The base versions handle constant folding just fine, use them
directly. The replacements are OK in the include/ files as they are
not exported to userspace so we don't need the __ prefixed versions.
This patch does not affect code generation at all.
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ehca_plpar_hcall9() takes an unsigned long array, so make all callers
pass that in. This fixes warnings introduced by commit fe333321
("powerpc: Change u64/s64 to a long long integer type"), which changed
u64 from unsigned long to unsigned long long.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit fe333321 ("powerpc: Change u64/s64 to a long long integer
type") changed u64 from unsigned long to unsigned long long, which
means that printk formats for printing u64 values should use "ll"
instead of "l" to avoid warnings. Fix all the places affected by this
in ehca.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When IPoIB tries to join a multicast group, and the SA module's SM
address handle is NULL (because of an SM change, etc), the join
returns with -EAGAIN status. In that case, don't print an error
message unless multicast debugging is enabled.
Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The current work request posting code writes the LSO segment before
writing any data segments. This leaves a window where the LSO segment
overwrites the stamping in one cacheline that the HCA prefetches
before the rest of the cacheline is filled with the correct data
segments. When the HCA processes this work request, a local
protection error may result.
Fix this by saving the LSO header size field off and writing it only
after all data segments are written. This fix is a cleaned-up version
of a patch from Jack Morgenstein <jackm@dev.mellanox.co.il>.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1383>.
Reported-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fix a deadlock between child interface creation/deletion and ipoib
start/stop. The former takes vlan_mutex, and then might take RTNL via
register_netdev()/unregister_netdev(). The latter is executed with
RTNL held, and tries to take vlan_mutex, which can lead to an AB-BA
deadlock.
Fix this by having the child interface creation/deletion code take the
RTNL first so vlan_mutex always nests inside RTNL. We can use
register_netdevice() for child interfaces because we form the
interface name from the parent interface and hence don't need the '%'
expansion of register_netdev().
Reported-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
After commit fe25c561 ("IPoIB: Don't enable NAPI when it's already
enabled"), if an interface is brought up but the corresponding P_Key
never appears, then ipoib_stop() will hang in napi_disable(), because
ipoib_open() returns before it does napi_enable().
Fix this by changing ipoib_open() to call napi_enable() even if the
P_Key isn't present.
Reported-by: Yossi Etigin <yosefe@Voltaire.COM>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/iser: Add dependency on INFINIBAND_ADDR_TRANS
IPoIB: Do not join broadcast group if interface is brought down
RDMA/nes: Fix for NIPQUAD removal
IPoIB: Fix loss of connectivity after bonding failover on both sides
IB/mlx4: Don't register IB device for adapters with no IB ports
mlx4_core: Fix warning from min()
IB/ehca: spin_lock_irqsave() takes an unsigned long
Fix ib_iser build to depend on INFINIBAND_ADDR_TRANS; if INET=y but
IPV6=n, then the RDMA CM is not built but INFINIBAND_ISER can be
enabled, leading to:
ERROR: "rdma_destroy_id" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
ERROR: "rdma_connect" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
ERROR: "rdma_destroy_qp" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
ERROR: "rdma_create_id" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
ERROR: "rdma_create_qp" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
ERROR: "rdma_resolve_route" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
ERROR: "rdma_disconnect" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
ERROR: "rdma_resolve_addr" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Because the ipoib_workqueue is not flushed when ipoib interface is
brought down, ipoib_mcast_join() may trigger a join to the broadcast
group after priv->broadcast was set to NULL (during cleanup). This
will cause the system to be a member of the broadcast group when
interface is down. As a side effect, this breaks the optimization of
setting the Q_key only when joining the broadcast group.
Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit 63779436 ("drivers: replace NIPQUAD()") accidentally replaced
some HIPQUAD()s, causing IP addresses to be printed in reverse order.
Add temporary local vars until the byteswapping can be pushed further
up the stack.
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fix bonding failover in the case both peers failover and the
gratuitous ARP is lost. In that case, the sender side will create an
ipoib_neigh and issue a path request with the old GID first. When
skb->dst->neighbour->ha changes due to ARP refresh, this ipoib_neigh
will not be added to the path->list of the path of the new GID,
because the ipoib_neigh already exists. It will not have an AH
either, because of sender-side failover. Therefore, it will not get
an AH when the path is resolved.
