Permalink
Commits on Feb 4, 2010
  1. Merge branch 'pf-configs'

    Oleksandr Natalenko committed Feb 4, 2010
  2. pf-configs: update config for Dell Inspiron 1525

    Oleksandr Natalenko committed Feb 4, 2010
  3. Merge branch 'pf-version'

    Oleksandr Natalenko committed Feb 4, 2010
  4. pf-version: bump to v2.6.32-pf14

    Oleksandr Natalenko committed Feb 4, 2010
  5. Merge remote branch 'tuxonice'

    Oleksandr Natalenko committed Feb 4, 2010
Commits on Feb 2, 2010
  1. Bump version to 3.0.99.47.

    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Feb 2, 2010
  2. Re-add LZF support.

    While we're seeking to find the cause of problems with reading the image,
    reinstate the LZF cryptoapi support. It has a long history of reliability
    that should help eliminate one potential cause of issues (though it shouldn't
    be necessary when the previous patch - implementing support for checking
    decompressability - is used).
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Feb 2, 2010
  3. Add support for checking compression.

    Add a new /sys/power/tuxonice/compress/check sysfs entry which defaults to
    zero. If set non-zero, when writing an image, we check that the results of
    compression, verifying that the result can be decompressed back to the
    original contents. This will be used to help diagnose the cause of the
    problems with reading an image that are sometimes seen.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Feb 2, 2010
  4. Fix result string typo.

    Fix a typo in the result strings, displayed in the debugging info and
    logged using printk. Thanks to Nix for alerting me to it.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Feb 2, 2010
Commits on Feb 1, 2010
  1. Allow the user to limit the number of threads used.

    Add support for limiting the number of threads we use for
    doing I/O, providing a new sysfs entry that can be used
    to set the value. The default of zero provides the old
    behaviour - one per cpu.
    
    The impetus for this patch is the report from Nix that
    on his eight core box, throughput was much lower than
    expected. Let's see if it's due to lock contention.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Feb 1, 2010
  2. Fix missing write_image_header label.

    I somehow managed to muck up the write_image_header patch, omitting
    the label write_image_header_abort_no_cleanup. Readd it.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Feb 1, 2010
  3. Add image header versioning.

    Add a version number to the image header, so that we can avoid future
    issues with backwards compatability. Since previous versions haven't
    had a version number and the bytes we're going to use haven't been
    overwritten, use an additional copy of the binary signature so we can
    be sure we really have a version number and not whatever might have
    been in those bytes before the page was used for the image signature.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Feb 1, 2010
  4. Improve error handling in write_image_header.

    Improve the error handling in write_image_header. We weren't checking the
    return codes of a couple of routines, and an error in writing the header
    structure would result in a failure to free that page that's allocated for
    it.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Feb 1, 2010
Commits on Jan 31, 2010
  1. Init & cleanup compression per pageset.

    Rather than initialising compression once at the start of
    hibernating or resuming, and cleaning up once, do this at the
    beginning and end of reading or writing a pageset. The main
    reason for doing things less efficiently is because if the
    compressor being used isn't the default (lzo), then prior to this
    change, we try and init lzo anyway before reading the name of
    the compressor we'll actually use from the image header. If lzo
    isn't compiled in, we'll error out. With this change, we'll never
    even look at the lzo compression algo.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Jan 31, 2010
  2. UUID debugging kernel parameter.

    Rather than making users modify the #if at line 5, implement a new boot time
    parameter "uuid_debug=1" that can be used to enable the debugging code. I could
    also provide a means to toggle the setting post-boot, but will worry about that
    later.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Jan 31, 2010
Commits on Jan 30, 2010
  1. Merge remote branch 'tuxonice'

    Oleksandr Natalenko committed Jan 30, 2010
  2. Handle old TuxOnIce signatures.

    Current TuxOnIce versions haven't been recognising signatures from 3.0.1.
    Fix that, adding the ability to recognise old signatures and give the user
    the option of rebooting into a kernel with 3.0.1 or removing the image.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Jan 30, 2010
Commits on Jan 29, 2010
  1. Fix memory allocation error handling in UUID code.

    The UUID code was not checking for allocation failures. Add this
    support, returning -ENOMEM and printing an error message in such
    cases.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Jan 29, 2010
  2. fix merge conflict

    Oleksandr Natalenko committed Jan 29, 2010
  3. pf-version: bump to v2.6.32-pf13

    Oleksandr Natalenko committed Jan 29, 2010
  4. Merge remote branch 'tuxonice'

    Oleksandr Natalenko committed Jan 29, 2010
  5. fix merge conflict

    Oleksandr Natalenko committed Jan 29, 2010
  6. Fix gfp_flags used in getting ps1 load addresses.

