7858
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Fri Jun 4 02:54:13 2010 +0900

        aufs: bugfix, the dentry paramter for security funcs

        Pass the correct parameter.

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 92172f59e363f630ab44217e78135cdeff90ba0d
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Fri Jun 4 02:09:23 2010 +0900

        aufs: tiny, fake type-cast by union

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 9f8ad8ca00cf7677e9d982c9e98d4b7586e759f6
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Wed Jun 2 23:38:37 2010 +0900

        aufs: possible bugfix, revalidate inode race between readdir and lookup

        Both of readdir and lookup operation need to assign the aufs inode
        number, but it requires other aufs locks including xi_nondir_mtx which
        prevents hard-linked inode number from race condition. There can happen
        a violation of the order of these locks.

        They acquire these locks.
        aufs_readdir("./dirA")
        + si_read_lock
        + fi_write_lock
        + di_write_lock
        + ii_write_lock for dirA
        + xi_nondir_mtx for non-dir

        aufs_lookup("./dirA/fileB")
        + si_read_lock
        + di_write_lock
        + xi_nondir_mtx for non-dir
        + ii_write_lock_nested for fileB

        Here the fileB may be in copy-up operation which acquires the parent's
        dentry-info and inode->info lock. So aufs_lookup() waits for the
        completion of copy-up, aufs_readdir() waits for xi_nondir_mtx, and the
        copy-up waits for the parent, but it is held by readdir.
        This is very complicated situation and I am afraid the design of aufs
        inode assignment is not godd. But I don't have other idea.

        This commit refines xi_nondir_mtx and releases it before
        "ii_write_lock_nested for fileB."

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

    commit 907f03a11f5a681de402d733b6e80532adece324
    Author: J. R. Okajima <hooanon05@yahoo.co.jp>
    Date:   Tue Jun 1 01:41:43 2010 +0900

        aufs: tiny, follow the changes in linux-2.6.35-rcN

        The dentry parameter is removed from ->fsync().

Signed-off-by: J. R. Okajima <hooanon05@yahoo.co.jp>

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: AUFS -- add BOM and automated update script
Andy Whitcroft [Fri, 2 Jul 2010 17:45:50 +0000 (18:45 +0100)]
UBUNTU: ubuntu: AUFS -- add BOM and automated update script

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: Yama: verify inode is symlink to avoid bind mounts
Kees Cook [Tue, 13 Jul 2010 21:54:56 +0000 (14:54 -0700)]
UBUNTU: SAUCE: Yama: verify inode is symlink to avoid bind mounts

The inode_follow_link LSM hook is called in bind mount situations as
well as for symlink situations, so we must explicitly check for the
inode being a symlink to not reject bind mounts in 1777 directories,
which seems to be a common NFSv4 configuration.

BugLink: https://bugs.launchpad.net/bugs/604407

[submitted upstream to security-next]

Signed-off-by: Kees Cook <kees.cook@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: Yama: search for PTRACE exceptions via thread group leader
Kees Cook [Tue, 13 Jul 2010 16:04:08 +0000 (09:04 -0700)]
UBUNTU: SAUCE: Yama: search for PTRACE exceptions via thread group leader

When looking for PTRACE exceptions, the thread group leader must be used
otherwise the exceptions can be missed for threaded processes.

BugLink: https://bugs.launchpad.net/bugs/603716

Signed-off-by: Kees Cook <kees.cook@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: Yama: check PTRACE using thread group leader
Kees Cook [Sat, 10 Jul 2010 19:12:20 +0000 (12:12 -0700)]
UBUNTU: SAUCE: Yama: check PTRACE using thread group leader

When examining process ancestory, we must use the thread group leader
or else we end up missing potential matches.

[submitted upstream to security-next]

Signed-off-by: Kees Cook <kees.cook@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: make ndiswrapper available on X86 only
Nicolas Pitre [Tue, 13 Jul 2010 04:58:12 +0000 (00:58 -0400)]
UBUNTU: SAUCE: make ndiswrapper available on X86 only

There are several functions in ntoskernel.* with inline x86 assembler.
Thus enable ndiswrapper for X86 only, especially when it gets enabled by
default in Kconfig.

Signed-off-by: Nicolas Pitre <nicolas.pitre@linaro.org>
Acked-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: fix build error with CONFIG_BLK_DEV_INITRD=n
Tim Gardner [Tue, 13 Jul 2010 12:41:16 +0000 (06:41 -0600)]
UBUNTU: SAUCE: fix build error with CONFIG_BLK_DEV_INITRD=n

The original aynchronous boot patch (UBUNTU: SAUCE: Make populate_rootfs asynchronous)
did not take into consideration the case when CONFIG_BLK_DEV_INITRD=n,
e.g., populate_rootfs_domain becomes undefined. Therefore, add it to
noinitramfs.c where its use is benign.

Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Original-patch-by: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Surbhi Palande <surbhi.palande@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: security: unconditionally chain to Yama LSM
Kees Cook [Tue, 29 Jun 2010 18:07:44 +0000 (11:07 -0700)]
UBUNTU: SAUCE: security: unconditionally chain to Yama LSM

This patch forces the LSM to always chain through the Yama LSM
regardless of which LSM is selected as the primary LSM.

Signed-off-by: Kees Cook <kees.cook@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: Yama: add PTRACE exception tracking and interface
Kees Cook [Tue, 29 Jun 2010 05:36:00 +0000 (22:36 -0700)]
UBUNTU: SAUCE: Yama: add PTRACE exception tracking and interface

Some application suites have external crash handlers that depend on
being able to use PTRACE to generate crash reports (KDE, Chromium, etc).
Since the inferior process generally knows the PID of the debugger,
it can use PR_SET_PTRACER to allow a specific PID and its descendants
to perform the PTRACE instead of only a direct ancestor.

Signed-off-by: Kees Cook <kees.cook@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: security: create task_free security callback
Kees Cook [Tue, 29 Jun 2010 05:34:04 +0000 (22:34 -0700)]
UBUNTU: SAUCE: security: create task_free security callback

The current LSM interface to cred_free is not sufficient for allowing
an LSM to track the life and death of a task.  This patch adds the
task_free hook so that an LSM can clean up resources on task death.

Signed-off-by: Kees Cook <kees.cook@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: [Upstream] security: Yama LSM
Kees Cook [Mon, 21 Jun 2010 20:07:07 +0000 (13:07 -0700)]
UBUNTU: [Upstream] security: Yama LSM

This adds the Yama Linux Security Module to collect several security
features (symlink, hardlink, and PTRACE restrictions) that have existed
in various forms over the years and have been carried outside the mainline
kernel by other Linux distributions like Openwall and grsecurity.

