stands out; by patch count lots of fixes to the mlx4 driver plus some
cleanups and fixes to the core and other drivers.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)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=DbLx
-----END PGP SIGNATURE-----
Merge tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
Pull InfiniBand/RDMA changes for the 3.4 merge window from Roland Dreier:
"Nothing big really stands out; by patch count lots of fixes to the
mlx4 driver plus some cleanups and fixes to the core and other
drivers."
* tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (28 commits)
mlx4_core: Scale size of MTT table with system RAM
mlx4_core: Allow dynamic MTU configuration for IB ports
IB/mlx4: Fix info returned when querying IBoE ports
IB/mlx4: Fix possible missed completion event
mlx4_core: Report thermal error events
mlx4_core: Fix one more static exported function
IB: Change CQE "csum_ok" field to a bit flag
RDMA/iwcm: Reject connect requests if cmid is not in LISTEN state
RDMA/cxgb3: Don't pass irq flags to flush_qp()
mlx4_core: Get rid of redundant ext_port_cap flags
RDMA/ucma: Fix AB-BA deadlock
IB/ehca: Fix ilog2() compile failure
IB: Use central enum for speed instead of hard-coded values
IB/iser: Post initial receive buffers before sending the final login request
IB/iser: Free IB connection resources in the proper place
IB/srp: Consolidate repetitive sysfs code
IB/srp: Use pr_fmt() and pr_err()/pr_warn()
IB/core: Fix SDR rates in sysfs
mlx4: Enforce device max FMR maps in FMR alloc
IB/mlx4: Set bad_wr for invalid send opcode
...
When destroying a listening cmid, the iwcm first marks the state of
the cmid as DESTROYING, then releases the lock and calls into the
iWARP provider to destroy the endpoint. Since the cmid is not locked,
its possible for the iWARP provider to pass a connection request event
to the iwcm, which will be silently dropped by the iwcm. This causes
the iWARP provider to never free up the resources from this connection
because the assumption is the iwcm will accept or reject this connection.
The solution is to reject these connection requests.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
When we destroy a cm_id, we must purge associated events from the
event queue. If the cm_id is for a listen request, we also purge
corresponding pending connect requests. This requires destroying
the cm_id's associated with the connect requests by calling
rdma_destroy_id(). rdma_destroy_id() blocks until all outstanding
callbacks have completed.
The issue is that we hold file->mut while purging events from the
event queue. We also acquire file->mut in our event handler. Calling
rdma_destroy_id() while holding file->mut can lead to a deadlock,
since the event handler callback cannot acquire file->mut, which
prevents rdma_destroy_id() from completing.
Fix this by moving events to purge from the event queue to a temporary
list. We can then release file->mut and call rdma_destroy_id()
outside of holding any locks.
Bug report by Or Gerlitz <ogerlitz@mellanox.com>:
[ INFO: possible circular locking dependency detected ]
3.3.0-rc5-00008-g79f1e43-dirty #34 Tainted: G I
tgtd/9018 is trying to acquire lock:
(&id_priv->handler_mutex){+.+.+.}, at: [<ffffffffa0359a41>] rdma_destroy_id+0x33/0x1f0 [rdma_cm]
but task is already holding lock:
(&file->mut){+.+.+.}, at: [<ffffffffa02470fe>] ucma_free_ctx+0xb6/0x196 [rdma_ucm]
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&file->mut){+.+.+.}:
