7 years ago- patches.arch/s390-fate311926-ap-type10-toleration.patch:
John Jolly [Tue, 19 Jul 2011 04:58:58 +0000 (06:58 +0200)]
- patches.arch/s390-fate311926-ap-type10-toleration.patch:
  ap: Toleration for ap bus devices with device type 10

suse-commit: 283e18454af490739709a644ea17f936fb742eff

7 years agoMerge branch 'SLE11-SP2' of into SLE11-SP2
Jeff Mahoney [Mon, 18 Jul 2011 22:18:39 +0000 (18:18 -0400)]
Merge branch 'SLE11-SP2' of into SLE11-SP2

suse-commit: 999a2fca7d1a2a025a21079b7dab658eaec56ebf

7 years ago- patches.drivers/cxgb3-fix-device-opening-error-path: cxgb3:
Jeff Mahoney [Mon, 18 Jul 2011 22:18:22 +0000 (18:18 -0400)]
- patches.drivers/cxgb3-fix-device-opening-error-path: cxgb3:
  fix device opening error path (FATE#311470 bnc#694966).
- patches.drivers/cxgb3-fix-panic-in-free_tx_desc: cxgb3: Fix
  panic in free_tx_desc() (FATE#311470 bnc#694966).
- patches.drivers/cxgb3-skb_record_rx_queue-now-records-the-queue-index-relative-to-the-net_device:
  cxgb3: skb_record_rx_queue now records the queue index relative
  to the net_device (FATE#311470 bnc=694966).

suse-commit: 135104aeb4d02f2b03e2040b11a381e1129f64a5

7 years agoMerge branch 'SLE11-SP2' of ssh:// into SLE11-SP2
Jan Kara [Mon, 18 Jul 2011 17:08:22 +0000 (19:08 +0200)]
Merge branch 'SLE11-SP2' of ssh:// into SLE11-SP2

suse-commit: 5258fde086ce1cb23d2ff281fbfeee911db0313b

7 years ago- patches.fixes/vfs-fix_double_put_in_sync_filesystems.patch:
Jan Kara [Mon, 18 Jul 2011 17:07:58 +0000 (19:07 +0200)]
- patches.fixes/vfs-fix_double_put_in_sync_filesystems.patch:
  Fix double put in sync_filesystems() (bnc#706472).

suse-commit: e54f118aa31bd7a4fdefd28d0732bf9fd8f05c43

7 years ago- Fix config files after 5810156.
Michal Marek [Mon, 18 Jul 2011 16:31:40 +0000 (18:31 +0200)]
- Fix config files after 5810156.

suse-commit: 05b3b371ee84c14a3572b526c82f12294eb023d2

7 years ago- Enable the writeback patches again, the performance regression is
Michal Marek [Mon, 18 Jul 2011 16:24:42 +0000 (18:24 +0200)]
- Enable the writeback patches again, the performance regression is
  solved in SP2 (bnc#703786#c2)
- patches.suse/laptop-mode-Make-flushes-per-device.patch: Refresh.

suse-commit: 74ebdff4a68ebe02a55120aa2a0f61c804fbfadd

7 years agoMerge branch 'scripts' into SLE11-SP2
Michal Marek [Mon, 18 Jul 2011 15:49:18 +0000 (17:49 +0200)]
Merge branch 'scripts' into SLE11-SP2

suse-commit: 33735280717873a0e56ba3c956a33fc905b9daf0

7 years agoMerge branch 'SP2-packaging' into SLE11-SP2
Michal Marek [Mon, 18 Jul 2011 15:42:51 +0000 (17:42 +0200)]
Merge branch 'SP2-packaging' into SLE11-SP2

suse-commit: c17c2f0e39e7d6b8001c04faec4741dc225b8c09

7 years agoMerge branch 'SLE11-SP2' of into SLE11-SP2
Torsten Duwe [Mon, 18 Jul 2011 15:39:53 +0000 (17:39 +0200)]
Merge branch 'SLE11-SP2' of into SLE11-SP2

suse-commit: 18db9f7616649ed5976cffdf5a758d40b4540a4e

7 years ago- supported.conf: mark pseries_energy.ko as supported (by IBM)
Torsten Duwe [Mon, 18 Jul 2011 15:39:42 +0000 (17:39 +0200)]
- supported.conf: mark pseries_energy.ko as supported (by IBM)
  (bnc#706473 / FATE 311643)

suse-commit: 9889630d0af43ee0dbca3db93c036613f730f91f

7 years ago- Update config files.
Jeff Mahoney [Mon, 18 Jul 2011 15:38:09 +0000 (11:38 -0400)]
- Update config files.
- patches.drivers/bna-add-appropriate-linux-prefetch-h-include-for-prefetch-users:
  Add appropriate <linux/prefetch.h> include for prefetch users
  (FATE#311451 bnc#689230).
- patches.drivers/bna-brocade-10gb-ethernet-device-driver: bna:
  Brocade 10Gb Ethernet device driver (FATE#311451 bnc#689230).
- patches.drivers/bna-check-for-null-before-deref-in-bnad_cb_tx_cleanup:
  bna: Check for NULL before deref in bnad_cb_tx_cleanup
  (FATE#311451 bnc#689230).
- patches.drivers/bna-compatibility-fixes-for-sle11sp2-backport:
  bna: Compatibility fixes for SLE11SP2 backport (FATE#311451
- patches.drivers/bna-delete-get_flags-and-set_flags-ethtool-methods:
  bna: Delete get_flags and set_flags ethtool methods (FATE#311451
- patches.drivers/bna-drivers-net-return-operator-cleanup:
  drivers/net: return operator cleanup (FATE#311451 bnc#689230).
- patches.drivers/bna-enable-pure-priority-tagged-packet-reception-and-rxf-uninit-cleanup-fix:
  bna: Enable pure priority tagged packet reception and rxf
  uninit cleanup fix (FATE#311451 bnc#689230).
- patches.drivers/bna-ethtool-cosmetic-use-ethtool-ethtool_cmd_speed-api:
  ethtool: cosmetic: Use ethtool ethtool_cmd_speed API
  (FATE#311451 bnc#689230).
- patches.drivers/bna-ethtool-use-full-32-bit-speed-range-in-ethtool-s-set_settings:
  ethtool: Use full 32 bit speed range in ethtool's set_settings
  (FATE#311451 bnc#689230).
- patches.drivers/bna-fix-call-trace-when-interrupts-are-disabled-while-sleeping-function-kzalloc-is-called:
  Fix call trace when interrupts are disabled while sleeping
  function kzalloc is called (FATE#311451 bnc#689230).
- patches.drivers/bna-fix-common-misspellings: Fix common
  misspellings (FATE#311451 bnc#689230).
- patches.drivers/bna-fix-ethtool-register-dump-and-reordered-an-api:
  bna: Fix ethtool register dump and reordered an API (FATE#311451
- patches.drivers/bna-fix-for-clean-fw-re-initialization: bna:
  fix for clean fw re-initialization (FATE#311451 bnc#689230).
- patches.drivers/bna-fix-for-handling-firmware-heartbeat-failure:
  bna: Fix for handling firmware heartbeat failure (FATE#311451
- patches.drivers/bna-fix-for-tx-queue: bna: Fix for TX queue
  (FATE#311451 bnc#689230).
- patches.drivers/bna-fix-interrupt-handling: bna: fix interrupt
  handling (FATE#311451 bnc#689230).
- patches.drivers/bna-fix-memory-leak-during-rx-path-cleanup: bna:
  fix memory leak during RX path cleanup (FATE#311451 bnc#689230).
- patches.drivers/bna-fix-set-but-unused-variables: bna: Fix
  set-but-unused variables (FATE#311451 bnc#689230).
- patches.drivers/bna-fix-stats-handling: bna: fix stats handling
  (FATE#311451 bnc#689230).
- patches.drivers/bna-fixed-build-break-for-allyesconfig: bna:
  Fixed build break for allyesconfig (FATE#311451 bnc#689230).
- patches.drivers/bna-ioc-failure-auto-recovery-fix: bna: IOC
  failure auto recovery fix (FATE#311451 bnc#689230).
- patches.drivers/bna-ioc-uninit-check-and-misc-cleanup: bna:
  IOC uninit check and misc cleanup (FATE#311451 bnc#689230).
- patches.drivers/bna-net-bna-fix-compile-warning-of-bfa_ioc_smem_pgoff-defined-but-not-used:
  net:bna: fix compile warning of ‘bfa_ioc_smem_pgoff’
  defined but not used (FATE#311451 bnc#689230).
- patches.drivers/bna-net-bna-fix-lock-imbalance: NET: bna,
  fix lock imbalance (FATE#311451 bnc#689230).
- patches.drivers/bna-off-by-one: bna: off by one (FATE#311451
- patches.drivers/bna-port-enable-disable-sync-and-txq-priority-fix:
  bna: Port enable disable sync and txq priority fix (FATE#311451
- patches.drivers/bna-remove-unnecessary-memset-0: bna: Remove
  unnecessary memset(,0,) (FATE#311451 bnc#689230).
- patches.drivers/bna-removed-unused-code: bna: Removed unused
  code (FATE#311451 bnc#689230).
- patches.drivers/bna-restore-vlan-filter-table: bna: Restore
  VLAN filter table (FATE#311451 bnc#689230).
- patches.drivers/bna-scope-and-dead-code-cleanup: bna: scope
  and dead code cleanup (FATE#311451 bnc#689230).
- patches.drivers/bna-txrx-and-datapath-fix: bna: TxRx and
  datapath fix (FATE#311451 bnc#689230).
- patches.drivers/bna-update-the-driver-version-to-2-3-2-3: bna:
  Update the driver version to (FATE#311451 bnc#689230).
- patches.drivers/bna-use-device-model-dma-api: bna: use device
  model DMA API (FATE#311451 bnc#689230).
- patches.drivers/bna-vlan-don-t-check-for-vlan-group-before-vlan_tx_tag_present:
  vlan: Don't check for vlan group before vlan_tx_tag_present
  (FATE#311451 bnc#689230).

suse-commit: 71b742296d49f4ddcd96a05dca801d9c94dfc29f

7 years ago- patches.suse/laptop-mode-Make-flushes-per-device.patch: Fix build.
Michal Marek [Mon, 18 Jul 2011 14:48:20 +0000 (16:48 +0200)]
- patches.suse/laptop-mode-Make-flushes-per-device.patch: Fix build.

suse-commit: 2d7d699524670d1c24c46297916e7160d6d47f23

7 years agoMerge branch 'SLE11-SP2' of into SLE11-SP2
Torsten Duwe [Mon, 18 Jul 2011 14:26:05 +0000 (16:26 +0200)]
Merge branch 'SLE11-SP2' of into SLE11-SP2

suse-commit: 6ab4f7b85a912066d5f8ef977a5badc861dcb026

7 years ago- patches.suse/zcrypt-feed-hwrandom: s390: provide hardware
Torsten Duwe [Mon, 18 Jul 2011 14:25:24 +0000 (16:25 +0200)]
- patches.suse/zcrypt-feed-hwrandom: s390: provide hardware
  randomness from zcrypt card to /dev/random (fate 310591).

suse-commit: 5413ef9c777e243ac8c05cba174aabd70977c233

7 years ago- bnx2fc update to 1.0.4:
Hannes Reinecke [Mon, 18 Jul 2011 13:20:01 +0000 (15:20 +0200)]
- bnx2fc update to 1.0.4:
  * patches.drivers/bnx2fc-Do-not-arm-CQ-when-there-are-no-CQEs.patch:
    bnx2fc: Do not arm CQ when there are no CQEs (bnc#698053,
  * patches.drivers/bnx2fc-Enable-REC-CONF-support-for-the-session.patch:
    bnx2fc: Enable REC & CONF support for the session (bnc#698053,
  * patches.drivers/bnx2fc-Enable-support-for-sequence-level-error-reco.patch:
    bnx2fc: Enable support for sequence level error recovery
    (bnc#698053, FATE#311462).
  * patches.drivers/bnx2fc-Fix-kernel-panic-when-deleting-NPIV-ports.patch:
    bnx2fc: Fix kernel panic when deleting NPIV ports (bnc#698053,
  * patches.drivers/bnx2fc-HSI-changes-for-tape.patch: bnx2fc:
    HSI changes for tape (bnc#698053, FATE#311462).
  * patches.drivers/bnx2fc-Handle-REC_TOV-error-code-from-firmware.patch:
    bnx2fc: Handle REC_TOV error code from firmware (bnc#698053,
  * patches.drivers/bnx2fc-Introduce-interface-structure-for-each-vlan.patch:
    bnx2fc: Introduce interface structure for each vlan interface
    (bnc#698053, FATE#311462).
  * patches.drivers/bnx2fc-REC-SRR-link-service-request-and-response-ha.patch:
    bnx2fc: REC/SRR link service request and response handling
    (bnc#698053, FATE#311462).
  * patches.drivers/bnx2fc-Replace-printks-with-KERN_ALERT-to-KERN_ERR.patch:
    bnx2fc: Replace printks with KERN_ALERT to KERN_ERR/KERN_INFO
    (bnc#698053, FATE#311462).
  * patches.drivers/bnx2fc-Support-sequence-cleanup-task.patch:
    bnx2fc: Support 'sequence cleanup' task (bnc#698053, FATE#311462).
  * patches.drivers/bnx2fc-Tx-BDs-cache-in-write-tasks.patch:
    bnx2fc: Tx BDs cache in write tasks (bnc#698053, FATE#311462).
  * patches.drivers/bnx2fc-Update-copyright-and-bump-version-to-1.0.4.patch:
    bnx2fc: Update copyright and bump version to 1.0.4 (bnc#698053,
  * patches.drivers/bnx2fc-hold-tgt-lock-when-calling-cmd_release.patch:
    bnx2fc: hold tgt lock when calling cmd_release (bnc#698053,
  * patches.drivers/bnx2fc-host-stats-show-the-link-speed-unknown-on.patch:
    bnx2fc: host stats show the link speed 'unknown' on NIC
    partitioned interfaces (bnc#698053, FATE#311462).
  * patches.drivers/bnx2fc-scsi_dma_unmap-not-invoked-on-IO-completio.patch:
    bnx2fc: scsi_dma_unmap() not invoked on IO completions
    (bnc#698053, FATE#311462).

suse-commit: 5f56d3a9b52c31b567636ca3c0a325057e20e2da

7 years agoMerge branch 'SLE11-SP1' into SLE11-SP2
Michal Marek [Mon, 18 Jul 2011 13:11:18 +0000 (15:11 +0200)]
Merge branch 'SLE11-SP1' into SLE11-SP2