The solution here is to compare GIDs in ipoib_start_xmit() even if
neigh->ah is invalid. Comparing with an uninitialized value of
neigh->dgid should be fine, since a spurious match is harmless (and
astronomically unlikely too).
Signed-off-by: Moni Shoua <monis@voltaire.com>
Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If the mlx4_ib driver finds an adapter that has only ethernet ports, the
current code will register an IB device with 0 ports. Nothing useful or
sensible can be done with such a device, so just skip registering it.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When I review ocfs2 code, find there are 2 typos to "successfull". After
doing grep "successfull " in kernel tree, 22 typos found totally -- great
minds always think alike :)
This patch fixes all the similar typos. Thanks for Randy's ack and comments.
Signed-off-by: Coly Li <coyli@suse.de>
Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
Acked-by: Roland Dreier <rolandd@cisco.com>
Cc: Jeremy Kerr <jk@ozlabs.org>
Cc: Jeff Garzik <jeff@garzik.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Vlad Yasevich <vladislav.yasevich@hp.com>
Cc: Sridhar Samudrala <sri@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (24 commits)
trivial: chack -> check typo fix in main Makefile
trivial: Add a space (and a comma) to a printk in 8250 driver
trivial: Fix misspelling of "firmware" in docs for ncr53c8xx/sym53c8xx
trivial: Fix misspelling of "firmware" in powerpc Makefile
trivial: Fix misspelling of "firmware" in usb.c
trivial: Fix misspelling of "firmware" in qla1280.c
trivial: Fix misspelling of "firmware" in a100u2w.c
trivial: Fix misspelling of "firmware" in megaraid.c
trivial: Fix misspelling of "firmware" in ql4_mbx.c
trivial: Fix misspelling of "firmware" in acpi_memhotplug.c
trivial: Fix misspelling of "firmware" in ipw2100.c
trivial: Fix misspelling of "firmware" in atmel.c
trivial: Fix misspelled firmware in Kconfig
trivial: fix an -> a typos in documentation and comments
trivial: fix then -> than typos in comments and documentation
trivial: update Jesper Juhl CREDITS entry with new email
trivial: fix singal -> signal typo
trivial: Fix incorrect use of "loose" in event.c
trivial: printk: fix indentation of new_text_line declaration
trivial: rtc-stk17ta8: fix sparse warning
...
The flags argument to spin_lock_irqsave() should really be unsigned
long. This will also help prevent some warnings when we change u64 to
unsigned long long.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
... and don't bother in callers. Don't bother with zeroing i_blocks,
while we are at it - it's already been zeroed.
i_mode is not worth the effort; it has no common default value.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/mlx4: Fix reading SL field out of cqe->sl_vid
RDMA/addr: Fix build breakage when IPv6 is disabled
Commit f780a9f1 ("mlx4_core: Add ethernet fields to CQE struct")
introduced a bug in how wc->sl is set in mlx4_ib_poll_one() -- since
cqe->sl_vid is a big-endian value, the shift must be done after
converting to host endianness.