Currently accepted in upstream security tree:
git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6.git next

Signed-off-by: Kees Cook <kees.cook@canonical.com>
[kees: resolved AppArmor conflicts for Maverick merge]
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ARM: Adding regulator supply for vdds_sdi.
Mathieu J. Poirier [Sun, 27 Jun 2010 20:08:58 +0000 (14:08 -0600)]
UBUNTU: ARM: Adding regulator supply for vdds_sdi.

The omapfb driver couldn't locate its display sink because of
an initialisation error in the DSS subsystem.  This error was
caused by a missing 'sdi' entry in the board power regulator list.

BugLink: https://bugs.launchpad.net/bugs/597904

Signed-off-by: Mathieu Poirier <mathieu.poirier@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: [Upstream] HID: magicmouse: add param for scroll speed
Chase Douglas [Thu, 24 Jun 2010 15:37:29 +0000 (11:37 -0400)]
UBUNTU: [Upstream] HID: magicmouse: add param for scroll speed

The new scroll_speed param takes an integer value from 0 to 63, where 0
is slowest and 63 is fastest. The default of 32 remains the same. This
parameter also affects scroll acceleration linearly.

A second part of this change is a tightly coupled modification to the
scroll acceleration. Previously, scroll acceleration could be reset
without lifting the scroll finger. This is rather unintuitive and hard
to control in the case where a user wants faster scrolling, but wants to
hold the scroll touch for longer than a moment.

Note that scroll acceleration levels are now 1-7, where 7 is slowest. In
the previous implementation, there were 8 levels defined, but it was
impossible to start at the slowest level. In order to keep the default
scroll speed unchanged, only 7 levels are used now.

Signed-off-by: Chase Douglas <chase.douglas@canonical.com>
Acked-by: Michael Poole <mdpoole@troilus.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
(cherry picked from commit 0b778e76c1e7ccf49f8980b594e72f984095fd26
from git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid.git)
Acked-by: Steve Conklin <steve.conklin@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: [Upstream] HID: magicmouse: properly account for scroll movement in state
Chase Douglas [Thu, 24 Jun 2010 15:37:28 +0000 (11:37 -0400)]
UBUNTU: [Upstream] HID: magicmouse: properly account for scroll movement in state

Before this change, sequential scroll events would take a variable
amount of movement due to incorrect accounting. This change ensures all
scroll movements require a deterministic touch movement for an action to
occur.

Signed-off-by: Chase Douglas <chase.douglas@canonical.com>
Acked-by: Michael Poole <mdpoole@troilus.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
(cherry picked from commit 8d93efb27ab8927ffc7a357f1b2d10039de50ed4
from git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid.git)
Acked-by: Steve Conklin <steve.conklin@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: [Upstream] HID: magicmouse: disable and add module param for scroll acceleration
Chase Douglas [Thu, 24 Jun 2010 15:37:27 +0000 (11:37 -0400)]
UBUNTU: [Upstream] HID: magicmouse: disable and add module param for scroll acceleration

Scroll acceleration is unique to the magicmouse driver, and is
unintuitive to a user who is unaware of the functionality. Thus, disable
it by default, but add a module parameter to enable it for power users
who want it.

Signed-off-by: Chase Douglas <chase.douglas@canonical.com>
Acked-by: Michael Poole <mdpoole@troilus.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
(cherry picked from commit 9846f350ef4d4108c1154acfc125fe8d8630ef84
from git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid.git)
Acked-by: Steve Conklin <steve.conklin@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: [Upstream] HID: magicmouse: scroll on entire surface, not just middle of...
Chase Douglas [Thu, 24 Jun 2010 15:37:26 +0000 (11:37 -0400)]
UBUNTU: [Upstream] HID: magicmouse: scroll on entire surface, not just middle of mouse

Previously, scroll events only occurred when the user moved a touch
along the middle of the touch surface. This is unintuitive for a normal
user who is not aware of this. The device has a uniform surface, so the
distinction is artificial. This change removes the touch area check for
a scroll event, which replicates the OS X behavior.

Signed-off-by: Chase Douglas <chase.douglas@canonical.com>
Acked-by: Michael Poole <mdpoole@troilus.org>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
(cherry picked from commit ef566d30a702cc9b49d24edc4ad45c62208a4f5d
from git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid.git)
Acked-by: Steve Conklin <steve.conklin@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoperf probe: Add kernel source path option
Chase Douglas [Thu, 17 Jun 2010 21:08:11 +0000 (17:08 -0400)]
perf probe: Add kernel source path option

The probe plugin requires access to the source code for some operations.  The
source code must be in the exact same location as specified by the DWARF tags,
but sometimes the location is an absolute path that cannot be replicated by a
normal user. This change adds the -s|--source option to allow the user to
specify the root of the kernel source tree.

Cc: Ingo Molnar <mingo@elte.hu>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Masami Hiramatsu <mhiramat@redhat.com>
LKML-Reference: <1276543590-10486-1-git-send-email-chase.douglas@canonical.com>
Signed-off-by: Chase Douglas <chase.douglas@canonical.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
(cherry picked from commit 9ed7e1b85cd55dc46cb9410a23086bdaa2ff3eb9 from
git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6.git)
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: add option to hand off all kernel parameters to init
Andy Whitcroft [Thu, 17 Jun 2010 09:46:47 +0000 (10:46 +0100)]
UBUNTU: SAUCE: add option to hand off all kernel parameters to init

BugLink: http://bugs.launchpad.net/bugs/586386

Some init packages such as upstart find having all of the kernel parameters
passed in useful.  Currently they have to open up /proc/cmdline and
reparse that to obtain this information.  Add a kernel configuration
option to enable passing of all options.

Note, enabling this option will reduce the chances that a fallback from
/sbin/init to /bin/bash or /bin/sh will succeed.  Though it should be
noted that there are commonly unknown options present which would already
break this fallback.  init=/bin/foo provides explicit control over options
which is unaffected by this change.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: [Upstream] USB: option: Remove duplicate AMOI_VENDOR_ID
Leann Ogasawara [Thu, 10 Jun 2010 21:51:54 +0000 (14:51 -0700)]
UBUNTU: [Upstream] USB: option: Remove duplicate AMOI_VENDOR_ID

AMOI_VENDOR_ID is defined twice.  Remove the duplicate entry and move
the AMOI_PRODUCT_9508 definition to be grouped with the other AMOI
product definitions.

Sent upstream:
http://thread.gmane.org/gmane.linux.usb.general/32545

Originally-by: Ben Collins <ben.collins@ubuntu.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: ndiswrapper: convert multicast list to list_head
Leann Ogasawara [Thu, 3 Jun 2010 20:34:15 +0000 (13:34 -0700)]
UBUNTU: SAUCE: ndiswrapper: convert multicast list to list_head

The following upstream commit removed mc_count from struct net_device
and converts multicast lists to list_head:

  commit 22bedad3ce112d5ca1eaf043d4990fa2ed698c87
  Author: Jiri Pirko <jpirko@redhat.com>
  Date:   Thu Apr 1 21:22:57 2010 +0000

    net: convert multicast list to list_head

Thus, we need to convert ndiscrapper to incorporate the same changes to
avoid build failures.

Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: AUFS -- track changes to the arguements to fop fsync()
Andy Whitcroft [Thu, 3 Jun 2010 07:40:41 +0000 (08:40 +0100)]
UBUNTU: ubuntu: AUFS -- track changes to the arguements to fop fsync()

The file operation fsync dropped an unused struct dentry parameter,
apply this change to aufs2:

  commit 7ea8085910ef3dd4f3cad6845aaa2b580d39b115
  Author: Christoph Hellwig <hch@lst.de>
  Date:   Wed May 26 17:53:25 2010 +0200

    drop unused dentry argument to ->fsync

WARNING: this currently removes support for syncing some forms of aufs
directory as we are now required to have a filp to sync against and aufs
does not always have them.  I have added a WARN_ON to try and track the
triggers for this mode if any.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: AUFS -- update to standalone 2.6.35-rcN as at 20100601
Andy Whitcroft [Thu, 3 Jun 2010 07:34:46 +0000 (08:34 +0100)]
UBUNTU: ubuntu: AUFS -- update to standalone 2.6.35-rcN as at 20100601

BugLink: http://bugs.launchpad.net/bugs/587888

Update to current aufs2 for v2.6.35-rcN:

  commit a9be01e5e9688018ebe9ef46ec5414bb356bc556
  Author: J. R. Okajima <hooanon05@yahoo.co.jp>
  Date:   Sun May 30 04:40:49 2010 +0900

    aufs: fi_hdir, rename au_do_refresh_file() to au_do_refresh_dir()

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: ndiswrapper: rename usb_buffer_alloc() and usb_buffer_free()
Leann Ogasawara [Thu, 3 Jun 2010 03:42:24 +0000 (20:42 -0700)]
UBUNTU: SAUCE: ndiswrapper: rename usb_buffer_alloc() and usb_buffer_free()

usb_buffer_alloc() is renamed to usb_alloc_coherent()
usb_buffer_free()  is renamed to usb_free_coherent()

This is due to the removal of the usb_buffer_alloc() and
usb_buffer_free() macros:

  commit e26bcf37234c67624f62d9fc95f922b8dbda1363
  Author: Greg Kroah-Hartman <gregkh@suse.de>
  Date:   Thu Apr 29 15:46:07 2010 -0700

    USB: remove unused usb_buffer_alloc and usb_buffer_free macros

Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: AUFS -- update to standalone 2.6.34 as at 20100601
Andy Whitcroft [Tue, 1 Jun 2010 13:23:50 +0000 (14:23 +0100)]
UBUNTU: ubuntu: AUFS -- update to standalone 2.6.34 as at 20100601

BugLink: http://bugs.launchpad.net/bugs/587888

Update to current aufs2 for v2.6.34:

  commit a9be01e5e9688018ebe9ef46ec5414bb356bc556
  Author: J. R. Okajima <hooanon05@yahoo.co.jp>
  Date:   Sun May 30 04:40:49 2010 +0900

    aufs: fi_hdir, rename au_do_refresh_file() to au_do_refresh_dir()

Also adds ubuntu/aufs/include to the include path for aufs so it can
find its types.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: AUFS -- aufs2 base patch for linux-2.6.34
Andy Whitcroft [Tue, 1 Jun 2010 13:10:55 +0000 (14:10 +0100)]
UBUNTU: ubuntu: AUFS -- aufs2 base patch for linux-2.6.34

BugLink: http://bugs.launchpad.net/bugs/587888

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: iscsitarget -- version 1.4.20.1
Leann Ogasawara [Wed, 26 May 2010 02:31:39 +0000 (19:31 -0700)]
UBUNTU: ubuntu: iscsitarget -- version 1.4.20.1

Update iscsitarget to the latest stable version (v1.4.20.1) from
sourceforge.

ExternalDriver: iscsi_trgt
Url: http://sourceforge.net/projects/iscsitarget/files/
Version: 1.4.20.1

Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: Updated ndiswrapper to 1.56
Tim Gardner [Wed, 19 May 2010 15:17:26 +0000 (09:17 -0600)]
UBUNTU: SAUCE: Updated ndiswrapper to 1.56

BugLink: http://bugs.launchpad.net/bugs/582555

Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: rtl8192se -- update to version 0015.0127.2010
Tim Gardner [Fri, 23 Apr 2010 16:55:46 +0000 (10:55 -0600)]
UBUNTU: ubuntu: rtl8192se -- update to version 0015.0127.2010

BugLink: http://bugs.launchpad.net/bugs/567016

Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Acked-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) dma-mapping: Remove WARN_ON in dma_free_coherent
Stefan Bader [Mon, 29 Mar 2010 17:12:36 +0000 (19:12 +0200)]
UBUNTU: SAUCE: (no-up) dma-mapping: Remove WARN_ON in dma_free_coherent

BugLink: http://bugs.launchpad.net/bugs/458201

Triggered by the following backtrace:

WARNING: at
/build/buildd/linux-2.6.32/arch/x86/include/asm/dma-mapping.h:154
___free_dma_mem_cluster+0x102/0x110()

 [<ffffffff81064f9b>] warn_slowpath_common+0x7b/0xc0
 [<ffffffff81064ff4>] warn_slowpath_null+0x14/0x20
 [<ffffffff8139a2a2>] ___free_dma_mem_cluster+0x102/0x110
 [<ffffffff8139a072>] __sym_mfree+0xd2/0x100
 [<ffffffff8139a109>] __sym_mfree_dma+0x69/0x100
 [<ffffffff8139245f>] sym_hcb_free+0x8f/0x1f0

This patch never will be accepted upstream because the WARN_ON
is supposed to perevent driver development which is only
compatible with x86 on x86 (ARM can sleep in that function).
The right way to fix it would be to make the offending function
use locks in the right way but that requires careful implementation.

Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Acked-by: Andy Whitcroft <apw@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: ubuntu: dm-raid4-5 -- update to compile with 2.6.34-rc4
Leann Ogasawara [Wed, 14 Apr 2010 15:27:05 +0000 (08:27 -0700)]
UBUNTU: ubuntu: dm-raid4-5 -- update to compile with 2.6.34-rc4

Need to include linux/slab.h to prevent implicit declaration of
functions which otherwise result in build failures.

Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agodrm/radeon/bo: add some fallback placements for VRAM only objects.
Dave Airlie [Fri, 19 Mar 2010 00:32:10 +0000 (10:32 +1000)]
drm/radeon/bo: add some fallback placements for VRAM only objects.

