GitHub/MotorolaMobilityLLC/kernel-slsi.git
13 years agotarget: remove TRANSPORT_DEFERRED_CMD state
Christoph Hellwig [Mon, 17 Oct 2011 17:56:51 +0000 (13:56 -0400)]
target: remove TRANSPORT_DEFERRED_CMD state

We never check for this state, and it makes testing for a completed
state much harder given that it overrides the existing state.

Also remove the unused deferred_t_state which is related to it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: remove the TRANSPORT_REMOVE state
Christoph Hellwig [Mon, 17 Oct 2011 17:56:50 +0000 (13:56 -0400)]
target: remove the TRANSPORT_REMOVE state

We never queue an command with this state, and only set it in a completely
bogus place in tcm_fc.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: move depth_left manipulation out of transport_generic_request_failure
Christoph Hellwig [Mon, 17 Oct 2011 17:56:49 +0000 (13:56 -0400)]
target: move depth_left manipulation out of transport_generic_request_failure

We only need to decrement dev->depth_left if failing a command from
__transport_execute_tasks.  Instead of doing it first thing in
transport_generic_request_failure and requiring a pseudo-flag argument
for it just opencode the decrement in the two callers (which should
be factored into a single one anyway)

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: stop task timers earlier
Christoph Hellwig [Mon, 17 Oct 2011 17:56:48 +0000 (13:56 -0400)]
target: stop task timers earlier

Currently we stop the timers for all tasks in a command fairly late during
I/O completion, which is fairly pointless and requires all kinds of safety
checks.

Instead delete pending timers early on in transport_complete_task, thus
ensuring no new timers firest after that.  We take t_state_lock a bit later
in that function thus making sure currenly running timers are out of the
criticial section.  To be completely sure the timer has finished we also
add another del_timer_sync call when freeing the task.

This also allows removing TF_TIMER_RUNNING as it would be equivalent
to TF_ACTIVE now.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: remove TF_TIMER_STOP
Christoph Hellwig [Mon, 17 Oct 2011 17:56:47 +0000 (13:56 -0400)]
target: remove TF_TIMER_STOP

TF_TIMER_STOP is useless as it only helps to mitigate a tiny race during
deleting the timer.  But given that we have cleared TF_ACTIVE at this point
we already have another mitigation a few lines down the function.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: factor some duplicate code for stopping a task
Christoph Hellwig [Mon, 17 Oct 2011 17:56:46 +0000 (13:56 -0400)]
target: factor some duplicate code for stopping a task

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: fix list walking in transport_free_dev_tasks
Christoph Hellwig [Mon, 17 Oct 2011 17:56:45 +0000 (13:56 -0400)]
target: fix list walking in transport_free_dev_tasks

list_for_each_entry_safe only protects against deletions from the list,
but not against any concurrent modifications.  Given that we drop
t_state_lock inside the loop it is not safe in transport_free_dev_tasks.

Instead of use a local dispose_list that we move all tasks that are
to be deleted to.  This is safe because we never do list_emptry checks
on t_list to check if a command is on the list anywhere.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: use transport_cmd_check_stop_to_fabric consistently
Christoph Hellwig [Mon, 17 Oct 2011 17:56:44 +0000 (13:56 -0400)]
target: use transport_cmd_check_stop_to_fabric consistently

Change one remaining user of transport_cmd_check_stop(cmd, 2, 0) to the
transport_cmd_check_stop_to_fabric wrapper.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: do not pass the queue object to transport_remove_cmd_from_queue
Christoph Hellwig [Mon, 17 Oct 2011 17:56:43 +0000 (13:56 -0400)]
target: do not pass the queue object to transport_remove_cmd_from_queue

We always operated on the same queue, so move finding it into the function,
just like we do for all other helpers operating on it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: remove SCF_EMULATE_QUEUE_FULL
Christoph Hellwig [Mon, 17 Oct 2011 17:56:42 +0000 (13:56 -0400)]
target: remove SCF_EMULATE_QUEUE_FULL

Add a new boolean at_head parameter to transport_add_cmd_to_queue and thus
obsolete the SCF_EMULATE_QUEUE_FULL flag.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: remove the transport_qf_callback se_cmd callback
Christoph Hellwig [Mon, 17 Oct 2011 17:56:41 +0000 (13:56 -0400)]
target: remove the transport_qf_callback se_cmd callback

Remove the need for the transport_qf_callback callback by making
sure we have specific states with specific handlers for the two
queue full cases.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: clean up the backend interface to caching parameters
Christoph Hellwig [Fri, 14 Oct 2011 11:30:17 +0000 (07:30 -0400)]
target: clean up the backend interface to caching parameters

Remove the dpo_emulated, fua_write_emulated, fua_read_emulated and
write_cache_emulated methods, and replace them with a simple bitfields in
se_subsystem_api in those cases where they ever returned one.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: make iblock_emulate_sync_cache asynchronous
Christoph Hellwig [Fri, 14 Oct 2011 11:29:58 +0000 (07:29 -0400)]
target: make iblock_emulate_sync_cache asynchronous

Do not block the submitting thread when handling a SYNCHRONIZE CACHE command,
but implement it asynchronously by sending the FLUSH command ourself and
calling transport_complete_sync_cache from the completion handler.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Fix REPORT TARGET PORT GROUPS handling with small allocation length
Nicholas Bellinger [Wed, 19 Oct 2011 06:48:04 +0000 (23:48 -0700)]
target: Fix REPORT TARGET PORT GROUPS handling with small allocation length

This patch fixes a bug with the handling of REPORT TARGET PORT GROUPS
containing a smaller allocation length than the payload requires causing
memory writes beyond the end of the buffer.  This patch checks for the
minimum 4 byte length for the response payload length, and also checks
upon each loop of T10_ALUA(su_dev)->tg_pt_gps_list to ensure the Target
port group and Target port descriptor list is able to fit into the
remaining allocation length.

If the response payload exceeds the allocation length length, then rd_len
is still increments to indicate to the initiator that the payload has
been truncated.

Reported-by: Roland Dreier <roland@purestorage.com>
Cc: stable@kernel.org
Signed-off-by: Nicholas Bellinger <nab@risingtidesystems.com>
13 years agotarget: remove the ->transport_split_cdb callback in se_cmd
Christoph Hellwig [Wed, 12 Oct 2011 15:09:13 +0000 (11:09 -0400)]
target: remove the ->transport_split_cdb callback in se_cmd

Add a switch statement implementing the CDB LBA/len update directly
in target_get_task_cdb and remove the old ->transport_split_cdb
callback and all its implementations.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: replace ->get_cdb with a target_get_task_cdb helper
Christoph Hellwig [Wed, 12 Oct 2011 15:09:12 +0000 (11:09 -0400)]
target: replace ->get_cdb with a target_get_task_cdb helper

Instead of calling out to the backends from the core to get a per-task
CDB and then modify it for the LBA/len pair used for this CDB provide
a helper that writes the adjusted CDB into a provided buffer and call
this method from ->do_task in pscsi.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: make the ->get_cdb method optional
Christoph Hellwig [Wed, 12 Oct 2011 15:09:11 +0000 (11:09 -0400)]
target: make the ->get_cdb method optional

The most commonly used file, iblock and rd backends have no use for
a per-task CDB and thus don't need a method to copy it into their
otherwise unused CDB fields.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: pack struct se_task more tightly
Christoph Hellwig [Wed, 12 Oct 2011 15:07:07 +0000 (11:07 -0400)]
target: pack struct se_task more tightly

Rearrange the fields in se_task to avoid holes.  Also increase the
flags field to 16 bits as we have the space for it, and this makes
adding new flags safer.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Remove unnecessary se_task members
Christoph Hellwig [Wed, 12 Oct 2011 15:07:04 +0000 (11:07 -0400)]
target: Remove unnecessary se_task members

This is a squashed version of the following unnecessary se_task structure
member removal patches:

target: remove the task_execute_queue field in se_task

    Instead of using a separate flag we can simply do list_emptry checks
    on t_execute_list if we make sure to always use list_del_init to remove
    a task from the list.  Also factor some duplicate code into a new
    __transport_remove_task_from_execute_queue helper.

target: remove the read-only task_no field in se_task

    The task_no field never was initialized and only used in debug printks,
    so kill it.

target: remove the task_padded_sg field in se_task

    This field is only check in one place and not actually needed there.

