i) by the time DM core calls the postsuspend hook the dm_noflush flag
has been cleared.  So the old thin_postsuspend did nothing.  We need to
use the presuspend hook instead.

ii) There was a race between bios leaving DM core and arriving in the
deferred queue.

thin_presuspend now sets a 'requeue' flag causing all bios destined for
that thin to be requeued back to DM core.  Then it requeues all held IO,
and all IO on the deferred queue (destined for that thin).  Finally
postsuspend clears the 'requeue' flag.
Signed-off-by: NJoe Thornber <ejt@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>

The spin lock in requeue_io() was held for too long, allowing deadlock.
Don't worry, due to other issues addressed in the following "dm thin:
fix noflush suspend IO queueing" commit, this code was never called.

Fix this by taking the spin lock for a much shorter period of time.
Signed-off-by: NJoe Thornber <ejt@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>

Ideally a thin pool would never run out of data space; the low water
mark would trigger userland to extend the pool before we completely run
out of space.  However, many small random IOs to unprovisioned space can
consume data space at an alarming rate.  Adjust your low water mark if
you're frequently seeing "out-of-data-space" mode.

Before this fix, if data space ran out the pool would be put in
PM_READ_ONLY mode which also aborted the pool's current metadata
transaction (data loss for any changes in the transaction).  This had a
side-effect of needlessly compromising data consistency.  And retry of
queued unserviceable bios, once the data pool was resized, could
initiate changes to potentially inconsistent pool metadata.

Now when the pool's data space is exhausted transition to a new pool
mode (PM_OUT_OF_DATA_SPACE) that allows metadata to be changed but data
may not be allocated.  This allows users to remove thin volumes or
discard data to recover data space.

The pool is no longer put in PM_READ_ONLY mode in response to the pool
running out of data space.  And PM_READ_ONLY mode no longer aborts the
pool's current metadata transaction.  Also, set_pool_mode() will now
notify userspace when the pool mode is changed.
Signed-off-by: NJoe Thornber <ejt@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>

If a thin metadata operation fails the current transaction will abort,
whereby causing potential for IO layers up the stack (e.g. filesystems)
to have data loss.  As such, set THIN_METADATA_NEEDS_CHECK_FLAG in the
thin metadata's superblock which:
1) requires the user verify the thin metadata is consistent (e.g. use
   thin_check, etc)
2) suggests the user verify the thin data is consistent (e.g. use fsck)

The only way to clear the superblock's THIN_METADATA_NEEDS_CHECK_FLAG is
to run thin_repair.

On metadata operation failure: abort current metadata transaction, set
pool in read-only mode, and now set the needs_check flag.

As part of this change, constraints are introduced or relaxed:
* don't allow a pool to transition to write mode if needs_check is set
* don't allow data or metadata space to be resized if needs_check is set
* if a thin pool's metadata space is exhausted: the kernel will now
  force the user to take the pool offline for repair before the kernel
  will allow the metadata space to be extended.

Also, update Documentation to include information about when the thin
provisioning target commits metadata, how it handles metadata failures
and running out of space.
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Signed-off-by: NJoe Thornber <ejt@redhat.com>

Commit b5330655 ("dm thin: handle metadata failures more consistently")
increased potential for the pool's mode to be changed in response to
metadata operation failures.

When the pool mode is changed it isn't synchronized with the mode in
pool_features stored in the target's context (ti->private) that is used
as the basis for (re)establishing the pool mode during resume via
bind_control_target.

It is important that we synchronize the pool mode when it is changed
otherwise the pool may experience and unexpected mode transition on the
next resume (especially if there was no new table load).
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Acked-by: NJoe Thornber <ejt@redhat.com>

Commit 55494bf2 ("dm snapshot: use dm-bufio") broke snapshots.
Before that 3.14-rc1 commit, loading a snapshot's list of exceptions
involved reading exception areas one by one into ps->area and inserting
those exceptions into the hash table.  Commit 55494bf2 changed
it so that dm-bufio with prefetch is used to load exceptions in batchs.
Exceptions are loaded correctly, but ps->area is left uninitialized.
When a new exception is allocated, it is stored in this uninitialized
ps->area which will be written to the disk.  This causes metadata
corruption.

Fix this corruption by copying the last area that was read via dm-bufio
into ps->area.
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>

Since DM_DEBUG_BLOCK_STACK_TRACING is a DM_PERSISTENT_DATA config option
move it from drivers/md/Kconfig to drivers/md/persistent-data/Kconfig.

