1. 09 Jan, 2009 19 commits
  2. 08 Jan, 2009 21 commits
    • Linus Torvalds's avatar
      Merge branch 'docs-next' of git://git.lwn.net/linux-2.6 · 1df2d017
      Linus Torvalds authored
      * 'docs-next' of git://git.lwn.net/linux-2.6:
        Fix a typo in the development process document.
        Document handling of bad memory
        Document RCU and unloadable modules
      1df2d017
    • David John's avatar
      RTC: Remove the BKL. · 048cd588
      David John authored
      Remove calls to the BKL since concurrent access is protected
      by the spin lock rtc_lock.
      Signed-off-by: default avatarDavid John <davidjon@xenontk.org>
      Cc: Ingo Molnar <mingo@elte.hu>
      Signed-off-by: default avatarJonathan Corbet <corbet@lwn.net>
      048cd588
    • Jonathan Corbet's avatar
    • Linus Torvalds's avatar
      Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/lrg/voltage-2.6 · 022992ee
      Linus Torvalds authored
      * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/lrg/voltage-2.6:
        regulator: fix kernel-doc warnings
        regulator: catch some registration errors
        regulator: Add basic DocBook manual
        regulator: Fix some kerneldoc rendering issues
        regulator: Add missing kerneldoc
        regulator: Clean up kerneldoc warnings
        regulator: Remove extraneous kerneldoc annotations
        regulator: init/link earlier
        regulator: move set_machine_constraints after regulator device initialization
        regulator: da903x: make da903x_is_enabled return 0 or 1
        regulator: da903x: add '\n' to error messages
        regulator: sysfs attribute reduction (v2)
        regulator: code shrink (v2)
        regulator: improved mode error checks
        regulator: enable/disable refcounting
        regulator: struct device - replace bus_id with dev_name(), dev_set_name()
      022992ee
    • Linus Torvalds's avatar
      Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6 · 5fbbf5f6
      Linus Torvalds authored
      * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (84 commits)
        wimax: fix kernel-doc for debufs_dentry member of struct wimax_dev
        net: convert pegasus driver to net_device_ops
        bnx2x: Prevent eeprom set when driver is down
        net: switch kaweth driver to netdevops
        pcnet32: round off carrier watch timer
        i2400m/usb: wrap USB power saving in #ifdef CONFIG_PM
        wimax: testing for rfkill support should also test for CONFIG_RFKILL_MODULE
        wimax: fix kconfig interactions with rfkill and input layers
        wimax: fix '#ifndef CONFIG_BUG' layout to avoid warning
        r6040: bump release number to 0.20
        r6040: warn about MAC address being unset
        r6040: check PHY status when bringing interface up
        r6040: make printks consistent with DRV_NAME
        gianfar: Fixup use of BUS_ID_SIZE
        mlx4_en: Returning real Max in get_ringparam
        mlx4_en: Consider inline packets on completion
        netdev: bfin_mac: enable bfin_mac net dev driver for BF51x
        qeth: convert to net_device_ops
        vlan: add neigh_setup
        dm9601: warn on invalid mac address
        ...
      5fbbf5f6
    • Linus Torvalds's avatar
      Merge master.kernel.org:/home/rmk/linux-2.6-arm · ce279e6e
      Linus Torvalds authored
      * master.kernel.org:/home/rmk/linux-2.6-arm: (23 commits)
        [ARM] fix pxa930_trkball build errors
        [ARM] fix netx
        [ARM] fix pnx4008
        [ARM] fix pxa
        [ARM] remove missed CLPS7500 defconfig
        [ARM] clps711x: fix warning in edb7211-mm.c
        [ARM] clps711x: fix warning in fortunet
        [ARM] fix AT91, davinci, h720x, ks8695, msm, mx2, mx3, netx, omap1, omap2, pxa, s3c
        [ARM] Fix realview build
        [ARM] 5357/1: Kirkwood: add missing ge01 tclk initialization
        [ARM] 5358/1: AT2440EVB: Use new include path of mci.h
        [ARM] 5361/1: mv78xx0: fix compilation error
        [ARM] 5360/1: Orion: fix compilation error
        [ARM] 5359/1: Kirkwood: fix compilation error
        [ARM] S3C64XX: Fix EINT group macro definition
        [ARM] Ensure CONFIG_SERIAL_SAMSUNG_UARTS is always set.
        [ARM] S3C24XX: Add gpio_to_irq implementation
        [ARM] S3C24XX: Add gpio_to_irq() facility
