Calling into the new API to reset the secondary bus results in a deadlock.
This occurs because the device/bus is already locked at probe time.
Reverting back to the old behavior while the API is improved.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=200985
Fixes: c6a44ba9 ("PCI: Rename pci_try_reset_bus() to pci_reset_bus()")
Fixes: 409888e0 ("IB/hfi1: Use pci_try_reset_bus() for initiating PCI Secondary Bus Reset")
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Reviewed-by: NMichael J. Ruhl <michael.j.ruhl@intel.com>
Cc: Sinan Kaya <okaya@codeaurora.org>

There are several blockable mmu notifiers which might sleep in
mmu_notifier_invalidate_range_start and that is a problem for the
oom_reaper because it needs to guarantee a forward progress so it cannot
depend on any sleepable locks.

Currently we simply back off and mark an oom victim with blockable mmu
notifiers as done after a short sleep.  That can result in selecting a new
oom victim prematurely because the previous one still hasn't torn its
memory down yet.

We can do much better though.  Even if mmu notifiers use sleepable locks
there is no reason to automatically assume those locks are held.  Moreover
majority of notifiers only care about a portion of the address space and
there is absolutely zero reason to fail when we are unmapping an unrelated
range.  Many notifiers do really block and wait for HW which is harder to
handle and we have to bail out though.

This patch handles the low hanging fruit.
__mmu_notifier_invalidate_range_start gets a blockable flag and callbacks
are not allowed to sleep if the flag is set to false.  This is achieved by
using trylock instead of the sleepable lock for most callbacks and
continue as long as we do not block down the call chain.

I think we can improve that even further because there is a common pattern
to do a range lookup first and then do something about that.  The first
part can be done without a sleeping lock in most cases AFAICS.

The oom_reaper end then simply retries if there is at least one notifier
which couldn't make any progress in !blockable mode.  A retry loop is
already implemented to wait for the mmap_sem and this is basically the
same thing.

The simplest way for driver developers to test this code path is to wrap
userspace code which uses these notifiers into a memcg and set the hard
limit to hit the oom.  This can be done e.g.  after the test faults in all
the mmu notifier managed memory and set the hard limit to something really
small.  Then we are looking for a proper process tear down.

[akpm@linux-foundation.org: coding style fixes]
[akpm@linux-foundation.org: minor code simplification]
Link: http://lkml.kernel.org/r/20180716115058.5559-1-mhocko@kernel.orgSigned-off-by: NMichal Hocko <mhocko@suse.com>
Acked-by: Christian König <christian.koenig@amd.com> # AMD notifiers
Acked-by: Leon Romanovsky <leonro@mellanox.com> # mlx and umem_odp
Reported-by: NDavid Rientjes <rientjes@google.com>
Cc: "David (ChunMing) Zhou" <David1.Zhou@amd.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: David Airlie <airlied@linux.ie>
Cc: Jani Nikula <jani.nikula@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Doug Ledford <dledford@redhat.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Mike Marciniszyn <mike.marciniszyn@intel.com>
Cc: Dennis Dalessandro <dennis.dalessandro@intel.com>
Cc: Sudeep Dutt <sudeep.dutt@intel.com>
Cc: Ashutosh Dixit <ashutosh.dixit@intel.com>
Cc: Dimitri Sivanich <sivanich@sgi.com>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: "Jérôme Glisse" <jglisse@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Felix Kuehling <felix.kuehling@amd.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Building UCM with CONFIG_INFINIBAND_USER_ACCESS=m results in a
set of link errors including:

drivers/infiniband/core/ucm.o: In function `ib_ucm_event_handler':
ucm.c:(.text+0x6dc): undefined reference to `ib_copy_path_rec_to_user'
drivers/infiniband/core/ucma.o: In function `ucma_event_handler':
ucma.c:(.text+0xdc0): undefined reference to `ib_copy_ah_attr_to_user'

To get it to build-test again, this makes the option itself a
tristate, which lets Kconfig figure out the dependency correctly.