Doing so fixes indentation for other DM config options.
Signed-off-by: NMike Snitzer <snitzer@redhat.com>

The memory allocated for the multiqueue policy's hash table doesn't need
to be physically contiguous.  Use vzalloc() instead of kzalloc().
Fedora has been carrying this fix since 10/10/2013.

Failure seen during creation of a 10TB cached device with a 2048 sector
block size and 411GB cache size:

 dmsetup: page allocation failure: order:9, mode:0x10c0d0
 CPU: 11 PID: 29235 Comm: dmsetup Not tainted 3.10.4 #3
 Hardware name: Supermicro X8DTL/X8DTL, BIOS 2.1a       12/30/2011
  000000000010c0d0 ffff880090941898 ffffffff81387ab4 ffff880090941928
  ffffffff810bb26f 0000000000000009 000000000010c0d0 ffff880090941928
  ffffffff81385dbc ffffffff815f3840 ffffffff00000000 000002000010c0d0
 Call Trace:
  [<ffffffff81387ab4>] dump_stack+0x19/0x1b
  [<ffffffff810bb26f>] warn_alloc_failed+0x110/0x124
  [<ffffffff81385dbc>] ? __alloc_pages_direct_compact+0x17c/0x18e
  [<ffffffff810bda2e>] __alloc_pages_nodemask+0x6c7/0x75e
  [<ffffffff810bdad7>] __get_free_pages+0x12/0x3f
  [<ffffffff810ea148>] kmalloc_order_trace+0x29/0x88
  [<ffffffff810ec1fd>] __kmalloc+0x36/0x11b
  [<ffffffffa031eeed>] ? mq_create+0x1dc/0x2cf [dm_cache_mq]
  [<ffffffffa031efc0>] mq_create+0x2af/0x2cf [dm_cache_mq]
  [<ffffffffa0314605>] dm_cache_policy_create+0xa7/0xd2 [dm_cache]
  [<ffffffffa0312530>] ? cache_ctr+0x245/0xa13 [dm_cache]
  [<ffffffffa031263e>] cache_ctr+0x353/0xa13 [dm_cache]
  [<ffffffffa012b916>] dm_table_add_target+0x227/0x2ce [dm_mod]
  [<ffffffffa012e8e4>] table_load+0x286/0x2ac [dm_mod]
  [<ffffffffa012e65e>] ? dev_wait+0x8a/0x8a [dm_mod]
  [<ffffffffa012e324>] ctl_ioctl+0x39a/0x3c2 [dm_mod]
  [<ffffffffa012e35a>] dm_ctl_ioctl+0xe/0x12 [dm_mod]
  [<ffffffff81101181>] vfs_ioctl+0x21/0x34
  [<ffffffff811019d3>] do_vfs_ioctl+0x3b1/0x3f4
  [<ffffffff810f4d2e>] ? ____fput+0x9/0xb
  [<ffffffff81050b6c>] ? task_work_run+0x7e/0x92
  [<ffffffff81101a68>] SyS_ioctl+0x52/0x82
  [<ffffffff81391d92>] system_call_fastpath+0x16/0x1b
Signed-off-by: NHeinz Mauelshagen <heinzm@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org

When remapping a block to the cache's fast device that is larger than
2TB we must not truncate the destination sector to 32bits.  The 32bit
temporary result of from_cblock() was being overflowed in
remap_to_cache() due to the logical left shift.

Use an intermediate 64bit type to store the 32bit from_cblock() result
to fix the overflow.
Signed-off-by: NHeinz Mauelshagen <heinzm@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org

It was always intended that a user could provide a thin metadata device
that is larger than the max supported by the on-disk format.  The extra
space would just go unused.

Unfortunately that never worked.  If the user attempted to use a larger
metadata device on creation they would get an error like the following:

 device-mapper: space map common: space map too large
 device-mapper: transaction manager: couldn't create metadata space map
 device-mapper: thin metadata: tm_create_with_sm failed
 device-mapper: table: 252:17: thin-pool: Error creating metadata object
 device-mapper: ioctl: error adding target to table

Fix this by allowing the initial metadata space map creation to cap its
size at the max number of blocks supported (DM_SM_METADATA_MAX_BLOCKS).
get_metadata_dev_size() must also impose DM_SM_METADATA_MAX_BLOCKS (via
THIN_METADATA_MAX_SECTORS), otherwise extending metadata would cap at
THIN_METADATA_MAX_SECTORS_WARNING (which is larger than supported).

Also, the calculation for THIN_METADATA_MAX_SECTORS didn't account for
the sizeof the disk_bitmap_header.  So the supported maximum metadata
size is a bit smaller (reduced from 33423360 to 33292800 sectors).