    Rationale:
    - transport_do_task_sg_chain asserts that we have task_sg_chaining
      set early on
    - we only make use of the sg_prev_nents field we calculate based on it
      if there is another sg list that gets chained onto this one, which
      never happens for the last (or only) task.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: make more use of the task_flags field in se_task
Christoph Hellwig [Wed, 12 Oct 2011 15:07:03 +0000 (11:07 -0400)]
target: make more use of the task_flags field in se_task

Replace various atomic_t variables that were mostly under t_state_lock
with new flags in task_flags.  Note that the execution error path
didn't take t_state_lock before, so add it there.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Cleanup unused se_task bits
Christoph Hellwig [Wed, 12 Oct 2011 15:07:00 +0000 (11:07 -0400)]
target: Cleanup unused se_task bits

This is a squashed version of the following se_task cleanup patches:

    target: remove the unused task_state_flags field in se_task
    target: remove the unused se_obj_ptr field in se_task
    target: remove the se_dev field in se_task

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Cleanup unused target_core_base.h bits
Christoph Hellwig [Wed, 12 Oct 2011 15:06:56 +0000 (11:06 -0400)]
target: Cleanup unused target_core_base.h bits

This is a squashed version of the following target_core_base.h
cleanup patches:

    target: remove the unused SHUTDOWN_SIGS defintion
    target: remove unused se_mem leftovers
    target: remove the unused map_func_t typedef
    target: move TRANSPORT_IOV_DATA_BUFFER to the iscsi-specific code

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agoloopback: Prevent uninitialized use of tl_tpg in tcm_loop_queuecommand
Nicholas Bellinger [Tue, 11 Oct 2011 02:44:05 +0000 (19:44 -0700)]
loopback: Prevent uninitialized use of tl_tpg in tcm_loop_queuecommand

This patch fixes a bug with tcm_loop where performing a scsi_host rescan was
causing an oops due to a received scsi_cmnd->device->id value not matching a
previously configured tcm_loop_tpg entry in tcm_loop_hba->tl_hba_tpgs[]
obtained from within tcm_loop_queuecommand() code.

This fix adds an explict check for tcm_loop_tpg->tl_hba in order to ensure
tcm_loop_make_naa_tpg() has already been invoked to initialize a given
tcm_loop_tpg entry, and also adds an explict clear of tcm_loop_tpg->tl_hba
from within the tcm_loop_drop_naa_tpg() release path.

This bug was manifesting itself with the following OOPs:

[176289.430909] BUG: unable to handle kernel NULL pointer dereference at 0000000000000090
[176289.431337] IP: [<ffffffffa0395617>] transport_processing_thread+0x1e3/0x794 [target_core_mod]
[176289.431399] PGD 22e9b067 PUD 23375067 PMD 0
[176289.431399] Oops: 0000 [#1] SMP
[176289.431815] CPU 1
[176289.431815] Modules linked in: tcm_loop target_core_stgt target_core_pscsi target_core_file target_core_iblock target_core_mod crc32c ib_cm ib_sa ib_mad ib_core qla2xxx scsi_tgt configfs fcoe libfcoe libfc scsi_transport_fc ipv6 iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi sr_mod cdrom sd_mod ata_piix libata e1000 mptspi mptscsih mptbase [last unloaded: target_core_mod]
[176289.431815]
[176289.431815] Pid: 12339, comm: LIO_iblock Tainted: G        W   3.1.0-rc8+
[176289.431815] RIP: 0010:[<ffffffffa0395617>]  [<ffffffffa0395617>] transport_processing_thread+0x1e3/0x794 [target_core_mod]
[176289.431815] RSP: 0018:ffff880023bfbe10  EFLAGS: 00010283
[176289.431815] RAX: 0000000000000000 RBX: ffff88002d600040 RCX: ffff88002d600108
[176289.431815] RDX: ffff88000c9e50bc RSI: 0000000000000246 RDI: 0000000000000246
[176289.431815] RBP: ffff880023bfbee0 R08: ffff88002d600108 R09: 0000000000000000
[176289.431815] R10: ffff88002fc8cc80 R11: ffffffff81671b60 R12: ffff88002d600108
[176289.431815] R13: ffff88000c9e4f38 R14: ffff88000c9e50b8 R15: 0000000000000000
[176289.431815] FS:  0000000000000000(0000) GS:ffff88002fc80000(0000) knlGS:0000000000000000
[176289.431815] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[176289.431815] CR2: 0000000000000090 CR3: 000000002a33f000 CR4: 00000000000006e0
[176289.431815] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[176289.431815] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[176289.431815] Process LIO_iblock (pid: 12339, threadinfo ffff880023bfa000, task ffff88002a2e0000)
[176289.431815] Stack:
[176289.431815]  0000000000011280 0000000000000246 ffff88002a2e0000 ffff880023a58900
[176289.431815]  ffff880023bfbed0 ffff880023bfa000 ffff880023bfa000 ffff88000c9e50d0
[176289.431815]  ffff88000c9e50c0 ffff88000c9e50bc ffff880023bfa000 ffff880023bfbfd8
[176289.431815] Call Trace:
[176289.431815]  [<ffffffff81056657>] ? wake_up_bit+0x25/0x25
[176289.431815]  [<ffffffffa0395434>] ? transport_handle_cdb_direct+0x92/0x92 [target_core_mod]
[176289.431815]  [<ffffffff8105619a>] kthread+0x7d/0x85
[176289.431815]  [<ffffffff813cbcb4>] kernel_thread_helper+0x4/0x10
[176289.431815]  [<ffffffff8105611d>] ? kthread_worker_fn+0x16d/0x16d
[176289.431815]  [<ffffffff813cbcb0>] ? gs_change+0x13/0x13
[176289.431815] Code: 67 05 00 00 41 8b 84 24 4c ff ff ff ff c8 83 f8 11 0f 87 f0 04 00 00 89 c0 ff 24 c5 b0 c6 39 a0 0f 0b eb fe 48 8b 83 d8 00 00 00
[176289.431815] RIP  [<ffffffffa0395617>] transport_processing_thread+0x1e3/0x794 [target_core_mod]
[176289.431815]  RSP <ffff880023bfbe10>
[176289.431815] CR2: 0000000000000090
[176295.041004] ---[ end trace 85dc6865b23b8f3e ]---

Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Remove legacy + unused device active I/O shutdown code
Nicholas Bellinger [Sun, 9 Oct 2011 09:19:01 +0000 (02:19 -0700)]
target: Remove legacy + unused device active I/O shutdown code

This patch removes the legacy device active I/O shutdown code that was
originally called from transport_processing_thread() context during shutdown
including transport_processing_shutdown() and transport_release_all_cmds().

This is due to the fact that in modern configfs control plane code by the
time shutdown of an se_device instance in transport_processing_thread()
is allowed to occur via:

rmdir /sys/kernel/config/target/core/$HBA/$DEV

all active I/O will already have been ceased while removing active configfs
fabric Port/LUN symlinks.  Eg: the removal of an active se_device is protected
by inter-module VFS references from active Port/LUN symlinks.

Two WARN_ON() checks have been added in their place before exiting
transport_processing_thread() to watch out for any leaked descriptors.

Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Merge transport_cmd_finish_abort_tmr into transport_cmd_finish_abort
Nicholas Bellinger [Sun, 9 Oct 2011 09:02:51 +0000 (02:02 -0700)]
target: Merge transport_cmd_finish_abort_tmr into transport_cmd_finish_abort

This patch merges transport_cmd_finish_abort_tmr() logic into a single
transport_cmd_finish_abort() function by adding a cmd->se_tmr_req check
around transport_lun_remove_cmd(), and updates the single caller within
core_tmr_drain_tmr_list().

Reported-by: Christoph Hellwig <hch@lst.de>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agoiscsi-target: Remove SCF_SE_LUN_CMD flag abuses
Nicholas Bellinger [Sun, 9 Oct 2011 08:48:14 +0000 (01:48 -0700)]
iscsi-target: Remove SCF_SE_LUN_CMD flag abuses

This patch removes a number of SCF_SE_LUN_CMD flag abuses within iscsi-target
code to determine when iscsit_release_cmd() or transport_generic_free_cmd()
should be called while releasing an individual iscsi_cmd descriptor.