Fixes: 486edfb1 ("IB/ucm: Fix compiling ucm.c")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

If the system BIOS does not supply NUMA node information to the
PCI devices, the NUMA node is selected by choosing the current
node.

This can lead to the following crash:

divide error: 0000 SMP
CPU: 0 PID: 4 Comm: kworker/0:0 Tainted: G          IOE
------------   3.10.0-693.21.1.el7.x86_64 #1
Hardware name: Intel Corporation S2600KP/S2600KP, BIOS
SE5C610.86B.01.01.0005.101720141054 10/17/2014
Workqueue: events work_for_cpu_fn
task: ffff880174480fd0 ti: ffff880174488000 task.ti: ffff880174488000
RIP: 0010: [<ffffffffc020ac69>] hfi1_dev_affinity_init+0x129/0x6a0 [hfi1]
RSP: 0018:ffff88017448bbf8  EFLAGS: 00010246
RAX: 0000000000000011 RBX: ffff88107ffba6c0 RCX: ffff88085c22e130
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff880824ad0000
RBP: ffff88017448bc48 R08: 0000000000000011 R09: 0000000000000002
R10: ffff8808582b6ca0 R11: 0000000000003151 R12: ffff8808582b6ca0
R13: ffff8808582b6518 R14: ffff8808582b6010 R15: 0000000000000012
FS:  0000000000000000(0000) GS:ffff88085ec00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007efc707404f0 CR3: 0000000001a02000 CR4: 00000000001607f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Call Trace:
 hfi1_init_dd+0x14b3/0x27a0 [hfi1]
 ? pcie_capability_write_word+0x46/0x70
 ? hfi1_pcie_init+0xc0/0x200 [hfi1]
 do_init_one+0x153/0x4c0 [hfi1]
 ? sched_clock_cpu+0x85/0xc0
 init_one+0x1b5/0x260 [hfi1]
 local_pci_probe+0x4a/0xb0
 work_for_cpu_fn+0x1a/0x30
 process_one_work+0x17f/0x440
 worker_thread+0x278/0x3c0
 ? manage_workers.isra.24+0x2a0/0x2a0
 kthread+0xd1/0xe0
 ? insert_kthread_work+0x40/0x40
 ret_from_fork+0x77/0xb0
 ? insert_kthread_work+0x40/0x40

If the BIOS is not supplying NUMA information:
  - set the default table count to 1 for all possible nodes
  - select node 0 (instead of current NUMA) node to get consistent
    performance
  - generate an error indicating that the BIOS should be upgraded
Reviewed-by: NGary Leshner <gary.s.leshner@intel.com>
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NMichael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

hns bitmap allocation functions return 0 on success and -1 on failure.
Callers of these functions wrongly used their return value as an errno,
fix that by making a proper conversion.

Fixes: a598c6f4 ("IB/hns: Simplify function of pd alloc and qp alloc")
Signed-off-by: NGal Pressman <pressmangal@gmail.com>
Acked-by: NLijun Ou <oulijun@huawei.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Filter functions returns either 0 or 1, therefore better change their
return type from int to bool to reflect the same.  Additionally some
filter functions have suffix of _filter some doesn't.  Make all filter
function consistent to have __filter suffix to improve code readability.
Signed-off-by: NParav Pandit <parav@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Update all GID table entries of the netdevice whose MAC address changed.
Signed-off-by: NParav Pandit <parav@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Currently following issues exist:
1. Default GIDs of the lower (slave) netdevice if the bond netdevice is
   added. Rather default GID should be of bond master netdevice.
2. Due to this, when failover event occurs FAILOVER event handler attempts
   to delete the GID of the upper device and tries to add the default GID
   of the lower device. This is incorrect behavior.