Lastly, remove the "excess space will not be used" warning message from
get_metadata_dev_size(); it resulted in printing the warning multiple
times.  Factor out warn_if_metadata_device_too_big(), call it from
pool_ctr() and maybe_resize_metadata_dev().
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Acked-by: NJoe Thornber <ejt@redhat.com>

An invalid ioctl will never be valid, irrespective of whether multipath
has active paths or not.  So for invalid ioctls we do not have to wait
for multipath to activate any paths, but can rather return an error
code immediately.  This fix resolves numerous instances of:

 udevd[]: worker [] unexpectedly returned with status 0x0100

that have been seen during testing.
Signed-off-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org

dm_pool_close_thin_device() must be called if dm_set_target_max_io_len()
fails in thin_ctr().  Otherwise __pool_destroy() will fail because the
pool will still have an open thin device:

 device-mapper: thin metadata: attempt to close pmd when 1 device(s) are still open
 device-mapper: thin: __pool_destroy: dm_pool_metadata_close() failed.

Also, must establish error code if failing thin_ctr() because the pool
is in fail_io mode.
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Acked-by: NJoe Thornber <ejt@redhat.com>
Cc: stable@vger.kernel.org

When restoring bi_end_io, increase bi_remaining before retrying the bio
to avoid BUG_ON(atomic_read(&bio->bi_remaining) <= 0) in bio_endio().
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>

Commit 003b5c57 ("block: Convert drivers
to immutable biovecs") broke dm-mirror due to dm-io breakage.

dm-io had three possible iterators (DM_IO_PAGE_LIST, DM_IO_BVEC,
DM_IO_VMA) that iterate over pages where the I/O should be performed.

The switch to immutable biovecs changed the DM_IO_BVEC iterator to
DM_IO_BIO.  Before this change the iterator stored the pointer to a bio
vector in the dpages structure.  The iterator incremented the pointer in
the dpages structure as it advanced over the pages.  After the immutable
biovecs change, the DM_IO_BIO iterator stores a pointer to the bio in
the dpages structure and uses bio_advance to change the bio as it
advances.

The problem is that the function dispatch_io stores the content of the
dpages structure into the variable old_pages and restores it before
issuing I/O to each of the devices.  Before the change, the statement
"*dp = old_pages;" restored the iterator to its starting position.
After the change, struct dpages holds a pointer to the bio, thus the
statement "*dp = old_pages;" doesn't restore the iterator.

Consequently, in the context of dm-mirror: only the first mirror leg is
written correctly, the kernel locks up when trying to write the other
mirror legs because the number of sectors to write in the where->count
variable doesn't match the number of sectors returned by the iterator.

This patch fixes the bug by partially reverting the original patch - it
changes the code so that struct dpages holds a pointer to the bio vector,
so that the statement "*dp = old_pages;" restores the iterator correctly.

The field "context_u" holds the offset from the beginning of the current
bio vector entry, just like the "bio->bi_iter.bi_bvec_done" field.
Signed-off-by: NMikulas Patocka <mpatocka@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>

Commit 905e51b3 ("dm thin: commit outstanding data every second")
introduced a periodic commit.  This commit occurs regardless of whether
any thin devices have made changes.

Fix the periodic commit to check if any of a pool's thin devices have
changed using dm_pool_changed_this_transaction().
Reported-by: NAlexander Larsson <alexl@redhat.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Acked-by: NJoe Thornber <ejt@redhat.com>
Cc: stable@vger.kernel.org

When completing an overwrite bio, in overwrite_endio(), the associated
migration should not be added to the 'completed_migrations' until the
bio's fields are restored with dm_unhook_bio().

Otherwise, do_worker() can race to process 'completed_migrations' before
dm_unhook_bio() -- so the bio's bi_end_io is incorrect.  This is
unlikely to cause any problems given the current code but should be
fixed on the basis of correctness.

Also, the cache's spinlock only needs to be held when manipulating the
'completed_migrations' list -- other changes don't need protection.
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Acked-by: NJoe Thornber <ejt@redhat.com>

Commit c9d28d5d ("dm cache: promotion optimisation for writes")
incorrectly placed the 'hook_info' member in the writethrough-only
portion of the per_bio_data structure.

Given that the overwrite optimization may be used for writeback the
'hook_info' member must be placed above the 'cache' member of the
per_bio_data structure.  Any members above 'cache' are available from
both writeback and writethrough modes' per_bio_data structure.
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Acked-by: NJoe Thornber <ejt@redhat.com>
Cc: stable@vger.kernel.org # 3.13+

Currently, of_match_node compares each given match against all node's
compatible strings with of_device_is_compatible.