        [ARM] footbridge: dc21285.c warning fixes
        [ARM] footbridge: add isa_init_irq() to common header
        ...
      ce279e6e
    • Linus Torvalds's avatar
      Merge branch 'for-linus' of git://neil.brown.name/md · 894bcdfb
      Linus Torvalds authored
      * 'for-linus' of git://neil.brown.name/md:
        md: don't retry recovery of raid1 that fails due to error on source drive.
        md: Allow md devices to be created by name.
        md: make devices disappear when they are no longer needed.
        md: centralise all freeing of an 'mddev' in 'md_free'
        md: move allocation of ->queue from mddev_find to md_probe
        md: need another print_sb for mdp_superblock_1
        md: use list_for_each_entry macro directly
        md: raid0: make hash_spacing and preshift sector-based.
        md: raid0: Represent the size of strip zones in sectors.
        md: raid0 create_strip_zones(): Add KERN_INFO/KERN_ERR to printk's.
        md: raid0 create_strip_zones(): Make two local variables sector-based.
        md: raid0: Represent zone->zone_offset in sectors.
        md: raid0: Represent device offset in sectors.
        md: raid0_make_request(): Replace local variable block by sector.
        md: raid0_make_request(): Remove local variable chunk_size.
        md: raid0_make_request(): Replace chunksize_bits by chunksect_bits.
        md: use sysfs_notify_dirent to notify changes to md/sync_action.
        md: fix bitmap-on-external-file bug.
      894bcdfb
    • Russell King's avatar
      Merge branch 'devel' · 97b5fe8e
      Russell King authored
      97b5fe8e
    • Linus Torvalds's avatar
      Merge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev · a419df8a
      Linus Torvalds authored
      * 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
        libata: clean up the SFF code for coding style
        libata: Add 32bit PIO support
        [libata] ahci: Withdraw IGN_SERR_INTERNAL for SB800 SATA
        pata_hpt366: reimplement mode programming
        [libata] pata_hpt3x3: correct _freeze() function declaration
        libata: Add special ata_pio_need_iordy() handling for Compact Flash.
        pata_platform: __pata_platform_remove() shouldn't be in discard section
        sata_sil24: remove unused sil24_port_multiplier
        [libata] ahci: Add SATA GEN3 related messages
        ata_piix: save, use saved and restore IOCFG
        pata_ali: Fix and workaround for FIFO DMA bug
        pata_ali: force initialise a few bits
        pata_hpt3x3: Workarounds for chipset
      a419df8a
    • Russell King's avatar
    • Alan Cox's avatar
      libata: clean up the SFF code for coding style · 0fe40ff8
      Alan Cox authored
      Signed-off-by: default avatarAlan Cox <alan@redhat.com>
      Signed-off-by: default avatarJeff Garzik <jgarzik@redhat.com>
      0fe40ff8
    • Alan Cox's avatar
      libata: Add 32bit PIO support · 871af121
      Alan Cox authored
      This matters for some controllers and in one or two cases almost doubles
      PIO performance. Add a bmdma32 operations set we can inherit and activate
      it for some controllers
      Signed-off-by: default avatarAlan Cox <alan@redhat.com>
      Signed-off-by: default avatarJeff Garzik <jgarzik@redhat.com>
      871af121
    • Shane Huang's avatar
      [libata] ahci: Withdraw IGN_SERR_INTERNAL for SB800 SATA · e427fe04
      Shane Huang authored
      There is an issue in ATI SB600/SB700 SATA that PxSERR.E should not be
      set on some conditions, which will lead to many SATA ODD error messages.
      commit 55a61604 is the workaround.
      Since SB800 fixed this HW issue, IGN_SERR_INTERNAL should be withdrawn
      for SB800.
      Signed-off-by: default avatarShane Huang <shane.huang@amd.com>
      Signed-off-by: default avatarJeff Garzik <jgarzik@redhat.com>
      e427fe04
    • NeilBrown's avatar
      md: don't retry recovery of raid1 that fails due to error on source drive. · 4044ba58
      NeilBrown authored
      If a raid1 has only one working drive and it has a sector which
      gives an error on read, then an attempt to recover onto a spare will
      fail, but as the single remaining drive is not removed from the
      array, the recovery will be immediately re-attempted, resulting
      in an infinite recovery loop.
      