To have simple and correct code:
(a) Split default GIDs addition out of add_netdev_ips().  This allows
    easier removal in future if RoCE default GIDs are removed.
(b) Add default GIDs of the bond master device by using right filter and
    callback function.
(c) Remove unused function enum_netdev_default_gids().
Signed-off-by: NParav Pandit <parav@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Now that we correctly delete the default GIDs of lower devices during
CHANGEUPPER event, add default GIDs of the bonding master device.
Signed-off-by: NParav Pandit <parav@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

When NETDEV_CHANGEUPPER event occurs, lower device is not yet established
as slave of the master, and when upper device is bond device, default GID
entries not deleted.

Due to this, when bond device is fully configured, default GID entries of
bond device cannot be added as default GID entries are occupied by the
lower netdevice. This is incorrect.

Default GID entries should really be of bond netdevice because in all RoCE
GIDs (default or IP), MAC address of the bond device will be used.  It is
confusing to have default GID of netdevice which is not really used for
any purpose.

Therefore, as first step, implement
(a) filter function which filters if a CHANGEUPPER event netdevice and
    associated upper device is master device or not.
(b) callback function which deletes the default GIDs of lower (event
    netdevice).
Signed-off-by: NParav Pandit <parav@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Currently bond_delete_netdev_default_gids() is called by two callers.
(a) del_netdev_default_ips_join()
(b) del_netdev_default_ips()

Both above functions changes the argument order while calling
bond_delete_netdev_default_gids().  This required silly
del_netdev_default_ips() wrapper.

Additionally, del_netdev_default_ips() deletes default GIDs not IP based
GIDs.  del_netdev_default_ips() having _ips suffix is confusing.

Therefore, get rid of confusing del_netdev_default_ips() and simplify
bond_delete_netdev_default_gids() to follow same argument order as its
caller.
Signed-off-by: NParav Pandit <parav@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Add comment for handling CHANGEUPPER netevent handling.
To improve code readability,
(a) move cmd definitions to its respective if-else branches,
(b) avoid single line structure definitions.
Signed-off-by: NParav Pandit <parav@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

This patch adds support for SRQ's created in user space and update
qedr_affiliated_event to deal with general SRQ events.
Signed-off-by: NMichal Kalderon <michal.kalderon@cavium.com>
Signed-off-by: NYuval Bason <yuval.bason@cavium.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Implement the SRQ specific verbs and update the poll_cq verb to deal with
SRQ completions.
Signed-off-by: NMichal Kalderon <michal.kalderon@cavium.com>
Signed-off-by: NYuval Bason <yuval.bason@cavium.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Today, we are using idr mechanism for QP's only.
This patch prepares the qedr_idr stuctures and the idr routines for
both QP's and SRQ's.
Signed-off-by: NYuval Bason <yuval.bason@cavium.com>
Signed-off-by: NMichal Kalderon <michal.kalderon@cavium.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

mlx5_ib_create_qp_resp was never initialized and only the first 4 bytes
were written.

Fixes: 41d902cb ("RDMA/mlx5: Fix definition of mlx5_ib_create_qp_resp")
Cc: <stable@vger.kernel.org>
Acked-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Even though this interface is marked CONFIG_BROKEN we still expect it to
compile, at least until we delete it completely.

Also mark INFINIBAND_USER_ACCESS_UCM with COMPILE_TEST so these situations
can be detected.

Fixes: e7ff98ae ("RDMA/cma: Constify path record, ib_cm_event, listen_id pointers")
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Now that the ioctl path and uobjects are converted to use uverbs_api, it
is now safe to remove the disassociation protection from the common ioctl
code.

This completes the work to make destroy functions continue to work even
after device disassociation.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Everything now uses the uverbs_uapi data structure.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Convert the ioctl method syscall path to use the uverbs_api data
structures. The new uapi structure includes all the same information, just
in a different and more optimal way.