    There was one problem with the previous patch to the code for
    getting pageset 1 load addresses: we were always using __GFP_HIGHMEM
    in allocations, with the result being that we were allocating too
    many highmem pages on systems with them. Fix this by removing the
    highmem flag for the last $low_needed allocations, thereby ensuring
    that those ones get lowmem pages.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    Nigel Cunningham committed Jan 29, 2010
  7. Merge branch 'origin' into combined

    Nigel Cunningham committed Jan 29, 2010
Commits on Jan 28, 2010
  1. Linux 2.6.32.7

    gregkh committed Jan 28, 2010
  2. x86, msr/cpuid: Pass the number of minors when unregistering MSR and …

    …CPUID drivers.
    
    commit da48247 upstream.
    
    Pass the number of minors when unregistering MSR and CPUID drivers.
    
    Reported-by: Dean Nelson <dnelson@redhat.com>
    Signed-off-by: Dean Nelson <dnelson@redhat.com>
    LKML-Reference: <20100127023722.GA22305@sgi.com>
    Signed-off-by: Russ Anderson <rja@sgi.com>
    Signed-off-by: H. Peter Anvin <hpa@zytor.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    Russ Anderson committed with gregkh Jan 27, 2010
  3. fnctl: f_modown should call write_lock_irqsave/restore

    commit b04da8b upstream.
    
    Commit 7036251 exposed that f_modown()
    should call write_lock_irqsave instead of just write_lock_irq so that
    because a caller could have a spinlock held and it would not be good to
    renable interrupts.
    
    Cc: Eric W. Biederman <ebiederm@xmission.com>
    Cc: Al Viro <viro@ZenIV.linux.org.uk>
    Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
    Cc: Tavis Ormandy <taviso@google.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
    gregkh committed Jan 26, 2010
  4. iwlwifi: Fix throughput stall issue in HT mode for 5000

    commit 1152dcc upstream
    
    Similar to 6000 and 1000 series, RTS/CTS is the recommended protection
    mechanism for 5000 series in HT mode based on the HW design.
    
    Using RTS/CTS will better protect the inner exchange from interference,
    especially in highly-congested environment, it also prevent uCode encounter
    TX FIFO underrun and other HT mode related performance issues.
    
    Signed-off-by: Wey-Yi Guy <wey-yi.w.guy@intel.com>
    Signed-off-by: Reinette Chatre <reinette.chatre@intel.com>
    Signed-off-by: John W. Linville <linville@tuxdriver.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    Wey-Yi Guy committed with gregkh Jan 26, 2010
  5. ACPI: enable C2 and Turbo-mode on Nehalem notebooks on A/C

    upstream in 2.6.33-rc:  5d76b6f
    
    Refreshed here for 2.6.32.y, applies w/ offset back to 2.6.29.y.
    
    Linux has always ignored ACPI BIOS C2 with exit latency > 100 usec,
    and the ACPI spec is clear that is correct FADT-supplied C2.
    
    However, the ACPI spec explicitly states that _CST-supplied C-states
    have no latency limits.
    
    So move the 100usec C2 test out of the code shared
    by FADT and _CST code-paths, and into the FADT-specific path.
    
    This bug has not been visible until Nehalem, which advertises
    a CPU-C2 worst case exit latency on servers of 205usec.
    That (incorrect) figure is being used by BIOS writers
    on mobile Nehalem systems for the AC configuration.
    Thus, Linux ignores C2 leaving just C1, which is
    saves less power, and also impacts performance
    by preventing the use of turbo mode.
    
    http://bugzilla.kernel.org/show_bug.cgi?id=15064
    
    Tested-by: Alex Chiang <achiang@hp.com>
    Signed-off-by: Len Brown <len.brown@intel.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    lenb committed with gregkh Jan 26, 2010
  6. x86: Reenable TSC sync check at boot, even with NONSTOP_TSC

    commit 6c56cce upstream.
    
    Commit 83ce400 did the following change
    If the TSC is constant and non-stop, also set it reliable.
    
    But, there seems to be few systems that will end up with TSC warp across
    sockets, depending on how the cpus come out of reset. Skipping TSC sync
    test on such systems may result in time inconsistency later.
    
    So, reenable TSC sync test even on constant and non-stop TSC systems.
    Set, sched_clock_stable to 1 by default and reset it in
    mark_tsc_unstable, if TSC sync fails.
    