In the place of SCF_SE_LUN_CMD checks, this patch converts existing code to
use a new iscsit_free_cmd() that inspects iscsi_cmd->iscsi_opcode types to
determine which of the above functions should be invoked.  It also removes the
now unnecessary special case checking in iscsit_release_commands_from_conn().

(hch: Use iscsit_free_cmd instead of open-coded alternative)

Reported-by: Christoph Hellwig <hch@lst.de>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Convert ->transport_wait_for_tasks usage to transport_generic_free_cmd
Nicholas Bellinger [Sun, 9 Oct 2011 08:00:58 +0000 (01:00 -0700)]
target: Convert ->transport_wait_for_tasks usage to transport_generic_free_cmd

This patch converts se_cmd->transport_wait_for_tasks(se_cmd, 1) usage to use
transport_generic_free_cmd() directly in target-core and iscsi-target fabric
usage.  The includes:

*) Removal of the optional transport_generic_free_cmd() call from within
   transport_generic_wait_for_tasks()
*) Usage of existing SCF_SUPPORTED_SAM_OPCODE to determine when
   transport_generic_wait_for_tasks() processing may occur instead of
   checking se_cmd->transport_wait_for_tasks()
*) Move transport_generic_wait_for_tasks() call ahead of core_dec_lacl_count()
   and transport_lun_remove_cmd() in transport_generic_free_cmd() to follow
   existing logic for iscsi-target w/ se_cmd->transport_wait_for_tasks(se_cmd, 1)
*) Removal of se_cmd->transport_wait_for_tasks() function pointer
*) Rename transport_generic_wait_for_tasks() -> transport_wait_for_tasks(), and
   add docbook comment.
*) Add EXPORT_SYMBOL for transport_wait_for_tasks()

For the case in iscsi_target_erl2.c:iscsit_prepare_cmds_for_realligance()
where se_cmd->transport_wait_for_tasks(se_cmd, 0) is called, this patch
adds a direct call to transport_wait_for_tasks().

(hch: Fix transport_generic_free_cmd() usage in iscsit_release_commands_from_conn)
(nab: Add patch: Ensure that TMRs hit wait_for_tasks logic during release)

Reported-by: Christoph Hellwig <hch@lst.de>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Have core_tmr_alloc_req() take an explicit GFP_xxx flag
Roland Dreier [Thu, 6 Oct 2011 16:56:16 +0000 (09:56 -0700)]
target: Have core_tmr_alloc_req() take an explicit GFP_xxx flag

Testing in_interrupt() to know when sleeping is allowed is not really
reliable (since eg it won't be true if the caller is holding a spinlock).
Instead have the caller tell core_tmr_alloc_req() what GFP_xxx to use;
every caller except tcm_qla2xxx can use GFP_KERNEL.

Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Make pscsi_create_virtdevice use ERR_CAST
Dan Carpenter [Sat, 1 Oct 2011 22:59:13 +0000 (01:59 +0300)]
target: Make pscsi_create_virtdevice use ERR_CAST

This patch changes pscsi_create_virtdevice() to properly return ERR_CAST
instead of a raw pointer upon scsi_host_lookup() failure.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotcm_fc: remove custom hex_to_bin in ft_parse_wwn
Andy Shevchenko [Fri, 30 Sep 2011 11:45:40 +0000 (14:45 +0300)]
tcm_fc: remove custom hex_to_bin in ft_parse_wwn

This patch converts ft_parse_wwn() to use hex_to_bin() instead of custom
conversion code.

(Andy: Re-add missing strict && isupper(c) check)

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agoiscsi-target: use native hex2bin for chap_string_to_hex
Andy Shevchenko [Fri, 30 Sep 2011 11:39:54 +0000 (14:39 +0300)]
iscsi-target: use native hex2bin for chap_string_to_hex

This patch converts chap_string_to_hex() to use hex2bin() instead of
the internal chap_asciihex_to_binaryhex().

(nab: Fix up minor compile breakage + typo)

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: remove unused se_subsystem_api methods
Christoph Hellwig [Sun, 25 Sep 2011 18:56:43 +0000 (14:56 -0400)]
target: remove unused se_subsystem_api methods

The cdb_none, map_data_SG and map_control_SG methods have no callers left
and can be removed now.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: cleanup pscsi request submission
Christoph Hellwig [Sun, 25 Sep 2011 18:56:32 +0000 (14:56 -0400)]
target: cleanup pscsi request submission

Move the entirely request allocation, mapping and submission into ->do_task.
This

 a) avoids blocking the I/O submission thread unessecarily, and
 b) simplifies the code greatly

Note that the code seems to have various error handling issues, mostly
related to bidi handling in the current form.  I've added comments about
those but not tried to fix them in this commit.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: cleanup iblock bio submission
Christoph Hellwig [Sun, 25 Sep 2011 18:56:24 +0000 (14:56 -0400)]
target: cleanup iblock bio submission

Move the entirely bio allocation, mapping and submission into ->do_task.
This

 a) avoids blocking the I/O submission thread unessecarily, and
 b) simplifies the code greatly

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: simplify target_parse_naa_6h_vendor_specific()
Andy Shevchenko [Mon, 19 Sep 2011 08:11:21 +0000 (11:11 +0300)]
target: simplify target_parse_naa_6h_vendor_specific()

This patch adds a minor simplfication in target_parse_naa_6h_vendor_specific()
to remove direct isxdigit() + ctype.h usage.

(nab: Fix next assignment breakage in for loop)

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Nicholas Bellinger <nab@linux-iscsi.org>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Remove session_reinstatement parameter from ->transport_wait_for_tasks
Nicholas Bellinger [Sat, 8 Oct 2011 20:59:52 +0000 (13:59 -0700)]
target: Remove session_reinstatement parameter from ->transport_wait_for_tasks

This patch removes the unnecessary session_reinstatement parameter from
se_cmd->transport_wait_for_tasks(), logic in transport_generic_wait_for_tasks,
and usage within iscsi-target code.

This also includes the removal of the 'bool' return from transport_put_cmd() +
transport_generic_free_cmd() that is no longer necessary.

Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: push session reinstatement out of transport_generic_free_cmd
Christoph Hellwig [Tue, 13 Sep 2011 21:09:01 +0000 (23:09 +0200)]
target: push session reinstatement out of transport_generic_free_cmd

Push session reinstatement out of transport_generic_free_cmd into the only
caller that actually needs it.  Clean up transport_generic_free_cmd a bit,
and remove the useless comment.  I'd love to add a more useful kerneldoc
comment for it, but as this point I'm still a bit confused in where it
stands in the command release stack.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: remove transport_generic_remove
Christoph Hellwig [Tue, 13 Sep 2011 21:08:50 +0000 (23:08 +0200)]
target: remove transport_generic_remove

All callers that never have the session_reinstatement flag set can trivially
be converted to transport_put_cmd.  Opencode the session reinstatement code
in transport_generic_free_cmd, which was the only caller ever asking for it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: simplify transport_put_cmd
Christoph Hellwig [Tue, 13 Sep 2011 21:08:42 +0000 (23:08 +0200)]
target: simplify transport_put_cmd

Inline two simple functions only used by it, and replace a goto
with a simple if else construct.

Note that the code moved from transport_dec_and_check seems fairly
buggy - the atomic_read check on a variable where we'd do an
atomic_dec_and_test looks racy if we'll ever get someone increment
it without the lock held around them (which it looks like we do),
and not decrementing the second counter if the first one doesn't
hit zero also at least needs an explanation.