      So detect this situation and don't retry recovery once an error
      on the lone remaining drive is detected.
      
      Allow recovery to be retried once every time a spare is added
      in case the problem wasn't actually a media error.
      Signed-off-by: default avatarNeilBrown <neilb@suse.de>
      4044ba58
    • NeilBrown's avatar
      md: Allow md devices to be created by name. · efeb53c0
      NeilBrown authored
      Using sequential numbers to identify md devices is somewhat artificial.
      Using names can be a lot more user-friendly.
      
      Also, creating md devices by opening the device special file is a bit
      awkward.
      
      So this patch provides a new option for creating and naming devices.
      
      Writing a name such as "md_home" to
          /sys/modules/md_mod/parameters/new_array
      will cause an array with that name to be created.  It will appear in
      /sys/block/ /proc/partitions and /proc/mdstat as 'md_home'.
      It will have an arbitrary minor number allocated.
      
      md devices that a created by an open are destroyed on the last
      close when the device is inactive.
      For named md devices, they will not be destroyed until the array
      is explicitly stopped, either with the STOP_ARRAY ioctl or by
      writing 'clear' to /sys/block/md_XXXX/md/array_state.
      
      The name of the array must start 'md_' to avoid conflict with
      other devices.
      Signed-off-by: default avatarNeilBrown <neilb@suse.de>
      efeb53c0
    • NeilBrown's avatar
      md: make devices disappear when they are no longer needed. · d3374825
      NeilBrown authored
      Currently md devices, once created, never disappear until the module
      is unloaded.  This is essentially because the gendisk holds a
      reference to the mddev, and the mddev holds a reference to the
      gendisk, this a circular reference.
      
      If we drop the reference from mddev to gendisk, then we need to ensure
      that the mddev is destroyed when the gendisk is destroyed.  However it
      is not possible to hook into the gendisk destruction process to enable
      this.
      
      So we drop the reference from the gendisk to the mddev and destroy the
      gendisk when the mddev gets destroyed.  However this has a
      complication.
      Between the call
         __blkdev_get->get_gendisk->kobj_lookup->md_probe
      and the call
         __blkdev_get->md_open
      
      there is no obvious way to hold a reference on the mddev any more, so
      unless something is done, it will disappear and gendisk will be
      destroyed prematurely.
      
      Also, once we decide to destroy the mddev, there will be an unlockable
      moment before the gendisk is unlinked (blk_unregister_region) during
      which a new reference to the gendisk can be created.  We need to
      ensure that this reference can not be used.  i.e. the ->open must
      fail.
      
      So:
       1/  in md_probe we set a flag in the mddev (hold_active) which
           indicates that the array should be treated as active, even
           though there are no references, and no appearance of activity.
           This is cleared by md_release when the device is closed if it
           is no longer needed.
           This ensures that the gendisk will survive between md_probe and
           md_open.
      
       2/  In md_open we check if the mddev we expect to open matches
           the gendisk that we did open.
           If there is a mismatch we return -ERESTARTSYS and modify
           __blkdev_get to retry from the top in that case.
           In the -ERESTARTSYS sys case we make sure to wait until
           the old gendisk (that we succeeded in opening) is really gone so
           we loop at most once.
      
      Some udev configurations will always open an md device when it first
      appears.   If we allow an md device that was just created by an open
      to disappear on an immediate close, then this can race with such udev
      configurations and result in an infinite loop the device being opened
      and closed, then re-open due to the 'ADD' even from the first open,
      and then close and so on.
      So we make sure an md device, once created by an open, remains active
      at least until some md 'ioctl' has been made on it.  This means that
      all normal usage of md devices will allow them to disappear promptly
      when not needed, but the worst that an incorrect usage will do it
      cause an inactive md device to be left in existence (it can easily be
      removed).
      