[<ffffffff810682f3>] lock_acquire+0xf0/0x116
[<ffffffff8135f179>] mutex_lock_nested+0x64/0x2e6
[<ffffffffa0247636>] ucma_event_handler+0x148/0x1dc [rdma_ucm]
[<ffffffffa035a79a>] cma_ib_handler+0x1a7/0x1f7 [rdma_cm]
[<ffffffffa0333e88>] cm_process_work+0x32/0x119 [ib_cm]
[<ffffffffa03362ab>] cm_work_handler+0xfb8/0xfe5 [ib_cm]
[<ffffffff810423e2>] process_one_work+0x2bd/0x4a6
[<ffffffff810429e2>] worker_thread+0x1d6/0x350
[<ffffffff810462a6>] kthread+0x84/0x8c
[<ffffffff81369624>] kernel_thread_helper+0x4/0x10
-> #0 (&id_priv->handler_mutex){+.+.+.}:
[<ffffffff81067b86>] __lock_acquire+0x10d5/0x1752
[<ffffffff810682f3>] lock_acquire+0xf0/0x116
[<ffffffff8135f179>] mutex_lock_nested+0x64/0x2e6
[<ffffffffa0359a41>] rdma_destroy_id+0x33/0x1f0 [rdma_cm]
[<ffffffffa024715f>] ucma_free_ctx+0x117/0x196 [rdma_ucm]
[<ffffffffa0247255>] ucma_close+0x77/0xb4 [rdma_ucm]
[<ffffffff810df6ef>] fput+0x117/0x1cf
[<ffffffff810dc76e>] filp_close+0x6d/0x78
[<ffffffff8102b667>] put_files_struct+0xbd/0x17d
[<ffffffff8102b76d>] exit_files+0x46/0x4e
[<ffffffff8102d057>] do_exit+0x299/0x75d
[<ffffffff8102d599>] do_group_exit+0x7e/0xa9
[<ffffffff8103ae4b>] get_signal_to_deliver+0x536/0x555
[<ffffffff81001717>] do_signal+0x39/0x634
[<ffffffff81001d39>] do_notify_resume+0x27/0x69
[<ffffffff81361c03>] retint_signal+0x46/0x83
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(&file->mut);
lock(&id_priv->handler_mutex);
lock(&file->mut);
lock(&id_priv->handler_mutex);
*** DEADLOCK ***
1 lock held by tgtd/9018:
#0: (&file->mut){+.+.+.}, at: [<ffffffffa02470fe>] ucma_free_ctx+0xb6/0x196 [rdma_ucm]
stack backtrace:
Pid: 9018, comm: tgtd Tainted: G I 3.3.0-rc5-00008-g79f1e43-dirty #34
Call Trace:
[<ffffffff81029e9c>] ? console_unlock+0x18e/0x207
[<ffffffff81066433>] print_circular_bug+0x28e/0x29f
[<ffffffff81067b86>] __lock_acquire+0x10d5/0x1752
[<ffffffff810682f3>] lock_acquire+0xf0/0x116
[<ffffffffa0359a41>] ? rdma_destroy_id+0x33/0x1f0 [rdma_cm]
[<ffffffff8135f179>] mutex_lock_nested+0x64/0x2e6
[<ffffffffa0359a41>] ? rdma_destroy_id+0x33/0x1f0 [rdma_cm]
[<ffffffff8106546d>] ? trace_hardirqs_on_caller+0x11e/0x155
[<ffffffff810654b1>] ? trace_hardirqs_on+0xd/0xf
[<ffffffffa0359a41>] rdma_destroy_id+0x33/0x1f0 [rdma_cm]
[<ffffffffa024715f>] ucma_free_ctx+0x117/0x196 [rdma_ucm]
[<ffffffffa0247255>] ucma_close+0x77/0xb4 [rdma_ucm]
[<ffffffff810df6ef>] fput+0x117/0x1cf
[<ffffffff810dc76e>] filp_close+0x6d/0x78
[<ffffffff8102b667>] put_files_struct+0xbd/0x17d
[<ffffffff8102b5cc>] ? put_files_struct+0x22/0x17d
[<ffffffff8102b76d>] exit_files+0x46/0x4e
[<ffffffff8102d057>] do_exit+0x299/0x75d
[<ffffffff8102d599>] do_group_exit+0x7e/0xa9
[<ffffffff8103ae4b>] get_signal_to_deliver+0x536/0x555
[<ffffffff810654b1>] ? trace_hardirqs_on+0xd/0xf
[<ffffffff81001717>] do_signal+0x39/0x634
[<ffffffff8135e037>] ? printk+0x3c/0x45
[<ffffffff8106546d>] ? trace_hardirqs_on_caller+0x11e/0x155
[<ffffffff810654b1>] ? trace_hardirqs_on+0xd/0xf
[<ffffffff81361803>] ? _raw_spin_unlock_irq+0x2b/0x40
[<ffffffff81039011>] ? set_current_blocked+0x44/0x49
[<ffffffff81361bce>] ? retint_signal+0x11/0x83
[<ffffffff81001d39>] do_notify_resume+0x27/0x69
[<ffffffff8118a1fe>] ? trace_hardirqs_on_thunk+0x3a/0x3f
[<ffffffff81361c03>] retint_signal+0x46/0x83
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
The kernel IB stack uses one enumeration for IB speed, which wasn't
explicitly specified in the verbs header file. Add that enum, and use
it all over the code.
