In case the discovery session is carried over iser, we can't
access the assumed network portal since the default portal is
used. In this case we don't really need to allocate the fastreg
pool, just prepare to the text pdu that will follow.
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Reported-by: Alex Tabachnik <alext@mellanox.com>
Cc: stable@vger.kernel.org # 3.15+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch fixes a iscsi_queue_req memory leak when ABORT_TASK response
has been queued by TFO->queue_tm_rsp() -> lio_queue_tm_rsp() after a
long standing I/O completes, but the connection has already reset and
waiting for cleanup to complete in iscsit_release_commands_from_conn()
-> transport_generic_free_cmd() -> transport_wait_for_tasks() code.

It moves iscsit_free_queue_reqs_for_conn() after the per-connection command
list has been released, so that the associated se_cmd tag can be completed +
released by target-core before freeing any remaining iscsi_queue_req memory
for the connection generated by lio_queue_tm_rsp().

Cc: Thomas Glanzmann <thomas@glanzmann.de>
Cc: Charalampos Pournaris <charpour@gmail.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch fixes a bug where multiple waiters on ->t_transport_stop_comp
occurs due to a concurrent ABORT_TASK and session reset both invoking
transport_wait_for_tasks(), while waiting for the associated se_cmd
descriptor backend processing to complete.

For this case, complete_all() should be invoked in order to wake up
both waiters in core_tmr_abort_task() + transport_generic_free_cmd()
process contexts.

Cc: Thomas Glanzmann <thomas@glanzmann.de>
Cc: Charalampos Pournaris <charpour@gmail.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch fixes a bug where se_cmd descriptors associated with a
Task Management Request (TMR) where not setting CMD_T_ACTIVE before
being dispatched into target_tmr_work() process context.

This is required in order for transport_generic_free_cmd() ->
transport_wait_for_tasks() to wait on se_cmd->t_transport_stop_comp
if a session reset event occurs while an ABORT_TASK is outstanding
waiting for another I/O to complete.

Cc: Thomas Glanzmann <thomas@glanzmann.de>
Cc: Charalampos Pournaris <charpour@gmail.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

Convert "x == true" to "x" and "x == false" to "!x".
Signed-off-by: Christophe Vu-Brugier <cvubrugier@yahoo.fr>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

Signed-off-by: Christophe Vu-Brugier <cvubrugier@yahoo.fr>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch changes ft_queue_status() to set SAM_STAT_TASK_SET_FULL
status upon lport->tt.seq_send( failure, and return -EAGAIN to notify
target-core to attempt to requeue the response.

It also does the same for a fc_frame_alloc() failures, in order to
signal the initiator that it should try to reduce it's current
queue_depth, to lower the number of outstanding I/Os on the wire.
Reported-by: Vasu Dev <vasu.dev@linux.intel.com>
Reviewed-by: Vasu Dev <vasu.dev@linux.intel.com>
Cc: Jun Wu <jwu@stormojo.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch changes ft_queue_data_in() to set SAM_STAT_TASK_SET_FULL
status upon a lport->tt.seq_send() failure, where it will now stop
sending subsequent DataIN, and immediately attempt to send the
response with exception status.

Sending a response with SAM_STAT_TASK_SET_FULL status is useful in
order to signal the initiator that it should try to reduce it's
current queue_depth, to lower the number of outstanding I/Os on
the wire.

Also, add a check to skip sending DataIN if TASK_SET_FULL status
has already been set due to a response lport->tt.seq_send()
failure, that has asked target-core to requeue a response.
Reported-by: Vasu Dev <vasu.dev@linux.intel.com>
Reviewed-by: Vasu Dev <vasu.dev@linux.intel.com>
Cc: Jun Wu <jwu@stormojo.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch adds an explicit check in chap_server_compute_md5() to ensure
the CHAP_C value received from the initiator during mutual authentication
does not match the original CHAP_C provided by the target.

This is in line with RFC-3720, section 8.2.1:

   Originators MUST NOT reuse the CHAP challenge sent by the Responder
   for the other direction of a bidirectional authentication.
   Responders MUST check for this condition and close the iSCSI TCP
   connection if it occurs.
Reported-by: Tejas Vaykole <tejas.vaykole@calsoftinc.com>
Cc: stable@vger.kernel.org # 3.1+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch removes a no-op iscsit_clear_tpg_np_login_threads() call
in iscsit_tpg_del_portal_group(), which is unnecessary because
iscsit_tpg_del_portal_group() can only ever be removed from configfs
once all of the child network portals have been released.