 - Use attr_bkey instead of 2 level radix trees for everything related to
   attributes. This includes the attribute storage, presence, and
   detection of missing mandatory attributes.
 - Avoid iterating over all attribute storage at finish, instead use
   find_first_bit with the attr_bkey to locate only those attrs that need
   cleanup.
 - Organize things to always run, and always rely on, cleanup. This
   avoids a bunch of tricky error unwind cases.
 - Locate the method using the radix tree, and locate the attributes
   using a very efficient incremental radix tree lookup
 - Use the precomputed destroy_bkey to handle uobject destruction
 - Use the precomputed allocation sizes and precomputed 'need_stack'
   to avoid maths in the fast path. This is optimal if userspace
   does not pass (many) unsupported attributes.

Overall this results in much better codegen for the attribute accessors,
everything is now stored in bitmaps or linear arrays indexed by attr_bkey.
The compiler can compute attr_bkey values at compile time for all method
attributes, meaning things like uverbs_attr_is_valid() now compile into
single instruction bit tests.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Several handlers need temporary allocations for the life of the method,
switch them to use the uverbs_alloc allocator.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Reviewed-by: NLeon Romanovsky <leonro@mellanox.com>

This is similar in spirit to devm, it keeps track of any allocations
linked to this method call and ensures they are all freed when the method
exits. Further, if there is space in the internal/onstack buffer then the
allocator will hand out that memory and avoid an expensive call to
kalloc/kfree in the syscall path.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Reviewed-by: NLeon Romanovsky <leonro@mellanox.com>

Memory in the bundle is valuable, do not waste it holding an 8 byte
pointer for the rare case of writing to a PTR_OUT. We can compute the
pointer by storing a small 1 byte array offset and the base address of the
uattr memory in the bundle private memory.

This also means we can access the kernel's copy of the ib_uverbs_attr, so
drop the copy of flags as well.

Since the uattr base should be private bundle information this also
de-inlines the already too big uverbs_copy_to inline and moves
create_udata into uverbs_ioctl.c so they can see the private struct
definition.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Reviewed-by: NLeon Romanovsky <leonro@mellanox.com>

This already existed as the anonymous 'ctx' structure, but this was not
really a useful form. Hoist this struct into bundle_priv and rework the
internal things to use it instead.

Move a bunch of the processing internal state into the priv and reduce the
excessive use of function arguments.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Reviewed-by: NLeon Romanovsky <leonro@mellanox.com>

Currently the struct uverbs_obj_type stored in the ib_uobject is part of
the .rodata segment of the module that defines the object. This is a
problem if drivers define new uapi objects as we will be left with a
dangling pointer after device disassociation.

Switch the uverbs_obj_type for struct uverbs_api_object, which is
allocated memory that is part of the uverbs_api and is guaranteed to
always exist. Further this moves the 'type_class' into this memory which
means access to the IDR/FD function pointers is also guaranteed. Drivers
cannot define new types.

This makes it safe to continue to use all uobjects, including driver
defined ones, after disassociation.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

This radix tree datastructure is intended to replace the 'hash' structure
used today for parsing ioctl methods during system calls. This first
commit introduces the structure and builds it from the existing .rodata
descriptions.

The so-called hash arrangement is actually a 5 level open coded radix tree.
This new version uses a 3 level radix tree built using the radix tree
library.

Overall this is much less code and much easier to build as the radix tree
API allows for dynamic modification during the building. There is a small
memory penalty to pay for this, but since the radix tree is allocated on
a per device basis, a few kb of RAM seems immaterial considering the
gained simplicity.

The radix tree is similar to the existing tree, but also has a 'attr_bkey'
concept, which is a small value'd index for each method attribute. This is
used to simplify and improve performance of everything in the next
patches.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Reviewed-by: NLeon Romanovsky <leonro@mellanox.com>
Reviewed-by: NMichael J. Ruhl <michael.j.ruhl@intel.com>

There is no reason for drivers to do this, the core code should take of
everything. The drivers will provide their information from rodata to
describe their modifications to the core's base uapi specification.