      As an array can be stopped by writing to a sysfs attribute
        echo clear > /sys/block/mdXXX/md/array_state
      we need to use scheduled work for deleting the gendisk and other
      kobjects.  This allows us to wait for any pending gendisk deletion to
      complete by simply calling flush_scheduled_work().
      Signed-off-by: default avatarNeilBrown <neilb@suse.de>
      d3374825
    • NeilBrown's avatar
      md: centralise all freeing of an 'mddev' in 'md_free' · a21d1504
      NeilBrown authored
      md_free is the .release handler for the md kobj_type.
      So it makes sense to release all the objects referenced by
      the mddev in there, rather than just prior to calling kobject_put
      for what we think is the last time.
      Signed-off-by: default avatarNeilBrown <neilb@suse.de>
      a21d1504
    • NeilBrown's avatar
      md: move allocation of ->queue from mddev_find to md_probe · 8b765398
      NeilBrown authored
      It is more balanced to just do simple initialisation in mddev_find,
      which allocates and links a new md device, and leave all the
      more sophisticated allocation to md_probe (which calls mddev_find).
      md_probe already allocated the gendisk.  It should allocate the
      queue too.
      Signed-off-by: default avatarNeilBrown <neilb@suse.de>
      8b765398
    • Cheng Renquan's avatar
      md: need another print_sb for mdp_superblock_1 · cd2ac932
      Cheng Renquan authored
      md_print_devices is called in two code path: MD_BUG(...), and md_ioctl
      with PRINT_RAID_DEBUG.  it will dump out all in use md devices
      information;
      
      However, it wrongly processed two types of superblock in one:
      
      The header file <linux/raid/md_p.h> has defined two types of superblock,
      struct mdp_superblock_s (typedefed with mdp_super_t) according to md with
      metadata 0.90, and struct mdp_superblock_1 according to md with metadata
      1.0 and later,
      
      These two types of superblock are very different,
      
      The md_print_devices code processed them both in mdp_super_t, that would
      lead to wrong informaton dump like:
      
      	[ 6742.345877]
      	[ 6742.345887] md:	**********************************
      	[ 6742.345890] md:	* <COMPLETE RAID STATE PRINTOUT> *
      	[ 6742.345892] md:	**********************************
      	[ 6742.345896] md1: <ram7><ram6><ram5><ram4>
      	[ 6742.345907] md: rdev ram7, SZ:00065472 F:0 S:1 DN:3
      	[ 6742.345909] md: rdev superblock:
      	[ 6742.345914] md:  SB: (V:0.90.0) ID:<42ef13c7.598c059a.5f9f1645.801e9ee6> CT:4919856d
      	[ 6742.345918] md:     L5 S00065472 ND:4 RD:4 md1 LO:2 CS:65536
      	[ 6742.345922] md:     UT:4919856d ST:1 AD:4 WD:4 FD:0 SD:0 CSUM:b7992907 E:00000001
      	[ 6742.345924]      D  0:  DISK<N:0,(1,8),R:0,S:6>
      	[ 6742.345930]      D  1:  DISK<N:1,(1,10),R:1,S:6>
      	[ 6742.345933]      D  2:  DISK<N:2,(1,12),R:2,S:6>
      	[ 6742.345937]      D  3:  DISK<N:3,(1,14),R:3,S:6>
      	[ 6742.345942] md:     THIS:  DISK<N:3,(1,14),R:3,S:6>
      	...
      	[ 6742.346058] md0: <ram3><ram2><ram1><ram0>
      	[ 6742.346067] md: rdev ram3, SZ:00065472 F:0 S:1 DN:3
      	[ 6742.346070] md: rdev superblock:
      	[ 6742.346073] md:  SB: (V:1.0.0) ID:<369aad81.00000000.00000000.00000000> CT:9a322a9c
      	[ 6742.346077] md:     L-1507699579 S976570180 ND:48 RD:0 md0 LO:65536 CS:196610
      	[ 6742.346081] md:     UT:00000018 ST:0 AD:131048 WD:0 FD:8 SD:0 CSUM:00000000 E:00000000
      	[ 6742.346084]      D  0:  DISK<N:-1,(-1,-1),R:-1,S:-1>
      	[ 6742.346089]      D  1:  DISK<N:-1,(-1,-1),R:-1,S:-1>
      	[ 6742.346092]      D  2:  DISK<N:-1,(-1,-1),R:-1,S:-1>
      	[ 6742.346096]      D  3:  DISK<N:-1,(-1,-1),R:-1,S:-1>
      	[ 6742.346102] md:     THIS:  DISK<N:0,(0,0),R:0,S:0>
      	...
      	[ 6742.346219] md:	**********************************
      	[ 6742.346221]
      
      Here md1 is metadata 0.90.0, and md0 is metadata 1.2
      
      After some more code to distinguish these two types of superblock, in this patch,
      
      it will generate dump information like:
      