(nab: Fix transport_put_cmd breakage)

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: simplify transport_generic_remove
Christoph Hellwig [Tue, 13 Sep 2011 21:08:32 +0000 (23:08 +0200)]
target: simplify transport_generic_remove

Instead of duplicating the code from transport_release_fe_cmd re-use it by
allowing transport_release_fe_cmd to return wether it actually freed the
command or not.  Also rename transport_release_fe_cmd to transport_put_cmd
and add a kerneldoc comment for it to make the use case more obvious.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: remove transport_free_se_cmd
Christoph Hellwig [Tue, 13 Sep 2011 21:08:19 +0000 (23:08 +0200)]
target: remove transport_free_se_cmd

It is only called by transport_release_cmd, so inline it there.  Also add
a kerneldoc comment for transport_release_cmd while we are at it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: don't opencode transport_release_cmd in transport_release_fe_cmd
Christoph Hellwig [Tue, 13 Sep 2011 21:08:11 +0000 (23:08 +0200)]
target: don't opencode transport_release_cmd in transport_release_fe_cmd

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: remove transport_generic_handle_cdb
Christoph Hellwig [Mon, 12 Sep 2011 19:51:14 +0000 (21:51 +0200)]
target: remove transport_generic_handle_cdb

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agoiscsi-target: always call transport_handle_cdb_direct
Christoph Hellwig [Mon, 12 Sep 2011 19:50:56 +0000 (21:50 +0200)]
iscsi-target: always call transport_handle_cdb_direct

iscsit_task_reassign_complete is always called from the TX thread, so
handle the CDB directly instead of offloading it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotfm_fc: use transport_handle_cdb_direct
Christoph Hellwig [Mon, 12 Sep 2011 19:50:49 +0000 (21:50 +0200)]
tfm_fc: use transport_handle_cdb_direct

ft_send_work is always called from workqueue context, which means we can
handle the CDB directly instead of doing another context switch.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Cc: Kiran Patil <kiran.patil@intel.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agotarget: Prevent transport_send_task_abort when CHECK_CONDITION status
Nicholas Bellinger [Thu, 29 Sep 2011 21:22:13 +0000 (14:22 -0700)]
target: Prevent transport_send_task_abort when CHECK_CONDITION status

This patch fixes a bug where transport_send_task_abort() could be called
during LUN_RESET to return SAM_STAT_TASK_ABORTED + tfo->queue_status(), when
SCF_SENT_CHECK_CONDITION -> tfo->queue_status() has already been sent from
within another context via transport_send_check_condition_and_sense().

Cc: stable@kernel.org
Signed-off-by: Nicholas Bellinger <nab@risingtidesystems.com>
13 years agotarget: Fix transport_cmd_finish_abort queue removal bug
Nicholas Bellinger [Thu, 29 Sep 2011 08:01:35 +0000 (01:01 -0700)]
target: Fix transport_cmd_finish_abort queue removal bug

This patch fixes a bug in LUN_RESET operation with transport_cmd_finish_abort()
where transport_remove_cmd_from_queue() was incorrectly being called, causing
descriptors with t_state == TRANSPORT_FREE_CMD_INTR to be incorrectly removed
from qobj->qobj_list during process context release.  This change ensures the
descriptor is only removed via transport_remove_cmd_from_queue() when doing a
direct release via transport_generic_remove().

Cc: stable@kernel.org
Signed-off-by: Nicholas Bellinger <nab@risingtidesystems.com>
13 years agotarget: Prevent TRANSPORT_FREE_CMD_INTR processing in core_tmr_drain_cmd_list
Nicholas Bellinger [Tue, 11 Oct 2011 06:02:48 +0000 (06:02 +0000)]
target: Prevent TRANSPORT_FREE_CMD_INTR processing in core_tmr_drain_cmd_list

This patch contains a bugfix for TMR LUN_RESET related to TRANSPORT_FREE_CMD_INTR
operation, where core_tmr_drain_cmd_list() will now skip processing for this
case to prevent an ABORT_TASK status from being returned for descriptors that
are already queued up to be released by processing thread context.

Cc: Roland Dreier <roland@purestorage.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: stable@kernel.org
Signed-off-by: Nicholas Bellinger <nab@risingtidesystems.com>
13 years agotarget: Re-org of core_tmr_lun_reset
Nicholas Bellinger [Thu, 29 Sep 2011 04:37:29 +0000 (21:37 -0700)]
target: Re-org of core_tmr_lun_reset

This patch is a re-orginzation of core_tmr_lun_reset() logic to properly
scan the active tmr_list, dev->state_task_list and qobj->qobj_list w/ the
relivent locks held, and performing a list_move_tail onto seperate local
scope lists before performing the full drain.

This involves breaking out the code into three seperate list specific
functions: core_tmr_drain_tmr_list(), core_tmr_drain_task_list() and
core_tmr_drain_cmd_list().

(nab: Include target: Remove non-active tasks from execute list during
      LUN_RESET patch to address original breakage)

Reported-by: Roland Dreier <roland@purestorage.com>
Cc: Roland Dreier <roland@purestorage.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: stable@kernel.org
Signed-off-by: Nicholas Bellinger <nab@risingtidesystems.com>
13 years agotarget: Prevent cmd->se_queue_node double add
Roland Dreier [Thu, 29 Sep 2011 05:12:07 +0000 (22:12 -0700)]
target: Prevent cmd->se_queue_node double add

This patch addresses a bug with the lio-core-2.6.git conversion of
transport_add_cmd_to_queue() to use a single embedded list_head, instead
of individual struct se_queue_req allocations allowing a single se_cmd to
be added to the queue mulitple times.  This was changed in the following:

commit 2a9e4d5ca5d99f4c600578d6285d45142e7e5208
Author: Andy Grover <agrover@redhat.com>
Date:   Tue Apr 26 17:45:51 2011 -0700

    target: Embed qr in struct se_cmd

The problem is that some target code still assumes performing multiple
adds is allowed via transport_add_cmd_to_queue(), which ends up causing
list corruption in qobj->qobj_list code.  This patch addresses this
by removing an existing struct se_cmd from the list before the add, and
removes an unnecessary list walk in transport_remove_cmd_from_queue()

It also changes cmd->t_transport_queue_active to use explict sets intead
of increment/decrement to prevent confusion during exception path handling.

Signed-off-by: Roland Dreier <roland@purestorage.com>
Cc: Andy Grover <agrover@redhat.com>
Cc: stable@kernel.org
Signed-off-by: Nicholas Bellinger <nab@risingtidesystems.com>
13 years agotarget: Remove unneeded version.h includes
Jesper Juhl [Mon, 1 Aug 2011 21:29:11 +0000 (23:29 +0200)]
target: Remove unneeded version.h includes

It was pointed out by 'make versioncheck' that some includes of
linux/version.h are not needed in drivers/target/.
This patch removes them.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
13 years agoMerge git://git.samba.org/sfrench/cifs-2.6
Linus Torvalds [Mon, 10 Oct 2011 02:53:11 +0000 (14:53 +1200)]
Merge git://git.samba.org/sfrench/cifs-2.6

* git://git.samba.org/sfrench/cifs-2.6:
  [CIFS] Fix first time message on mount, ntlmv2 upgrade delayed to 3.2

13 years agoMerge branch 'fixes' of git://git.linaro.org/people/arnd/arm-soc
Linus Torvalds [Mon, 10 Oct 2011 02:48:27 +0000 (14:48 +1200)]
Merge branch 'fixes' of git://git.linaro.org/people/arnd/arm-soc

* 'fixes' of git://git.linaro.org/people/arnd/arm-soc:
  ARM: mach-ux500: enable fix for ARM errata 754322
  ARM: OMAP: musb: Remove a redundant omap4430_phy_init call in usb_musb_init
  ARM: OMAP: Fix i2c init for twl4030
  ARM: OMAP4: MMC: fix power and audio issue, decouple USBC1 from MMC1

13 years agoARM: tegra: fix compilation error due to mach/hardware.h removal
Marc Dietrich [Fri, 7 Oct 2011 15:31:41 +0000 (08:31 -0700)]
ARM: tegra: fix compilation error due to mach/hardware.h removal

This fixes a compilation error in cpu-tegra.c which was introduced in
dc8d966bccde ("ARM: convert PCI defines to variables") which removed the
now obsolete mach/hardware.h from the mach-tegra subtree.

Signed-off-by: Marc Dietrich <marvin24@gmx.de>
Signed-off-by: Olof Johansson <olof@lixom.net>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Mon, 10 Oct 2011 02:43:06 +0000 (14:43 +1200)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/radeon/kms: use hardcoded dig encoder to transmitter mapping for DCE4.1
  drm/radeon/kms: fix dp_detect handling for DP bridge chips
  drm/radeon/kms: retry aux transactions if there are status flags

13 years agoMAINTAINERS: Update tegra maintainer information
Olof Johansson [Fri, 7 Oct 2011 20:27:48 +0000 (13:27 -0700)]
MAINTAINERS: Update tegra maintainer information

A couple of changes to the Tegra maintainership setup:

I'm very glad to bring on Stephen Warren on board as a maintainer. The
work he has done so far is excellent, and the fact that he works for
Nvidia means he has long-term interest in the platform.