Also, go ahed and make iscsit_clear_tpg_np_login_threads() declared
as static.
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

The target is failing to handle list of CHAP_A key-value pair form
initiator.The target is expecting CHAP_A=5 always. In other cases,
where initiator sends list (for example) CHAP_A=6,5 target is failing
the security negotiation. Which is incorrect.

This patch handles the case (RFC 3720 section 11.1.4).
where in the initiator may send list of CHAP_A values and target replies
with appropriate CHAP_A value in response

(Drop whitespaces + rename to chap_check_algorithm + save original
 pointer + add explicit check for CHAP_A key - nab)
Signed-off-by: Tejas Vaykole <tejas.vaykole@calsoftinc.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

If the message "Unable to allocate…" pops up, it's useful to know
whether the problem is that the system is genuinely out of memory, or
that some bug has led to a crazy allocation length.

In particular this helped debug a corruption of login headers in
iscsi_login_non_zero_tsih_s1().
Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch changes iscsi_target_handle_csg_zero() to explicitly
reject login requests in SecurityNegotiation with a zero-length
payload, following the language in RFC-3720 Section 8.2:

  Whenever an iSCSI target gets a response whose keys, or their
  values, are not according to the step definition, it MUST answer
  with a Login reject with the "Initiator Error" or "Missing Parameter"
  status.

Previously when a zero-length login request in CSG=0 was received,
the target would send a login response with CSG=0 + T_BIT=0 asking
the initiator to complete authentication, and not fail the login
until MAX_LOGIN_PDUS was reached.  This change will now immediately
fail the login attempt with ISCSI_STATUS_CLS_INITIATOR_ERR status.
Reported-by: Tejas Vaykole <tejas.vaykole@calsoftinc.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch converts qla2xxx target code to use generic percpu_ida
tag allocation provided by target-core, thus removing the original
kmem_cache_zalloc() for each struct qla_tgt_cmd descriptor in the
incoming ATIO packet fast-path.

This includes the conversion of qlt_handle_cmd_for_atio() to perform
qla_tgt_sess lookup before dispatching a command descriptor into
qla_tgt_wq process context, along with handling the case where no
active session exists, and subsequently kicking off a seperate
process context for qlt_create_sess_from_atio() to create a new one.

It also includes moving tag allocation into generic code within
qlt_get_tag(), so that the same logic can be shared between
qlt_handle_cmd_for_atio() + qlt_create_sess_from_atio() contexts.
Also, __qlt_do_work() has been made generic between both normal
process context in qlt_do_work() + qlt_create_sess_from_atio().

Next, update qlt_free_cmd() to release the percpu-ida tags, and
drop the now-unused global qla_tgt_cmd_cachep.

Finally in tcm_qla2xxx code, tcm_qla2xxx_check_initiator_node_acl()
has been updated to use transport_init_session_tags() along with a
hardcoded TCM_QLA2XXX_DEFAULT_TAGS=2088 as the number of qla_tgt_cmd
descriptors to pre-allocate per qla_tgt_sess instance.

(Use ha->fw_xcb_count if available to calculate num_tags, and
 also factor in extra pad tags - Quinn)

Cc: Saurav Kashyap <saurav.kashyap@qlogic.com>
Cc: Quinn Tran <quinn.tran@qlogic.com>
Cc: Giridhar Malavali <giridhar.malavali@qlogic.com>
Cc: Chad Dupuis <chad.dupuis@qlogic.com>
Cc: Roland Dreier <roland@kernel.org>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch updates virtscsi_probe() to setup necessary Scsi_Host
level protection resources. (currently hardcoded to 1)

It changes virtscsi_add_cmd() to attach outgoing / incoming
protection SGLs preceeding the data payload, and is using the
new virtio_scsi_cmd_req_pi->pi_bytes[out,in] field to signal
to signal to vhost/scsi bytes to expect for protection data.

(Add missing #include <linux/blkdev.h> for blk_integrity - sfr + nab)
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Sagi Grimberg <sagig@dev.mellanox.co.il>
Cc: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch updates vhost_scsi_handle_vq() to check for the existance
of virtio_scsi_cmd_req_pi comparing vq->iov[0].iov_len in order to
calculate seperate data + protection SGLs from data_num.