The IB speed/width notation is also used by iWARP and IBoE HW drivers,
which use the convention of rate = speed * width to advertise their
port link rate.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Commit 71eeba16 ("IB: Add new InfiniBand link speeds") introduced a bug
where eg the rate for IB 4X SDR links iss displayed as "8.5 Gb/sec"
instead of "10 Gb/sec" as it used to be. Fix that.
Reported-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Davem considers that the argument list of this interface is getting
out of control. This patch tries to address this issue following
his proposal:
struct netlink_dump_control c = { .dump = dump, .done = done, ... };
netlink_dump_start(..., &c);
Suggested by David S. Miller.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Set up a response with appropriate error status and send it for MADs
that are not supported by a specific class/version.
Reviewed-by: Hal Rosenstock <hal@mellanox.com>
Signed-off-by: Swapna Thete <swapna.thete@qlogic.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
After reporting a new connection request to user space, the rdma_ucm
will discard subsequent events until the user has associated a user
space idenfier with the kernel cm_id. This is needed to avoid
reporting a reject/disconnect event to the user for a request that
they may not have processed.
The user space identifier is set once the user tries to accept the
connection request. However, the following race exists in ucma_accept():
ctx->uid = cmd.uid;
<events may be reported now>
ret = rdma_accept(ctx->cm_id, ...);
Once ctx->uid has been set, new events may be reported to the user.
While the above mentioned race is avoided, there is an issue that the
user _may_ receive a reject/disconnect event if rdma_accept() fails,
depending on when the event is processed. To simplify the use of
rdma_accept(), discard all events unless rdma_accept() succeeds.
This problem was discovered based on questions from Roland Dreier
<roland@purestorage.com>.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
We have just been investigating kernel panics related to
cq->ibcq.event_handler() completion calls. The problem is that
ib_destroy_qp() fails with -EBUSY.
Further investigation revealed qp->usecnt is not initialized. This
counter was introduced in linux-3.2 by commit 0e0ec7e063
("RDMA/core: Export ib_open_qp() to share XRC TGT QPs") but it only
gets initialized for IB_QPT_XRC_TGT, but it is checked in
ib_destroy_qp() for any QP type.
Fix this by initializing qp->usecnt for every QP we create.
Signed-off-by: Bernd Schubert <bernd.schubert@itwm.fraunhofer.de>
Signed-off-by: Sven Breuner <sven.breuner@itwm.fraunhofer.de>
[ Initialize qp->usecnt in uverbs too. - Sean ]
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Now we must provide the IP destination address, and a reference has
to be dropped when we're done with the entry.
Signed-off-by: David S. Miller <davem@davemloft.net>
* 'for-linus2' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (165 commits)
reiserfs: Properly display mount options in /proc/mounts
vfs: prevent remount read-only if pending removes
vfs: count unlinked inodes
vfs: protect remounting superblock read-only
vfs: keep list of mounts for each superblock
vfs: switch ->show_options() to struct dentry *
vfs: switch ->show_path() to struct dentry *
vfs: switch ->show_devname() to struct dentry *
vfs: switch ->show_stats to struct dentry *
switch security_path_chmod() to struct path *
vfs: prefer ->dentry->d_sb to ->mnt->mnt_sb
vfs: trim includes a bit
switch mnt_namespace ->root to struct mount
vfs: take /proc/*/mounts and friends to fs/proc_namespace.c
vfs: opencode mntget() mnt_set_mountpoint()
vfs: spread struct mount - remaining argument of next_mnt()
vfs: move fsnotify junk to struct mount
vfs: move mnt_devname
vfs: move mnt_list to struct mount
vfs: switch pnode.h macros to struct mount *
...