Erik Gilling did an astounding amount of work on getting things up and
running but has been a silent partner on the maintainership side for a
while, and is stepping down. Thanks for your contributions so far, Erik.

Finally, update the git URL since I'll take over running the main repo
for a while.

Overall maintainership model isn't changing much at this time: We'll all
three review patches as appropriate, and one of us will collect the main
repo (me at this time).

Signed-off-by: Olof Johansson <olof@lixom.net>
Cc: Erik Gilling <konkers@android.com>
Acked-by: Colin Cross <ccross@android.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoMerge branch 'upstream' of git://git.linux-mips.org/pub/scm/upstream-linus
Linus Torvalds [Mon, 10 Oct 2011 02:39:03 +0000 (14:39 +1200)]
Merge branch 'upstream' of git://git.linux-mips.org/upstream-linus

* 'upstream' of git://git.linux-mips.org/pub/scm/upstream-linus: (29 commits)
  MIPS: Call oops_enter, oops_exit in die
  staging/octeon: Software should check the checksum of no tcp/udp packets
  MIPS: Octeon: Enable C0_UserLocal probing.
  MIPS: No branches in delay slots for huge pages in handle_tlbl
  MIPS: Don't clobber CP0_STATUS value for CONFIG_MIPS_MT_SMTC
  MIPS: Octeon: Select CONFIG_HOLES_IN_ZONE
  MIPS: PM: Use struct syscore_ops instead of sysdevs for PM (v2)
  MIPS: Compat: Use 32-bit wrapper for compat_sys_futex.
  MIPS: Do not use EXTRA_CFLAGS
  MIPS: Alchemy: DB1200: Disable cascade IRQ in handler
  SERIAL: Lantiq: Set timeout in uart_port
  MIPS: Lantiq: Fix setting the PCI bus speed on AR9
  MIPS: Lantiq: Fix external interrupt sources
  MIPS: tlbex: Fix build error in R3000 code.
  MIPS: Alchemy: Include Au1100 in PM code.
  MIPS: Alchemy: Fix typo in MAC0 registration
  MIPS: MSP71xx: Fix build error.
  MIPS: Handle __put_user() sleeping.
  MIPS: Allow forced irq threading
  MIPS: i8259: Mark cascade interrupt non-threaded
  ...

13 years agoMerge branch 'omap/fixes-for-3.1' into fixes
Arnd Bergmann [Sat, 8 Oct 2011 20:21:07 +0000 (22:21 +0200)]
Merge branch 'omap/fixes-for-3.1' into fixes

13 years ago[CIFS] Fix first time message on mount, ntlmv2 upgrade delayed to 3.2
Steve French [Fri, 7 Oct 2011 04:14:07 +0000 (23:14 -0500)]
[CIFS] Fix first time message on mount, ntlmv2 upgrade delayed to 3.2

Microsoft has a bug with ntlmv2 that requires use of ntlmssp, but
we didn't get the required information on when/how to use ntlmssp to
old (but once very popular) legacy servers (various NT4 fixpacks
for example) until too late to merge for 3.1.  Will upgrade
to NTLMv2 in NTLMSSP in 3.2

Signed-off-by: Steve French <smfrench@gmail.com>
Reviewed-by: Jeff Layton <jlayton@redhat.com>
13 years agoARM: mach-ux500: enable fix for ARM errata 754322
srinidhi kasagar [Tue, 20 Sep 2011 05:45:46 +0000 (11:15 +0530)]
ARM: mach-ux500: enable fix for ARM errata 754322

This applies ARM errata fix 754322 for all ux500 platforms.

Cc: stable@kernel.org
Signed-off-by: srinidhi kasagar <srinidhi.kasagar@stericsson.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
13 years agoMerge git://github.com/davem330/net
Linus Torvalds [Thu, 6 Oct 2011 23:15:10 +0000 (16:15 -0700)]
Merge git://github.com/davem330/net

* git://github.com/davem330/net:
  net: fix typos in Documentation/networking/scaling.txt
  bridge: leave carrier on for empty bridge
  netfilter: Use proper rwlock init function
  tcp: properly update lost_cnt_hint during shifting
  tcp: properly handle md5sig_pool references
  macvlan/macvtap: Fix unicast between macvtap interfaces in bridge mode

13 years agox86/PCI: use host bridge _CRS info on ASUS M2V-MX SE
Paul Menzel [Wed, 31 Aug 2011 15:07:10 +0000 (17:07 +0200)]
x86/PCI: use host bridge _CRS info on ASUS M2V-MX SE

In summary, this DMI quirk uses the _CRS info by default for the ASUS
M2V-MX SE by turning on `pci=use_crs` and is similar to the quirk
added by commit 2491762cfb47 ("x86/PCI: use host bridge _CRS info on
ASRock ALiveSATA2-GLAN") whose commit message should be read for further
information.

Since commit 3e3da00c01d0 ("x86/pci: AMD one chain system to use pci
read out res") Linux gives the following oops:

    parport0: PC-style at 0x378, irq 7 [PCSPP,TRISTATE]
    HDA Intel 0000:20:01.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17
    HDA Intel 0000:20:01.0: setting latency timer to 64
    BUG: unable to handle kernel paging request at ffffc90011c08000
    IP: [<ffffffffa0578402>] azx_probe+0x3ad/0x86b [snd_hda_intel]
    PGD 13781a067 PUD 13781b067 PMD 1300ba067 PTE 800000fd00000173
    Oops: 0009 [#1] SMP
    last sysfs file: /sys/module/snd_pcm/initstate
    CPU 0
    Modules linked in: snd_hda_intel(+) snd_hda_codec snd_hwdep snd_pcm_oss snd_mixer_oss snd_pcm snd_seq_midi snd_rawmidi snd_seq_midi_event tpm_tis tpm snd_seq tpm_bios psmouse parport_pc snd_timer snd_seq_device parport processor evdev snd i2c_viapro thermal_sys amd64_edac_mod k8temp i2c_core soundcore shpchp pcspkr serio_raw asus_atk0110 pci_hotplug edac_core button snd_page_alloc edac_mce_amd ext3 jbd mbcache sha256_generic cryptd aes_x86_64 aes_generic cbc dm_crypt dm_mod raid1 md_mod usbhid hid sg sd_mod crc_t10dif sr_mod cdrom ata_generic uhci_hcd sata_via pata_via libata ehci_hcd usbcore scsi_mod via_rhine mii nls_base [last unloaded: scsi_wait_scan]
    Pid: 1153, comm: work_for_cpu Not tainted 2.6.37-1-amd64 #1 M2V-MX SE/System Product Name
    RIP: 0010:[<ffffffffa0578402>]  [<ffffffffa0578402>] azx_probe+0x3ad/0x86b [snd_hda_intel]
    RSP: 0018:ffff88013153fe50  EFLAGS: 00010286
    RAX: ffffc90011c08000 RBX: ffff88013029ec00 RCX: 0000000000000006
    RDX: 0000000000000000 RSI: 0000000000000246 RDI: 0000000000000246
    RBP: ffff88013341d000 R08: 0000000000000000 R09: 0000000000000040
    R10: 0000000000000286 R11: 0000000000003731 R12: ffff88013029c400
    R13: 0000000000000000 R14: 0000000000000000 R15: ffff88013341d090
    FS:  0000000000000000(0000) GS:ffff8800bfc00000(0000) knlGS:00000000f7610ab0
    CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
    CR2: ffffc90011c08000 CR3: 0000000132f57000 CR4: 00000000000006f0
    DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
    DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
    Process work_for_cpu (pid: 1153, threadinfo ffff88013153e000, task ffff8801303c86c0)
    Stack:
     0000000000000005 ffffffff8123ad65 00000000000136c0 ffff88013029c400
     ffff8801303c8998 ffff88013341d000 ffff88013341d090 ffff8801322d9dc8
     ffff88013341d208 0000000000000000 0000000000000000 ffffffff811ad232
    Call Trace:
     [<ffffffff8123ad65>] ? __pm_runtime_set_status+0x162/0x186
     [<ffffffff811ad232>] ? local_pci_probe+0x49/0x92
     [<ffffffff8105afc5>] ? do_work_for_cpu+0x0/0x1b
     [<ffffffff8105afc5>] ? do_work_for_cpu+0x0/0x1b
     [<ffffffff8105afd0>] ? do_work_for_cpu+0xb/0x1b
     [<ffffffff8105fd3f>] ? kthread+0x7a/0x82
     [<ffffffff8100a824>] ? kernel_thread_helper+0x4/0x10
     [<ffffffff8105fcc5>] ? kthread+0x0/0x82
     [<ffffffff8100a820>] ? kernel_thread_helper+0x0/0x10
    Code: f4 01 00 00 ef 31 f6 48 89 df e8 29 dd ff ff 85 c0 0f 88 2b 03 00 00 48 89 ef e8 b4 39 c3 e0 8b 7b 40 e8 fc 9d b1 e0 48 8b 43 38 <66> 8b 10 66 89 14 24 8b 43 14 83 e8 03 83 f8 01 77 32 31 d2 be
    RIP  [<ffffffffa0578402>] azx_probe+0x3ad/0x86b [snd_hda_intel]
     RSP <ffff88013153fe50>
    CR2: ffffc90011c08000
    ---[ end trace 8d1f3ebc136437fd ]---

Trusting the ACPI _CRS information (`pci=use_crs`) fixes this problem.