Refresh patches.fixes/013-numa-hotadd.patch to apply cleanly.

suse-commit: a3834d7cca92cd3a4e4fb8805b19b0dee9934d5a

7 years ago- patches.drivers/alsa-sp1-hda-117-add-Intel-Panther-Point-HDMI-codec-id:
Takashi Iwai [Mon, 18 Jul 2011 09:13:41 +0000 (11:13 +0200)]
- patches.drivers/alsa-sp1-hda-117-add-Intel-Panther-Point-HDMI-codec-id:
  Refresh; added the missing module-alias definition for PPT.

suse-commit: 18d4fdc406d93a395bbbd9bb7b072e6f8c0cfe2e

7 years ago- patches.fixes/mm-do_fault-preallocate-cow-page.patch:
Michal Hocko [Fri, 15 Jul 2011 09:26:30 +0000 (11:26 +0200)]
- patches.fixes/mm-do_fault-preallocate-cow-page.patch:
  mm: preallocate page before lock_page() at filemap COW

suse-commit: e75e37de18da4a391f575981bf5c2b663bee9353

7 years ago- patches.suse/intel-perf-0001-pmu.patch: perf, x86: Add PEBS
Tony Jones [Fri, 15 Jul 2011 06:29:51 +0000 (08:29 +0200)]
- patches.suse/intel-perf-0001-pmu.patch: perf, x86: Add PEBS
  infrastructure (fate#311939, bnc#704917).
- patches.suse/intel-perf-0002-pmu.patch: perf, x86: Implement
  simple LBR support (fate#311939, bnc#704917).
- patches.suse/intel-perf-0003-pmu.patch: perf, x86: use LBR
  for PEBS IP+1 fixup (fate#311939, bnc#704917).
- patches.suse/intel-perf-0004-pmu.patch: perf, x86: Clean up
  IA32_PERF_CAPABILITIES usage (fate#311939, bnc#704917).
- patches.suse/intel-perf-0005-pmu.patch: perf, x86: Expose
  the full PEBS record using PERF_SAMPLE_RAW (fate#311939,
- patches.suse/intel-perf-0006-pmu.patch: perf, x86: Fix silly
  bug in data store buffer allocation (fate#311939, bnc#704917).
- patches.suse/intel-perf-0007-pmu.patch: perf, x86: Disable
  PEBS on clovertown chips (fate#311939, bnc#704917).
- patches.suse/intel-perf-0008-pmu.patch: perf, x86: Clear the
  LBRs on init (fate#311939, bnc#704917).
- patches.suse/intel-perf-0009-pmu.patch: perf, x86: Robustify
  PEBS fixup (fate#311939, bnc#704917).
- patches.suse/intel-perf-0010-pmu.patch: perf, x86: Avoid double
  disable on throttle vs ioctl(PERF_IOC_DISABLE) (fate#311939,
- patches.suse/intel-perf-0011-pmu.patch: perf, x86: Fix pebs
  drains (fate#311939, bnc#704917).
- patches.suse/intel-perf-0012-pmu.patch: perf, x86: Fix PEBS
  enable/disable vs cpuc->enabled (fate#311939, bnc#704917).
- patches.suse/intel-perf-0013-pmu.patch: perf, x86: Fix LBR
  enable/disable vs cpuc->enabled (fate#311939, bnc#704917).
- patches.suse/intel-perf-0014-pmu.patch: perf, x86: Reorder
  intel_pmu_enable_all() (fate#311939, bnc#704917).
- patches.suse/intel-perf-0015-pmu.patch: perf, x86: Deal with
  multiple state bits for pebs-fmt1 (fate#311939, bnc#704917).
- patches.suse/intel-perf-0016-pmu.patch: perf, x86: Fix silly bug
  in intel_pmu_pebs_{enable,disable} (fate#311939, bnc#704917).
- patches.suse/intel-perf-0017-pmu.patch: perf, x86: Don't reset
  the LBR as frequently (fate#311939, bnc#704917).
- patches.suse/intel-perf-0018-pmu.patch: perf, x86: Remove
  checking_{wr,rd}msr() usage (fate#311939, bnc#704917).
- patches.suse/intel-perf-0019-pmu.patch: perf, x86: Fixup the
  PEBS handler for Core2 cpus (fate#311939, bnc#704917).
- patches.suse/intel-perf-0020-pmu.patch: perf, x86: Fix LBR
  read-out (fate#311939, bnc#704917).
- patches.suse/intel-perf-0021-pmu.patch: perf, x86: Fix the
  !CONFIG_CPU_SUP_INTEL build (fate#311939, bnc#704917).
- patches.suse/intel-perf-0022-pmu.patch: perf: Introduce new
  perf_fetch_caller_regs() for hot regs snapshot (fate#311939,
- patches.suse/intel-perf-0023-pmu.patch: perf: export
  perf_trace_regs and perf_arch_fetch_caller_regs (fate#311939,
- patches.suse/intel-perf-0024-pmu.patch: perf, x86: Implement
  initial P4 PMU driver (fate#311939, bnc#704917).
- patches.suse/intel-perf-0025-pmu.patch: x86, perf: Fix NULL
  deref on not assigned x86_pmu (fate#311939, bnc#704917).
- patches.suse/intel-perf-0026-pmu.patch: x86, perf: Unmask
  LVTPC only if we have APIC supported (fate#311939, bnc#704917).
- patches.suse/intel-perf-0027-pmu.patch: perf, x86: Enable
  not tagged retired instruction counting on P4s (fate#311939,
- patches.suse/intel-perf-0028-pmu.patch: perf: Fix unexported
  generic perf_arch_fetch_caller_regs (fate#311939, bnc#704917).
- patches.suse/intel-perf-0029-pmu.patch: perf, x86: Report
  error code that returned from x86_pmu.hw_config() (fate#311939,
- patches.suse/intel-perf-0030-pmu.patch: perf/core, x86:
  Reduce number of CONFIG_X86_LOCAL_APIC macros (fate#311939,
- patches.suse/intel-perf-0031-pmu.patch: perf/core, x86: Remove
  cpu_hw_events.interrupts (fate#311939, bnc#704917).
- patches.suse/intel-perf-0032-pmu.patch: perf/core, x86: Remove
  duplicate perf_event_mask variable (fate#311939, bnc#704917).
- patches.suse/intel-perf-0033-pmu.patch: x86, perf: Use
  apic_write unconditionally (fate#311939, bnc#704917).
- patches.suse/intel-perf-0034-pmu.patch: perf, x86: Add a key
  to simplify template lookup in Pentium-4 PMU (fate#311939,
- patches.suse/intel-perf-0035-pmu.patch: perf, x86: Add cache
  events for the Pentium-4 PMU (fate#311939, bnc#704917).
- patches.suse/intel-perf-0036-pmu.patch: perf_events: Fix
  resource leak in x86 __hw_perf_event_init() (fate#311939,
- patches.suse/intel-perf-0037-pmu.patch: x86, perf: Fix
  few cosmetic dabs for P4 pmu (comments and constantify)
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0038-pmu.patch: x86, perf: Add raw
  events support for the P4 PMU (fate#311939, bnc#704917).
- patches.suse/intel-perf-0039-pmu.patch: perf, x86: Clean up
  debugctlmsr bit definitions (fate#311939, bnc#704917).
- patches.suse/intel-perf-0040-pmu.patch: perf, x86: Add Nehelem
  PMU programming errata workaround (fate#311939, bnc#704917).
- patches.suse/intel-perf-0041-pmu.patch: perf: Use hot regs with
  software sched switch/migrate events (fate#311939, bnc#704917).
- patches.suse/intel-perf-0042-pmu.patch: perf, x86: Undo some
  some *_counter* -> *_event* renames (fate#311939, bnc#704917).
- patches.suse/intel-perf-0043-pmu.patch: perf, x86: implement
  ARCH_PERFMON_EVENTSEL bit masks (fate#311939, bnc#704917).
- patches.suse/intel-perf-0044-pmu.patch: perf, x86: Fix up the
  ANY flag stuff (fate#311939, bnc#704917).
- patches.suse/intel-perf-0045-pmu.patch: perf, x86: Fix
  __initconst vs const (fate#311939, bnc#704917).
- patches.suse/intel-perf-0046-pmu.patch: perf, x86: Add Nehalem
  programming quirk to Westmere (fate#311939, bnc#704917).
- patches.suse/intel-perf-0047-pmu.patch: perf: Drop the frame
  reliablity check (fate#311939, bnc#704917).
- patches.suse/intel-perf-0048-pmu.patch: perf: Enhance perf to
  allow for guest statistic collection from host (fate#311939,
- patches.suse/intel-perf-0049-pmu.patch: perf & kvm: Clean up
  some of the guest profiling callback API details (fate#311939,
- patches.suse/intel-perf-0050-pmu.patch: perf, x86: Move perfctr
  init code to x86_setup_perfctr() (fate#311939, bnc#704917).
- patches.suse/intel-perf-0051-pmu.patch: perf, x86: Move
  x86_setup_perfctr() (fate#311939, bnc#704917).
- patches.suse/intel-perf-0052-pmu.patch: perf, x86: Call
  x86_setup_perfctr() from .hw_config() (fate#311939, bnc#704917).
- patches.suse/intel-perf-0053-pmu.patch: perf, x86: Pass enable
  bit mask to __x86_pmu_enable_event() (fate#311939, bnc#704917).
- patches.suse/intel-perf-0054-pmu.patch: perf, x86: Use weight
  instead of cmask in for_each_event_constraint() (fate#311939,
- patches.suse/intel-perf-0055-pmu.patch: perf, x86: Remove PEBS
  SAMPLE_RAW support (fate#311939, bnc#704917).
- patches.suse/intel-perf-0056-pmu.patch: perf, x86: Consolidate
  some code repetition (fate#311939, bnc#704917).
- patches.suse/intel-perf-0057-pmu.patch: perf, x86: Improve
  the PEBS ABI (fate#311939, bnc#704917).
- patches.suse/intel-perf-0058-pmu.patch: perf, x86: implement
  group scheduling transactional APIs (fate#311939, bnc#704917).
- patches.suse/intel-perf-0059-pmu.patch: x86, perf: P4 PMU --
  configure predefined events (fate#311939, bnc#704917).
- patches.suse/intel-perf-0060-pmu.patch: x86, perf: P4 PMU --
  protect sensible procedures from preemption (fate#311939,
- patches.suse/intel-perf-0061-pmu.patch: x86, perf: P4 PMU --
  Get rid of redundant check for array index (fate#311939,
- patches.suse/intel-perf-0062-pmu.patch: x86, perf: P4 PMU --
  check for proper event index in RAW events (fate#311939,
- patches.suse/intel-perf-0063-pmu.patch: x86, perf: P4 PMU --
  use hash for p4_get_escr_idx() (fate#311939, bnc#704917).
- patches.suse/intel-perf-0064-pmu.patch: x86, perf: P4 PMU -
  fix counters management logic (fate#311939, bnc#704917).
- patches.suse/intel-perf-0065-pmu.patch: perf, x86: P4 PMU --
  handle unflagged events (fate#311939, bnc#704917).
- patches.suse/intel-perf-0066-pmu.patch: perf, x86: P4 PMU --
  fix typo in unflagged NMI handling (fate#311939, bnc#704917).
- patches.suse/intel-perf-0067-pmu.patch: perf, x86: P4 PMU --
  do a real check for ESCR address being in hash (fate#311939,
- patches.suse/intel-perf-0068-pmu.patch: perf, x86:
  P4_pmu_schedule_events -- use smp_processor_id instead of raw_
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0069-pmu.patch: perf, trace: Fix !x86
  build bug (fate#311939, bnc#704917).
- patches.suse/intel-perf-0070-pmu.patch: perf_events: Fix event
  scheduling issues introduced by transactional API (fate#311939,
- patches.suse/intel-perf-0071-pmu.patch: perf, x86: Make a
  second write to performance counter if needed (fate#311939,
- patches.suse/intel-perf-0072-pmu.patch: perf_events: Fix Intel
  Westmere event constraints (fate#311939, bnc#704917).
- patches.suse/intel-perf-0073-pmu.patch: perf, x86: Fix incorrect
  branches event on AMD CPUs (fate#311939, bnc#704917).
- patches.suse/intel-perf-0074-pmu.patch: perf, x86: P4 PMU --
  redesign cache events (fate#311939, bnc#704917).
- patches.suse/intel-perf-0075-pmu.patch: perf, x86: P4 PMU
  -- update nmi irq statistics and unmask lvt entry properly
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0076-pmu.patch: perf, x86: Fix Intel-nhm
  PMU programming errata workaround (fate#311939, bnc#704917).
- patches.suse/intel-perf-0077-pmu.patch: perf: Drop unappropriate
  tests on arch callchains (fate#311939, bnc#704917).
- patches.suse/intel-perf-0078-pmu.patch: perf, x86, Pentium4:
  Clear the P4_CCCR_FORCE_OVF flag (fate#311939, bnc#704917).
- patches.suse/intel-perf-0079-pmu.patch: perf, x86, Pentium4:
  Add RAW events verification (fate#311939, bnc#704917).
- patches.suse/intel-perf-0080-pmu.patch: perf, x86: Fix
  accidentally ack'ing a second event on intel perf counter
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0081-pmu.patch: perf, x86: Fix
  handle_irq return values (fate#311939, bnc#704917).
- patches.suse/intel-perf-0082-pmu.patch: perf, x86: Try to handle
  unknown nmis with an enabled PMU (fate#311939, bnc#704917).
- patches.suse/intel-perf-0083-pmu.patch: perf_events: Fix
  BTS interrupt handling to avoid being dazed by NMI (v2)
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0084-pmu.patch: perf, x86: Catch
  spurious interrupts after disabling counters (fate#311939,
- patches.suse/intel-perf-0085-pmu.patch: perf, x86: Handle in
  flight NMIs on P4 platform (fate#311939, bnc#704917).
- patches.suse/intel-perf-0086-pmu.patch: perf, gcc-4.6: Fix
  set but unused variable (fate#311939, bnc#704917).
- patches.suse/intel-perf-0087-pmu.patch: perf_events: Fix bogus
  AMD64 generic TLB events (fate#311939, bnc#704917).
- patches.suse/intel-perf-0088-pmu.patch: perf, x86: Extract
  PEBS/BTS buffer free routines (fate#311939, bnc#704917).
- patches.suse/intel-perf-0089-pmu.patch: perf, x86: Extract
  PEBS/BTS allocation functions (fate#311939, bnc#704917).
- patches.suse/intel-perf-0090-pmu.patch: perf, x86: Extract DS
  alloc/free functions (fate#311939, bnc#704917).
- patches.suse/intel-perf-0091-pmu.patch: perf, x86: Fixup the
  precise_ip computation (fate#311939, bnc#704917).
- patches.suse/intel-perf-0092-pmu.patch: perf, x86: Less
  disastrous PEBS/BTS buffer allocation failure (fate#311939,
- patches.suse/intel-perf-0093-pmu.patch: perf, x86: Clean up
  reserve_ds_buffers() signature (fate#311939, bnc#704917).
- patches.suse/intel-perf-0094-pmu.patch: perf, x86: Use NUMA
  aware allocations for PEBS/BTS/DS allocations (fate#311939,
- patches.suse/intel-perf-0095-pmu.patch: x86, perf, nmi: Disable
  perf if counters are not accessible (fate#311939, bnc#704917).
- patches.suse/intel-perf-0096-pmu.patch: perf: Introduce
  is_sampling_event() (fate#311939, bnc#704917).
- patches.suse/intel-perf-0097-pmu.patch: perf, x86: Detect
  broken BIOSes that corrupt the PMU (fate#311939, bnc#704917).
- patches.suse/intel-perf-0098-pmu.patch: perf, x86: Provide a
  PEBS capable cycle event (fate#311939, bnc#704917).
- patches.suse/intel-perf-0099-pmu.patch: perf, x86: P4 PMU -
  Fix unflagged overflows handling (fate#311939, bnc#704917).
- patches.suse/intel-perf-0100-pmu.patch: perf: Fix Pentium4
  raw event validation (fate#311939, bnc#704917).
- patches.suse/intel-perf-0101-pmu.patch: perf, x86: P4 PMU:
  Fix spurious NMI messages (fate#311939, bnc#704917).
- patches.suse/intel-perf-0102-pmu.patch: perf, x86: Use helper
  function in x86_pmu_enable_all() (fate#311939, bnc#704917).
- patches.suse/intel-perf-0103-pmu.patch: perf, x86: Calculate
  perfctr msr addresses in helper functions (fate#311939,
- patches.suse/intel-perf-0104-pmu.patch: perf, x86: Add new
  AMD family 15h msrs to perfctr reservation code (fate#311939,
- patches.suse/intel-perf-0105-pmu.patch: perf, x86: Store
  perfctr msr addresses in config_base/event_base (fate#311939,
- patches.suse/intel-perf-0106-pmu.patch: perf, x86: Add support
  for AMD family 15h core counters (fate#311939, bnc#704917).
- patches.suse/intel-perf-0107-pmu.patch: perf, x86: Add Intel
  SandyBridge CPU support (fate#311939, bnc#704917).
- patches.suse/intel-perf-0108-plus-pmu.patch: perf, arch: Cleanup
  perf-pmu init vs lockup-detector (fate#311939, bnc#704917).
- patches.suse/intel-perf-0108-pmu.patch: perf_events: Update
  PEBS event constraints (fate#311939, bnc#704917).
- patches.suse/intel-perf-0109-pmu.patch: perf: Add support for
  supplementary event registers (fate#311939, bnc#704917).
- patches.suse/intel-perf-0110-pmu.patch: perf: Fix LLC-* events
  on Intel Nehalem/Westmere (fate#311939, bnc#704917).
- patches.suse/intel-perf-0111-pmu.patch: perf: Avoid the
  percore allocations if the CPU is not HT capable (fate#311939,
- patches.suse/intel-perf-0112-pmu.patch: perf, x86: Clean up
  SandyBridge PEBS events (fate#311939, bnc#704917).
- patches.suse/intel-perf-0113-pmu.patch: perf, x86:
  Use INTEL_*_CONSTRAINT() for all PEBS event constraints
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0114-pmu.patch: perf, x86: Fix Intel
  fixed counters base initialization (fate#311939, bnc#704917).
- patches.suse/intel-perf-0115-pmu.patch: perf, x86: P4 PMU
  - Read proper MSR register to catch unflagged overflows
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0116-pmu.patch: perf, x86: Complain
  louder about BIOSen corrupting CPU/PMU state and continue
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0117-pmu.patch: perf, x86: P4 PMU -
  clean up the code a bit (fate#311939, bnc#704917).
- patches.suse/intel-perf-0118-pmu.patch: perf, x86: Fix
  pre-defined cache-misses event for AMD family 15h cpus
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0119-pmu.patch: perf, x86: Fix AMD
  family 15h FPU event constraints (fate#311939, bnc#704917).
- patches.suse/intel-perf-0120-pmu.patch: perf, x86:
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0121-pmu.patch: perf: Support Xeon
  E7's via the Westmere PMU driver (fate#311939, bnc#704917).
- patches.suse/intel-perf-0122-pmu.patch: x86, perf event:
  Turn off unstructured raw event access to offcore registers
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0123-pmu.patch: perf, x86: P4 PMU --
  Use perf_sample_data_init helper (fate#311939, bnc#704917).
- patches.suse/intel-perf-0124-pmu.patch: perf, x86: Update/fix
  Intel Nehalem cache events (fate#311939, bnc#704917).
- patches.suse/intel-perf-0125-pmu.patch: perf events, x86, P4:
  Fix typo in comment (fate#311939, bnc#704917).
- patches.suse/intel-perf-0126-pmu.patch: perf, x86: Fix BTS
  condition (fate#311939, bnc#704917).
- patches.suse/intel-perf-0127-pmu.patch: perf events, x86:
  Work around the Nehalem AAJ80 erratum (fate#311939, bnc#704917).
- patches.suse/intel-perf-0128-pmu.patch: perf events: Add
  stalled cycles generic event - PERF_COUNT_HW_STALLED_CYCLES
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0129-pmu.patch: perf events, x86:
  Mark constrant tables read mostly (fate#311939, bnc#704917).
- patches.suse/intel-perf-0130-pmu.patch: perf, x86, nmi: Move
  LVT un-masking into irq handlers (fate#311939, bnc#704917).
- patches.suse/intel-perf-0131-pmu.patch: perf event, x86:
  Use better stalled cycles metric (fate#311939, bnc#704917).
- patches.suse/intel-perf-0132-pmu.patch: perf events: Add
  generic front-end and back-end stalled cycle event definitions
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0133-pmu.patch: perf, x86: Add new
  stalled cycles events for Intel and AMD CPUs (fate#311939,
- patches.suse/intel-perf-0134-pmu.patch: perf events, x86: Add
  Westmere stalled-cycles-frontend/backend events (fate#311939,
- patches.suse/intel-perf-0135-pmu.patch: perf events, x86:
  Add SandyBridge stalled-cycles-frontend/backend events
  (fate#311939, bnc#704917).
- patches.suse/intel-perf-0136-pmu.patch: perf events, x86: Fix
  Intel Nehalem and Westmere last level cache event definitions
  (fate#311939, bnc#704917).