Clean up sparse warnings in the rdma core layer.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Fix endianness bugs reported by sparse in the RDMA core stack. Note
that these are real bugs, but don't affect any existing code to the
best of my knowledge. The mlid issue would only affect kernel users
of rdma_join_multicast which have the rdma_cm attach/detach its QP.
There are no current in tree users that do this. (rdma_join_multicast
may be used called by user space applications, which does not have
this issue.) And the pkey setting is simply returned as
informational.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Add a missing 16-bit reserved field between ap_status and info fields.
Signed-off-by: Eli Cohen <eli@mellanox.com>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Userspace verbs multicast attach/detach operations on a QP are done
while holding the rwsem of the QP for reading. That's not sufficient
since a reader lock allows more than one reader to acquire the
lock. However, multicast attach/detach does list manipulation that
can corrupt the list if multiple threads run in parallel.
Fix this by acquiring the rwsem as a writer to serialize attach/detach
operations. Add idr_write_qp() and put_qp_write() to encapsulate
this.
This fixes oops seen when running applications that perform multicast
joins/leaves.
Reported by: Mike Dubman <miked@mellanox.com>
Signed-off-by: Eli Cohen <eli@mellanox.com>
Cc: <stable@kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
both callers of device_get_devnode() are only interested in lower 16bits
and nobody tries to return anything wider than 16bit anyway.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Conflicts:
net/bluetooth/l2cap_core.c
Just two overlapping changes, one added an initialization of
a local variable, and another change added a new local variable.
Signed-off-by: David S. Miller <davem@davemloft.net>
private_data_len is defined as a u8. If the user specifies a large
private_data size (> 220 bytes), we will calculate a total length that
exceeds 255, resulting in private_data_len wrapping back to 0. This
can lead to overwriting random kernel memory. Avoid this by verifying
that the resulting size fits into a u8.
Reported-by: B. Thery <benjamin.thery@bull.net>
Addresses: <http://bugs.openfabrics.org/bugzilla/show_bug.cgi?id=2335>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
IPV4 should do exactly what the IPV6 code does here, which is
use the neighbour obtained via the dst entry.
And now that the two code paths do the same thing, use a common
helper function to perform the operation.
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Roland Dreier <roland@purestorage.com>
To reflect the fact that a refrence is not obtained to the
resulting neighbour entry.
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Roland Dreier <roland@purestorage.com>
Commit f2c31e32b3 ("net: fix NULL dereferences in check_peer_redir()")
forgot to take care of infiniband uses of dst neighbours.
Many thanks to Marc Aurele who provided a nice bug report and feedback.
Reported-by: Marc Aurele La France <tsi@ualberta.ca>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David Miller <davem@davemloft.net>
Cc: <stable@kernel.org>
Signed-off-by: Roland Dreier <roland@purestorage.com>
C assignment can handle struct in6_addr copying.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* 'modsplit-Oct31_2011' of git://git.kernel.org/pub/scm/linux/kernel/git/paulg/linux: (230 commits)
Revert "tracing: Include module.h in define_trace.h"
irq: don't put module.h into irq.h for tracking irqgen modules.
bluetooth: macroize two small inlines to avoid module.h
ip_vs.h: fix implicit use of module_get/module_put from module.h
nf_conntrack.h: fix up fallout from implicit moduleparam.h presence
include: replace linux/module.h with "struct module" wherever possible
include: convert various register fcns to macros to avoid include chaining
crypto.h: remove unused crypto_tfm_alg_modname() inline
uwb.h: fix implicit use of asm/page.h for PAGE_SIZE
pm_runtime.h: explicitly requires notifier.h
linux/dmaengine.h: fix implicit use of bitmap.h and asm/page.h
miscdevice.h: fix up implicit use of lists and types
stop_machine.h: fix implicit use of smp.h for smp_processor_id
of: fix implicit use of errno.h in include/linux/of.h
of_platform.h: delete needless include <linux/module.h>
acpi: remove module.h include from platform/aclinux.h
miscdevice.h: delete unnecessary inclusion of module.h
device_cgroup.h: delete needless include <linux/module.h>
net: sch_generic remove redundant use of <linux/module.h>
net: inet_timewait_sock doesnt need <linux/module.h>
...