    $ dmesg | grep -i crs # with the quirk
    PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug

The match has to be against the DMI board entries though since the vendor entries are not populated.

    DMI: System manufacturer System Product Name/M2V-MX SE, BIOS 0304    10/30/2007

This quirk should be removed when `pci=use_crs` is enabled for machines
from 2006 or earlier or some other solution is implemented.

Using coreboot [1] with this board the problem does not exist but this
quirk also does not affect it either. To be safe though the check is
tightened to only take effect when the BIOS from American Megatrends is
used.

        15:13 < ruik> but coreboot does not need that
        15:13 < ruik> because i have there only one root bus
        15:13 < ruik> the audio is behind a bridge

        $ sudo dmidecode
        BIOS Information
                Vendor: American Megatrends Inc.
                Version: 0304
                Release Date: 10/30/2007

[1] http://www.coreboot.org/

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=30552

Cc: stable@kernel.org (2.6.34)
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: x86@kernel.org
Signed-off-by: Paul Menzel <paulepanter@users.sourceforge.net>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agonet: fix typos in Documentation/networking/scaling.txt
Benjamin Poirier [Tue, 4 Oct 2011 04:00:30 +0000 (04:00 +0000)]
net: fix typos in Documentation/networking/scaling.txt

The second hunk fixes rps_sock_flow_table but has to re-wrap the paragraph.

Signed-off-by: Benjamin Poirier <benjamin.poirier@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agobridge: leave carrier on for empty bridge
stephen hemminger [Mon, 3 Oct 2011 18:14:45 +0000 (18:14 +0000)]
bridge: leave carrier on for empty bridge

This resolves a regression seen by some users of bridging.
Some users use the bridge like a dummy device.
They expect to be able to put an IPv6 address on the device
with no ports attached. Although there are better ways of doing
this, there is no reason to not allow it.

Note: the bridge still will reflect the state of ports in the
bridge if there are any added.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoMerge branch 'for-linus' of http://people.redhat.com/agk/git/linux-dm
Linus Torvalds [Thu, 6 Oct 2011 15:31:47 +0000 (08:31 -0700)]
Merge branch 'for-linus' of people.redhat.com/agk/git/linux-dm

* 'for-linus' of http://people.redhat.com/agk/git/linux-dm:
  dm crypt: always disable discard_zeroes_data
  dm: raid fix write_mostly arg validation
  dm table: avoid crash if integrity profile changes
  dm: flakey fix corrupt_bio_byte error path

13 years agoMerge branch 'for-linus' of git://neil.brown.name/md
Linus Torvalds [Thu, 6 Oct 2011 15:30:03 +0000 (08:30 -0700)]
Merge branch 'for-linus' of git://neil.brown.name/md

* 'for-linus' of git://neil.brown.name/md:
  md: Avoid waking up a thread after it has been freed.

13 years agodrm/radeon/kms: use hardcoded dig encoder to transmitter mapping for DCE4.1
Alex Deucher [Wed, 5 Oct 2011 22:36:50 +0000 (18:36 -0400)]
drm/radeon/kms: use hardcoded dig encoder to transmitter mapping for DCE4.1

The encoders are supposedly fully routeable, but changing the mapping
doesn't always seem to take.  Using a hardcoded mapping is much more
reliable.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=41366

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Tested-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agonetfilter: Use proper rwlock init function
Thomas Gleixner [Wed, 5 Oct 2011 03:24:43 +0000 (03:24 +0000)]
netfilter: Use proper rwlock init function

Replace the open coded initialization with the init function.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Hans Schillstrom <hans.schillstrom@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoMerge branch 'for-linus' of git://github.com/dtor/input
Linus Torvalds [Wed, 5 Oct 2011 16:22:38 +0000 (09:22 -0700)]
Merge branch 'for-linus' of git://github.com/dtor/input

* 'for-linus' of git://github.com/dtor/input:
  Input: wacom - revert "Cintiq 21UX2 does not have menu strips"

13 years agoMerge git://bedivere.hansenpartnership.com/git/scsi-rc-fixes-2.6
Linus Torvalds [Wed, 5 Oct 2011 16:16:11 +0000 (09:16 -0700)]
Merge git://bedivere.hansenpartnership.com/git/scsi-rc-fixes-2.6

* git://bedivere.hansenpartnership.com/git/scsi-rc-fixes-2.6:
  [SCSI] libsas: fix panic when single phy is disabled on a wide port
  [SCSI] qla2xxx: Fix crash in qla2x00_abort_all_cmds() on unload

13 years agodrm/radeon/kms: fix dp_detect handling for DP bridge chips
Alex Deucher [Tue, 4 Oct 2011 16:23:24 +0000 (12:23 -0400)]
drm/radeon/kms: fix dp_detect handling for DP bridge chips

The HPD pin is not reliable for detecting whether a monitor
is connected or not.  Skip HPD and just use DDC or load
detection.

Fixes phantom VGA connected bugs.

[Michel: fixes phantom VGA bugs on his llano system.]

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Tested-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agodrm/radeon/kms: retry aux transactions if there are status flags
Alex Deucher [Tue, 4 Oct 2011 21:23:15 +0000 (17:23 -0400)]
drm/radeon/kms: retry aux transactions if there are status flags

If there are error flags in the aux status, retry the transaction.
This makes aux much more reliable, especially on llano systems.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agoInput: wacom - revert "Cintiq 21UX2 does not have menu strips"
Jason Gerecke [Wed, 5 Oct 2011 05:50:45 +0000 (22:50 -0700)]
Input: wacom - revert "Cintiq 21UX2 does not have menu strips"

This reverts commit 71c86ce59791bcd67af937bbea719a508079d7c2.
The 21UX2 does have touchstrips, but they are in a somewhat-
hidden location.

Signed-off-by: Jason Gerecke <killertofu@gmail.com>
Acked-by: Ping Cheng <pinglinux@gmail.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
13 years agotcp: properly update lost_cnt_hint during shifting
Yan, Zheng [Sun, 2 Oct 2011 04:21:50 +0000 (04:21 +0000)]
tcp: properly update lost_cnt_hint during shifting

lost_skb_hint is used by tcp_mark_head_lost() to mark the first unhandled skb.
lost_cnt_hint is the number of packets or sacked packets before the lost_skb_hint;
When shifting a skb that is before the lost_skb_hint, if tcp_is_fack() is ture,
the skb has already been counted in the lost_cnt_hint; if tcp_is_fack() is false,
tcp_sacktag_one() will increase the lost_cnt_hint. So tcp_shifted_skb() does not
need to adjust the lost_cnt_hint by itself. When shifting a skb that is equal to
lost_skb_hint, the shifted packets will not be counted by tcp_mark_head_lost().
So tcp_shifted_skb() should adjust the lost_cnt_hint even tcp_is_fack(tp) is true.

Signed-off-by: Zheng Yan <zheng.z.yan@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agotcp: properly handle md5sig_pool references
Yan, Zheng [Thu, 29 Sep 2011 17:10:10 +0000 (17:10 +0000)]
tcp: properly handle md5sig_pool references

tcp_v4_clear_md5_list() assumes that multiple tcp md5sig peers
only hold one reference to md5sig_pool. but tcp_v4_md5_do_add()
increases use count of md5sig_pool for each peer. This patch
makes tcp_v4_md5_do_add() only increases use count for the first
tcp md5sig peer.