Also update tcm_vhost_submission_work() to pass the pre-allocated
cmd->tvc_prot_sgl[] memory into target_submit_cmd_map_sgls(), and
update vhost_scsi_get_tag() parameters to accept scsi_tag, lun, and
task_attr.

Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Sagi Grimberg <sagig@dev.mellanox.co.il>
Cc: H. Peter Anvin <hpa@zytor.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch adds vhost_scsi_map_iov_to_prot() to perform the mapping of
T10 data integrity memory between virtio iov + struct scatterlist using
get_user_pages_fast() following existing code.

As with vhost_scsi_map_iov_to_sgl(), this does sanity checks against the
total prot_sgl_count vs. pre-allocated SGLs, and loops across protection
iovs using vhost_scsi_map_to_sgl() to perform the actual memory mapping.

Also update tcm_vhost_release_cmd() to release associated tvc_prot_sgl[]
struct page.

Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Sagi Grimberg <sagig@dev.mellanox.co.il>
Cc: H. Peter Anvin <hpa@zytor.com>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch updates tcm_vhost_make_nexus() to pre-allocate per descriptor
tcm_vhost_cmd->tvc_prot_sgl[] used to expose protection SGLs from within
virtio-scsi guest memory to vhost-scsi.

Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

Move the overflow check for sgl_count > TCM_VHOST_PREALLOC_SGLS into
vhost_scsi_map_iov_to_sgl() so that it's based on the total number
of SGLs for all IOVs, instead of single IOVs.

Also, rename TCM_VHOST_PREALLOC_PAGES -> TCM_VHOST_PREALLOC_UPAGES
to better describe pointers to user-space pages.

Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch adds a virtio_scsi_cmd_req_pi header as recommened by
Paolo that contains pi_bytesout + pi_bytesin elements used for
signaling when protection information buffers (in bytes) are
expected to preceed the data payload buffers.

Also add new VIRTIO_SCSI_F_T10_PI feature bit to be used to signal
host support.

Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Sagi Grimberg <sagig@dev.mellanox.co.il>
Cc: H. Peter Anvin <hpa@zytor.com>
Acked-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

In case user chose to set T10-PI enable on the target while
the IB device does not support it, gracefully reject the request.
Reported-by: Slava Shwartsman <valyushash@gmail.com>
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.15+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

disconnected_handler works are scheduled on system_wq.
When attempting to unload, first make sure all works
have cleaned up.
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

There are 4 RDMA_CM events that all basically mean that
the user should teardown the IB connection:
- DISCONNECTED
- ADDR_CHANGE
- DEVICE_REMOVAL
- TIMEWAIT_EXIT

Only in DISCONNECTED/ADDR_CHANGE it makes sense to
call rdma_disconnect (send DREQ/DREP to our initiator).
So we keep the same teardown handler for all of them
but only indicate calling rdma_disconnect for the relevant
events.

This patch also removes redundant debug prints for each single
event.

v2 changes:
 - Call isert_disconnected_handler() for DEVICE_REMOVAL (Or + Sag)
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

In ungraceful teardowns isert close flows seem racy such that
isert_wait_conn hangs as RDMA_CM_EVENT_DISCONNECTED never
gets invoked (no one called rdma_disconnect).

Both graceful and ungraceful teardowns will have rx flush errors
(isert posts a batch once connection is established). Once all
flush errors are consumed we invoke isert_wait_conn and it will
be responsible for calling rdma_disconnect. This way it can be
sure that rdma_disconnect was called and it won't wait forever.

This patch also removes the logout_posted indicator. either the
logout completion was consumed and no problem decrementing the
post_send_buf_count, or it was consumed as a flush error. no point
of keeping it for isert_wait_conn as there is no danger that
isert_conn will be accidentally removed while it is running.

(Drop unnecessary sleep_on_conn_wait_comp check in
 isert_cq_rx_comp_err - nab)
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

In case np_thread state is in RESET/SHUTDOWN/EXIT states,
no point for isert to stall there as we may get a hang in
case no one will wake it up later.
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

On each processed XCOPY command, two "kmalloc-512" memory objects are
leaked. These represent two allocations of struct xcopy_pt_cmd in
target_core_xcopy.c.

The reason for the memory leak is that the cmd_kref field is not
initialized (thus, it is zero because the allocations were done with
kzalloc). When we decrement zero kref in target_put_sess_cmd, the result
is not zero, thus target_release_cmd_kref is not called.