Fix up trivial conflicts (other header files, and removal of the ab3550 mfd driver) in
- drivers/media/dvb/frontends/dibx000_common.c
- drivers/media/video/{mt9m111.c,ov6650.c}
- drivers/mfd/ab3550-core.c
- include/linux/dmaengine.h
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (62 commits)
mlx4_core: Deprecate log_num_vlan module param
IB/mlx4: Don't set VLAN in IBoE WQEs' control segment
IB/mlx4: Enable 4K mtu for IBoE
RDMA/cxgb4: Mark QP in error before disabling the queue in firmware
RDMA/cxgb4: Serialize calls to CQ's comp_handler
RDMA/cxgb3: Serialize calls to CQ's comp_handler
IB/qib: Fix issue with link states and QSFP cables
IB/mlx4: Configure extended active speeds
mlx4_core: Add extended port capabilities support
IB/qib: Hold links until tuning data is available
IB/qib: Clean up checkpatch issue
IB/qib: Remove s_lock around header validation
IB/qib: Precompute timeout jiffies to optimize latency
IB/qib: Use RCU for qpn lookup
IB/qib: Eliminate divide/mod in converting idx to egr buf pointer
IB/qib: Decode path MTU optimization
IB/qib: Optimize RC/UC code by IB operation
IPoIB: Use the right function to do DMA unmap pages
RDMA/cxgb4: Use correct QID in insert_recv_cqe()
RDMA/cxgb4: Make sure flush CQ entries are collected on connection close
...
Some kernel components pin user space memory (infiniband and perf) (by
increasing the page count) and account that memory as "mlocked".
The difference between mlocking and pinning is:
A. mlocked pages are marked with PG_mlocked and are exempt from
swapping. Page migration may move them around though.
They are kept on a special LRU list.
B. Pinned pages cannot be moved because something needs to
directly access physical memory. They may not be on any
LRU list.
I recently saw an mlockalled process where mm->locked_vm became
bigger than the virtual size of the process (!) because some
memory was accounted for twice:
Once when the page was mlocked and once when the Infiniband
layer increased the refcount because it needt to pin the RDMA
memory.
This patch introduces a separate counter for pinned pages and
accounts them seperately.
Signed-off-by: Christoph Lameter <cl@linux.com>
Cc: Mike Marciniszyn <infinipath@qlogic.com>
Cc: Roland Dreier <roland@kernel.org>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
These were getting it implicitly via device.h --> module.h but
we are going to stop that when we clean up the headers.
Fix these in advance so the tree remains biscect-clean.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
They had been getting it implicitly via device.h but we can't
rely on that for the future, due to a pending cleanup so fix
it now.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
They get it via module.h (via device.h) but we want to clean that up.
When we do, we'll get things like:
CC [M] drivers/infiniband/core/sysfs.o
sysfs.c:361: error: 'S_IRUGO' undeclared here (not in a function)
sysfs.c:654: error: 'S_IWUSR' undeclared here (not in a function)
so add in the stat header it is using explicitly in advance.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Allow processes that share the same XRC domain to open an existing
shareable QP. This permits those processes to receive events on the
shared QP and transfer ownership, so that any process may modify the
QP. The latter allows the creating process to exit, while a remaining
process can still transition it for path migration purposes.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
XRC TGT QPs are shared resources among multiple processes. Since the
creating process may exit, allow other processes which share the same
XRC domain to open an existing QP. This allows us to transfer
ownership of an XRC TGT QP to another process.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Because an XRC TGT QP can end up being shared among multiple
processes, don't have the ib_cm automatically send a DREQ when the
userspace process that owns the ib_cm_id exits. Disconnect can be
initiated by the user directly; otherwise, the owner of the XRC INI QP
controls the connection.