Signed-off-by: Zheng Yan <zheng.z.yan@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agomacvlan/macvtap: Fix unicast between macvtap interfaces in bridge mode
David Ward [Sun, 18 Sep 2011 12:53:20 +0000 (12:53 +0000)]
macvlan/macvtap: Fix unicast between macvtap interfaces in bridge mode

Packets should always be forwarded to the lowerdev using dev_forward_skb.
vlan->forward is for packets being forwarded directly to another macvlan/
macvtap device (used for multicast in bridge mode).

Reported-and-tested-by: Shlomo Pongratz <shlomop@mellanox.com>
Signed-off-by: David Ward <david.ward@ll.mit.edu>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agoLinux 3.1-rc9
Linus Torvalds [Wed, 5 Oct 2011 01:11:50 +0000 (18:11 -0700)]
Linux 3.1-rc9

13 years agoMerge git://github.com/davem330/net
Linus Torvalds [Tue, 4 Oct 2011 17:37:06 +0000 (10:37 -0700)]
Merge git://github.com/davem330/net

* git://github.com/davem330/net:
  pch_gbe: Fixed the issue on which a network freezes
  pch_gbe: Fixed the issue on which PC was frozen when link was downed.
  make PACKET_STATISTICS getsockopt report consistently between ring and non-ring
  net: xen-netback: correctly restart Tx after a VM restore/migrate
  bonding: properly stop queuing work when requested
  can bcm: fix incomplete tx_setup fix
  RDSRDMA: Fix cleanup of rds_iw_mr_pool
  net: Documentation: Fix type of variables
  ibmveth: Fix oops on request_irq failure
  ipv6: nullify ipv6_ac_list and ipv6_fl_list when creating new socket
  cxgb4: Fix EEH on IBM P7IOC
  can bcm: fix tx_setup off-by-one errors
  MAINTAINERS: tehuti: Alexander Indenbaum's address bounces
  dp83640: reduce driver noise
  ptp: fix L2 event message recognition

13 years agoMerge branch 'fix/asoc' of git://github.com/tiwai/sound
Linus Torvalds [Tue, 4 Oct 2011 16:59:22 +0000 (09:59 -0700)]
Merge branch 'fix/asoc' of git://github.com/tiwai/sound

* 'fix/asoc' of git://github.com/tiwai/sound:
  ASoC: omap_mcpdm_remove cannot be __devexit
  ASoC: Fix setting update bits for WM8753_LADC and WM8753_RADC
  ASoC: use a valid device for dev_err() in Zylonite

13 years agoMerge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux
Linus Torvalds [Tue, 4 Oct 2011 16:54:18 +0000 (09:54 -0700)]
Merge branch 'drm-fixes' of git://people.freedesktop.org/~airlied/linux

* 'drm-fixes' of git://people.freedesktop.org/~airlied/linux:
  drm/radeon/kms: fix channel_remap setup (v2)
  drm/radeon: Set cursor x/y to 0 when x/yorigin > 0.
  drm/radeon: Update AVIVO cursor coordinate origin before x/yorigin calculation.
  drm/radeon: Simplify cursor x/yorigin calculation.
  drm/radeon/kms: fix cursor image off-by-one error
  drm/radeon/kms: Fix logic error in DP HPD handler
  drm/radeon/kms: add retry limits for native DP aux defer
  drm/radeon/kms: fix regression in DP aux defer handling

13 years agoMerge branch 'spi/merge' of git://git.secretlab.ca/git/linux-2.6
Linus Torvalds [Tue, 4 Oct 2011 16:52:56 +0000 (09:52 -0700)]
Merge branch 'spi/merge' of git://git.secretlab.ca/git/linux-2.6

* 'spi/merge' of git://git.secretlab.ca/git/linux-2.6:
  spi-topcliff-pch: Fix overrun issue
  spi-topcliff-pch: Add recovery processing in case FIFO overrun error occurs
  spi-topcliff-pch: Fix CPU read complete condition issue
  spi-topcliff-pch: Fix SSN Control issue
  spi-topcliff-pch: add tx-memory clear after complete transmitting

13 years agoPCI: Disable MPS configuration by default
Jon Mason [Mon, 3 Oct 2011 14:50:20 +0000 (09:50 -0500)]
PCI: Disable MPS configuration by default

Add the ability to disable PCI-E MPS turning and using the BIOS
configured MPS defaults.  Due to the number of issues recently
discovered on some x86 chipsets, make this the default behavior.

Also, add the option for peer to peer DMA MPS configuration.  Peer to
peer DMA is outside the scope of this patch, but MPS configuration could
prevent it from working by having the MPS on one root port different
than the MPS on another.  To work around this, simply make the system
wide MPS the smallest possible value (128B).

Signed-off-by: Jon Mason <mason@myri.com>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agodrm/radeon/kms: fix channel_remap setup (v2)
Alex Deucher [Tue, 4 Oct 2011 14:46:34 +0000 (10:46 -0400)]
drm/radeon/kms: fix channel_remap setup (v2)

Most asics just use the hw default value which requires
no explicit programming.  For those that need a different
value, the vbios will program it properly.  As such,
there's no need to program these registers explicitly
in the driver.  Changing MC_SHARED_CHREMAP requires a reload
of all data in vram otherwise its contents will be scambled.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=40103

v2: drop now unused channel_remap functions.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agospi-topcliff-pch: Fix overrun issue
Tomoya MORINAGA [Tue, 6 Sep 2011 08:16:38 +0000 (17:16 +0900)]
spi-topcliff-pch: Fix overrun issue

We found that adding load, Rx data sometimes drops.(with DMA transfer mode)
The cause is that before starting Rx-DMA processing, Tx-DMA processing starts.
This causes FIFO overrun occurs.

This patch fixes the issue by modifying FIFO tx-threshold and DMA descriptor
size like below.

                      Current                   this patch
Rx-descriptor   4Byte+12Byte*341    -->    12Byte*340-4Byte-12Byte
Rx-threshold                   (Not modified)
Tx-descriptor   4Byte+12Byte*341    -->    16Byte-12Byte*340
Rx-threshold    12Byte              -->    2Byte

Signed-off-by: Tomoya MORINAGA <tomoya-linux@dsn.okisemi.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agospi-topcliff-pch: Add recovery processing in case FIFO overrun error occurs
Tomoya MORINAGA [Tue, 6 Sep 2011 08:16:37 +0000 (17:16 +0900)]
spi-topcliff-pch: Add recovery processing in case FIFO overrun error occurs

Add recovery processing in case FIFO overrun error occurs with DMA transfer mode.

Signed-off-by: Tomoya MORINAGA <tomoya-linux@dsn.okisemi.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agospi-topcliff-pch: Fix CPU read complete condition issue
Tomoya MORINAGA [Tue, 6 Sep 2011 08:16:36 +0000 (17:16 +0900)]
spi-topcliff-pch: Fix CPU read complete condition issue

We found Rx data sometimes drops.(with non-DMA transfer mode)
The cause is read complete condition is not true.

This patch fixes the issue.

Signed-off-by: Tomoya MORINAGA <tomoya-linux@dsn.okisemi.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agospi-topcliff-pch: Fix SSN Control issue
Tomoya MORINAGA [Tue, 6 Sep 2011 08:16:35 +0000 (17:16 +0900)]
spi-topcliff-pch: Fix SSN Control issue

During processing 1 command/data series,
SSN should keep LOW.
However, currently, SSN becomes HIGH.
This patch fixes the issue.

Signed-off-by: Tomoya MORINAGA <tomoya-linux@dsn.okisemi.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agospi-topcliff-pch: add tx-memory clear after complete transmitting
Tomoya MORINAGA [Tue, 6 Sep 2011 08:16:34 +0000 (17:16 +0900)]
spi-topcliff-pch: add tx-memory clear after complete transmitting

Currently, in case of reading date from SPI flash,
command is sent twice.
The cause is that tx-memory clear processing is missing .
This patch adds the tx-momory clear processing.

Signed-off-by: Tomoya MORINAGA <tomoya-linux@dsn.okisemi.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
13 years agolis3: fix regression of HP DriveGuard with 8bit chip
Takashi Iwai [Tue, 4 Oct 2011 01:09:14 +0000 (18:09 -0700)]
lis3: fix regression of HP DriveGuard with 8bit chip

Commit 2a7fade7e03 ("hwmon: lis3: Power on corrections") caused a
regression on HP laptops with 8bit chip.  Writing CTRL2_BOOT_8B bit seems
clearing the BIOS setup, and no proper interrupt for DriveGuard will be
triggered any more.