To achieve multiple compatible strings per node with ordering from
specific to generic, this requires given matches to be ordered from
specific to generic. For most of the drivers this is not true and also
an alphabetical ordering is more sane there.

Therefore, this patch introduces a function to match each of the node's
compatible strings against all given compatible matches without type and
name first, before checking the next compatible string. This implies
that node's compatibles are ordered from specific to generic while
given matches can be in any order. If we fail to find such a match
entry, then fall-back to the old method in order to keep compatibility.

Cc: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Signed-off-by: NKevin Hao <haokexin@gmail.com>
Tested-by: NStephen Chivers <schivers@csc.com>
Signed-off-by: NRob Herring <robh@kernel.org>

Because the offload mechanism can fall back to a standard transfer,
having two seperate initialization states is unfortunate. Let's just
have one state which does things consistently. This fixes a bug where
some preparation was missing when the fallback happened. And it makes
the code much easier to follow. To implement this, we put the check
if offload is possible at the top of the offload setup function.
Signed-off-by: NWolfram Sang <wsa@the-dreams.de>
Tested-by: NGregory CLEMENT <gregory.clement@free-electrons.com>
Cc: stable@vger.kernel.org # v3.12+
Fixes: 930ab3d4 (i2c: mv64xxx: Add I2C Transaction Generator support)

This reverts commit 10535314.
Stephen Chivers reported this is broken as we will get a match
entry '.type = "serial"' instead of the '.compatible = "ns16550"'
in the following scenario:
	serial0: serial@4500 {
		compatible = "fsl,ns16550", "ns16550";
	}

	struct of_device_id of_platform_serial_table[] = {
		{ .compatible = "ns8250",   .data = (void *)PORT_8250, },
		{ .compatible = "ns16450",  .data = (void *)PORT_16450, },
		{ .compatible = "ns16550a", .data = (void *)PORT_16550A, },
		{ .compatible = "ns16550",  .data = (void *)PORT_16550, },
		{ .compatible = "ns16750",  .data = (void *)PORT_16750, },
		{ .compatible = "ns16850",  .data = (void *)PORT_16850, },
		...
		{ .type = "serial",         .data = (void *)PORT_UNKNOWN, },
		{ /* end of list */ },
	};

So just revert this patch, we will use another implementation to find
the best compatible match in a follow-on patch.
Reported-by: NStephen N Chivers <schivers@csc.com.au>
Cc: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Signed-off-by: NKevin Hao <haokexin@gmail.com>
Signed-off-by: NRob Herring <robh@kernel.org>

This reverts commit f0de8e04, it is
incorrect, a future patch will fix this up properly.

Cc: Maxime Ripard <maxime.ripard@free-electrons.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

We should use rdma_vlan_dev_real_dev() instead of using vlan_dev_real_dev()
when building the GID table for a vlan interface.
Signed-off-by: NDevesh Sharma <devesh.sharma@emulex.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

Use correct value for obtaining traffic class from device
response for Query QP request.
Signed-off-by: NDevesh Sharma <devesh.sharma@emulex.com>
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

We use "tx_desc" again after we free it.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Acked-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

Fix a possible NULL pointer dereference in disconnection flow. This
can happen if the target disconnected/rejected the connection request,
e.g before the binding stage between iscsi connection to the transport
connection.
Signed-off-by: NAlex Tabachnik <alext@mellanox.com>
Signed-off-by: NRoi Dayan <roid@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

Error reported at http://marc.info/?l=linux-rdma&m=138995755801039&w=2

Fix short to int cast for big endian systems.
Signed-off-by: NUpinder Malhi <umalhi@cisco.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

We're using edac_mc_workq_setup() both on the init path, when
we load an edac driver and when we change the polling period
(edac_mc_reset_delay_period) through /sys/.../edac_mc_poll_msec.

On that second path we don't need to init the workqueue which has been
initialized already.

Thanks to Tejun for workqueue insights.
Signed-off-by: NBorislav Petkov <bp@suse.de>
Link: http://lkml.kernel.org/r/1391457913-881-1-git-send-email-prarit@redhat.com
Cc: <stable@vger.kernel.org>

Sanitize code even more to accept unsigned longs only and to not allow
polling intervals below 1 second as this is unnecessary and doesn't make
much sense anyway for polling errors.
Signed-off-by: NBorislav Petkov <bp@suse.de>
Link: http://lkml.kernel.org/r/1391457913-881-1-git-send-email-prarit@redhat.com
Cc: Doug Thompson <dougthompson@xmission.com>
Cc: <stable@vger.kernel.org>

Without that patch, a user can't select the imxfb driver when the i.MX25
and/or the i.MX27 device tree board are selected and that no boards that
selects IMX_HAVE_PLATFORM_IMX_FB are compiled in.