suse-commit: a74c421e4223bf4803803492cb813f7f2aa06c4c

7 years ago- patches.fixes/mm-avoid-null-pointer-deref-in-sync_mm_rss.patch:
Bruce Rogers [Fri, 15 Jul 2011 03:16:03 +0000 (21:16 -0600)]
- patches.fixes/mm-avoid-null-pointer-deref-in-sync_mm_rss.patch:
  mm: avoid null-pointer deref in sync_mm_rss() (bnc#704163).

suse-commit: 07eb337df78503490d03ccead73f3d732d937ba3

7 years agoBoxboro-EX: Memory/CPU hot add - Fate#311831:
Thomas Renninger [Fri, 15 Jul 2011 01:41:50 +0000 (03:41 +0200)]
Boxboro-EX: Memory/CPU hot add - Fate#311831:
- patches.arch/acpi_processor_delay_cpuidle_init_on_hotplug.patch:
  acpi processor hotplug: Delay most initialization to cpu online
- patches.arch/acpi_processor_fix_cpuidle_unregister.patch:
  acpi processor: Avoid WARN message on processor driver removal
- patches.arch/acpi_processor_split_add.patch: acpi processor:
  Split up acpi_processor_add() (fate#311831,bnc#703100).
- patches.arch/intel_idle_register_per_cpu.patch: intel_idle:
  Split up and provide per CPU initialization func
- patches.fixes/001-numa-hotadd.patch: memory-hotplug:
  create /sys/firmware/memmap entry for new memory
- patches.fixes/002-numa-hotadd.patch: slab: add memory hotplug
  support (fate#311831,bnc#703100).
- patches.fixes/003-numa-hotadd.patch: x86, mm: Separate
  x86_64 vmalloc_sync_all() into separate functions
- patches.fixes/004-numa-hotadd.patch: cpu/mem hotplug: enable
  CPUs online before local memory online (fate#311831,bnc#703100).
- patches.fixes/005a-numa-hotadd.patch: mm: do not iterate over
  NR_CPUS in __zone_pcp_update() (fate#311831,bnc#703100).
- patches.fixes/006-numa-hotadd.patch: mem-hotplug: avoid
  multiple zones sharing same boot strapping boot_pageset
- patches.fixes/007-numa-hotadd.patch: mem-hotplug: fix
  potential race while building zonelist for new populated zone
- patches.fixes/008-numa-hotadd.patch: x86-64, mem: Update
  all PGDs for direct mapping and vmemmap mapping changes
- patches.fixes/009-numa-hotadd.patch: x86, mtrr: Use stop machine
  context to rendezvous all the cpu's (fate#311831,bnc#703100).
- patches.fixes/010-numa-hotadd.patch: x86, mtrr:
  lock stop machine during MTRR rendezvous sequence
- patches.fixes/011-numa-hotadd.patch: mem-hotplug: introduce
  {un}lock_memory_hotplug() (fate#311831,bnc#703100).
- patches.fixes/012-numa-hotadd.patch: memory hotplug: one more
  lock on memory hotplug (fate#311831,bnc#703100).
- patches.fixes/013-numa-hotadd.patch: mm, mem-hotplug:
  recalculate lowmem_reserve when memory hotplug occurs
- patches.fixes/014-numa-hotadd.patch: mm/memory_hotplug.c:
  fix building of node hotplug zonelist (fate#311831,bnc#703100).
- patches.fixes/015-numa-hotadd.patch:  mm, hotplug: fix error
  handling in mem_online_node() (fate#311831,bnc#703100).
- patches.fixes/016-numa-hotadd.patch: mm, hotplug: protect
  zonelist building with zonelists_mutex (fate#311831,bnc#703100).
- patches.fixes/017-numa-hotadd.patch: hugetlb: fix section
  mismatches (fate#311831,bnc#703100).
- patches.fixes/018-numa-hotadd.patch: x86, mm: fix uninitialized
  addr in kernel_physical_mapping_init() (fate#311831,bnc#703100).
- patches.xen/xen3-auto-common.diff: Refresh.
- patches.xen/xen3-patch-2.6.19: Refresh.
- patches.xen/xen3-patch-2.6.23: Refresh.
- patches.xen/xen3-patch-2.6.24: Refresh.
- patches.xen/xen3-patch-2.6.27: Refresh.
- patches.xen/xen3-patch-2.6.32: Refresh.

suse-commit: 9aec54a2fed345fa9f9844b286c6c9331d1cd221

7 years ago- patches.arch/x86-efi-00-introduce-free_boot_mem_late,
Raymund Will [Thu, 14 Jul 2011 19:26:16 +0000 (21:26 +0200)]
- patches.arch/x86-efi-00-introduce-free_boot_mem_late,
  Backport of a series of patches for 3.0, which significantly improve
  support for recent UEFI systems.   (bnc#568848, bnc#655434, bnc#681242)

suse-commit: c5b11eef88c62bd1e0f0f5454a1e7eaeab208082

7 years ago- patches.fixes/arch-mm-filter-disallowed-nodes-from-arch-specific-s.patch:
Michal Hocko [Thu, 14 Jul 2011 17:57:44 +0000 (19:57 +0200)]
- patches.fixes/arch-mm-filter-disallowed-nodes-from-arch-specific-s.patch:
  arch, mm: filter disallowed nodes from arch specific show_mem
  functions (bnc#705433).
- patches.fixes/lib-arch-add-filter-argument-to-show_mem-and-fix-pri.patch:
  lib, arch: add filter argument to show_mem and fix private
  implementations (bnc#705433).
- patches.fixes/oom-suppress-nodes-that-are-not-allowed-from-meminfo-on-page-alloc.patch:
  oom: suppress nodes that are not allowed from meminfo on page
  alloc failure (bnc#705433).
- patches.fixes/oom-suppress-nodes-that-are-not-allowed-from-meminfo.patch:
  oom: suppress nodes that are not allowed from meminfo on oom
  kill (bnc#705433).
- patches.fixes/oom-suppress-show_mem-for-many-nodes-in-irq-context-.patch:
  oom: suppress show_mem() for many nodes in irq context on page
  alloc failure (bnc#705433).
- patches.suse/SoN-08-mm-page_alloc-emerg.patch: Refresh.

suse-commit: 5a9602d7060d46693c90d2bd07b7b205a579575a

7 years ago- patches.suse/staging-hv-fixup-reference-counting-for-net-and-stor.patch:
Olaf Hering [Thu, 14 Jul 2011 17:17:51 +0000 (19:17 +0200)]
- patches.suse/staging-hv-fixup-reference-counting-for-net-and-stor.patch:
  Fixup reference counting for both net and stor devices.
- patches.suse/staging-hv-vmbus-init-on-dmi-match.patch:
  only load the driver in a hyper-v guest (bnc#704957).
- patches.suse/staging-hv-vmbus-modalias-namespace.patch:
- patches.suse/staging-hv-vmbus-modalias-show.patch: hv:
  show the modalias files in /sys/bus/vmbus/devices/*/ (fate#311487).