This patch fixes the bug by moving kref initialization from
target_get_sess_cmd to transport_init_se_cmd (this function is called from
target_core_xcopy.c, so it will correctly initialize cmd_kref). It can be
easily verified that all code that calls target_get_sess_cmd also calls
transport_init_se_cmd earlier, thus moving kref_init shouldn't introduce
any new problems.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org	# 3.12+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

Just like for pSCSI, if the transport sets get_write_cache, then it is
not valid to enable write cache emulation for it. Return an error.

see https://bugzilla.redhat.com/show_bug.cgi?id=1082675Reviewed-by: Chris Leech <cleech@redhat.com>
Signed-off-by: Andy Grover <agrover@redhat.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch explicitly disables Immediate + Unsolicited Data for ISER
connections during login in iscsi_login_zero_tsih_s2() when protection
has been enabled for the session by the underlying hardware.

This is currently required because protection / signature memory regions
(MRs) expect T10 PI to occur on RDMA READs + RDMA WRITEs transfers, and
not on a immediate data payload associated with ISCSI_OP_SCSI_CMD, or
unsolicited data-out associated with a ISCSI_OP_SCSI_DATA_OUT.

v2 changes:
  - Add TARGET_PROT_DOUT_INSERT check (Sagi)
  - Add pr_debug noisemaker (Sagi)
  - Add goto to avoid early return from MRDSL check (nab)

Cc: Sagi Grimberg <sagig@mellanox.com>
Cc: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch fixes a free-after-use regression in ft_free_cmd(), where
ft_sess_put() is called with cmd->sess after percpu_ida_free() has
already released the tag.

Fix this bug by saving the ft_sess pointer ahead of percpu_ida_free(),
and pass it directly to ft_sess_put().

The regression was originally introduced in v3.13-rc1 commit:

  commit 5f544cfa
  Author: Nicholas Bellinger <nab@daterainc.com>
  Date:   Mon Sep 23 12:12:42 2013 -0700

      tcm_fc: Convert to per-cpu command map pre-allocation of ft_cmd
Reported-by: Jun Wu <jwu@stormojo.com>
Cc: Mark Rustad <mark.d.rustad@intel.com>
Cc: Robert Love <robert.w.love@intel.com>
Cc: <stable@vger.kernel.org> #3.13+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

This patch changes an incorrect use of BUG_ON to instead generate a
REJECT + PROTOCOL_ERROR in iscsit_process_nop_out() code.  This case
can occur with traditional TCP where a flood of zeros in the data
stream can reach this block for what is presumed to be a NOP-OUT with
a solicited reply, but without a valid iscsi_cmd pointer.

This incorrect BUG_ON was introduced during the v3.11-rc timeframe
with the following commit:

commit 778de368
Author: Nicholas Bellinger <nab@linux-iscsi.org>
Date:   Fri Jun 14 16:07:47 2013 -0700

    iscsi/isert-target: Refactor ISCSI_OP_NOOP RX handling
Reported-by: Arshad Hussain <arshad.hussain@calsoftinc.com>
Cc: stable@vger.kernel.org # 3.11+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

When the target is in stop stage, iSER transport initiates RDMA disconnects.
The iSER initiator may wish to establish a new connection over the
still existing network portal. In this case iSER transport should not
accept and resume new RDMA connections. In order to learn that, iscsi_np
is added with enabled flag so the iSER transport can check when deciding
weather to accept and resume a new connection request.

The iscsi_np is enabled after successful transport setup, and disabled
before iscsi_np login threads are cleaned up.
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

RDMA CM and iSCSI target flows are asynchronous and completely
uncorrelated. Relying on the fact that iscsi_accept_np will be called
after CM connection request event and will wait for it is a mistake.

When attempting to login to a few targets this flow is racy and
unpredictable, but for parallel login to dozens of targets will
race and hang every time.

The correct synchronizing mechanism in this case is pending on
a semaphore rather than a wait_for_event. We keep the pending
interruptible for iscsi_np cleanup stage.