Cc: Eric Bénard <eric@eukrea.com>
Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com>
Cc: Sascha Hauer <kernel@pengutronix.de>
Cc: Tomi Valkeinen <tomi.valkeinen@ti.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-fbdev@vger.kernel.org
Signed-off-by: NDenis Carikli <denis@eukrea.com>
Acked-by: NShawn Guo <shawn.guo@linaro.org>
Acked-by: NJean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Acked-by: NSascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

Enable S6E8AX0 LCD driver only if LCD_CLASS_DEVICE is a built-in driver.
Else we get the following errors due to missing symbols:
drivers/built-in.o: In function `s6e8ax0_probe':
:(.text+0x51aec): undefined reference to `lcd_device_register'
:(.text+0x51c44): undefined reference to `lcd_device_unregister'
Signed-off-by: NSachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: NTomi Valkeinen <tomi.valkeinen@ti.com>

Remove Kconfig dependency of mlx5_ib/mlx5_core on X86, since there is
no such dependency in reality.
Signed-off-by: NEli Cohen <eli@mellanox.com>
Signed-off-by: NMichael Neuling <mikey@neuling.org>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

The ntc thermistor code was doing math whose temporary result might
have overflowed 32-bits.  We need some casts in there to make it safe.

In one example I found:
- pullup_uV: 1800000
- result of iio_read_channel_raw: 3226
- 1800000 * 3226 => 0x15a1cbc80
Signed-off-by: NDoug Anderson <dianders@chromium.org>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: NGuenter Roeck <linux@roeck-us.net>

Research has shown that commit a77fcf89 ("IB/qib: Use a single
txselect module parameter for serdes tuning") missed a key serdes init
sequence.

This patch add that sequence.

Cc: <stable@vger.kernel.org>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

Signed-off-by: NKumar Sanghvi <kumaras@chelsio.com>
Signed-off-by: NHariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

For userspace RoCE UD QPs we need to know the GID format that the
kernel uses, e.g when working over older kernels. For that end, add a
new port capability IB_PORT_IP_BASED_GIDS and report it when query
port is issued.
Signed-off-by: NMoni Shoua <monis@mellanox.co.il>
Signed-off-by: NMatan Barak <matanb@mellanox.com>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

When scanning netdevices we need to check a few more conditions and
cases to build the IBoE GID table properly.  For example, under
bonding we must make sure that when a port is down, the bond IP
address isn't programmed as a GID, since doing so will cause failure
with IB core flows that selects ports by GID.
Signed-off-by: NMoni Shoua <monis@mellanox.co.il>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

The IBoE code used to reset the GID table did it for all Ethernet
ports of the device.  Since the whole architecture of generating GIDs
and responding to events is port-based, this is inefficient and can
lead to wrong content in the GID table.  Change the reset flow to be
per-port.
Signed-off-by: NMoni Shoua <monis@mellanox.co.il>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

Updating the GID table under IBoE requires read/write from/to shared
data structures.  These data structures are protected with the device
iboe lock.  The flows that modify the GID table start from

    1. Initializing the GID table
    2. NETDEV events
    3. INET or INET6 events

This patch makes sure that the flow of initializing the GID table is
consistent with the other two flows w.r.t on what step the lock is taken.
Signed-off-by: NMoni Shoua <monis@mellanox.co.il>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

On the one hand, the invocation of netdev_master_upper_dev_get()
within mlx4_ib_scan_netdevs() must be done with rtnl lock held.  On
the other hand, it's wrong to call rtnl_lock() from within this
function since it's also called by our netdev notifier callback.
Therefore move the locking to mlx4_ib_add() so that both cases are
covered.
Signed-off-by: NMoni Shoua <monis@mellanox.co.il>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>

Make sure that for Ethernet ports, the port GID table index 0 is always
occupied with a default GID of the relevant IPv6 link-local adderss.

This provides better user experience for legacy applications that don't use
the RDMA CM and were working on index 0 prior to the IP addressing change.

Also, as GIDs are generated from IP addresses of the network devices that
are associated with the port, it's basically possible that the GID table
will be empty if no IP address was assigned.  This doesn't comply with the
IB spec section 4.1.1 "GID usage and properties".
Signed-off-by: NMoni Shoua <monis@mellanox.co.il>
Signed-off-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NRoland Dreier <roland@purestorage.com>