suse-commit: ca4f4658c7833b623c844589bd8880285491b6b8

7 years ago- patches.drivers/intel-add-partner-point-pch-support.patch:
Jiri Kosina [Thu, 14 Jul 2011 14:59:57 +0000 (16:59 +0200)]
- patches.drivers/intel-add-partner-point-pch-support.patch:
  Core Driver Enabling for Intel Panther Point PCH (bnc#682755

suse-commit: 35b6c9f0eb0311aa7dc1edc4571573882a69d1c6

7 years agoMerge branch 'SLE11-SP1' of ssh:// into SLE11-SP1
Jan Kara [Thu, 14 Jul 2011 12:47:44 +0000 (14:47 +0200)]
Merge branch 'SLE11-SP1' of ssh:// into SLE11-SP1


suse-commit: 7da281afbd78855773911f3b44d7d45efa04ec20

7 years ago- patches.fixes/fs-d_validate-fixes.patch: fs: d_validate fixes
Miklos Szeredi [Thu, 14 Jul 2011 12:26:01 +0000 (14:26 +0200)]
- patches.fixes/fs-d_validate-fixes.patch: fs: d_validate fixes

suse-commit: 1f1e5e88f7c81a0e4936349731910655ccf51ff8

7 years ago- Disable:
Jan Kara [Thu, 14 Jul 2011 12:19:26 +0000 (14:19 +0200)]
- Disable:
  as they cause sync(1) to have larger overhead (bnc#703786)

suse-commit: e5cac1e7b72d9752ee2ed17462f9beb7a200dc11

7 years ago- patches.suse/mm-thp-fix-dev-zero.patch: Add CVE and bugzilla information
Mel Gorman [Thu, 14 Jul 2011 11:44:05 +0000 (12:44 +0100)]
- patches.suse/mm-thp-fix-dev-zero.patch: Add CVE and bugzilla information

suse-commit: 5bcc15b6a6b8abbe71981a49f1dbab9b618a35f7

7 years ago- patches.suse/mm-thp-fix-dev-zero.patch: Correct git commit ID
Mel Gorman [Thu, 14 Jul 2011 10:17:29 +0000 (11:17 +0100)]
- patches.suse/mm-thp-fix-dev-zero.patch: Correct git commit ID

suse-commit: 4941c54769df88566276462049f1bdfaa611c800

7 years agoMerge branch 'memcg-backports' into SLE11-SP2
Michal Hocko [Thu, 14 Jul 2011 09:21:08 +0000 (11:21 +0200)]
Merge branch 'memcg-backports' into SLE11-SP2

suse-commit: 44b5d4708f5e76edc62a8115b61923ee8484952a

7 years agoMerge commit 'origin/SLE11-SP2' into SLE11-SP2
Jiri Slaby [Thu, 14 Jul 2011 08:57:27 +0000 (10:57 +0200)]
Merge commit 'origin/SLE11-SP2' into SLE11-SP2

suse-commit: 9d205ac06464a72ef556e1c9a07110b63f8b8888

7 years agoMerge branch 'SLE11-SP2' of into...
Bruce Rogers [Wed, 13 Jul 2011 22:48:24 +0000 (16:48 -0600)]
Merge branch 'SLE11-SP2' of into SLE11-SP2

suse-commit: ecd2693d6897c0215b36fc5a135028e2ff824dc6

7 years ago- patches.suse/0001-9p-connect-fixes.patch: 9p connect fixes
Bruce Rogers [Wed, 13 Jul 2011 22:47:32 +0000 (16:47 -0600)]
- patches.suse/0001-9p-connect-fixes.patch: 9p connect fixes
- patches.suse/0002-Kill-9p-readlink.patch: Kill 9p readlink()
- patches.suse/0003-net-9p-fix-virtio-transport-to-correctly-update-stat.patch:
  net/9p: fix virtio transport to correctly update status on
  connect (FATE#311639).
- patches.suse/0004-net-9p-fail-when-user-specifies-a-transport-which-we.patch:
  net/9p: fail when user specifies a transport which we can't find
- patches.suse/0005-net-9p-fix-statsize-inside-twstat.patch:
  net/9p: fix statsize inside twstat (FATE#311639).
- patches.suse/0006-9p-Include-fsync-support-for-9p-client.patch:
  9p: Include fsync support for 9p client (FATE#311639).
- patches.suse/0007-9p-fix-option-parsing.patch: 9p: fix option
  parsing (FATE#311639).
- patches.suse/0008-9p-Fix-the-kernel-crash-on-a-failed-mount.patch:
  9p: Fix the kernel crash on a failed mount (FATE#311639).
- patches.suse/0009-9p-fix-memory-leak-in-v9fs_parse_options.patch:
  9p: fix memory leak in v9fs_parse_options() (FATE#311639).
- patches.suse/0010-9p-fix-p9_client_destroy-unconditional-calling-v9fs_.patch:
  9p: fix p9_client_destroy unconditional calling v9fs_put_trans
- patches.suse/0011-net-9p-Add-multi-channel-support.patch:
  net/9p: Add multi channel support (FATE#311639).
- patches.suse/0012-net-9p-Remove-MAX_9P_CHAN-limit.patch:
  net/9p: Remove MAX_9P_CHAN limit (FATE#311639).
- patches.suse/0013-net-9p-Handle-mount-errors-correctly.patch:
  net/9p: Handle mount errors correctly (FATE#311639).
- patches.suse/0014-9P2010.L-handshake-Add-VFS-flags.patch:
  9P2010.L handshake: Add VFS flags (FATE#311639).
- patches.suse/0015-9P2010.L-handshake-Add-mount-option.patch:
  9P2010.L handshake: Add mount option (FATE#311639).
- patches.suse/0016-9P2010.L-handshake-Remove-dotu-variable.patch:
  9P2010.L handshake: Remove "dotu" variable (FATE#311639).
- patches.suse/0017-9P2010.L-handshake-.L-protocol-negotiation.patch:
  9P2010.L handshake: .L protocol negotiation (FATE#311639).
- patches.suse/0018-fs-9p-Add-hardlink-support-to-.u-extension.patch:
  fs/9p: Add hardlink support to .u extension (FATE#311639).
- patches.suse/0019-net-9p-Use-the-tag-name-in-the-config-space-for-iden.patch:
  net/9p: Use the tag name in the config space for identifying
  mount point (FATE#311639).
- patches.suse/0020-net-9p-Add-sysfs-mount_tag-file-for-virtio-9P-device.patch:
  net/9p: Add sysfs mount_tag file for virtio 9P device
- patches.suse/0021-fs-9p-re-init-the-wstat-in-readdir-loop.patch:
  fs/9p: re-init the wstat in readdir loop (FATE#311639).
- patches.suse/0022-9p-Change-the-name-of-new-protocol-from-9p2010.L-to-.patch:
  9p: Change the name of new protocol from 9p2010.L to 9p2000.L
- patches.suse/0023-9p-Fixes-a-simple-bug-enabling-writes-beyond-2GB.patch:
  9p: Fixes a simple bug enabling writes beyond 2GB (FATE#311639).
- patches.suse/0024-9p-Fix-setting-of-protocol-flags-in-v9fs_session_inf.patch:
  9p: Fix setting of protocol flags in v9fs_session_info structure
- patches.suse/0025-9p-documentation-update.patch: 9p:
  documentation update (FATE#311639).
- patches.suse/0026-fs-9p-Clunk-the-fid-resulting-from-partial-walk-of-t.patch:
  fs/9p: Clunk the fid resulting from partial walk of the name
- patches.suse/0027-9p-drop-nlink-remove.patch: 9p: drop nlink
  remove (FATE#311639).
- patches.suse/0028-9p-Make-sure-we-are-able-to-clunk-the-cached-fid-on-.patch:
  9p: Make sure we are able to clunk the cached fid on umount
- patches.suse/0029-9p-Creating-files-with-names-too-long-should-fail-wi.patch:
  9p: Creating files with names too long should fail with
- patches.suse/0030-9p-return-on-mutex_lock_interruptible.patch:
  9p: return on mutex_lock_interruptible() (FATE#311639).
- patches.suse/0031-9p-saving-negative-to-unsigned-char.patch:
  9p: saving negative to unsigned char (FATE#311639).
- patches.suse/0032-9p-add-bdi-backing-to-mount-session.patch:
  9p: add bdi backing to mount session (FATE#311639).
- patches.suse/0033-Backport-fix-for-bdi-patch-to-make-it-work-with-SLES.patch:
  Backport fix for bdi patch to make it work with SLES kernel
- patches.suse/0034-9p-VFS-switches-for-9p2000.L-protocol-and-client-cha.patch:
  9p: VFS switches for 9p2000.L: protocol and client changes
- patches.suse/0035-9p-VFS-switches-for-9p2000.L-VFS-switches.patch:
  9p: VFS switches for 9p2000.L: VFS switches (FATE#311639).
- patches.suse/0036-9p-add-9P2000.L-statfs-operation.patch:
  9p: add 9P2000.L statfs operation (FATE#311639).
- patches.suse/0037-9p-add-9P2000.L-rename-operation.patch:
  9p: add 9P2000.L rename operation (FATE#311639).
- patches.suse/0038-fs-9p-Make-sure-we-properly-instantiate-dentry.patch:
  fs/9p: Make sure we properly instantiate dentry (FATE#311639).
- patches.suse/0039-9p-Add-mksock-support.patch:     9p: Add
  mksock support (FATE#311639).
- patches.suse/0040-9p-cleanup-remove-unneeded-assignment.patch:
  9p: cleanup: remove unneeded assignment (FATE#311639).
- patches.suse/0041-9p-Optimize-TCREATE-by-eliminating-a-redundant-fid-c.patch:
  9p: Optimize TCREATE by eliminating a redundant fid clone
- patches.suse/0042-9p-Pass-the-correct-end-of-buffer-to-p9stat_read.patch:
  9p: Pass the correct end of buffer to p9stat_read (FATE#311639).
- patches.suse/0043-virtio_9p.h-needs-linux-types.h.patch:
  virtio_9p.h needs <linux/types.h> (FATE#311639).
- patches.suse/0044-9p-Make-use-of-iounit-for-read-write.patch:
  9p: Make use of iounit for read/write (FATE#311639).
- patches.suse/0045-9p-readdir-implementation-for-9p2000.L.patch:
  9p: readdir implementation for 9p2000.L (FATE#311639).
- patches.suse/0046-net-9p-Handle-the-server-returned-error-properly.patch:
  net/9p: Handle the server returned error properly (FATE#311639).
- patches.suse/0047-fs-9p-Pass-the-correct-user-credentials-during-attac.patch:
  fs/9p: Pass the correct user credentials during attach
- patches.suse/0048-9p-getattr-client-implementation-for-9P2000.L-protoc.patch:
  9p: getattr client implementation for 9P2000.L protocol
- patches.suse/0049-9p-Implement-client-side-of-setattr-for-9P2000.L-pro.patch:
  9p: Implement client side of setattr for 9P2000.L protocol
- patches.suse/0050-9p-Define-and-implement-TLINK-for-9P2000.L.patch:
  9p: Define and implement TLINK for 9P2000.L (FATE#311639).
- patches.suse/0051-9p-Define-and-implement-TLINK-for-9P2000.L.patch:
  9p: Define and implement TLINK for 9P2000.L (FATE#311639).
- patches.suse/0052-9p-Define-and-implement-TSYMLINK-for-9P2000.L.patch:
  9p: Define and implement TSYMLINK for 9P2000.L (FATE#311639).
- patches.suse/0053-9p-Implement-TMKNOD.patch:     9p: Implement
  TMKNOD (FATE#311639).
- patches.suse/0054-9p-Implement-TMKDIR.patch:     9p: Implement
  TMKDIR (FATE#311639).
- patches.suse/0055-fs-9p-This-patch-implements-TLCREATE-for-9p2000.L-pr.patch:
  fs/9p: This patch implements TLCREATE for 9p2000.L protocol
- patches.suse/0056-9p-Implement-LOPEN.patch:     9p: Implement
  LOPEN (FATE#311639).
- patches.suse/0057-fs-9p-Prevent-parallel-rename-when-doing-fid_lookup.patch:
  fs/9p: Prevent parallel rename when doing fid_lookup
- patches.suse/0058-fs-9p-destroy-fid-on-failed-remove.patch:
  fs/9p: destroy fid on failed remove (FATE#311639).
- patches.suse/0059-fs-9p-remove-sparse-warning-in-vfs_inode.patch:
  fs/9p: remove sparse warning in vfs_inode (FATE#311639).
- patches.suse/0060-virtio_9p.h-include-linux-types.h.patch:
  virtio_9p.h: include linux/types.h (FATE#311639).
- patches.suse/0061-9p-potential-ERR_PTR-dereference.patch:
  9p: potential ERR_PTR() dereference (FATE#311639).
- patches.suse/0062-net-9p-trans_fd.c-Fix-unsigned-return-type.patch:
  net/9p/trans_fd.c: Fix unsigned return type (FATE#311639).
- patches.suse/0063-fs-9p-net-9p-memory-leak-fixes.patch:
  fs/9p, net/9p: memory leak fixes (FATE#311639).
- patches.suse/0064-fs-9p-Fix-error-handling-in-v9fs_get_sb.patch:
  fs/9p: Fix error handling in v9fs_get_sb (FATE#311639).
- patches.suse/0065-9p-Check-for-NULL-fid-in-v9fs_dir_release.patch:
  9p: Check for NULL fid in v9fs_dir_release() (FATE#311639).
- patches.suse/0066-fs-9p-Use-the-correct-dentry-operations.patch:
  fs/9p: Use the correct dentry operations (FATE#311639).
- patches.suse/0067-fs-9p-Don-t-use-dotl-version-of-mknod-for-dotu-inode.patch:
  fs/9p: Don't use dotl version of mknod for dotu inode operations
- patches.suse/0068-net-9p-fix-memory-handling-allocation-in-rdma_reques.patch:
  net/9p: fix memory handling/allocation in rdma_request()
- patches.suse/0069-net-9p-Mount-only-matching-virtio-channels.patch:
  net/9p: Mount only matching virtio channels (FATE#311639).
- patches.suse/0070-9p-client-code-cleanup.patch:     9p: client
  code cleanup (FATE#311639).
- patches.suse/0071-9p-remove-unneeded-checks.patch:     9p:
  remove unneeded checks (FATE#311639).
- patches.suse/0072-fs-9p-setrlimit-fix-for-9p-write.patch:
  fs/9p: setrlimit fix for 9p write (FATE#311639).
- patches.suse/0073-9p-Pass-the-correct-end-of-buffer-to-p9dirent_read.patch:
  9p: Pass the correct end of buffer to p9dirent_read
- patches.suse/0074-fs-9p-mkdir-fix-for-setting-S_ISGID-bit-as-per-paren.patch:
  fs/9p: mkdir fix for setting S_ISGID bit as per parent directory
- patches.suse/0075-net-9p-This-patch-implements-TLERROR-RLERROR-on-the-.patch:
  net/9p: This patch implements TLERROR/RLERROR on the 9P client
- patches.suse/0076-9p-Add-a-Direct-IO-support-for-non-cached-operations.patch:
  9p: Add a Direct IO support for non-cached operations
- patches.suse/0077-fs-9p-Remove-the-redundant-rsize-calculation-in-v9fs.patch:
  fs/9p: Remove the redundant rsize calculation in
  v9fs_file_write() (FATE#311639).
- patches.suse/0078-net-9p-Add-a-Warning-to-catch-NULL-fids-passed-to-p9.patch:
  net/9p: Add a Warning to catch NULL fids passed to
  p9_client_clunk() (FATE#311639).
- patches.suse/0079-fs-9p-Add-file_operations-for-cached-mode-in-dotl-pr.patch:
  [fs/9p] Add file_operations for cached mode in dotl protocol
- patches.suse/0080-9p-Introduce-client-side-TFSYNC-RFSYNC-for-dotl.patch:
  [9p] Introduce client side TFSYNC/RFSYNC for dotl (FATE#311639).
- patches.suse/0081-Backport-fix-change-fsync-signature-to-make-it-compa.patch:
  Backport fix: change fsync signature to make it compatible
  with SLES kernel (FATE#311639).
- patches.suse/0082-9p-Implement-TLOCK.patch:     9p: Implement
  TLOCK (FATE#311639).
- patches.suse/0083-9p-Implement-TGETLOCK.patch:     9p: Implement
  TGETLOCK (FATE#311639).
- patches.suse/0084-9p-Use-V9FS_MAGIC-in-statfs.patch:     9p:
  Use V9FS_MAGIC in statfs (FATE#311639).
- patches.suse/0085-9p-Implement-TREADLINK-operation-for-9p2000.L.patch:
  9p: Implement TREADLINK operation for 9p2000.L (FATE#311639).
- patches.suse/0086-net-9p-Serialize-virtqueue-operations-to-make-VirtIO.patch:
  [net/9p]Serialize virtqueue operations to make VirtIO transport
  SMP safe (FATE#311639).
- patches.suse/0087-net-9p-Add-waitq-to-VirtIO-transport.patch:
  net/9p: Add waitq to VirtIO transport (FATE#311639).
- patches.suse/0088-fs-9p-Use-mknod-9p-operation-on-create-without-open-.patch:
  fs/9p: Use mknod 9p operation on create without open request
- patches.suse/0089-fs-9p-Add-missing-iput-in-v9fs_vfs_lookup.patch:
  fs/9p: Add missing iput in v9fs_vfs_lookup (FATE#311639).
- patches.suse/0090-fs-9p-Use-generic_file_open-with-lookup_instantiate_.patch:
  fs/9p: Use generic_file_open with lookup_instantiate_filp
- patches.suse/0091-net-9p-Return-error-if-we-fail-to-encode-protocol-da.patch:
  net/9p: Return error if we fail to encode protocol data
- patches.suse/0092-9p-Add-datasync-to-client-side-TFSYNC-RFSYNC-for-dot.patch:
  9p: Add datasync to client side TFSYNC/RFSYNC for dotl
- patches.suse/0093-net-9p-Return-error-on-read-with-NULL-buffer.patch:
  net/9p: Return error on read with NULL buffer (FATE#311639).
- patches.suse/0094-net-9p-protocol.c-Remove-duplicated-macros.patch:
  net/9p/protocol.c: Remove duplicated macros (FATE#311639).
- patches.suse/0095-fs-9p-fix-menu-presentation.patch:     fs/9p:
  fix menu presentation (FATE#311639).
- patches.suse/0096-fs-9p-Move-dotl-inode-operations-into-a-seperate-fil.patch:
  fs/9p: Move dotl inode operations into a seperate file
- patches.suse/0097-fs-9p-Simplify-the-.L-create-operation.patch:
  fs/9p: Simplify the .L create operation (FATE#311639).
- patches.suse/0098-net-9p-Use-proper-data-types.patch:
  net/9p: Use proper data types (FATE#311639).
- patches.suse/0099-fs-9p-TREADLINK-bugfix.patch:     fs/9p:
  TREADLINK bugfix (FATE#311639).
- patches.suse/0100-fs-9p-fix-spelling-typo.patch:     fs/9p:
  fix spelling typo (FATE#311639).
- patches.suse/0101-fs-9p-Don-t-set-dentry-d_op-in-create-routines.patch:
  fs/9p: Don't set dentry->d_op in create routines (FATE#311639).
- patches.suse/0102-Backport-fix-set-root-d_op-and-add-dummy-revalidate-.patch:
  Backport fix: set root->d_op and add dummy revalidate func
- patches.suse/0103-net-9p-Preparation-and-helper-functions-for-zero-cop.patch:
  [net/9p] Preparation and helper functions for zero copy
- patches.suse/0104-net-9p-Assign-type-of-transaction-to-tc-pdu-id-which.patch:
  [net/9p] Assign type of transaction to tc->pdu->id which is
  otherwise unsed (FATE#311639).
- patches.suse/0105-net-9p-Add-gup-zero_copy-support-to-VirtIO-transport.patch:
  [net/9p] Add gup/zero_copy support to VirtIO transport layer
- patches.suse/0106-net-9p-Add-preferences-to-transport-layer.patch:
  [net/9p] Add preferences to transport layer (FATE#311639).
- patches.suse/0107-net-9p-Read-side-zerocopy-changes-for-9P2000.L-proto.patch:
  [net/9p] Read side zerocopy changes for 9P2000.L protocol
- patches.suse/0108-net-9p-Write-side-zerocopy-changes-for-9P2000.L-prot.patch:
  [net/9p] Write side zerocopy changes for 9P2000.L protocol
- patches.suse/0109-net-9p-readdir-zerocopy-changes-for-9P2000.L-protoco.patch:
  [net/9p] readdir zerocopy changes for 9P2000.L protocol
- patches.suse/0110-net-9p-Handle-Zero-Copy-TREAD-RERROR-case-in-dotl-ca.patch:
  [net/9p] Handle Zero Copy TREAD/RERROR case in !dotl case
- patches.suse/0111-net-9p-Small-non-IO-PDUs-for-zero-copy-supporting-tr.patch:
  [net/9p] Small non-IO PDUs for zero-copy supporting transports