On constrained r100 systems compiz would fail to start due to a lack
of memory, we can just fallback place the objects rather than completely
failing it works a lot better.

Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>

(cherry picked from commit 924788672f9c802a4c8e6bc8d88fbc2c9b2e3089 drm-2.6)
BugLink: http://launchpad.net/bugs/507148

Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>
Acked-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: ubuntu: dm-raid4-5 -- update to compile with 2.6.34-rc2
Leann Ogasawara [Fri, 26 Mar 2010 22:15:51 +0000 (15:15 -0700)]
UBUNTU: ubuntu: dm-raid4-5 -- update to compile with 2.6.34-rc2

The following commit removes unused params from dm_get_device().
Convert dm-raid4-5 to match:

  commit 8215d6ec5fee1e76545decea2cd73717efb5cb42
  Author: Nikanth Karthikesan <knikanth@novell.com>
  Date:   Sat Mar 6 02:32:27 2010 +0000

    dm table: remove unused dm_get_device range parameters

Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: iscsitarget -- rename daddr inet_sock field
Leann Ogasawara [Wed, 24 Mar 2010 19:37:34 +0000 (12:37 -0700)]
UBUNTU: ubuntu: iscsitarget -- rename daddr inet_sock field

The following commit renames some inet_sock fields including daddr.
Update daddr to inet_daddr for iscsitarget.

  commit c720c7e8383aff1cb219bddf474ed89d850336e3
  Author: Eric Dumazet <eric.dumazet@gmail.com>
  Date:   Thu Oct 15 06:30:45 2009 +0000

    inet: rename some inet_sock fields

Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: ubuntu: dm-raid4-5 -- update to compile with 2.6.33
Leann Ogasawara [Wed, 24 Mar 2010 17:36:13 +0000 (10:36 -0700)]
UBUNTU: ubuntu: dm-raid4-5 -- update to compile with 2.6.33

The following commit adds a fn callback arguement to dm_dirty_log_create().
convert dm-raid4-5 to match:

    commit 87a8f240e9bcf025ba45e4563c842b0d59c5e8ef
    Author: Mikulas Patocka <mpatocka@redhat.com>
    Date:   Thu Dec 10 23:52:01 2009 +0000

        dm log: add flush callback fn

Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) cdrom -- default to not locking the tray when in use
Andy Whitcroft [Thu, 11 Mar 2010 10:33:36 +0000 (10:33 +0000)]
UBUNTU: SAUCE: (no-up) cdrom -- default to not locking the tray when in use

BugLink: http://bugs.launchpad.net/bugs/397734

It seems that users are have a high expectation that the eject button
on their CDROM drive will eject the disk regardless of whether it is in
use or not.  To this end we are now changing the default LOCK mode for
mounted CDROMS to 0 to allow ejects.  This however does not handle the
direct open cases like music and video players.  From the launchpad bug
commentary:

    So, according to the upstream discussion David Zeuthen recommended
    to just not lock CD-ROM trays by default. Kernel/userspace already
    handles prematurely removed USB storage devices reasonably, and with
    read-only devices like CD-ROMs it is even less of an issue. So we
    should just set /proc/sys/dev/cdrom/lock to 0 by default.

Note that we still will have the drive mounted after the eject.  There is a
media change uevent generated and this will be used to trigger the unmount
of the drive in udisks.  The burner software will also have to be looked
at to ensure they are explicitly locking the drive closed during the burn.
This will all be handled under the bug above.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Tim Gardner <tim.gardner@canonical.com>
Acked-by: Colin King <colin.king@canonical.com>

13 years agoUBUNTU: ubuntu: rtl8192se -- version 2010-0115,0014
Manoj Iyer [Tue, 2 Mar 2010 06:01:41 +0000 (00:01 -0600)]
UBUNTU: ubuntu: rtl8192se -- version 2010-0115,0014

ExternalDriver: rtl8192se_pci
Description:    Driver for RTL8191/92 or RTL8171 wifi nic
Url:            http://www.realtek.com/
Mask:
Version:        2010-0115,0014

Buglink: http://launchpad.net/bugs/530275

Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) PM report driver and device suspend/resume times.
Manoj Iyer [Fri, 18 Dec 2009 00:57:47 +0000 (01:57 +0100)]
UBUNTU: SAUCE: (no-up) PM report driver and device suspend/resume times.

Based on a patch from Rafael J. Wysocki. This patch prints suspend/resume
information for each driver/device to dmesg.

[apw@canonical.com: rejigged to split config updates]
[apw@canonical.com: move the configuration item fix build when full PM
 is not enabled]
Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: ubuntu: fsam7400: Depend on CHECK_SIGNATURE
Colin Watson [Thu, 25 Feb 2010 13:23:30 +0000 (13:23 +0000)]
UBUNTU: ubuntu: fsam7400: Depend on CHECK_SIGNATURE

Needed so that 'make defconfig && make' works.

Signed-off-by: Colin Watson <cjwatson@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: ubuntu: dm-raid4-5: Depend on XOR_BLOCKS
Colin Watson [Thu, 25 Feb 2010 13:21:02 +0000 (13:21 +0000)]
UBUNTU: ubuntu: dm-raid4-5: Depend on XOR_BLOCKS

Needed so that 'make defconfig && make' works.

Signed-off-by: Colin Watson <cjwatson@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: khubd -- switch USB product/manufacturer/serial handling to RCU
Andy Whitcroft [Wed, 17 Feb 2010 10:13:42 +0000 (10:13 +0000)]
UBUNTU: SAUCE: khubd -- switch USB product/manufacturer/serial handling to RCU

BugLink: http://bugs.launchpad.net/bugs/510937

With the introduction of wireless USB hubs the product, manufacturer,
and serial number are now mutable.  This necessitates new locking in the
consumers of these values including the sysfs read routines in order to
prevent use-after-free acces to these values.  These extra locks create
significant lock contention leading to increased boot times (0.3s for an
example Atom based system).  Move update of these values to RCU based
locking.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agox86, mce: Make xeon75xx memory driver dependent on PCI
Andi Kleen [Sat, 23 Jan 2010 11:33:59 +0000 (12:33 +0100)]
x86, mce: Make xeon75xx memory driver dependent on PCI

commit 757fd770c649b0dfa6eeefc2d5e2ea3119b6be9c upstream (linux-2.6-tip)

Found by Ingo Molnar's automated tester.