Note that as a result of the process exiting, the ib_cm will stop
tracking the XRC connection on the target side. For the purposes of
disconnecting, this isn't a big deal. The ib_cm will respond to the
DREQ appropriately. For other messages, mainly LAP, the CM will
reject the request, since there's no one available to route the
request to.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Allow users to connect XRC QPs through the rdma_cm.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Allow the user to indicate the QP type separately from the port space
when allocating an rdma_cm_id. With RDMA_PS_IB, there is no longer a
1:1 relationship between the QP type and port space, so we need to
switch on the QP type to select between UD and connected QPs.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Add RDMA_PS_IB. XRC QP types will use the IB port space when operating
over the RDMA CM. For the 'IP protocol' field value, we select 0x3F,
which is listed as being for 'any local network'.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
The XRC annex was updated to have XRC behave more like RD. Specifically,
the XRC TGT QPN moves from the local QPN to local EECN field. Lookup of
SRQN is done using the REQ/REP protocol.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Update the REQ and REP messages to support XRC connection setup
according to the XRC Annex. Several existing fields must be set to 0 or
1 when connecting XRC QPs, and a reserved field is changed to an
extended transport type.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Allow user space to operate on XRC TGT QPs the same way as other types
of QPs, with one notable exception: since XRC TGT QPs may be shared
among multiple processes, the XRC TGT QP is allowed to exist beyond the
lifetime of the creating process.
The process that creates the QP is allowed to destroy it, but if the
process exits without destroying the QP, then the QP will be left bound
to the lifetime of the XRCD.
TGT QPs are not associated with CQs or a PD.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
XRC INI QPs are similar to send only RC QPs. Allow user space to create
INI QPs. Note that INI QPs do not require receive CQs.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
We require additional information to create XRC SRQs than we can
exchange using the existing create SRQ ABI. Provide an enhanced create
ABI for extended SRQ types.
Based on patches by Jack Morgenstein <jackm@dev.mellanox.co.il>
and Roland Dreier <roland@purestorage.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Allow user space to create XRC domains. Because XRCDs are expected to
be shared among multiple processes, we use inodes to identify an XRCD.
Based on patches by Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
XRC TGT QPs are intended to be shared among multiple users and
processes. Allow the destruction of an XRC TGT QP to be done explicitly
through ib_destroy_qp() or when the XRCD is destroyed.
To support destroying an XRC TGT QP, we need to track TGT QPs with the
XRCD. When the XRCD is destroyed, all tracked XRC TGT QPs are also
cleaned up.
To avoid stale reference issues, if a user is holding a reference on a
TGT QP, we increment a reference count on the QP. The user releases the
reference by calling ib_release_qp. This releases any access to the QP
from a user above verbs, but allows the QP to continue to exist until
destroyed by the XRCD.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
XRC ("eXtended reliable connected") is an IB transport that provides
better scalability by allowing senders to specify which shared receive
queue (SRQ) should be used to receive a message, which essentially
allows one transport context (QP connection) to serve multiple
destinations (as long as they share an adapter, of course).
XRC communication is between an initiator (INI) QP and a target (TGT)
QP. Target QPs are associated with SRQs through an XRCD. An XRC TGT QP
behaves like a receive-only RD QP. XRC INI QPs behave similarly to RC
QPs, except that work requests posted to an XRC INI QP must specify the
remote SRQ that is the target of the work request.
We define two new QP types for XRC, to distinguish between INI and TGT
QPs, and update the core layer to support XRC QPs.
This patch is derived from work by Jack Morgenstein
<jackm@dev.mellanox.co.il>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
XRC ("eXtended reliable connected") is an IB transport that provides
better scalability by allowing senders to specify which shared receive
queue (SRQ) should be used to receive a message, which essentially
allows one transport context (QP connection) to serve multiple
destinations (as long as they share an adapter, of course).
XRC defines SRQs that are specifically used by XRC connections. Expand
the SRQ code to support XRC SRQs. An XRC SRQ is currently restricted to
only XRC use according to the IB XRC Annex.
Portions of this patch were derived from work by
Jack Morgenstein <jackm@dev.mellanox.co.il>.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
Linus Torvalds