suse-commit: 98f8eae44d50fc368e5e30a0cda4916f730db530

7 years agoMerge branch 'SLE11-SP1' into SLE11-SP2
Jiri Slaby [Wed, 13 Jul 2011 22:11:08 +0000 (00:11 +0200)]
Merge branch 'SLE11-SP1' into SLE11-SP2


suse-commit: 6392427f1599f3a1abf6f42f60791dda5b994790

7 years agoMerge branch 'SLE11-SP2' of into SLE11-SP2
Tony Jones [Wed, 13 Jul 2011 21:48:04 +0000 (23:48 +0200)]
Merge branch 'SLE11-SP2' of into SLE11-SP2

suse-commit: 7368d3cc0175fc18f6b274349cd6706e99ec8212

7 years agoMerge branch 'SLE11-SP2' of into SLE11-SP2
Tony Jones [Thu, 14 Jul 2011 02:38:24 +0000 (04:38 +0200)]
Merge branch 'SLE11-SP2' of into SLE11-SP2

suse-commit: 6348c725522ac2febf5b138eb21f30fdc13f154c

7 years ago- Update to
Jiri Slaby [Wed, 13 Jul 2011 19:12:08 +0000 (21:12 +0200)]
- Update to
- patches.fixes/PM-Hibernate-Avoid-hitting-OOM-during-preallocation-of-memory.patch:
- patches.fixes/PM-Hibernate-Fix-free_unnecessary_pages.patch:
- patches.fixes/ksm-fix-null-pointer-dereference-in-scan_get_next_rmap_item.patch:
- patches.fixes/tty-ldisc-do-not-close-until-there-are-readers.patch:
- patches.kabi/kabi-fix-struct-clocksource-breakage.patch: kabi:
  fix struct clocksource breakage.
- patches.kabi/revert-mm-prevent-concurrent-unmap_mapping_range-on-.patch:
  Revert "mm: prevent concurrent unmap_mapping_range() on the
  same inode".
- patches.fixes/mm-prevent-concurrent-unmap_mapping_range-on-.patch:
  mm: prevent concurrent unmap_mapping_range() on the same inode.
- patches.rpmify/fix-WARN_RATELIMIT-for-non-CONFIG_BUG.patch:

suse-commit: 40bd2245838a9a19d3b2e5d5294ad347836858f3

7 years ago- patches.fixes/proc_irq_ease_smp_affinity_setting.patch: bitmap,
Thomas Renninger [Wed, 13 Jul 2011 21:38:11 +0000 (23:38 +0200)]
- patches.fixes/proc_irq_ease_smp_affinity_setting.patch: bitmap,
  irq: add smp_affinity_list interface to /proc/irq (bnc#700448).

suse-commit: bf1f6e97dc2b0b6631951eac0fb5cc2b5909a9b7

7 years ago- patches.fixes/audit-filter-rules-cred-perf.patch: Update SP1 patch
Tony Jones [Wed, 13 Jul 2011 17:01:51 +0000 (10:01 -0700)]
- patches.fixes/audit-filter-rules-cred-perf.patch: Update SP1 patch
  with upstream commit info.  No code changes.

suse-commit: d31dee72e271d4a842cce7db98dc97aaf804a604

7 years ago- patches.fixes/audit-filter-rules-cred-perf.patch: Update SP1 patch
Tony Jones [Wed, 13 Jul 2011 16:52:23 +0000 (09:52 -0700)]
- patches.fixes/audit-filter-rules-cred-perf.patch: Update SP1 patch
  with upstream commit info.

suse-commit: c27cbaa1863de2e8beacb8d4434e3f6d593ddbe5

7 years agoRemove duplicate kernel-source.changes entry
Michal Marek [Wed, 13 Jul 2011 12:24:17 +0000 (14:24 +0200)]
Remove duplicate kernel-source.changes entry

suse-commit: 960cb821fbbade504156278383b18582b56e2724

7 years ago- patches.drivers/pmcraid-add-support-for-set-timestamp-comman.patch:
Hannes Reinecke [Wed, 13 Jul 2011 11:51:09 +0000 (13:51 +0200)]
- patches.drivers/pmcraid-add-support-for-set-timestamp-comman.patch:
  pmcraid: add support for set timestamp command and other fixes
- patches.drivers/pmcraid-disable-msix-and-expand-device-confi.patch:
  pmcraid: disable msix and expand device config entry

suse-commit: 13c33b063bb1083b7705baeb8890802d77e5b1c7

7 years agoMerge branch 'SLE11-SP2' of into SLE11-SP2
Hannes Reinecke [Wed, 13 Jul 2011 11:08:09 +0000 (13:08 +0200)]
Merge branch 'SLE11-SP2' of into SLE11-SP2

suse-commit: 0aa977f59e1fd35ecc568cd26e659fd2fa84fda8

7 years ago- patches.xen/xen3-auto-common.diff: Re-add removed Makefile
Hannes Reinecke [Wed, 13 Jul 2011 11:07:52 +0000 (13:07 +0200)]
- patches.xen/xen3-auto-common.diff: Re-add removed Makefile

suse-commit: db4b02355095bf90613f9112f4eb1a4169f669af

7 years agoMerge branch 'SLE11-SP1' into SLE11-SP2
Michal Marek [Wed, 13 Jul 2011 09:37:19 +0000 (11:37 +0200)]
Merge branch 'SLE11-SP1' into SLE11-SP2


suse-commit: 057a0243ecf21b92b89314c3ba8b61456bb9140f

7 years ago- patches.fixes/vgaarb-enable-legacy-resources-only-if-device-decodes.patch:
Egbert Eich [Wed, 13 Jul 2011 06:21:30 +0000 (08:21 +0200)]
- patches.fixes/vgaarb-enable-legacy-resources-only-if-device-decodes.patch:
  VGA arb: Only enable legacy resources if device really
  decodes them.

suse-commit: 46b48610e24aa1a56fe2656c64df33d3200789c6

7 years agoLinux
Greg Kroah-Hartman [Wed, 13 Jul 2011 03:29:43 +0000 (05:29 +0200)]

7 years agomm: prevent concurrent unmap_mapping_range() on the same inode
Miklos Szeredi [Wed, 23 Feb 2011 12:49:47 +0000 (13:49 +0100)]
mm: prevent concurrent unmap_mapping_range() on the same inode

commit 2aa15890f3c191326678f1bd68af61ec6b8753ec upstream.

Michael Leun reported that running parallel opens on a fuse filesystem
can trigger a "kernel BUG at mm/truncate.c:475"

Gurudas Pai reported the same bug on NFS.

The reason is, unmap_mapping_range() is not prepared for more than
one concurrent invocation per inode.  For example:

  thread1: going through a big range, stops in the middle of a vma and
     stores the restart address in vm_truncate_count.

  thread2: comes in with a small (e.g. single page) unmap request on
     the same vma, somewhere before restart_address, finds that the
     vma was already unmapped up to the restart address and happily
     returns without doing anything.

Another scenario would be two big unmap requests, both having to
restart the unmapping and each one setting vm_truncate_count to its
own value.  This could go on forever without any of them being able to

Truncate and hole punching already serialize with i_mutex.  Other
callers of unmap_mapping_range() do not, and it's difficult to get
i_mutex protection for all callers.  In particular ->d_revalidate(),
which calls invalidate_inode_pages2_range() in fuse, may be called
with or without i_mutex.

This patch adds a new mutex to 'struct address_space' to prevent
running multiple concurrent unmap_mapping_range() on the same mapping.

[ We'll hopefully get rid of all this with the upcoming mm
  preemptibility series by Peter Zijlstra, the "mm: Remove i_mmap_mutex
  lockbreak" patch in particular.  But that is for 2.6.39 ]

Signed-off-by: Miklos Szeredi <>
Reported-by: Michael Leun <>
Reported-by: Gurudas Pai <>
Tested-by: Gurudas Pai <>
Acked-by: Hugh Dickins <>
Signed-off-by: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoudp/recvmsg: Clear MSG_TRUNC flag when starting over for a new packet
Xufeng Zhang [Tue, 21 Jun 2011 10:43:40 +0000 (10:43 +0000)]
udp/recvmsg: Clear MSG_TRUNC flag when starting over for a new packet

[ Upstream commit 9cfaa8def1c795a512bc04f2aec333b03724ca2e ]

Consider this scenario: When the size of the first received udp packet
is bigger than the receive buffer, MSG_TRUNC bit is set in msg->msg_flags.
However, if checksum error happens and this is a blocking socket, it will
goto try_again loop to receive the next packet.  But if the size of the
next udp packet is smaller than receive buffer, MSG_TRUNC flag should not
be set, but because MSG_TRUNC bit is not cleared in msg->msg_flags before
receive the next packet, MSG_TRUNC is still set, which is wrong.