Reported-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andi Kleen <ak@linux.intel.com>
LKML-Reference: <20100123113359.GA29555@one.firstfloor.org>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agox86, mce: Rename cpu_specific_poll to mce_cpu_specific_poll
H. Peter Anvin [Fri, 22 Jan 2010 02:31:54 +0000 (18:31 -0800)]
x86, mce: Rename cpu_specific_poll to mce_cpu_specific_poll

commit f91c4d2649531cc36e10c6bc0f92d0f99116b209 upstream (linux-2.6-tip)

cpu_specific_poll is a global variable, and it should have a global
namespace name.  Since it is MCE-specific (it takes a struct mce *),
rename it mce_cpu_specific_poll.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Andi Kleen <andi@firstfloor.org>
LKML-Reference: <20100121221711.GA8242@basil.fritz.box>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agox86, mce: Xeon75xx specific interface to get corrected memory error information
Andi Kleen [Thu, 21 Jan 2010 22:17:12 +0000 (23:17 +0100)]
x86, mce: Xeon75xx specific interface to get corrected memory error information

commit c773f70fd6b53ee646727f871833e53649907264 upstream (linux-2.6-tip)

Xeon 75xx doesn't log physical addresses on corrected machine check
events in the standard architectural MSRs. Instead the address has to
be retrieved in a model specific way. This makes it impossible to do
predictive failure analysis.

Implement cpu model specific code to do this in mce-xeon75xx.c using a
new hook that is called from the generic poll code. The code retrieves
the physical address/DIMM of the last corrected error from the
platform and makes the address look like a standard architectural MCA
address for further processing.

In addition the DIMM information is retrieved and put into two new
aux0/aux1 fields in struct mce.  These fields are specific to a given
CPU.  These fields can then be decoded by mcelog into specific DIMM
information.  The latest mcelog version has support for this.

Longer term this will be likely in a different output format, but
short term that seemed like the least intrusive solution. Older mcelog
can deal with an extended record.

There's no code to print this information on a panic because this only
works for corrected errors, and corrected errors do not usually result
in panics.

The act of retrieving the DIMM/PA information can take some time, so
this code has a rate limit to avoid taking too much CPU time on a
error flood.

The whole thing can be loaded as a module and has suitable PCI-IDs so
that it can be auto-loaded by a distribution.  The code also checks
explicitely for the expected CPU model number to make sure this code
doesn't run anywhere else.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
LKML-Reference: <20100121221711.GA8242@basil.fritz.box>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: (no-up) arm -- enable ubuntu/ directory
Eric Miao [Mon, 25 Jan 2010 13:52:37 +0000 (21:52 +0800)]
UBUNTU: (no-up) arm -- enable ubuntu/ directory

Signed-off-by: Eric Miao <eric.miao@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: Make CONFIG_{OMNIBOOK, AVERATEC_5100P, PACKARDBELL_E5} depend on X86
Eric Miao [Wed, 27 Jan 2010 04:02:24 +0000 (04:02 +0000)]
UBUNTU: SAUCE: Make CONFIG_{OMNIBOOK, AVERATEC_5100P, PACKARDBELL_E5} depend on X86

There options are really X86 specific, and the source won't compile on
ARM platforms. Make them depend on X86 and less confusing when syncing
config options to ARM.

Signed-off-by: Eric Miao <eric.miao@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: acpi battery -- move first lookup asynchronous
Andy Whitcroft [Thu, 24 Dec 2009 02:09:03 +0000 (02:09 +0000)]
UBUNTU: SAUCE: acpi battery -- move first lookup asynchronous

BugLink: http://bugs.launchpad.net/bugs/507211

When instantiating the battery object on to the acpi bus in the kernel
we talk to the BIOS to get the current battery state.  This can take
a long time and holds the acpi bus object locked for the duration.
This leads to any other object wishing to add itself that bus blocking.
This leads to unpredicatable delays of up to .3s when initialising the
hpet during boot depending on execution order.  Make the first update of
the battery asynchronous.  Move the acpi bus handling back synchronous.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: i915 -- disable powersave by default
Andy Whitcroft [Mon, 18 Jan 2010 16:43:11 +0000 (16:43 +0000)]
UBUNTU: SAUCE: i915 -- disable powersave by default

BugLink: http://bugs.launchpad.net/bugs/492392

Powersave move seems to trigger both flickering during normal use and
apparent hard locks of a solid colour on the whole screen.  Disabling
powersave seems to avoid both issues.  Move the default to 0 for now.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: async_populate_rootfs: move rootfs init earlier
Andy Whitcroft [Wed, 23 Dec 2009 16:43:07 +0000 (16:43 +0000)]
UBUNTU: SAUCE: async_populate_rootfs: move rootfs init earlier

Check to see if the machine has more than one active CPU, if it does
then it is worth starting the decode of the rootfs earlier.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: Increase the default prealloc buffer for HDA audio devices (non-modem)
Brad Figg [Mon, 7 Dec 2009 20:24:47 +0000 (12:24 -0800)]
UBUNTU: SAUCE: Increase the default prealloc buffer for HDA audio devices (non-modem)

This allows for buffering of up to 4 seconds of audio. This is
intended to improve the audio experience. This patch is being
carried by Fedora but is not upstream.

Signed-off-by: Brad Figg <brad.figg@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: ubuntu: onmibook -- Added missing BOM file
Manoj Iyer [Thu, 10 Dec 2009 18:13:15 +0000 (12:13 -0600)]
UBUNTU: ubuntu: onmibook -- Added missing BOM file

Forgot to add BOM file when I added omnibook module to ubuntu/
this patch adds it.

Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: ubuntu: fsam7400 -- Cleanup Makefile
Tim Gardner [Mon, 14 Dec 2009 23:17:34 +0000 (16:17 -0700)]
UBUNTU: ubuntu: fsam7400 -- Cleanup Makefile

Ripped out the cruft from this Makefile so as not
to confuse folks.

Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: ubuntu: iscsitarget --- version 1.4.19
Manoj Iyer [Fri, 4 Dec 2009 01:54:20 +0000 (19:54 -0600)]
UBUNTU: ubuntu: iscsitarget --- version 1.4.19

BugLink: http://bugs.launchpad.net/bugs/494693

ExternalDriver: iscsitarget
Description: iSCSI storage system on Linux
Url: svn://svn.berlios.de/iscsitarget/trunk
Mask:
Version: 1.4.19

Use filemap_write_and_wait_range() in place of sync_page_range().
This appears to better preserve the original semantics.

Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: ubuntu: omnibook -- support Toshiba (HP) netbooks
Manoj Iyer [Tue, 8 Dec 2009 03:20:32 +0000 (21:20 -0600)]
UBUNTU: ubuntu: omnibook -- support Toshiba (HP) netbooks

ExternalDriver: omnibook
Description: driver for HP & Toshiba laptops
Url: https://omnibook.svn.sourceforge.net/svnroot/omnibook/omnibook/trunk
Mask:
Version: 2.20070211

Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>

13 years agoUBUNTU: UBUNTU: ubuntu: fsam7400 -- kill switch for Fujitsu Siemens Amilo M 7400
Manoj Iyer [Wed, 9 Dec 2009 17:54:32 +0000 (17:54 +0000)]
UBUNTU: UBUNTU: ubuntu: fsam7400 -- kill switch for Fujitsu Siemens Amilo M 7400

ExternalDriver: fsam7400
Description: SW RF kill switch for Fujitsu Siemens Amilo M7400/Maxdata
7000DX and compatible hardware.
Url: http://www.iragan.com/linux/fsam7400/
Mask:
Version: 0.5.2

Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: isapnp_init: make isa PNP scans occur async
Andy Whitcroft [Wed, 2 Dec 2009 14:41:53 +0000 (14:41 +0000)]
UBUNTU: SAUCE: isapnp_init: make isa PNP scans occur async

The results of scanning for devices is to trigger udev events therefore
we can push this processing async.