This bug was found using sparse endianness checking.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit 38617c64 ("RDMA/addr: Add support for translating IPv6
addresses") broke the build when CONFIG_IPV6=n, because the ib_addr
module unconditionally attempted to call ipv6_chk_addr() and other
IPv6 functions that are not defined when IPv6 is disabled. Fix this
by only building IPv6 support if CONFIG_IPV6 is turned on, and
add a Kconfig dependency to prevent the ib_addr code from being built
in when IPv6 is built modular.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Impact: cleanup
We're moving from handing around cpumask_t's to handing around struct
cpumask *'s. cpus_*, cpumask_t and cpu_*_map are deprecated: convert
to cpumask_*, cpu_*_mask.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Ralph Campbell <infinipath@qlogic.com>
Impact: cleanup
We're moving from handing around cpumask_t's to handing around struct
cpumask *'s. cpus_*, cpumask_t and cpu_*_map are deprecated: convert
to cpumask_*, cpu_*_mask.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Hoang-Nam Nguyen <hnguyen@de.ibm.com>
Tested-by: Hoang-Nam Nguyen <hnguyen@de.ibm.com>
Cc: Christoph Raisch <raisch@de.ibm.com>
Impact: cleanup
In future, accessing cpu numbers beyond nr_cpu_ids (the runtime limit)
will be undefined. We can avoid future problems by using
for_each_online_cpu() here.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Hoang-Nam Nguyen <hnguyen@de.ibm.com>
Tested-by: Hoang-Nam Nguyen <hnguyen@de.ibm.com>
Cc: Christoph Raisch <raisch@de.ibm.com>
We do not need to allocate a itt for data_out, so this
passes the opcode to the alloc_pdu callout.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
This just converts iser to new alloc_pdu api. It still
preallocates the pdu, so there is no difference.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6: (1429 commits)
net: Allow dependancies of FDDI & Tokenring to be modular.
igb: Fix build warning when DCA is disabled.
net: Fix warning fallout from recent NAPI interface changes.
gro: Fix potential use after free
sfc: If AN is enabled, always read speed/duplex from the AN advertising bits
sfc: When disabling the NIC, close the device rather than unregistering it
sfc: SFT9001: Add cable diagnostics
sfc: Add support for multiple PHY self-tests
sfc: Merge top-level functions for self-tests
sfc: Clean up PHY mode management in loopback self-test
sfc: Fix unreliable link detection in some loopback modes
sfc: Generate unique names for per-NIC workqueues
802.3ad: use standard ethhdr instead of ad_header
802.3ad: generalize out mac address initializer
802.3ad: initialize ports LACPDU from const initializer
802.3ad: remove typedef around ad_system
802.3ad: turn ports is_individual into a bool
802.3ad: turn ports is_enabled into a bool
802.3ad: make ntt bool
ixgbe: Fix set_ringparam in ixgbe to use the same memory pools.
...
Fixed trivial IPv4/6 address printing conflicts in fs/cifs/connect.c due
to the conversion to %pI (in this networking merge) and the addition of
doing IPv6 addresses (from the earlier merge of CIFS).
When we removed the network device argument from several
NAPI interfaces in 908a7a16b8
("net: Remove unused netdev arg from some NAPI interfaces.")
several drivers now started getting unused variable warnings.
This fixes those up.
Signed-off-by: David S. Miller <davem@davemloft.net>
When resizing a CQ, when copying over unpolled CQEs from the old CQE
buffer to the new buffer, the ownership bit must be set appropriately
for the new buffer, or the ownership bit in the new buffer gets
corrupted.
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
There is no lock protecting tx_free_list thus causing a system crash
when skb_dequeue() is called and the list is empty. Since it did not give
any performance boost under heavy load, remove it to simplify the code.
Replace get_free_pkt() with dev_alloc_skb() to allocate MAX_CM_BUFFER skb
for connection establishment/teardown as well as MPA request/response.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Handle AF_INET6 cases where required, and use struct sockaddr_storage
wherever an IPv6 address might be stored.
Signed-off-by: Aleksey Senin <aleksey@alst60.(none)>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add support for translating AF_INET6 addresses to the IB address
translation service. This requires using struct sockaddr_storage
instead of struct sockaddr wherever an IPv6 address might be stored,
and adding cases to handle IPv6 in addition to IPv4 to the various
translation functions.
Signed-off-by: Aleksey Senin <aleksey@alst60.(none)>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When the napi api was changed to separate its 1:1 binding to the net_device
struct, the netif_rx_[prep|schedule|complete] api failed to remove the now
vestigual net_device structure parameter. This patch cleans up that api by
properly removing it..
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When using MSI-X mode, create a completion event queue for each CPU.
Report the number of completion EQs in a new struct mlx4_caps member,
num_comp_vectors, and extend the mlx4_cq_alloc() interface with a
vector parameter so that consumers can specify which completion EQ
should be used to report events for the CQ being created.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
iSCSI/iSER targets may send PDUs without a prior request from the
initiator. RFC 5046 refers to these PDUs as "unexpected". NOP-In PDUs
with itt=RESERVED and Asynchronous Message PDUs occupy this category.