Since the init code there is basically only for embedded devices, put a
pdata check so that the problematic initialization will be skipped for
hp_accel stuff.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Cc: Eric Piel <eric.piel@tremplin-utc.net>
Cc: Samu Onkalo <samu.p.onkalo@nokia.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
13 years agoMerge branch 'hwmon-for-linus' of git://github.com/groeck/linux
Linus Torvalds [Mon, 3 Oct 2011 19:54:56 +0000 (12:54 -0700)]
Merge branch 'hwmon-for-linus' of git://github.com/groeck/linux

* 'hwmon-for-linus' of git://github.com/groeck/linux:
  hwmon: (coretemp) Avoid leaving around dangling pointer
  hwmon: (coretemp) Fixup platform device ID change

13 years agoMerge git://github.com/davem330/ide
Linus Torvalds [Mon, 3 Oct 2011 19:53:43 +0000 (12:53 -0700)]
Merge git://github.com/davem330/ide

* git://github.com/davem330/ide:
  ide-disk: Fix request requeuing

13 years agoMerge branch 'btrfs-3.0' of git://github.com/chrismason/linux
Linus Torvalds [Mon, 3 Oct 2011 19:17:44 +0000 (12:17 -0700)]
Merge branch 'btrfs-3.0' of git://github.com/chrismason/linux

* 'btrfs-3.0' of git://github.com/chrismason/linux:
  Btrfs: force a page fault if we have a shorty copy on a page boundary

13 years agoide-disk: Fix request requeuing
Borislav Petkov [Mon, 3 Oct 2011 18:28:18 +0000 (14:28 -0400)]
ide-disk: Fix request requeuing

Simon Kirby reported that on his RAID setup with idedisk underneath
the box OOMs after a couple of days of runtime. Running with
CONFIG_DEBUG_KMEMLEAK pointed to idedisk_prep_fn() which unconditionally
allocates an ide_cmd struct. However, ide_requeue_and_plug() can be
called more than once per request, either from the request issue or the
IRQ handler path and do blk_peek_request() ends up in idedisk_prep_fn()
repeatedly, allocating a struct ide_cmd everytime and "forgetting" the
previous pointer.

Make sure the code reuses the old allocated chunk.

Reported-and-tested-by: Simon Kirby <sim@hostway.ca>
Cc: <stable@kernel.org> [ 39.x, 3.0.x ]
Link: http://marc.info/?l=linux-kernel&m=131667641517919
Link: http://lkml.kernel.org/r/20110922072643.GA27232@hostway.ca
Signed-off-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agopch_gbe: Fixed the issue on which a network freezes
Toshiharu Okada [Sun, 25 Sep 2011 21:27:43 +0000 (21:27 +0000)]
pch_gbe: Fixed the issue on which a network freezes

The pch_gbe driver has an issue which a network stops,
when receiving traffic is high.
In the case, The link down and up are necessary to return a network.

This patch fixed this issue.

Signed-off-by: Toshiharu Okada <toshiharu-linux@dsn.okisemi.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agopch_gbe: Fixed the issue on which PC was frozen when link was downed.
Toshiharu Okada [Sun, 25 Sep 2011 21:27:42 +0000 (21:27 +0000)]
pch_gbe: Fixed the issue on which PC was frozen when link was downed.

When a link was downed during network use,
there is an issue on which PC freezes.

This patch fixed this issue.

Signed-off-by: Toshiharu Okada <toshiharu-linux@dsn.okisemi.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agomake PACKET_STATISTICS getsockopt report consistently between ring and non-ring
Willem de Bruijn [Fri, 30 Sep 2011 10:38:28 +0000 (10:38 +0000)]
make PACKET_STATISTICS getsockopt report consistently between ring and non-ring

This is a minor change.

Up until kernel 2.6.32, getsockopt(fd, SOL_PACKET, PACKET_STATISTICS,
...) would return total and dropped packets since its last invocation. The
introduction of socket queue overflow reporting [1] changed drop
rate calculation in the normal packet socket path, but not when using a
packet ring. As a result, the getsockopt now returns different statistics
depending on the reception method used. With a ring, it still returns the
count since the last call, as counts are incremented in tpacket_rcv and
reset in getsockopt. Without a ring, it returns 0 if no drops occurred
since the last getsockopt and the total drops over the lifespan of
the socket otherwise. The culprit is this line in packet_rcv, executed
on a drop:

drop_n_acct:
        po->stats.tp_drops = atomic_inc_return(&sk->sk_drops);

As it shows, the new drop number it taken from the socket drop counter,
which is not reset at getsockopt. I put together a small example
that demonstrates the issue [2]. It runs for 10 seconds and overflows
the queue/ring on every odd second. The reported drop rates are:
ring: 16, 0, 16, 0, 16, ...
non-ring: 0, 15, 0, 30, 0, 46, 0, 60, 0 , 74.

Note how the even ring counts monotonically increase. Because the
getsockopt adds tp_drops to tp_packets, total counts are similarly
reported cumulatively. Long story short, reinstating the original code, as
the below patch does, fixes the issue at the cost of additional per-packet
cycles. Another solution that does not introduce per-packet overhead
is be to keep the current data path, record the value of sk_drops at
getsockopt() at call N in a new field in struct packetsock and subtract
that when reporting at call N+1. I'll be happy to code that, instead,
it's just more messy.

[1] http://patchwork.ozlabs.org/patch/35665/
[2] http://kernel.googlecode.com/files/test-packetsock-getstatistics.c

Signed-off-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agonet: xen-netback: correctly restart Tx after a VM restore/migrate
David Vrabel [Fri, 30 Sep 2011 06:37:51 +0000 (06:37 +0000)]
net: xen-netback: correctly restart Tx after a VM restore/migrate

If a VM is saved and restored (or migrated) the netback driver will no
longer process any Tx packets from the frontend.  xenvif_up() does not
schedule the processing of any pending Tx requests from the front end
because the carrier is off.  Without this initial kick the frontend
just adds Tx requests to the ring without raising an event (until the
ring is full).

This was caused by 47103041e91794acdbc6165da0ae288d844c820b (net:
xen-netback: convert to hw_features) which reordered the calls to
xenvif_up() and netif_carrier_on() in xenvif_connect().

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Cc: Ian Campbell <ian.campbell@citrix.com>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agobonding: properly stop queuing work when requested
Andy Gospodarek [Fri, 23 Sep 2011 10:53:34 +0000 (10:53 +0000)]
bonding: properly stop queuing work when requested

During a test where a pair of bonding interfaces using ARP monitoring
were both brought up and torn down (with an rmmod) repeatedly, a panic
in the timer code was noticed.  I tracked this down and determined that
any of the bonding functions that ran as workqueue handlers and requeued
more work might not properly exit when the module was removed.

There was a flag protected by the bond lock called kill_timers that is
set when the interface goes down or the module is removed, but many of
the functions that monitor link status now unlock the bond lock to take
rtnl first.  There is a chance that another CPU running the rmmod could
get the lock and set kill_timers after the first check has passed.

This patch does not allow any function to queue work that will make
itself run unless kill_timers is not set.  I also noticed while doing
this work that bond_resend_igmp_join_requests did not have a check for
kill_timers, so I added the needed call there as well.

Signed-off-by: Andy Gospodarek <andy@greyhouse.net>
Reported-by: Liang Zheng <lzheng@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
13 years agodrm/radeon: Set cursor x/y to 0 when x/yorigin > 0.
Michel Dänzer [Fri, 30 Sep 2011 15:16:53 +0000 (17:16 +0200)]
drm/radeon: Set cursor x/y to 0 when x/yorigin > 0.

Apart from the obvious cleanup, this should make the line

cursor_end = x - xorigin + w;

correct now.

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agodrm/radeon: Update AVIVO cursor coordinate origin before x/yorigin calculation.
Michel Dänzer [Fri, 30 Sep 2011 15:16:52 +0000 (17:16 +0200)]
drm/radeon: Update AVIVO cursor coordinate origin before x/yorigin calculation.

Fixes cursor disappearing prematurely when moving off a top/left edge which
is not located at the desktop top/left edge.

Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>