Fix this problem by clearing MSG_TRUNC flag when starting over for a
new packet.

Signed-off-by: Xufeng Zhang <>
Signed-off-by: Paul Gortmaker <>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoipv6/udp: Use the correct variable to determine non-blocking condition
Xufeng Zhang [Tue, 21 Jun 2011 10:43:39 +0000 (10:43 +0000)]
ipv6/udp: Use the correct variable to determine non-blocking condition

[ Upstream commit 32c90254ed4a0c698caa0794ebb4de63fcc69631 ]

udpv6_recvmsg() function is not using the correct variable to determine
whether or not the socket is in non-blocking operation, this will lead
to unexpected behavior when a UDP checksum error occurs.

Consider a non-blocking udp receive scenario: when udpv6_recvmsg() is
called by sock_common_recvmsg(), MSG_DONTWAIT bit of flags variable in
udpv6_recvmsg() is cleared by "flags & ~MSG_DONTWAIT" in this call:

    err = sk->sk_prot->recvmsg(iocb, sk, msg, size, flags & MSG_DONTWAIT,
                   flags & ~MSG_DONTWAIT, &addr_len);

i.e. with udpv6_recvmsg() getting these values:

int noblock = flags & MSG_DONTWAIT
int flags = flags & ~MSG_DONTWAIT

So, when udp checksum error occurs, the execution will go to
csum_copy_err, and then the problem happens:

            if (flags & MSG_DONTWAIT)
                    return -EAGAIN;
            goto try_again;

But it will always go to try_again as MSG_DONTWAIT has been cleared
from flags at call time -- only noblock contains the original value
of MSG_DONTWAIT, so the test should be:

            if (noblock)
                    return -EAGAIN;

This is also consistent with what the ipv4/udp code does.

Signed-off-by: Xufeng Zhang <>
Signed-off-by: Paul Gortmaker <>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agonet/ipv4: Check for mistakenly passed in non-IPv4 address
Marcus Meissner [Thu, 2 Jun 2011 04:05:22 +0000 (21:05 -0700)]
net/ipv4: Check for mistakenly passed in non-IPv4 address

[ Upstream commit d0733d2e29b652b2e7b1438ececa732e4eed98eb ]

Check against mistakenly passing in IPv6 addresses (which would result
in an INADDR_ANY bind) or similar incompatible sockaddrs.

Signed-off-by: Marcus Meissner <>
Cc: Reinhard Max <>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoaf_packet: prevent information leak
Eric Dumazet [Tue, 7 Jun 2011 05:42:06 +0000 (22:42 -0700)]
af_packet: prevent information leak

[ Upstream commit 13fcb7bd322164c67926ffe272846d4860196dc6 ]

In 2.6.27, commit 393e52e33c6c2 (packet: deliver VLAN TCI to userspace)
added a small information leak.

Add padding field and make sure its zeroed before copy to user.

Signed-off-by: Eric Dumazet <>
CC: Patrick McHardy <>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agonet: filter: Use WARN_RATELIMIT
Joe Perches [Sat, 21 May 2011 07:48:40 +0000 (07:48 +0000)]
net: filter: Use WARN_RATELIMIT

[ Upstream commit 6c4a5cb219520c7bc937ee186ca53f03733bd09f ]

A mis-configured filter can spam the logs with lots of stack traces.

Rate-limit the warnings and add printout of the bogus filter information.

Original-patch-by: Ben Greear <>
Signed-off-by: Joe Perches <>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agobug.h: Add WARN_RATELIMIT
Joe Perches [Sat, 21 May 2011 07:48:39 +0000 (07:48 +0000)]

[ Upstream commit b3eec79b0776e5340a3db75b34953977c7e5086e ]

Add a generic mechanism to ratelimit WARN(foo, fmt, ...) messages
using a hidden per call site static struct ratelimit_state.

Also add an __WARN_RATELIMIT variant to be able to use a specific
struct ratelimit_state.

Signed-off-by: Joe Perches <>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoPM / Hibernate: Fix free_unnecessary_pages()
Rafael J. Wysocki [Wed, 6 Jul 2011 18:15:23 +0000 (20:15 +0200)]
PM / Hibernate: Fix free_unnecessary_pages()

commit 4d4cf23cdde2f8f9324f5684a7f349e182039529 upstream.

There is a bug in free_unnecessary_pages() that causes it to
attempt to free too many pages in some cases, which triggers the
BUG_ON() in memory_bm_clear_bit() for copy_bm.  Namely, if
count_data_pages() is initially greater than alloc_normal, we get
to_free_normal equal to 0 and "save" greater from 0.  In that case,
if the sum of "save" and count_highmem_pages() is greater than
alloc_highmem, we subtract a positive number from to_free_normal.
Hence, since to_free_normal was 0 before the subtraction and is
an unsigned int, the result is converted to a huge positive number
that is used as the number of pages to free.

Fix this bug by checking if to_free_normal is actually greater
than or equal to the number we're going to subtract from it.

Signed-off-by: Rafael J. Wysocki <>
Reported-and-tested-by: Matthew Garrett <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoPM / Hibernate: Avoid hitting OOM during preallocation of memory
Rafael J. Wysocki [Sat, 11 Sep 2010 18:58:27 +0000 (20:58 +0200)]
PM / Hibernate: Avoid hitting OOM during preallocation of memory

commit 6715045ddc7472a22be5e49d4047d2d89b391f45 upstream.

There is a problem in hibernate_preallocate_memory() that it calls
preallocate_image_memory() with an argument that may be greater than
the total number of available non-highmem memory pages.  If that's
the case, the OOM condition is guaranteed to trigger, which in turn
can cause significant slowdown to occur during hibernation.

To avoid that, make preallocate_image_memory() adjust its argument
before calling preallocate_image_pages(), so that the total number of
saveable non-highem pages left is not less than the minimum size of
a hibernation image.  Change hibernate_preallocate_memory() to try to
allocate from highmem if the number of pages allocated by
preallocate_image_memory() is too low.

Modify free_unnecessary_pages() to take all possible memory
allocation patterns into account.

Reported-by: KOSAKI Motohiro <>
Signed-off-by: Rafael J. Wysocki <>
Tested-by: M. Vefa Bicakci <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoinet_diag: fix inet_diag_bc_audit()
Eric Dumazet [Fri, 17 Jun 2011 20:25:39 +0000 (16:25 -0400)]
inet_diag: fix inet_diag_bc_audit()

[ Upstream commit eeb1497277d6b1a0a34ed36b97e18f2bd7d6de0d ]

A malicious user or buggy application can inject code and trigger an
infinite loop in inet_diag_bc_audit()

Also make sure each instruction is aligned on 4 bytes boundary, to avoid
unaligned accesses.

Reported-by: Dan Rosenberg <>
Signed-off-by: Eric Dumazet <>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agonetlink: Make nlmsg_find_attr take a const nlmsghdr*.
Nelson Elhage [Wed, 3 Nov 2010 16:35:40 +0000 (16:35 +0000)]
netlink: Make nlmsg_find_attr take a const nlmsghdr*.

commit 6b8c92ba07287578718335ce409de8e8d7217e40 upstream.

This will let us use it on a nlmsghdr stored inside a netlink_callback.

Signed-off-by: Nelson Elhage <>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoum: os-linux/mem.c needs sys/stat.h
Liu Aleaxander [Tue, 29 Jun 2010 22:05:40 +0000 (15:05 -0700)]
um: os-linux/mem.c needs sys/stat.h

commit fb967ecc584c20c74a007de749ca597068b0fcac upstream.

The os-linux/mem.c file calls fchmod function, which is declared in sys/stat.h
header file, so include it.  Fixes build breakage under FC13.

Signed-off-by: Liu Aleaxander <>
Acked-by: Boaz Harrosh <>
Cc: Jeff Dike <>
Signed-off-by: Andrew Morton <>
Signed-off-by: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agouml: fix CONFIG_STATIC_LINK=y build failure with newer glibc
Roland McGrath [Tue, 26 Oct 2010 21:22:19 +0000 (14:22 -0700)]
uml: fix CONFIG_STATIC_LINK=y build failure with newer glibc

commit aa5fb4dbfd121296ca97c68cf90043a7ea97579d upstream.

With glibc 2.11 or later that was built with --enable-multi-arch, the UML
link fails with undefined references to __rel_iplt_start and similar
symbols.  In recent binutils, the default linker script defines these
symbols (see ld --verbose).  Fix the UML linker scripts to match the new
defaults for these sections.

Signed-off-by: Roland McGrath <>
Cc: Jeff Dike <>
Cc: Al Viro <>
Signed-off-by: Andrew Morton <>
Signed-off-by: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoUSB: don't let the hub driver prevent system sleep
Alan Stern [Wed, 15 Jun 2011 20:29:16 +0000 (16:29 -0400)]
USB: don't let the hub driver prevent system sleep

commit cbb330045e5df8f665ac60227ff898421fc8fb92 upstream.

This patch (as1465) continues implementation of the policy that errors
during suspend or hibernation should not prevent the system from going
to sleep.

In this case, failure to turn on the Suspend feature for a hub port
shouldn't be reported as an error.  There are situations where this
does actually occur (such as when the device plugged into that port
was disconnected in the recent past), and it turns out to be harmless.
There's no reason for it to prevent a system sleep.

Also, don't allow the hub driver to fail a system suspend if the
downstream ports aren't all suspended.  This is also harmless (and
should never happen, given the change mentioned above); printing a
warning message in the kernel log is all we really need to do.

Signed-off-by: Alan Stern <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoUSB: don't let errors prevent system sleep
Alan Stern [Wed, 15 Jun 2011 20:27:43 +0000 (16:27 -0400)]
USB: don't let errors prevent system sleep

commit 0af212ba8f123c2eba151af7726c34a50b127962 upstream.

This patch (as1464) implements the recommended policy that most errors
during suspend or hibernation should not prevent the system from going
to sleep.  In particular, failure to suspend a USB driver or a USB
device should not prevent the sleep from succeeding:

Failure to suspend a device won't matter, because the device will
automatically go into suspend mode when the USB bus stops carrying
packets.  (This might be less true for USB-3.0 devices, but let's not
worry about them now.)

Failure of a driver to suspend might lead to trouble later on when the
system wakes up, but it isn't sufficient reason to prevent the system
from going to sleep.

Signed-off-by: Alan Stern <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agotaskstats: don't allow duplicate entries in listener mode
Vasiliy Kulikov [Mon, 27 Jun 2011 23:18:11 +0000 (16:18 -0700)]
taskstats: don't allow duplicate entries in listener mode

commit 26c4caea9d697043cc5a458b96411b86d7f6babd upstream.

Currently a single process may register exit handlers unlimited times.
It may lead to a bloated listeners chain and very slow process

kernel memory is stolen for the handlers chain and "time id" shows 2-7
seconds instead of normal 0.003.  It makes it possible to exhaust all
kernel memory and to eat much of CPU time by triggerring numerous exits
on a single CPU.

The patch limits the number of times a single process may register
itself on a single CPU to one.

One little issue is kept unfixed - as taskstats_exit() is called before
exit_files() in do_exit(), the orphaned listener entry (if it was not
explicitly deregistered) is kept until the next someone's exit() and
implicit deregistration in send_cpu_listeners().  So, if a process
registered itself as a listener exits and the next spawned process gets
the same pid, it would inherit taskstats attributes.

Signed-off-by: Vasiliy Kulikov <>
Cc: Balbir Singh <>
Signed-off-by: Andrew Morton <>
Signed-off-by: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>
7 years ago6pack,mkiss: fix lock inconsistency
Arnd Bergmann [Sat, 2 Jul 2011 00:30:00 +0000 (17:30 -0700)]
6pack,mkiss: fix lock inconsistency

commit 6e4e2f811bade330126d4029c88c831784a7efd9 upstream.

Lockdep found a locking inconsistency in the mkiss_close function:

> kernel: [ INFO: inconsistent lock state ]
> kernel: #3
> kernel: ---------------------------------
> kernel: inconsistent {IN-SOFTIRQ-R} -> {SOFTIRQ-ON-W} usage.
> kernel: ax25ipd/2813 [HC0[0]:SC0[0]:HE1:SE1] takes:
> kernel: (disc_data_lock){+++?.-}, at: [<ffffffffa018552b>] mkiss_close+0x1b/0x90 [mkiss]
> kernel: {IN-SOFTIRQ-R} state was registered at:

The message hints that disc_data_lock is aquired with softirqs disabled,
but does not itself disable softirqs, which can in rare circumstances
lead to a deadlock.
The same problem is present in the 6pack driver, this patch fixes both
by using write_lock_bh instead of write_lock.

Reported-by: Bernard F6BVP <>
Tested-by: Bernard F6BVP <>
Signed-off-by: Arnd Bergmann <>
Acked-by: Ralf Baechle<>
Signed-off-by: David S. Miller <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoSUNRPC: Ensure the RPC client only quits on fatal signals
Trond Myklebust [Fri, 17 Jun 2011 14:14:59 +0000 (10:14 -0400)]
SUNRPC: Ensure the RPC client only quits on fatal signals

commit 5afa9133cfe67f1bfead6049a9640c9262a7101c upstream.

Fix a couple of instances where we were exiting the RPC client on
arbitrary signals. We should only do so on fatal signals.

Signed-off-by: Trond Myklebust <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agomd: avoid endless recovery loop when waiting for fail device to complete.
NeilBrown [Tue, 28 Jun 2011 06:59:42 +0000 (16:59 +1000)]
md: avoid endless recovery loop when waiting for fail device to complete.

commit 4274215d24633df7302069e51426659d4759c5ed upstream.

If a device fails in a way that causes pending request to take a while
to complete, md will not be able to immediately remove it from the
array in remove_and_add_spares.
It will then incorrectly look like a spare device and md will try to
recover it even though it is failed.
This leads to a recovery process starting and instantly aborting over
and over again.

We should check if the device is faulty before considering it to be a
spare.  This will avoid trying to start a recovery that cannot

This bug was introduced in 2.6.26 so that patch is suitable for any
kernel since then.

Reported-by: Jim Paradis <>
Signed-off-by: NeilBrown <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoi2c-taos-evm: Fix log messages
Jean Delvare [Wed, 29 Jun 2011 09:36:10 +0000 (11:36 +0200)]
i2c-taos-evm: Fix log messages

commit 9b640f2e154268cb516efcaf9c434f2e73c6783e upstream.