The amount of active "unexpected" PDU's an iSER target may have at any
time is governed by the MaxOutstandingUnexpectedPDUs key, which is not
yet supported.
Currently when an iSER target sends an "unexpected" PDU, the
initiators recv buffer consumed by the PDU is not replaced. If over
initial_post_recv_bufs_num "unexpected" PDUs are received then the
receive queue will run out of receive work requests entirely.
This patch ensures recv buffers consumed by "unexpected" PDUs are
replaced in the next iser_post_receive_control() call.
Signed-off-by: David Disseldorp <ddiss@sgi.com>
Signed-off-by: Ken Sandars <ksandars@sgi.com>
Acked-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
vpage is checked not to be NULL just after it is initialized at the
beginning of each loop iteration.
A simplified version of the semantic patch that makes this change is
as follows: (http://www.emn.fr/x-info/coccinelle/)
// <smpl>
@r exists@
local idexpression x;
expression E;
position p1,p2;
@@
if (x@p1 == NULL || ...) { ... when forall
return ...; }
... when != \(x=E\|x--\|x++\|--x\|++x\|x-=E\|x+=E\|x|=E\|x&=E\|&x\)
(
x@p2 == NULL
|
x@p2 != NULL
)
// another path to the test that is not through p1?
@s exists@
local idexpression r.x;
position r.p1,r.p2;
@@
... when != x@p1
(
x@p2 == NULL
|
x@p2 != NULL
)
@fix depends on !s@
position r.p1,r.p2;
expression x,E;
statement S1,S2;
@@
(
- if ((x@p2 != NULL) || ...)
S1
|
- if ((x@p2 == NULL) && ...) S1
|
- BUG_ON(x@p2 == NULL);
)
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
With the latest flush error completion patch we introduced modulus
operation to calculate the next index within a qmap. Based on
comments from other mailing lists we decided to optimize this
operation by using an addition and an if-statement instead of modulus,
even though this is on the error path.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fixes timing race resulting in panic. Not a performance sensitive path.
Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ipath_piobufbase was a single value offset, but is multiple values on
newer chips, so use only the 32 bits for the 2K buffers (4K buffers
are currently used only by the driver).
Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Implement the ignoring of ibsymbol errors and linkrecover errors while
the link is at less than INIT (long needed), to get accurate counts.
Particularly an issue when doing non-IBTA DDR negotiation with chips
from vendors that do not support IBTA mode negotiation. If the driver
is unloaded, and there is a delta, the adjusted counters are written
back to the chip, so they stay adjusted across driver reload.
Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This fixes an obvious oversight where the return value is not checked
for error.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The PSN of the first packet after an RDMA read is based on the size of
the RDMA read request. This is calculated correctly for the WQE sent
after the first request message but not on subsequent requests if the
RDMA read is resent.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Wrap NES_DEBUG and assert macros with do while (0) to avoid ambiguous
else. No one is using sk_buff * returned from form_cm_frame(), so
drop the return. drop_packet() should not be incrementing reset
counter on receiving a FIN.
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Between the first empty list check and locking the list, the list can
change. Check it again after it is locked to make sure the list is
still not empty.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ANVL testing showed we are not handling all cm_node states during
connection establishment. Add missing state handlers and fix sequence
number send reset in handle_tcp_options().
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Under heavy traffic, there is a small windows when an APBVT entry is
not yet removed and a new connection is established. Packets for the
new connection are dropped until APBVT entry is removed. This patch
will forward the packets instead of dropping them.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
In passive open, after indicating MPA request to rdma_cm, an incoming
RST would fire a reset event to rdma_cm causing it to crash, since the
current state is not connected. The solution is to wait for
nes_accept() or nes_reject() before firing the reset event. If
nes_accept() or nes_reject() is already done, then the reset event
will be fired when RST is processed.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
While processing connected_nodes list, we would release the lock when
we need to send reset to remote partner. That created a window where
the list can be modified. Change this into a two step process: place
nodes that need processing on a local list then process the local list.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use nes_free_cqp_request() instead of open coding. Change some
continue to break in nes_cm_timer_tick, because send_entry used to be
a list processed in a loop (so continue went to the next item). Now
it is a single item, so using break is correct.