The core uses this to build up the runtime uapi for each device.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Reviewed-by: NMichael J. Ruhl <michael.j.ruhl@intel.com>
Reviewed-by: NLeon Romanovsky <leonro@mellanox.com>

This is missing a zeroing of the high bits of flags, and is also not
correct for big endian machines. Properly zero extend the 32 bit flags
into the 64 bit stack variable.
Reported-by: NMichael J. Ruhl <michael.j.ruhl@intel.com>
Fixes: bccd0622 ("IB/uverbs: Add UVERBS_ATTR_FLAGS_IN to the specs language")
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Reviewed-by: NMichael J. Ruhl <michael.j.ruhl@intel.com>

Every function that returns COMPST_ERROR must set wqe->status to another
value than IB_WC_SUCCESS before returning COMPST_ERROR. Fix the only code
path for which this is not yet the case.
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Cc: <stable@vger.kernel.org>
Reviewed-by: NYuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

This will allow FW to not send more data to TP (which would then need to
be buffered). Pass the negotiated TCP window scale to FW in the FLOWC WR.

Also refactor send_flowc() a bit to clean it up.
Signed-off-by: NSteve Wise <swise@opengridcomputing.com>
Signed-off-by: NPotnuri Bharat Teja <bharat@chelsio.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

[   61.182439] UBSAN: Undefined behaviour in drivers/infiniband/hw/mlx5/qp.c:5366:34
[   61.183673] shift exponent 4294967288 is too large for 32-bit type 'unsigned int'
[   61.185530] CPU: 0 PID: 639 Comm: qp Not tainted 4.18.0-rc1-00037-g4aa1d69a9c60-dirty #96
[   61.186981] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-2.fc27 04/01/2014
[   61.188315] Call Trace:
[   61.188661]  dump_stack+0xc7/0x13b
[   61.190427]  ubsan_epilogue+0x9/0x49
[   61.190899]  __ubsan_handle_shift_out_of_bounds+0x1ea/0x22f
[   61.197040]  mlx5_ib_create_wq+0x1c99/0x1d50
[   61.206632]  ib_uverbs_ex_create_wq+0x499/0x820
[   61.213892]  ib_uverbs_write+0x77e/0xae0
[   61.248018]  vfs_write+0x121/0x3b0
[   61.249831]  ksys_write+0xa1/0x120
[   61.254024]  do_syscall_64+0x7c/0x2a0
[   61.256178]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
[   61.259211] RIP: 0033:0x7f54bab70e99
[   61.262125] Code: 00 f3 c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89
[   61.268678] RSP: 002b:00007ffe1541c318 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[   61.271076] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f54bab70e99
[   61.273795] RDX: 0000000000000070 RSI: 0000000020000240 RDI: 0000000000000003
[   61.276982] RBP: 00007ffe1541c330 R08: 00000000200078e0 R09: 0000000000000002
[   61.280035] R10: 0000000000000000 R11: 0000000000000246 R12: 00000000004005c0
[   61.283279] R13: 00007ffe1541c420 R14: 0000000000000000 R15: 0000000000000000

Cc: <stable@vger.kernel.org> # 4.7
Fixes: 79b20a6c ("IB/mlx5: Add receive Work Queue verbs")
Cc: syzkaller <syzkaller@googlegroups.com>
Reported-by: NNoa Osherovich <noaos@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NKees Cook <keescook@chromium.org>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

sgid_attr is uninitialized on the stack, initialize it to NULL.

Fixes: 39839107 ("IB/cm: Replace members of sa_path_rec with 'struct sgid_attr *'")
Signed-off-by: NParav Pandit <parav@mellanox.com>
Reviewed-by: NYossi Itigin <yosefe@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>

Move all the checking for pkey and other validity to the __ipoib_vlan_add
function. This removes the last difference from the control flow
of the __ipoib_vlan_add to make the overall design simpler to
understand.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NErez Shitrit <erezsh@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

This fixes a bug in the netlink path where the vlan_rwsem was not
held around __ipoib_vlan_add causing the child_intfs to be manipulated
unsafely.