This reduces kernel initialisation time (the time from bootloader to
starting userspace) by several 10ths of a second x86 32bit systems.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: Make populate_rootfs asynchronous
Surbhi Palande [Tue, 8 Dec 2009 09:44:34 +0000 (11:44 +0200)]
UBUNTU: SAUCE: Make populate_rootfs asynchronous

The expansion of the initramfs is completely independant of other
boot activities.  The original data is already present at boot and the
filesystem is not required until we are ready to start init.  It is
therefore reasonable to populate the rootfs asynchronously.  Move this
processing to an async call.

This reduces kernel initialisation time (the time from bootloader to
starting userspace) by several 10ths of a second on a selection of test
hardware particularly SMP systems, although UP system also benefit.

Signed-off-by: Surbhi Palande <surbhi.palande@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) select a v7 CPU for versatile
Loic Minier [Thu, 4 Jun 2009 13:53:41 +0000 (15:53 +0200)]
UBUNTU: SAUCE: (no-up) select a v7 CPU for versatile

This allows us to use this kernel under QEMU while still using the
main Ubuntu ARMv7 userspace.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: vfs: Add a trace point in the mark_inode_dirty function
Arjan van de Ven [Sun, 25 Oct 2009 22:37:04 +0000 (15:37 -0700)]
UBUNTU: SAUCE: vfs: Add a trace point in the mark_inode_dirty function

PowerTOP would like to be able to show who is keeping the disk
busy by dirtying data. The most logical spot for this is in the vfs
in the mark_inode_dirty() function. Doing this on the block level
is not possible because by the time the IO hits the block layer the
guilty party can no longer be found ("kjournald" and "pdflush" are not
useful answers to "who caused this file to be dirty).

The trace point follows the same logic/style as the block_dump code
and pretty much dumps the same data, just not to dmesg (and thus to
/var/log/messages) but via the trace events streams.

Note: This patch was posted to lkml and might potentially go into 2.6.33 but I
have not seen which maintainer will take it.

Signed-of-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Amit Kucheria <amit.kucheria@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) set /proc/acpi/video/*/DOS to 4 by default
Andy Whitcroft [Mon, 9 Nov 2009 17:30:29 +0000 (17:30 +0000)]
UBUNTU: SAUCE: (no-up) set /proc/acpi/video/*/DOS to 4 by default

BugLink: http://bugs.launchpad.net/bugs/458982

acpi_video_bus_start_devices() in drivers/acpi/video.c sets the default
ACPI DOS value to 0, which lets the OS handle toggling of display output
but still leaves the BIOS handling brightness levels automatically when
connecting/disconnecting AC.

We want the OS to handle both where possible, so a better default would
be to set this to 4.

This likely will regress systems using proprietary video drivers which
will need to change this setting back using the sysfs interfaces.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: [Upstream] add local prefix to oss local change_bits
Andy Whitcroft [Wed, 11 Nov 2009 13:29:30 +0000 (13:29 +0000)]
UBUNTU: [Upstream] add local prefix to oss local change_bits

[Fixes powerpc FTBFS]

This collides with upstream powerpc exported functions.  Rename the
local copy in the oss soundblaster mixer and ad1848 driver.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) trace: add trace events for open(), exec() and uselib()
Scott James Remnant [Tue, 27 Oct 2009 10:05:32 +0000 (10:05 +0000)]
UBUNTU: SAUCE: (no-up) trace: add trace events for open(), exec() and uselib()

BugLink: http://bugs.launchpad.net/bugs/462111

This patch uses TRACE_EVENT to add tracepoints for the open(),
exec() and uselib() syscalls so that ureadahead can cheaply trace
the boot sequence to determine what to read to speed up the next.

It's not upstream because it will need to be rebased onto the syscall
trace events whenever that gets merged, and is a stop-gap.

Signed-off-by: Scott James Remnant <scott@ubuntu.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Acked-by: Andy Whitcroft <andy.whitcroft@canonical.com>
Signed-off-by: Stefan Bader <stefan.bader@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) raise the default console 'quiet' level to 2
Tim Gardner [Thu, 8 Oct 2009 17:42:36 +0000 (11:42 -0600)]
UBUNTU: SAUCE: (no-up) raise the default console 'quiet' level to 2

In the interests of providing a clean boot experience, i.e., a blank screen
before X starts, supress noisy driver messages.

See https://lists.ubuntu.com/archives/kernel-team/2009-October/007476.html
for the original diatribe.

Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: ubuntu: dm-raid-45 -- update to compile with 2.6.32
Andy Whitcroft [Tue, 6 Oct 2009 08:33:30 +0000 (09:33 +0100)]
UBUNTU: ubuntu: dm-raid-45 -- update to compile with 2.6.32

This commit changed bio_barrier() to direct flag accessors,
convert dm-raid-4-5 to match:

  commit 1f98a13f623e0ef666690a18c1250335fc6d7ef1
  Author: Jens Axboe <jens.axboe@oracle.com>
  Date:   Fri Sep 11 14:32:04 2009 +0200

    bio: first step in sanitizing the bio->bi_rw flag testing

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) disable adding scsi headers to linux-libc-dev
Andy Whitcroft [Fri, 26 Jun 2009 09:59:30 +0000 (10:59 +0100)]
UBUNTU: SAUCE: (no-up) disable adding scsi headers to linux-libc-dev

Currently scsi headers are generated by the kernel and by libc6-dev.
We need to coordinate any switch over to the kernel.  Temporarily disabled
these headers in the kernel package.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: Sony laptop: Some Sony Vaia laptops do not enable wwan power by default.
Tim Gardner [Tue, 21 Apr 2009 17:09:10 +0000 (11:09 -0600)]
UBUNTU: Sony laptop: Some Sony Vaia laptops do not enable wwan power by default.

Bug: https://bugs.launchpad.net/bugs/364678

Added quirk to enable wwan power based on DMI information already present in the module.
Ity appears that Vaio's do not enable power to wwan from a cold boot.

Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: SAUCE: hotkey quirks for various Zeptro Znote and Fujitsu Amilo laptops
Andy Whitcroft [Mon, 23 Feb 2009 18:23:53 +0000 (18:23 +0000)]
UBUNTU: SAUCE: hotkey quirks for various Zeptro Znote and Fujitsu Amilo laptops

OriginalAuthor: Marius Wenderoth <scale87>
Bug: #330259

Add model specific quirks for the hotkeys for the laptops below:

Zepto Znote 6615WD
Zepto Znote 6625WD
FUJITSU SIEMENS AMILO Xi 2428

In each case the volume keys and the mute key do not correctly produce
key release events.  Quirk these to force a key release. (Patch modified
to compile with 2.6.30)

[apw@canonical.com: Track the consolidation of release quirks in the
commit below:

    commit 000c2a35b8b0485f5a872c24c4f2d0d6579951c1
    Author: Herton Ronaldo Krzesinski <herton@mandriva.com.br>
    Date:   Fri Oct 16 16:13:59 2009 -0700

Input: atkbd - consolidate force release quirks for volume keys
]

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Tim Gardner <tim.gardner@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Stefan Bader <stefan.bader@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) nbd: Change default partitions per device to 15
Scott James Remnant [Mon, 16 Mar 2009 15:03:07 +0000 (15:03 +0000)]
UBUNTU: SAUCE: (no-up) nbd: Change default partitions per device to 15

This was previously changed by using an "options" line in a modprobe.d
file, however that practice is now deprecated.  This is because module
names, option names, their values and even their current defaults can
all change inside the kernel and module-init-tools has never been kept
in sync.

In addition, changing the kernel means that the option change will apply
if the module is built in by users or the OEM team.

Bug: #342563

Signed-off-by: Scott James Remnant <scott@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) swap: Add notify_swap_entry_free callback for compcache
Tim Gardner [Sun, 15 Mar 2009 19:38:17 +0000 (13:38 -0600)]
UBUNTU: SAUCE: (no-up) swap: Add notify_swap_entry_free callback for compcache

Code is required for ubuntu/compcache

Signed-off-by: Ben Collins <ben.collins@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) hostap: Change initial operation mode to managed (infra)
Scott James Remnant [Tue, 3 Mar 2009 14:24:52 +0000 (14:24 +0000)]
UBUNTU: SAUCE: (no-up) hostap: Change initial operation mode to managed (infra)

This was previously changed by using an "options" line in a modprobe.d
file, however that practice is now deprecated.  This is because module
names, option names, their values and even their current defaults can
all change inside the kernel and module-init-tools has never been kept
in sync.

In addition, changing the kernel means that the option change will apply
if the module is built in by users or the OEM team.

Signed-off-by: Scott James Remnant <scott@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) libata: Ignore HPA by default.
Scott James Remnant [Tue, 3 Mar 2009 14:20:01 +0000 (14:20 +0000)]
UBUNTU: SAUCE: (no-up) libata: Ignore HPA by default.

This was previously changed by using an "options" line in a modprobe.d
file, however that practice is now deprecated.  This is because module
names, option names, their values and even their current defaults can
all change inside the kernel and module-init-tools has never been kept
in sync.

In addition, changing the kernel means that the option change will apply
if the module is built in by users or the OEM team.

Signed-off-by: Scott James Remnant <scott@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) version: Implement version_signature proc file.
Andy Whitcroft [Thu, 27 Nov 2008 19:12:07 +0000 (19:12 +0000)]
UBUNTU: SAUCE: (no-up) version: Implement version_signature proc file.

Signed-off-by: Andy Whitcroft <apw@canonical.com>
Acked-by: Tim Gardener <tim.gardner@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) Modularize vesafb
Ben Collins [Tue, 14 Oct 2008 15:30:10 +0000 (11:30 -0400)]
UBUNTU: SAUCE: (no-up) Modularize vesafb

Signed-off-by: Ben Collins <ben.collins@canonical.com>

13 years agoUBUNTU: SAUCE: Enable speedstep for sonoma processors.
cking [Wed, 13 Feb 2008 19:02:10 +0000 (19:02 +0000)]
UBUNTU: SAUCE: Enable speedstep for sonoma processors.

OriginalAuthor: Christophe Dumez, repackaged by Wolfgang Tremmel
OriginalLocation: http://launchpadlibrarian.net/11575235/speedstep-sonoma.patch
Bug: 132271

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Ben Collins <ben.collins@canonical.com>

13 years agoUBUNTU: SAUCE: (no-up) Restore VT fonts on switch
Stefan Bader [Tue, 4 Mar 2008 22:10:36 +0000 (22:10 +0000)]
UBUNTU: SAUCE: (no-up) Restore VT fonts on switch

Not all X drivers save and restore fonts on text VTs. Add code to the
kernel to explicitly save and restore them on VT switches.

Signed-off-by: Matthew Garrett <mjg59@srcf.ucam.org>
Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Ben Collins <ben.collins@canonical.com>

13 years agoUBUNTU: SAUCE: [PATCH] Add extra headers to linux-libc-dev
Tim Gardner [Wed, 2 Jan 2008 15:57:35 +0000 (08:57 -0700)]
UBUNTU: SAUCE: [PATCH] Add extra headers to linux-libc-dev

OriginalAuthor: Soren Hansen
OriginalLocation: https://lists.ubuntu.com/archives/kernel-team/2007-November/001891.html

Added
include/linux/netfilter_ipv6/ip6t_mh.h
include/linux/netfilter/xt_quota.h
to linux-libc-dev to fix iptables build of those two extensions.

Signed-off-by: Soren Hansen <soren at ubuntu.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: ubuntu: rfkill drivers -- version 1.3
Ben Collins [Fri, 18 Jul 2008 15:42:11 +0000 (11:42 -0400)]
UBUNTU: ubuntu: rfkill drivers -- version 1.3

Version: 1.3

Signed-off-by: Ben Collins <ben.collins@canonical.com>

13 years agoUBUNTU: ubuntu: ndiswrapper -- version 1.55
Andy Whitcroft [Mon, 29 Jun 2009 10:36:42 +0000 (11:36 +0100)]
UBUNTU: ubuntu: ndiswrapper -- version 1.55

ExternalDriver: ndiswrapper
Description: Windows network driver wrapper
Url: http://sourceforge.net/project/showfiles.php?group_id=93482
Version: 1.55

Update ndiswrapper to the latest version (v1.55) from sourceforge,
and renable.

Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: ubuntu: dm-raid45 -- version 2009.04.24 (2.6.30-rc3)
Manoj Iyer [Wed, 1 Jul 2009 22:51:07 +0000 (17:51 -0500)]
UBUNTU: ubuntu: dm-raid45 -- version 2009.04.24 (2.6.30-rc3)

ExternalDriver: dm-raid45
Description: This software extends device-mapper by RAID4 and RAID5 mappings.
Url: http://people.redhat.com/~heinzm/sw/dm/dm-raid45/
Version: 2009.04.24 (2.6.30-rc3)

Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

13 years agoUBUNTU: ubuntu: dm-raid4-5 -- (no-up) Export dm_disk function of device-mapper
Stefan Bader [Thu, 14 Feb 2008 14:55:49 +0000 (09:55 -0500)]
UBUNTU: ubuntu: dm-raid4-5 -- (no-up) Export dm_disk function of device-mapper

This function is externally used by the dm-raid4-5 module.

Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Ben Collins <ben.collins@canonical.com>

13 years agoUBUNTU: ubuntu: compcache -- version 0.5.3
Manoj Iyer [Sun, 5 Jul 2009 23:48:01 +0000 (17:48 -0600)]
UBUNTU: ubuntu: compcache -- version 0.5.3

ExternalDriver: ramzswap xvmalloc
Description: This is a RAM based block device which acts as swap disk.
Url: http://code.google.com/p/compcache/
Version: 0.5.3

rtg - cleaned up Kconfig, enabled BLK_DEV_COMPCACHE for all arches/flavours.
apw - further cleaned up Kconfig

Signed-off-by: Manoj Iyer <manoj.iyer@canonical.com>
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>
Signed-off-by: Andy Whitcroft <apw@canonical.com>

13 years agoUBUNTU: (no-up) fold down debian for ubuntu-natty 2.6.36 rebase
Leann Ogasawara [Sat, 13 Mar 2010 01:13:25 +0000 (17:13 -0800)]
UBUNTU: (no-up) fold down debian for ubuntu-natty 2.6.36 rebase

Ignore: yes

Signed-off-by: Leann Ogasawara <leann.ogasawara@canonical.com>

13 years agoLinux 2.6.38.2
Greg Kroah-Hartman [Sun, 27 Mar 2011 18:37:20 +0000 (11:37 -0700)]
Linux 2.6.38.2

13 years agoext4: skip orphan cleanup if fs has unknown ROCOMPAT features
Amir Goldstein [Mon, 28 Feb 2011 05:53:45 +0000 (00:53 -0500)]
ext4: skip orphan cleanup if fs has unknown ROCOMPAT features

commit d39195c33bb1b5fdcb0f416e8a0b34bfdb07a027 upstream.

Orphan cleanup is currently executed even if the file system has some
number of unknown ROCOMPAT features, which deletes inodes and frees
blocks, which could be very bad for some RO_COMPAT features,
especially the SNAPSHOT feature.

This patch skips the orphan cleanup if it contains readonly compatible
features not known by this ext4 implementation, which would prevent
the fs from being mounted (or remounted) readwrite.

Signed-off-by: Amir Goldstein <amir73il@users.sf.net>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

13 years agodcdbas: force SMI to happen when expected
Stuart Hayes [Wed, 2 Mar 2011 12:42:05 +0000 (13:42 +0100)]
dcdbas: force SMI to happen when expected

commit dd65c736d1b5312c80c88a64bf521db4959eded5 upstream.

The dcdbas driver can do an I/O write to cause a SMI to occur.  The SMI handler
looks at certain registers and memory locations, so the SMI needs to happen
immediately.  On some systems I/O writes are posted, though, causing the SMI to
happen well after the "outb" occurred, which causes random failures.  Following
the "outb" with an "inb" forces the write to go through even if it is posted.

Signed-off-by: Stuart Hayes <stuart_hayes@yahoo.com>
Acked-by: Doug Warzecha <douglas_warzecha@dell.com>
Cc: Chuck Ebbert <cebbert@redhat.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

13 years agofs: call security_d_instantiate in d_obtain_alias V2
Josef Bacik [Fri, 19 Nov 2010 01:52:55 +0000 (20:52 -0500)]
fs: call security_d_instantiate in d_obtain_alias V2

commit 24ff6663ccfdaf088dfa7acae489cb11ed4f43c4 upstream.

While trying to track down some NFS problems with BTRFS, I kept noticing I was
getting -EACCESS for no apparent reason.  Eric Paris and printk() helped me
figure out that it was SELinux that was giving me grief, with the following
denial

type=AVC msg=audit(1290013638.413:95): avc:  denied  { 0x800000 } for  pid=1772
comm="nfsd" name="" dev=sda1 ino=256 scontext=system_u:system_r:kernel_t:s0
tcontext=system_u:object_r:unlabeled_t:s0 tclass=file

Turns out this is because in d_obtain_alias if we can't find an alias we create
one and do all the normal instantiation stuff, but we don't do the
security_d_instantiate.

Usually we are protected from getting a hashed dentry that hasn't yet run
security_d_instantiate() by the parent's i_mutex, but obviously this isn't an
option there, so in order to deal with the case that a second thread comes in
and finds our new dentry before we get to run security_d_instantiate(), we go
ahead and call it if we find a dentry already.  Eric assures me that this is ok
as the code checks to see if the dentry has been initialized already so calling
security_d_instantiate() against the same dentry multiple times is ok.  With
this patch I'm no longer getting errant -EACCESS values.

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Chuck Ebbert <cebbert@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

13 years agoSUNRPC: Never reuse the socket port after an xs_close()
Trond Myklebust [Tue, 22 Mar 2011 22:40:10 +0000 (18:40 -0400)]
SUNRPC: Never reuse the socket port after an xs_close()

commit 246408dcd5dfeef2df437ccb0ef4d6ee87805f58 upstream.

If we call xs_close(), we're in one of two situations:
 - Autoclose, which means we don't expect to resend a request
 - bind+connect failed, which probably means the port is in use

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

13 years agoNFS: Fix a hang/infinite loop in nfs_wb_page()
Trond Myklebust [Mon, 21 Mar 2011 19:37:01 +0000 (15:37 -0400)]
NFS: Fix a hang/infinite loop in nfs_wb_page()

commit b8413f98f997bb3ed7327e6d7117e7e91ce010c3 upstream.

When one of the two waits in nfs_commit_inode() is interrupted, it
returns a non-negative value, which causes nfs_wb_page() to think
that the operation was successful causing it to busy-loop rather
than exiting.
It also causes nfs_file_fsync() to incorrectly report the file as
being successfully committed to disk.

This patch fixes both problems by ensuring that we return an error
if the attempts to wait fail.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

13 years agoperf: Fix tear-down of inherited group events
Peter Zijlstra [Tue, 15 Mar 2011 13:37:10 +0000 (14:37 +0100)]
perf: Fix tear-down of inherited group events

commit 38b435b16c36b0d863efcf3f07b34a6fac9873fd upstream.

When destroying inherited events, we need to destroy groups too,
otherwise the event iteration in perf_event_exit_task_context() will
miss group siblings and we leak events with all the consequences.

Reported-and-tested-by: Vince Weaver <vweaver1@eecs.utk.edu>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1300196470.2203.61.camel@twins>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>