Skip to content
Commits on Dec 14, 2009
  1. Merge branch 'pf-version'

    Oleksandr Natalenko committed Dec 14, 2009
  2. Merge branch 'pf-configs'

    Oleksandr Natalenko committed Dec 14, 2009
  3. tuxonice: fix merge conflict

    Oleksandr Natalenko committed Dec 14, 2009
  4. Merge remote branch 'zen-sched'

    Oleksandr Natalenko committed Dec 14, 2009
  5. pf-version: bump to v2.6.32-pf2

    Oleksandr Natalenko committed Dec 14, 2009
  6. pf-configs: useless config update

    Oleksandr Natalenko committed Dec 14, 2009
  7. @NigelCunningham

    Fix & at start of comment line.

    Fix an ampersand, put in a multiline comment in place of an asterick.
    Thanks to Whoopie for noticing and alerting me.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 14, 2009
  8. @NigelCunningham

    Bump version to 3.0.99.42

    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 14, 2009
  9. @NigelCunningham

    Add support for scanning for an image.

    Add support for searching for an image if resume= is blank. This
    initial version extends the uuid code a little to enable finding
    partitions of a particular type. We use this code for swap,
    checking each swap partition found for an image header. The first
    partition found is remembered, and becomes a default resume=
    value if not image is found. If an image is found, that partition
    is used for resume= instead.
    
    The end result is that users should now be able to hibernate and
    resume without ever having to worry about setting resume=, if
    they have at least one swap partition. Even if that swap partition
    isn't used for hibernating, the cycle will still success (as it
    did before), because resume= just gives the location of the start
    of the image header, which may be on a completely different device
    again.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 14, 2009
Commits on Dec 12, 2009
  1. @NigelCunningham

    Bump version to 3.0.99.41

    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 13, 2009
  2. @NigelCunningham

    Fix bug 443 - generate_free_page_map bug

    Fix a bug in generate_free_page_map that caused the oops reported in
    bug 443.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 13, 2009
Commits on Dec 11, 2009
  1. @NigelCunningham

    Fix handling file allocator invalid targets.

    Fix modifying the file allocator target, so that setting it to an
    empty string or silly things like /dev/strerr doesn't cause an oops.
    If you're silly enough to try /dev/stderr, you'll hang indefinitely
    when you try to hibernate, but you won't cause an oops anymore by
    simply setting the target file to that value.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 12, 2009
  2. @NigelCunningham

    sizeof() is 'size_t' which may or may not be 'long'.

    Correct printk calls to format size_t values with '%zu'
    rather than '%ld'. Fixes the following three warnings
    when building an i686 tuxonice kernel:
    
    kernel/power/tuxonice_bio_core.c:559: warning: format '%ld' expects type 'long int', but argument 2 has type 'unsigned int'
    kernel/power/tuxonice_modules.c:81: warning: format '%ld' expects type 'long int', but argument 2 has type 'unsigned int'
    kernel/power/tuxonice_modules.c:84: warning: format '%ld' expects type 'long int', but argument 2 has type 'unsigned int'
    
    Signed-off-by: Mikael Pettersson <mikpe@it.uu.se>
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 12, 2009
Commits on Dec 10, 2009
  1. @NigelCunningham

    Bump version to 3.0.99.40.

    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 11, 2009
  2. @NigelCunningham

    Fix resume failure iff first extent one page only

    Fix a failure to resume that only occurs if the first extent on the
    first chain contains only one page. In this case, we need to read
    the second extent before we can set up the chain iteration state.
    If we don't do that, we get into the eof state (because we skip the
    first page, which was already read). This is almost certainly
    Kenneth's transient resume failure.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 10, 2009
  3. @NigelCunningham

    Bump the version number.

    I've gotten up to 3.0.99.39 - commit the new number to match before I
    push the changes to public repos.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 10, 2009
  4. @NigelCunningham

    Don't try to free storage when shouldn't.

    Prior to the atomic-restore (at resume time), we shouldn't try to invoke
    our allocators' storage freeing functions. Don't do it!
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 10, 2009
  5. @NigelCunningham

    Display storage usage in debugging info.

    Leverage the previous patch, adding support for recording in the image
    header how much storage from each device was actually used, then passing
    this info back to the restored kernel via the boot_kernel_data
    structure.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 10, 2009
  6. @NigelCunningham

    Add general post-atomic-restore module support.

    Replace the hardwared toi_ui_post_atomic_restore call with a
    more generic call to invoke post_atomic_restore functions that
    other modules might register as well. This will be used by
    the next patch.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 10, 2009
  7. @NigelCunningham

    Fix unaligned bios from uuid code.

    Correct a problem with submitting unaligned bios from the uuid
    code. by remembering that we're submitting page numbers, not
    sector numbers, and these need to be multiplied by 8.
    
    Additionally, fix up the read_bdev_page comment to match the
    contents.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 10, 2009
  8. @NigelCunningham

    Improve reliability of cancelling resuming.

    Improve the reliability of cancelling resuming by ensuring that
    we use the copy of the image header on disk post-atomic-restore
    (and not the version from when there was no image). In addition,
    make sure that we clean up the modules before trying to rewrite
    the signature, so as to avoid low-memory situations causing a
    failure to allocate a page for the signature and then an oops
    prior to rewriting the signature.
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 10, 2009
Commits on Dec 6, 2009
  1. Change this to match upstream

    Brandon Berhent committed Dec 6, 2009
Commits on Dec 5, 2009
  1. @NigelCunningham

    Remove a couple of debugging printks.

    Now that I've seen my poor, memory starved Karmic virtual machine
    successfully hibernate and resume while spewing the debugging
    printks, I'm pretty sure that Kenny's elusive issue isn't related
    to being low on memory (even more sure, knowing that his debug
    messages say he didn't hit this path). Remove these printks and
    keep hunting!
    
    Signed-off-by: Nigel Cunningham <nigel@tuxonice.net>
    NigelCunningham committed Dec 5, 2009
  2. Merge branch 'pf-configs'

    Oleksandr Natalenko committed Dec 5, 2009
  3. pf-configs: updated Dell config

    Oleksandr Natalenko committed Dec 5, 2009
  4. Merge branch 'pf-version'

    Oleksandr Natalenko committed Dec 5, 2009
  5. Merge branch 'pf-configs'

    Oleksandr Natalenko committed Dec 5, 2009
  6. Merge remote branch 'tuxonice'

    Oleksandr Natalenko committed Dec 5, 2009
  7. Merge remote branch 'bfq'

    Oleksandr Natalenko committed Dec 5, 2009
  8. fix bio_sync, 2.6.32 update

    Brandon Berhent committed Dec 5, 2009
  9. missing { in sched_bfs.c

    Brandon Berhent committed Dec 5, 2009
  10. remove extra } in sched_bfs.c

    Brandon Berhent committed Dec 5, 2009
  11. don't boost loop.c

    Brandon Berhent committed Dec 5, 2009
  12. bfq: few fixes

    Brandon Berhent committed Nov 6, 2009
  13. More correctly make bfq the default (oops) :P

    Brandon Berhent committed with Brandon Berhent Oct 10, 2009
Something went wrong with that request. Please try again.