* Print all error and information messages even when debugging is
* Don't use adapter device to log messages before it is ready.

Signed-off-by: Jean Delvare <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agocfq-iosched: fix a rcu warning
Shaohua Li [Mon, 27 Jun 2011 07:03:47 +0000 (09:03 +0200)]
cfq-iosched: fix a rcu warning

commit 3181faa85bda3dc3f5e630a1846526c9caaa38e3 upstream.

I got a rcu warnning at boot. the ioc->ioc_data is rcu_deferenced, but
doesn't hold rcu_read_lock.

Signed-off-by: Shaohua Li <>
Signed-off-by: Jens Axboe <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agocfq-iosched: fix locking around ioc->ioc_data assignment
Jens Axboe [Sun, 5 Jun 2011 04:01:13 +0000 (06:01 +0200)]
cfq-iosched: fix locking around ioc->ioc_data assignment

commit ab4bd22d3cce6977dc039664cc2d052e3147d662 upstream.

Since we are modifying this RCU pointer, we need to hold
the lock protecting it around it.

This fixes a potential reuse and double free of a cfq
io_context structure. The bug has been in CFQ for a long
time, it hit very few people but those it did hit seemed
to see it a lot.

Tracked in RH bugzilla here:

Credit goes to Paul Bolle for figuring out that the issue
was around the one-hit ioc->ioc_data cache. Thanks to his
hard work the issue is now fixed.

Signed-off-by: Jens Axboe <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agodebugobjects: Fix boot crash when kmemleak and debugobjects enabled
Marcin Slusarz [Sat, 28 May 2011 11:23:42 +0000 (13:23 +0200)]
debugobjects: Fix boot crash when kmemleak and debugobjects enabled

commit 161b6ae0e067e421b20bb35caf66bdb405c929ac upstream.

Order of initialization look like this:
...(lots of other subsystems)...
workqueues (through early initcall)

debugobjects use schedule_work for batch freeing of its data and kmemleak
heavily use debugobjects, so when it comes to freeing and workqueues were
not initialized yet, kernel crashes:

BUG: unable to handle kernel NULL pointer dereference at           (null)
IP: [<ffffffff810854d1>] __queue_work+0x29/0x41a
 [<ffffffff81085910>] queue_work_on+0x16/0x1d
 [<ffffffff81085abc>] queue_work+0x29/0x55
 [<ffffffff81085afb>] schedule_work+0x13/0x15
 [<ffffffff81242de1>] free_object+0x90/0x95
 [<ffffffff81242f6d>] debug_check_no_obj_freed+0x187/0x1d3
 [<ffffffff814b6504>] ? _raw_spin_unlock_irqrestore+0x30/0x4d
 [<ffffffff8110bd14>] ? free_object_rcu+0x68/0x6d
 [<ffffffff8110890c>] kmem_cache_free+0x64/0x12c
 [<ffffffff8110bd14>] free_object_rcu+0x68/0x6d
 [<ffffffff810b58bc>] __rcu_process_callbacks+0x1b6/0x2d9

because system_wq is NULL.

Fix it by checking if workqueues susbystem was initialized before using.

Signed-off-by: Marcin Slusarz <>
Cc: Catalin Marinas <>
Cc: Tejun Heo <>
Cc: Dipankar Sarma <>
Cc: Paul E. McKenney <>
Signed-off-by: Thomas Gleixner <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agowatchdog: mtx1-wdt: request gpio before using it
Florian Fainelli [Wed, 15 Jun 2011 17:15:23 +0000 (19:15 +0200)]
watchdog: mtx1-wdt: request gpio before using it

commit 9b19d40aa3ebaf1078779da10555da2ab8512422 upstream.

Otherwise, the gpiolib autorequest feature will produce a WARN_ON():

WARNING: at drivers/gpio/gpiolib.c:101 0x8020ec6c()
autorequest GPIO-215

Signed-off-by: Florian Fainelli <>
Signed-off-by: Wim Van Sebroeck <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agouvcvideo: Remove buffers from the queues when freeing
Sjoerd Simons [Tue, 24 May 2011 15:22:03 +0000 (12:22 -0300)]
uvcvideo: Remove buffers from the queues when freeing

commit 8ca2c80b170c47eeb55f0c2a0f2b8edf85f35d49 upstream.

When freeing memory for the video buffers also remove them from the
irq & main queues.

This fixes an oops when doing the following:

open ("/dev/video", ..)
close ()

As the second VIDIOC_REQBUFS will cause the list entries of the buffers
to be cleared while they still hang around on the main and irc queues

Signed-off-by: Sjoerd Simons <>
Acked-by: Laurent Pinchart <>
Signed-off-by: Mauro Carvalho Chehab <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agomm: fix negative commitlimit when gigantic hugepages are allocated
Rafael Aquini [Wed, 15 Jun 2011 22:08:39 +0000 (15:08 -0700)]
mm: fix negative commitlimit when gigantic hugepages are allocated

commit b0320c7b7d1ac1bd5c2d9dff3258524ab39bad32 upstream.

When 1GB hugepages are allocated on a system, free(1) reports less
available memory than what really is installed in the box.  Also, if the
total size of hugepages allocated on a system is over half of the total
memory size, CommitLimit becomes a negative number.

The problem is that gigantic hugepages (order > MAX_ORDER) can only be
allocated at boot with bootmem, thus its frames are not accounted to
'totalram_pages'.  However, they are accounted to hugetlb_total_pages()

What happens to turn CommitLimit into a negative number is this
calculation, in fs/proc/meminfo.c:

        allowed = ((totalram_pages - hugetlb_total_pages())
                * sysctl_overcommit_ratio / 100) + total_swap_pages;

A similar calculation occurs in __vm_enough_memory() in mm/mmap.c.

Also, every vm statistic which depends on 'totalram_pages' will render
confusing values, as if system were 'missing' some part of its memory.

Impact of this bug:

When gigantic hugepages are allocated and sysctl_overcommit_memory ==
OVERCOMMIT_NEVER.  In a such situation, __vm_enough_memory() goes through
the mentioned 'allowed' calculation and might end up mistakenly returning
-ENOMEM, thus forcing the system to start reclaiming pages earlier than it
would be ususal, and this could cause detrimental impact to overall
system's performance, depending on the workload.

Besides the aforementioned scenario, I can only think of this causing
annoyances with memory reports from /proc/meminfo and free(1).

[ standardize comment layout]
Reported-by: Russ Anderson <>
Signed-off-by: Rafael Aquini <>
Acked-by: Russ Anderson <>
Cc: Andrea Arcangeli <>
Cc: Christoph Lameter <>
Signed-off-by: Andrew Morton <>
Signed-off-by: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoath5k: fix memory leak when fewer than N_PD_CURVES are in use
Eugene A. Shatokhin [Wed, 29 Jun 2011 03:04:51 +0000 (23:04 -0400)]
ath5k: fix memory leak when fewer than N_PD_CURVES are in use

commit a0b8de350be458b33248e48b2174d9af8a4c4798 upstream.

We would free the proper number of curves, but in the wrong
slots, due to a missing level of indirection through
the pdgain_idx table.

It's simpler just to try to free all four slots, so do that.

Signed-off-by: Bob Copeland <>
Signed-off-by: John W. Linville <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoPM: Free memory bitmaps if opening /dev/snapshot fails
Michal Kubecek [Sat, 18 Jun 2011 18:34:01 +0000 (20:34 +0200)]
PM: Free memory bitmaps if opening /dev/snapshot fails

commit 8440f4b19494467883f8541b7aa28c7bbf6ac92b upstream.

When opening /dev/snapshot device, snapshot_open() creates memory
bitmaps which are freed in snapshot_release(). But if any of the
callbacks called by pm_notifier_call_chain() returns NOTIFY_BAD, open()
fails, snapshot_release() is never called and bitmaps are not freed.
Next attempt to open /dev/snapshot then triggers BUG_ON() check in
create_basic_memory_bitmaps(). This happens e.g. when vmwatchdog module
is active on s390x.

Signed-off-by: Michal Kubecek <>
Signed-off-by: Rafael J. Wysocki <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoxhci: Reject double add of active endpoints.
Sarah Sharp [Mon, 6 Jun 2011 06:10:04 +0000 (23:10 -0700)]
xhci: Reject double add of active endpoints.

commit fa75ac379e63c2864e9049b5e8615e40f65c1e70 upstream.

While trying to switch a UAS device from the BOT configuration to the UAS
configuration via the bConfigurationValue file, Tanya ran into an issue in
the USB core.  usb_disable_device() sets entries in udev->ep_out and
udev->ep_out to NULL, but doesn't call into the xHCI bandwidth management
functions to remove the BOT configuration endpoints from the xHCI host's
internal structures.

The USB core would then attempt to add endpoints for the UAS
configuration, and some of the endpoints had the same address as endpoints
in the BOT configuration.  The xHCI driver blindly added the endpoints
again, but the xHCI host controller rejected the Configure Endpoint
command because active endpoints were added without being dropped.

Make the xHCI driver reject calls to xhci_add_endpoint() that attempt to
add active endpoints without first calling xhci_drop_endpoint().

This should be backported to kernels as old as 2.6.31.

Signed-off-by: Sarah Sharp <>
Reported-by: Tanya Brokhman <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoTTY: ldisc, do not close until there are readers
Jiri Slaby [Sun, 5 Jun 2011 12:16:16 +0000 (14:16 +0200)]
TTY: ldisc, do not close until there are readers

commit 92f6fa09bd453ffe3351fa1f1377a1b7cfa911e6 upstream.

We restored tty_ldisc_wait_idle in 100eeae2c5c (TTY: restore
tty_ldisc_wait_idle). We used it in the ldisc changing path to fix the
case where there are tasks in n_tty_read waiting for data and somebody
tries to change ldisc.

Similar to the case above, there may be also tasks waiting in
n_tty_read while hangup is performed. As 65b770468e98 (tty-ldisc: turn
ldisc user count into a proper refcount) removed the wait-until-idle
from all paths, hangup path won't wait for them to disappear either
now. So add it back even to the hangup path.

There is a difference, we need uninterruptible sleep as there is
obviously HUP signal pending. So tty_ldisc_wait_idle now sleeps
without possibility to be interrupted. This is what original
tty_ldisc_wait_idle did. After the wait idle reintroduction
(100eeae2c5c), we have had interruptible sleeps for the ldisc changing
path. But as there is a 5s timeout anyway, we don't allow it to be
interrupted from now on. It's not worth the added complexity of
deciding what kind of sleep we want.

Before 65b770468e98 tty_ldisc_release was called also from
tty_ldisc_release. It is called from tty_release, so I don't think we
need to restore that one.

This is nicely reproducible after constifying the timing when
drivers/tty/n_tty.c is patched as follows ("TTY: ntty, add one more
sanity check" patch is needed to actually see it explode):
%% -1548,6 +1549,7 @@ static int n_tty_open(struct tty_struct *tty)

        /* These are ugly. Currently a malloc failure here can panic */
        if (!tty->read_buf) {
+               msleep(100);
                tty->read_buf = kzalloc(N_TTY_BUF_SIZE, GFP_KERNEL);
                if (!tty->read_buf)
                        return -ENOMEM;
%% -1785,6 +1788,7 @@ do_it_again:
                        timeout = schedule_timeout(timeout);
+                       msleep(20);
===== With a process: =====
    while (1) {
        int fd = open(argv[1], O_RDWR);
        read(fd, buf, sizeof(buf));
===== and its child: =====
        while (1) {
                int fd = open(tty, O_RDWR|O_NOCTTY);
                ioctl(fd, TIOCSCTTY, 1);
                usleep(100 * (10 + random() % 1000));
===== EOF =====

Signed-off-by: Jiri Slaby <>
Cc: Alan Cox <>
Cc: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoclocksource: Make watchdog robust vs. interruption
Thomas Gleixner [Thu, 16 Jun 2011 14:22:08 +0000 (16:22 +0200)]
clocksource: Make watchdog robust vs. interruption

commit b5199515c25cca622495eb9c6a8a1d275e775088 upstream.

The clocksource watchdog code is interruptible and it has been
observed that this can trigger false positives which disable the TSC.

The reason is that an interrupt storm or a long running interrupt
handler between the read of the watchdog source and the read of the
TSC brings the two far enough apart that the delta is larger than the
unstable treshold. Move both reads into a short interrupt disabled
region to avoid that.

Reported-and-tested-by: Vernon Mauery <>
Signed-off-by: Thomas Gleixner <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoxen: partially revert "xen: set max_pfn_mapped to the last pfn mapped"
Stefano Stabellini [Fri, 3 Jun 2011 09:51:34 +0000 (09:51 +0000)]
xen: partially revert "xen: set max_pfn_mapped to the last pfn mapped"

commit a91d92875ee94e4703fd017ccaadb48cfb344994 upstream.

We only need to set max_pfn_mapped to the last pfn mapped on x86_64 to
make sure that cleanup_highmap doesn't remove important mappings at

We don't need to do this on x86_32 because cleanup_highmap is not called
on x86_32. Besides lowering max_pfn_mapped on x86_32 has the unwanted
side effect of limiting the amount of memory available for the 1:1
kernel pagetable allocation.

This patch reverts the x86_32 part of the original patch.

Signed-off-by: Stefano Stabellini <>
Signed-off-by: Konrad Rzeszutek Wilk <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agomigrate: don't account swapcache as shmem
Andrea Arcangeli [Thu, 16 Jun 2011 19:56:19 +0000 (12:56 -0700)]
migrate: don't account swapcache as shmem

commit 99a15e21d96f6857dafab1e5167e5e8183215c9c upstream.

swapcache will reach the below code path in migrate_page_move_mapping,
and swapcache is accounted as NR_FILE_PAGES but it's not accounted as

Hugh pointed out we must use PageSwapCache instead of comparing
mapping to &swapper_space, to avoid build failure with CONFIG_SWAP=n.

Signed-off-by: Andrea Arcangeli <>
Acked-by: Hugh Dickins <>
Signed-off-by: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>
7 years agoksm: fix NULL pointer dereference in scan_get_next_rmap_item()
Hugh Dickins [Wed, 15 Jun 2011 22:08:58 +0000 (15:08 -0700)]
ksm: fix NULL pointer dereference in scan_get_next_rmap_item()

commit 2b472611a32a72f4a118c069c2d62a1a3f087afd upstream.