In the process this greatly simplifies the vlan_rwsem write side locking
to only cover a single non-sleeping statement.

This also further increases the safety of the removal ordering by holding
the netdev of the parent while the child is active to ensure most bugs
become either an oops on a NULL priv or a deadlock on the netdev refcount.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

Switching to priv_destructor and needs_free_netdev created a subtle
ordering problem in ipoib_remove_one.

Now that unregister_netdev frees the netdev and priv we must ensure that
the children are unregistered before trying to unregister the parent,
or child unregister will use after free.

The solution is to unregister the children, then parent, in the same batch
all while holding the rtnl_lock. This closes all the races where a new
child could have been added and ensures proper ordering.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

This mutex was introduced to deal with the deadlock formed by calling
unregister_netdev from within the sysfs callback of a netdev.

Now that we have priv_destructor and needs_free_netdev we can switch
to the more targeted solution of running the unregister from a
work queue. This avoids the deadlock and gets rid of the mutex.

The next patch in the series needs this mutex eliminated to create
atomicity of unregisteration.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

Now that the unregister_netdev flow for IPoIB no longer relies on external
code we can now introduce the use of priv_destructor and
needs_free_netdev.

The rdma_netdev flow is switched to use the netdev common priv_destructor
instead of the special free_rdma_netdev and the IPOIB ULP adjusted:
 - priv_destructor needs to switch to point to the ULP's destructor
   which will then call the rdma_ndev's in the right order
 - We need to be careful around the error unwind of register_netdev
   as it sometimes calls priv_destructor on failure
 - ULPs need to use ndo_init/uninit to ensure proper ordering
   of failures around register_netdev

Switching to priv_destructor is a necessary pre-requisite to using
the rtnl new_link mechanism.

The VNIC user for rdma_netdev should also be revised, but that is left for
another patch.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NDenis Drozdov <denisd@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

Now that we have a proper ndo_uninit, move code that naturally pairs
with the ndo_uninit into ndo_init. This allows the netdev core to natually
handle ordering.

This fixes the situation where register_netdev can fail before calling
ndo_init, in which case it wouldn't call ndo_uninit either.

Also move a bunch of duplicated init code that is shared between child
and parent for clarity. Now the child and parent register functions look
very similar.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

Currently uninit is sometimes done twice in error flows, and is sprinkled
a bit all over the place.

Improve the clarity of the design by moving all uninit only into
ndo_uinit.

Some duplication is removed:
 - Sometimes IPOIB_STOP_NEIGH_GC was done before unregister, but
   this duplicates the process in ipoib_neigh_hash_init
 - Flushing priv->wq was sometimes done before unregister,
   but that duplicates what has been done in ndo_uninit

Uniniting the IB event queue must remain before unregister_netdev as it
requires the RTNL lock to be dropped, this is moved to a helper to make
that flow really clear and remove some duplication in error flows.

If register_netdev fails (and ndo_init is NULL) then it almost always
calls ndo_uninit, which lets us remove all the extra code from the error
unwinds. The next patch in the series will close the 'almost always' hole
by pairing a proper ndo_init with ndo_uninit.
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

The neigh_reap_task is self restarting, but so long as we call
cancel_delayed_work_sync() it will be guaranteed to not be running and
never start again. Thus we don't need to have the racy
IPOIB_STOP_NEIGH_GC bit, or the confusing mismatch of places sometimes
calling flush_workqueue after the cancel.

This fixes a situation where the GC work could have been left running
in some rare situations.
Signed-off-by: NErez Shitrit <erezsh@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NJason Gunthorpe <jgg@mellanox.com>