    This change still gives perf benefit mentioned in 83ce400 for systems
    where TSC is reliable.
    
    Signed-off-by: Venkatesh Pallipadi <venkatesh.pallipadi@intel.com>
    Acked-by: Suresh Siddha <suresh.b.siddha@intel.com>
    LKML-Reference: <20091217202702.GA18015@linux-os.sc.intel.com>
    Signed-off-by: H. Peter Anvin <hpa@zytor.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    Pallipadi, Venkatesh committed with gregkh Dec 17, 2009
  7. IPoIB: Clear ipoib_neigh.dgid in ipoib_neigh_alloc()

    commit 0cd4d0f upstream.
    
    IPoIB can miss a change in destination GID under some conditions.  The
    problem is caused when ipoib_neigh->dgid contains a stale address.
    The fix is to set ipoib_neigh->dgid to zero in ipoib_neigh_alloc().
    
    This can happen when a system using bonding on its IPoIB interfaces
    has switched its active interface from interface A to B and back to A.
    The system that fails over will not correctly processes the 2nd
    address change, as described below.
    
    When an address has changed neighbor->ha is updated with the new
    address.  Each neighbor has an associated ipoib_neigh.
    ipoib_neigh->dgid also holds a copy of the remote node's hardware
    address.  When an address changes neighbor->ha is updated by the
    network layer (arp code) with the new address.  IPoIB detects this
    change in ipoib_start_xmit() by comparing neighbor->ha with
    ipoib_neigh->dgid.  The bug is that ipoib_neigh->dgid may already
    contain the new address (A) thus the change from B to A is missed by
    ipoib.  Here is the sequence of events:
    
        ipoib_neigh->dgid = A  and  neighbor->ha = A
    
    The address is switched to B (the first switch)
    
        neighbor->ha = B
    
    The change is seen in ipoib_start_xmit() -- neighbor->ha !=
    ipoib_neigh->dgid so ipoib_neigh is released, and a new one is
    allocated.
    
    The allocator may return the same chunk of memory that was just
    released, therefore ipoib_neigh->dgid still contains A at this point.
    
    ipoib_neigh->dgid should be updated in neigh_add_path(), but if the
    following conditions are true dgid is not updated:
    
            1) __path_find() returns a path
            2) path->ah is NULL
    
    The remote system now switches from address B to A, neighbor->ha is
    updated to A.
    
    Now we have again : ipoib_neigh->dgid = A  and  neighbor->ha = A
    
    Since the addresses are the same ipoib won't process the change in
    address.  Fix this by zeroing out the dgid field when allocating a new
    struct ipoib_neigh.
    
    Signed-off-by: David Wilder <dwilder@us.ibm.com>
    Signed-off-by: Roland Dreier <rolandd@cisco.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    DaveWilder committed with gregkh Dec 9, 2009
  8. KVM: only clear irq_source_id if irqchip is present

    commit e50212b upstream.
    
    Otherwise kvm might attempt to dereference a NULL pointer.
    
    Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
    Signed-off-by: Avi Kivity <avi@redhat.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    Marcelo Tosatti committed with gregkh Oct 29, 2009
  9. KVM: fix lock imbalance in kvm_*_irq_source_id()

    commit 0c6ddce upstream.
    
    Stanse found 2 lock imbalances in kvm_request_irq_source_id and
    kvm_free_irq_source_id. They omit to unlock kvm->irq_lock on fail paths.
    
    Fix that by adding unlock labels at the end of the functions and jump
    there from the fail paths.
    
    Signed-off-by: Jiri Slaby <jirislaby@gmail.com>
    Cc: Marcelo Tosatti <mtosatti@redhat.com>
    Signed-off-by: Avi Kivity <avi@redhat.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    jirislaby committed with gregkh Sep 25, 2009
  10. KVM: x86: Fix leak of free lapic date in kvm_arch_vcpu_init()

    commit 443c39b upstream.
    
    In function kvm_arch_vcpu_init(), if the memory malloc for
    vcpu->arch.mce_banks is fail, it does not free the memory
    of lapic date. This patch fixed it.
    
    Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
    Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    Wei Yongjun committed with gregkh Jan 22, 2010
  11. KVM: x86: Fix probable memory leak of vcpu->arch.mce_banks

    commit 36cb93f upstream.
    
    vcpu->arch.mce_banks is malloc in kvm_arch_vcpu_init(), but
    never free in any place, this may cause memory leak. So this
    patch fixed to free it in kvm_arch_vcpu_uninit().
    
    Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
    Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
    Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
    Wei Yongjun committed with gregkh Jan 22, 2010