(Squash patch to remove dead code into parent - nab)
Reported-by: Slava Shwartsman <valyushash@gmail.com>
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

Should be adding list_add_tail($new, $head) and not
the other way around.
Signed-off-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.10+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

Userspace tools assume if a value is read from configfs, it is valid
and will not cause an error if the same value is written back. The only
valid value for pi_prot_type for backends not supporting DIF is 0, so allow
this particular value to be set without returning an error.
Reported-by: Krzysztof Chojnowski <frirajder@gmail.com>
Signed-off-by: Andy Grover <agrover@redhat.com>
Reviewed-by: Sagi Grimberg <sagig@mellanox.com>
Cc: stable@vger.kernel.org # 3.14+
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>

The asm-generic, big-endian version of zero_bytemask creates a mask of
bytes preceding the first zero-byte by left shifting ~0ul based on the
position of the first zero byte.

Unfortunately, if the first (top) byte is zero, the output of
prep_zero_mask has only the top bit set, resulting in undefined C
behaviour as we shift left by an amount equal to the width of the type.
As it happens, GCC doesn't manage to spot this through the call to fls(),
but the issue remains if architectures choose to implement their shift
instructions differently.

An example would be arch/arm/ (AArch32), where LSL Rd, Rn, #32 results
in Rd == 0x0, whilst on arch/arm64 (AArch64) LSL Xd, Xn, #64 results in
Xd == Xn.

Rather than check explicitly for the problematic shift, this patch adds
an extra shift by 1, replacing fls with __fls. Since zero_bytemask is
never called with a zero argument (has_zero() is used to check the data
first), we don't need to worry about calling __fls(0), which is
undefined.

Cc: <stable@vger.kernel.org>
Cc: Victor Kamensky <victor.kamensky@linaro.org>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

This merges the patch to fix possible loss of dirty bit on munmap() or
madvice(DONTNEED).  If there are concurrent writers on other CPU's that
have the unmapped/unneeded page in their TLBs, their writes to the page
could possibly get lost if a third CPU raced with the TLB flush and did
a page_mkclean() before the page was fully written.

Admittedly, if you unmap() or madvice(DONTNEED) an area _while_ another
thread is still busy writing to it, you deserve all the lost writes you
could get.  But we kernel people hold ourselves to higher quality
standards than "crazy people deserve to lose", because, well, we've seen
people do all kinds of crazy things.

So let's get it right, just because we can, and we don't have to worry
about it.

* safe-dirty-tlb-flush:
  mm: split 'tlb_flush_mmu()' into tlb flushing and memory freeing parts

Pull btrfs fixes from Chris Mason.

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  Btrfs: limit the path size in send to PATH_MAX
  Btrfs: correctly set profile flags on seqlock retry
  Btrfs: use correct key when repeating search for extent item
  Btrfs: fix inode caching vs tree log
  Btrfs: fix possible memory leaks in open_ctree()
  Btrfs: avoid triggering bug_on() when we fail to start inode caching task
  Btrfs: move btrfs_{set,clear}_and_info() to ctree.h
  btrfs: replace error code from btrfs_drop_extents
  btrfs: Change the hole range to a more accurate value.
  btrfs: fix use-after-free in mount_subvol()

Pull arm fixes from Russell King:
 "A number of fixes for the PJ4/iwmmxt changes which arm-soc forced me
  to take during the merge window.  This stuff should have been better
  tested and sorted out *before* the merge window"

* 'fixes' of git://ftp.arm.linux.org.uk/~rmk/linux-arm:
  ARM: 8042/1: iwmmxt: allow to build iWMMXt on Marvell PJ4B
  ARM: 8041/1: pj4: fix cpu_is_pj4 check
  ARM: 8040/1: pj4: properly detect existence of iWMMXt coprocessor
  ARM: 8039/1: pj4: enable iWMMXt only if CONFIG_IWMMXT is set
  ARM: 8038/1: iwmmxt: explicitly check for supported architectures

Pull arm64 fixes from Catalin Marinas:
 - compat renameat2 syscall wiring and __NR_compat_syscalls fix
 - TLB fix for transparent huge pages following switch to generic
   mmu_gather
 - spinlock initialisation for init_mm's context
 - move of_clk_init() earlier
 - Kconfig duplicate entry fix

* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
  arm64: init: Move of_clk_init to time_init
  arm64: initialize spinlock for init_mm's context
  arm64: debug: remove noisy, pointless warning
  arm64: mm: Add THP TLB entries to general mmu_gather
  arm64: add renameat2 compat syscall
  ARM64: Remove duplicated Kconfig entry for "kernel/power/Kconfig"
  arm64: __NR_compat_syscalls fix