Andrea Righi reported a case where an exiting task can race against
ksmd::scan_get_next_rmap_item ( easily
triggering a NULL pointer dereference in ksmd.

ksm_scan.mm_slot == &ksm_mm_head with only one registered mm

CPU 1 (__ksm_exit) CPU 2 (scan_get_next_rmap_item)
  list_empty() is false
lock slot == &ksm_mm_head
(list now empty)
slot = list_entry(slot->
(list is empty, so slot is still ksm_mm_head)
slot->mm == NULL ... Oops

Close this race by revalidating that the new slot is not simply the list
head again.

Andrea's test case:

#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <sys/mman.h>

#define BUFSIZE getpagesize()

int main(int argc, char **argv)
void *ptr;

if (posix_memalign(&ptr, getpagesize(), BUFSIZE) < 0) {
if (madvise(ptr, BUFSIZE, MADV_MERGEABLE) < 0) {
*(char *)NULL = 0;

return 0;

Reported-by: Andrea Righi <>
Tested-by: Andrea Righi <>
Cc: Andrea Arcangeli <>
Signed-off-by: Hugh Dickins <>
Signed-off-by: Chris Wright <>
Signed-off-by: Andrew Morton <>
Signed-off-by: Linus Torvalds <>
Signed-off-by: Greg Kroah-Hartman <>
7 years ago- patches.fixes/md-bitmap-flush-when-readonly: Flush the md
NeilBrown [Wed, 13 Jul 2011 03:13:18 +0000 (13:13 +1000)]
- patches.fixes/md-bitmap-flush-when-readonly: Flush the md
  bitmap when switching to read-only (bnc#701977).

suse-commit: 920b60d5c97816cef23cbaeccee289a2f786d160

7 years agoMerge branch 'SLE11-SP2' of into SLE11-SP2
Tony Jones [Tue, 12 Jul 2011 20:51:43 +0000 (22:51 +0200)]
Merge branch 'SLE11-SP2' of into SLE11-SP2

suse-commit: fb24bcbc70b7e4c8211dcb7306f4cfd731eae6f6

7 years ago- patches.suse/add-target-mode-infrastructure.patch:
Hannes Reinecke [Tue, 12 Jul 2011 14:17:05 +0000 (16:17 +0200)]
- patches.suse/add-target-mode-infrastructure.patch:
  Fixup build failure on utsrelease.

suse-commit: 3c2ce70251c4277ba8ecb0a59848bcb048d92ddc

7 years ago- Update config files.
Hannes Reinecke [Tue, 12 Jul 2011 13:44:46 +0000 (15:44 +0200)]
- Update config files.
- patches.suse/add-target-mode-infrastructure.patch:
  drivers/target: Add target-mode infrastructure
- patches.suse/target-Backport-for-SLES11-SP2.patch: target:
  Backport for SLES11 SP2 (FATE#311801,bnc#687068).
- patches.xen/xen3-auto-common.diff: Refresh.
- supported.conf: Mark target core modules as supported.

suse-commit: 2ab0560efe95877d0bc14adf333df5ebcf933112

7 years agoMerge branch 'SLE11-SP1' into SLE11-SP2
Michal Marek [Tue, 12 Jul 2011 11:12:25 +0000 (13:12 +0200)]
Merge branch 'SLE11-SP1' into SLE11-SP2

Throw away patches.drivers/netxen-support-for-GbE-port-settings.patch
which is already applied in SP2.

suse-commit: 33efccc459a15010b2ff560239b0dd124f33def9

7 years ago- patches.fixes/memcg-bugfix-check-mem_cgroup_disabled-at-split-fixu.patch:
Michal Hocko [Tue, 12 Jul 2011 08:54:33 +0000 (10:54 +0200)]
- patches.fixes/memcg-bugfix-check-mem_cgroup_disabled-at-split-fixu.patch:
  memcg: bugfix check mem_cgroup_disabled() at split fixup
- patches.fixes/memcg-change-page_cgroup_zoneinfo-signature.patch:
  memcg: change page_cgroup_zoneinfo signature (bnc#704592).
- patches.fixes/memcg-charged-pages-always-have-valid-per-memcg-zone.patch:
  memcg: charged pages always have valid per-memcg zone info
- patches.fixes/memcg-condense-page_cgroup-to-page-lookup-points.patch:
  memcg: condense page_cgroup-to-page lookup points (bnc#704592).
- patches.fixes/memcg-count-the-soft_limit-reclaim-in-global-backgro.patch:
  memcg: count the soft_limit reclaim in global background reclaim
- patches.fixes/memcg-css_id-must-be-called-under-rcu_read_lock.patch:
  memcg: css_id() must be called under rcu_read_lock()
- patches.fixes/memcg-fix-css_id-RCU-locking-for-real.patch:
  memcg: fix css_id() RCU locking for real (bnc#704592).
- patches.fixes/memcg-fix-css_is_ancestor-RCU-locking.patch:
  memcg: fix css_is_ancestor() RCU locking (bnc#704592).
- patches.fixes/memcg-fix-false-positive-VM_BUG-on-non-SMP.patch:
  memcg: fix false positive VM_BUG on non-SMP (bnc#704592).
- patches.fixes/memcg-fix-memory-migration-of-shmem-swapcache.patch:
  memcg: fix memory migration of shmem swapcache (bnc#704592).
- patches.fixes/memcg-fix-mis-accounting-of-file-mapped-racy-with-mi.patch:
  memcg: fix mis-accounting of file mapped racy with migration
- patches.fixes/memcg-fix-off-by-one-when-calculating-swap-cgroup-ma.patch:
  memcg: fix off-by-one when calculating swap cgroup map length
- patches.fixes/memcg-fix-race-in-file_mapped-accounting.patch:
  memcg: fix race in file_mapped accounting (bnc#704592).
- patches.fixes/memcg-fix-race-in-file_mapped-accouting-flag-managem.patch:
  memcg: fix race in file_mapped accouting flag management
- patches.fixes/memcg-fix-wake-up-in-oom-wait-queue.patch: memcg:
  fix wake up in oom wait queue (bnc#704592).
- patches.fixes/memcg-give-current-access-to-memory-reserves-if-it-s.patch:
  memcg: give current access to memory reserves if it's trying
  to die (bnc#704592).
- patches.fixes/memcg-kill-unnecessary-initialization-in-mem_cgroup_.patch:
  memcg: kill unnecessary initialization in
  mem_cgroup_shrink_node_zone() (bnc#704592).
- patches.fixes/memcg-make-oom-killer-a-no-op-when-no-killable-task-.patch:
  memcg: make oom killer a no-op when no killable task can be
  found (bnc#704592).
- patches.fixes/memcg-mark-init_section_page_cgroup-properly.patch:
  memcg: mark init_section_page_cgroup() properly (bnc#704592).
- patches.fixes/memcg-more-mem_cgroup_uncharge-batching.patch:
  memcg: more mem_cgroup_uncharge() batching (bnc#704592).
- patches.fixes/memcg-move-memcg-reclaimable-page-into-tail-of-inact.patch:
  memcg: move memcg reclaimable page into tail of inactive list
- patches.fixes/memcg-move-page-freeing-code-out-of-lock.patch:
  memcg: move page-freeing code out of lock (bnc#704592).
- patches.fixes/memcg-never-OOM-when-charging-huge-pages.patch:
  memcg: never OOM when charging huge pages (bnc#704592).
- patches.fixes/memcg-no-uncharged-pages-reach-page_cgroup_zoneinfo.patch:
  memcg: no uncharged pages reach page_cgroup_zoneinfo
- patches.fixes/memcg-prevent-endless-loop-when-charging-huge-pages-.patch:
  memcg: prevent endless loop when charging huge pages to
  near-limit group (bnc#704592).
- patches.fixes/memcg-remove-NULL-check-from-lookup_page_cgroup-resu.patch:
  memcg: remove NULL check from lookup_page_cgroup() result
- patches.fixes/memcg-remove-direct-page_cgroup-to-page-pointer.patch:
  memcg: remove direct page_cgroup-to-page pointer (bnc#704592).
- patches.fixes/memcg-remove-memcg-reclaim_param_lock.patch:
  memcg: remove memcg->reclaim_param_lock (bnc#704592).
- patches.fixes/memcg-remove-unused-page-flag-bitfield-defines.patch:
  memcg: remove unused page flag bitfield defines (bnc#704592).
- patches.fixes/memcg-rework-usage-of-stats-by-soft-limit.patch:
  memcg: rework usage of stats by soft limit (bnc#704592).
- patches.fixes/memcg-sc.nr_to_reclaim-should-be-initialized.patch:
  memcg: sc.nr_to_reclaim should be initialized (bnc#704592).
- patches.fixes/memcg-soft-limit-reclaim-should-end-at-limit-not-bel.patch:
  memcg: soft limit reclaim should end at limit not below
- patches.fixes/memcg-use-for_each_mem_cgroup.patch: memcg:
  use for_each_mem_cgroup (bnc#704592).
- patches.fixes/mm-thp-optimize-memcg-charge-in-khugepaged.patch:
  mm: thp: optimize memcg charge in khugepaged (bnc#704592).
- patches.suse/cgroups-blkio-subsystem-as-module.patch: Refresh.

suse-commit: 65e6d85e30e4705e778cf093e38537db66e3efc4

7 years ago- patches.fixes/drm_i85x_watermark_fix.patch:
Egbert Eich [Tue, 12 Jul 2011 04:21:54 +0000 (04:21 +0000)]
- patches.fixes/drm_i85x_watermark_fix.patch:
  IS_MOBILE() catches 85x, so we'd always try to use the 9xx FIFO sizing;
  since there's an explicit 85x version, this seems wrong.
  v2: Handle 830m correctly too.
- patches.fixes/drm_i85x_chip_id_fix.patch:
  pci.ids and the datasheet both say it's 358e, not 35e8.
  (bnc #677443).

suse-commit: f9a54fdbf60c81f02d3960b7e737569d0c8227e9

7 years ago- patches.fixes/perf-eliminate-getimeofday.patch: perf record:
Tony Jones [Mon, 11 Jul 2011 21:25:35 +0000 (23:25 +0200)]
- patches.fixes/perf-eliminate-getimeofday.patch: perf record:
  remove unneeded gettimeofday() call (bnc#607339).

suse-commit: 97e67dc990b6a5586716afb1ff1a77be74937328

7 years ago- patches.suse/memcg-unify-charge-uncharge-quantities-to-units-of-pages.patch:
Michal Hocko [Mon, 11 Jul 2011 14:26:27 +0000 (16:26 +0200)]
- patches.suse/memcg-unify-charge-uncharge-quantities-to-units-of-pages.patch:
  Refresh. Use hpage_nr_pages rather than compound_order (bnc#704592)

suse-commit: 6d90f334b62aa90995da0bb8a544e4fb26f13e47

7 years agoCleanup: Remove patches already added by others or which are not needed:
Thomas Renninger [Mon, 11 Jul 2011 13:59:05 +0000 (15:59 +0200)]
Cleanup: Remove patches already added by others or which are not needed:
- patches.arch/012-intel_idle__Change_mode_755__644.patch: Delete.
- patches.drivers/2009-10-12-0001-mce-edac-Use-an-atomic-notifier-for-MCEs-decoding.patch:
- patches.drivers/2009-10-12-0003-x86-Export-k8-physical-topology.patch:
- patches.drivers/Modified-2010-03-08-0001-Driver-core-Constify-struct-sysfs_ops-in-struct-kob.patch:
- patches.drivers/Modified-2010-05-25-0001-kernel.h-add-pr_warn-for-symmetry-to-dev_warn-netd.patch:

suse-commit: 6ddb1ab21b9fd783524d3e241b49bc7d230c8042

7 years ago- patches.drivers/netxen-support-for-GbE-port-settings.patch:
Jean Delvare [Mon, 11 Jul 2011 13:51:07 +0000 (15:51 +0200)]
- patches.drivers/netxen-support-for-GbE-port-settings.patch:
  netxen: support for GbE port settings (bnc#703490).

suse-commit: 0aa118540539acb636787b69124a64b488baf80c

7 years ago- Remove unused patches:
Hannes Reinecke [Mon, 11 Jul 2011 13:12:41 +0000 (15:12 +0200)]
- Remove unused patches:
  * patches.suse/blk-queue-unprep-fn
  * patches.suse/gfs2-generalize-flags-for-blkdev_issue_fn-functio.patch
  * patches.suse/scsi-Retry-commands-with-UNIT_ATTENTION-sense-code.patch
  * patches.suse/staging-hv-host-lock-push-down.patch

suse-commit: 5350ae406f7135132e589d6aae4dcdc3ce942864

7 years ago- patches.arch/s390-sles11sp2-kmsg.patch: kmsg: add descriptions
John Jolly [Mon, 11 Jul 2011 05:30:22 +0000 (07:30 +0200)]
- patches.arch/s390-sles11sp2-kmsg.patch: kmsg: add descriptions
  for new messages (bnc#702651,LTC#72829).

suse-commit: c406731b5fc5486d1acbb6d3ab6437e8c1f07dd4

7 years agoMerge branch 'SLE11-SP1' into SLE11-SP2
Michal Marek [Sat, 9 Jul 2011 05:00:32 +0000 (07:00 +0200)]
Merge branch 'SLE11-SP1' into SLE11-SP2

suse-commit: d378e44d02e63ee43d4057765c4880025d281473

7 years ago- patches.fixes/sched-divide-by-zero-workaround-and-diag.diff:
Mike Galbraith [Sat, 9 Jul 2011 02:28:23 +0000 (04:28 +0200)]
- patches.fixes/sched-divide-by-zero-workaround-and-diag.diff:

suse-commit: 425886b2e0f44b7cb6691ecd1da693b70dc878cc

7 years ago- patches.suse/ppc-temporary-bugon-timer-workaround.patch:
Jiri Kosina [Fri, 8 Jul 2011 15:55:11 +0000 (17:55 +0200)]
- patches.suse/ppc-temporary-bugon-timer-workaround.patch:

suse-commit: d5f129789b2ae29ef6b2b830cfbe9d846d5a5ecc

7 years ago- patches.suse/ppc-temporary-bugon-timer-workaround.patch: ppc:
Jiri Kosina [Fri, 8 Jul 2011 15:15:09 +0000 (17:15 +0200)]
- patches.suse/ppc-temporary-bugon-timer-workaround.patch: ppc:
  temporary workaround for ppc timer BUG_ON() (bnc#702384).

suse-commit: 7741b4986aff0f233184daded3a9c4d19d30e197