      	[ 7906.755790]
      	[ 7906.755799] md:	**********************************
      	[ 7906.755802] md:	* <COMPLETE RAID STATE PRINTOUT> *
      	[ 7906.755804] md:	**********************************
      	[ 7906.755808] md1: <ram7><ram6><ram5><ram4>
      	[ 7906.755819] md: rdev ram7, SZ:00065472 F:0 S:1 DN:3
      	[ 7906.755821] md: rdev superblock (MJ:0):
      	[ 7906.755826] md:  SB: (V:0.90.0) ID:<3fca7a0d.a612bfed.5f9f1645.801e9ee6> CT:491989f3
      	[ 7906.755830] md:     L5 S00065472 ND:4 RD:4 md1 LO:2 CS:65536
      	[ 7906.755834] md:     UT:491989f3 ST:1 AD:4 WD:4 FD:0 SD:0 CSUM:00fb52ad E:00000001
      	[ 7906.755836]      D  0:  DISK<N:0,(1,8),R:0,S:6>
      	[ 7906.755842]      D  1:  DISK<N:1,(1,10),R:1,S:6>
      	[ 7906.755845]      D  2:  DISK<N:2,(1,12),R:2,S:6>
      	[ 7906.755849]      D  3:  DISK<N:3,(1,14),R:3,S:6>
      	[ 7906.755855] md:     THIS:  DISK<N:3,(1,14),R:3,S:6>
      	...
      	[ 7906.755972] md0: <ram3><ram2><ram1><ram0>
      	[ 7906.755981] md: rdev ram3, SZ:00065472 F:0 S:1 DN:3
      	[ 7906.755984] md: rdev superblock (MJ:1):
      	[ 7906.755989] md:  SB: (V:1) (F:0) Array-ID:<5fbcf158:55aa:5fbe:9a79:1e939880dcbd>
      	[ 7906.755990] md:    Name: "DG5:0" CT:1226410480
      	[ 7906.755998] md:       L5 SZ130944 RD:4 LO:2 CS:128 DO:24 DS:131048 SO:8 RO:0
      	[ 7906.755999] md:     Dev:00000003 UUID: 9194d744:87f7:a448:85f2:7497b84ce30a
      	[ 7906.756001] md:       (F:0) UT:1226410480 Events:0 ResyncOffset:-1 CSUM:0dbcd829
      	[ 7906.756003] md:         (MaxDev:384)
      	...
      	[ 7906.756113] md:	**********************************
      	[ 7906.756116]
      
      this md0 (metadata 1.2) information dumping is exactly according to struct
      mdp_superblock_1.
      Signed-off-by: default avatarCheng Renquan <crquan@gmail.com>
      Cc: Neil Brown <neilb@suse.de>
      Cc: Dan Williams <dan.j.williams@gmail.com>
      Signed-off-by: default avatarAndrew Morton <akpm@linux-foundation.org>
      Signed-off-by: default avatarNeilBrown <neilb@suse.de>
      cd2ac932
    • Cheng Renquan's avatar
      md: use list_for_each_entry macro directly · 159ec1fc
      Cheng Renquan authored
      The rdev_for_each macro defined in <linux/raid/md_k.h> is identical to
      list_for_each_entry_safe, from <linux/list.h>, it should be defined to
      use list_for_each_entry_safe, instead of reinventing the wheel.
      
      But some calls to each_entry_safe don't really need a safe version,
      just a direct list_for_each_entry is enough, this could save a temp
      variable (tmp) in every function that used rdev_for_each.
      
      In this patch, most rdev_for_each loops are replaced by list_for_each_entry,
      totally save many tmp vars; and only in the other situations that will call
      list_del to delete an entry, the safe version is used.
      Signed-off-by: default avatarCheng Renquan <crquan@gmail.com>
      Signed-off-by: default avatarNeilBrown <neilb@suse.de>
      159ec1fc
    • Andre Noll's avatar
      md: raid0: make hash_spacing and preshift sector-based. · ccacc7d2
      Andre Noll authored
      This patch renames the hash_spacing and preshift members of struct
      raid0_private_data to spacing and sector_shift respectively and
      changes the semantics as follows:
      
      We always have spacing = 2 * hash_spacing. In case
      sizeof(sector_t) > sizeof(u32) we also have sector_shift = preshift + 1
      while sector_shift = preshift = 0 otherwise.
      
      Note that the values of nb_zone and zone are unaffected by these changes
      because in the sector_div() preceeding the assignement of these two
      variables both arguments double.
      Signed-off-by: default avatarAndre Noll <maan@systemlinux.org>
      Signed-off-by: default avatarNeilBrown <neilb@suse.de>
      ccacc7d2