Signed-off-by: Faisal Latif <faisal.latif@intel.com>
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Receive work queue entries are checked for L_Key validity, and
pointers to the memory region structure are saved in an allocated
structure. For UD loopback packets, this structure is allocated and
freed for each packet. This patch changes that to allocate/free
during QP creation and destruction.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The return from lookup_one_len() is assigned to *dentry, so that's
what we should be checking with IS_ERR().
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
shca_list_lock is taken from softirq context in ehca_poll_eqs, so we
need to lock IRQ safe elsewhere. Found by lockdep.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When resizing a CQ, MTTs associated with the old CQE buffer were not
freed. As a result, if any app used resize CQ repeatedly, all MTTs
were eventually exhausted, which led to all memory registration
operations failing until the driver is reloaded.
Once the RESIZE_CQ command returns successfully from FW, FW no longer
accesses the old CQ buffer, so it is safe to deallocate the MTT
entries used by the old CQ buffer.
Finally, if the RESIZE_CQ command fails, the MTTs allocated for the
new CQEs buffer also need to be de-allocated.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1416>.
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This fix enables ehca device driver to generate flush work completions
even if the application doesn't request completions for all work
requests. The current implementation of ehca will generate flush work
completions for the wrong work requests if an application uses non
signaled work completions.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The error message printed when the eHCA driver prevents memory hotplug
is misleading -- the user might think that hot-removing the lhca,
hotplugging memory, then hot-adding the lhca again will work, but it
actually doesn't.
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This use of netdev->priv is wrong.
The right way is:
alloc_netdev() with no memory for private data.
make netdev->ml_priv to point to c2_dev.
Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com>
Acked-by: Roland Dreier <rolandd@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IPoIB: Fix crash in path_rec_completion()
IPoIB: Fix hang in ipoib_flush_paths()
IPoIB: Don't enable NAPI when it's already enabled
RDMA/cxgb3: Fix deadlock in iw_cxgb3 (hang when configuring interface)
IB/ehca: Remove reference to special QP in case of port activation failure
IB/mlx4: Set umem field to NULL in mlx4_ib_alloc_fast_reg_mr()
mlx4_core: Fix unused variable warning
RDMA/nes: Mitigate compatibility issue regarding PCIe write credits
RDMA/nes: Fix CQ allocation scheme for multicast receive queue apps
RDMA/nes: Correct handling of PBL resources
RDMA/nes: Reindent mis-indented spinlocks
RDMA/cxgb3: Fix too-big reserved field zeroing in iwch_post_zb_read()
IB/ipath: Fix RDMA write with immediate copy of last packet
Fix a crash in path_rec_completion() during an SM up/down loop. If
more than one path record request is issued, the first completion
releases path->done, allowing ipoib_flush_paths() to free the path,
and thus corrupting it for the second completion.
Commit ee1e2c82 ("IPoIB: Refresh paths instead of flushing them on SM
change events") added the field path->valid and changed the test "if
(!path)" to "if (!path || !path->valid)". This change made it
possible for a path with an outstanding query to pass the test and
issue another query on the same path. Having two queries on the same
path leads to a crash.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1325>.
Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ipoib_flush_paths() can hang during an SM up/down loop: if
path_rec_start() fails (for instance, because there is no sm_ah), the
path is still added to the path list by neigh_add_path(). Then,
ipoib_flush_paths() will wait for path->done, but it will never
complete because the request was not issued at all. Fix this by
completing path->done if issuing the query fails.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1329>.
Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If a P_Key is not present when an interface is created, ipoib_open()
will return after doing napi_enable(). ipoib_open() will be called
again from ipoib_pkey_poll() when the P_Key appears, after NAPI has
already been enabled, and try to enable it again. This triggers a
BUG_ON() in napi_enable().
Fix this by moving the call to napi_enable() to after the test for
P_Key presence.
Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When the iw_cxgb3 module's cxgb3_client "add" func gets called by the
cxgb3 module, the iwarp driver ends up calling the ethtool ops
get_drvinfo function in cxgb3 to get the fw version and other info.
Currently the iwarp driver grabs the rtnl lock around this down call
to serialize. As of 2.6.27 or so, things changed such that the rtnl
lock is held around the call to the netdev driver open function. Also
the cxgb3_client "add" function doesn't get called if the device is
down.
So, if you load cxgb3, then load iw_cxgb3, then ifconfig up the
device, the iw_cxgb3 add func gets called with the rtnl_lock held. If
you load cxgb3, ifconfig up the device, then load iw_cxgb3, the add
func gets called without the rtnl_lock held. The former causes the
deadlock, the latter does not.
In addition, there are iw_cxgb3 sysfs handlers that also can call down
into cxgb3 to gather the fw and hw versions. These can be called
concurrently on different processors and at any time. Thus we need to
push this serialization down in the cxgb3 driver get_drvinfo func.
The fix is to remove rtnl lock usage, and use a per-device lock in cxgb3.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Acked-by: Divy Le Ray <divy@chelsio.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If the initialization of a special QP (e.g. AQP1) fails due to a
software timeout, we have to remove the reference to that special QP
struct from the port struct to stop the driver from accessing the QP,
since it will be/has been destroyed by the caller, eg in this case
ib_mad.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Set mr->umem to NULL in mlx4_ib_alloc_fast_reg_mr(). Otherwise
ib_dereg_mr() may invoke ib_umem_release() on a random pointer value
and get an oops.
Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Under heavy load, there is an compatibility issue regarding PCIe write
credits with certain chipsets. It can be mitigated by limiting read
requests to 256 Bytes.
This workaround is always enabled for Tbird2 on Gladius. We also add
a module parameter to enable workaround for non-Gladius cards.
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fix CQ allocation for multicast receive queue applications. Before
this patch, the CQ was not lined up with the right NIC.
Signed-off-by: Vadim Makhervaks <vadim.makhervaks@intel.com>
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* Roll back allocated structures on failures.
* Use GFP_ATOMIC instead of GFP_KERNEL since we are holding a lock.
* Acquire nesadapter->pbl_lock when modifying PBL counters.
* Decrement PBL counters on deallocation.
Signed-off-by: Chien Tung <chien.tin.tung@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The array wqe->read.reserved has only two entries, but
iwch_post_zb_read() sets [0], [1], and [2], which is one too many.
This is harmless since it runs into the next field, rem_stag, which is
initialized correctly immediately after, but we might as well get
things right, especially since it makes the code smaller.
This was spotted by the Coverity checker (CID 2475).
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
As it is, all instances of ->release() for files that have ->fasync()
need to remember to evict file from fasync lists; forgetting that
creates a hole and we actually have a bunch that *does* forget.
So let's keep our lives simple - let __fput() check FASYNC in
file->f_flags and call ->fasync() there if it's been set. And lose that
crap in ->release() instances - leaving it there is still valid, but we
don't have to bother anymore.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Using NIPQUAD() with NIPQUAD_FMT, %d.%d.%d.%d or %u.%u.%u.%u
can be replaced with %pI4
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As a bonus, removes some unnecessary byteswapping.
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Replace all uses of IPOIB_GID_FMT, IPOIB_GID_RAW_ARG() and IPOIB_GID_ARG()
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This converts pretty much everything to print_mac. There were
a few things that had conflicts which I have just dropped for
now, no harm done.
I've built an allyesconfig with this and looked at the files
that weren't built very carefully, but it's a huge patch.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the last packet of a RDMA write with immediate is received, the
next receive work queue entry ID should be used to generate a completion
entry. The code was incorrectly resetting part of the state used to copy
the last packet.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/ehca: Reject dynamic memory add/remove when ehca adapter is present
IB/ehca: Fix reported max number of QPs and CQs in systems with >1 adapter
IPoIB: Set netdev offload features properly for child (VLAN) interfaces
IPoIB: Clean up ethtool support
mlx4_core: Add Ethernet PCI device IDs
mlx4_en: Add driver for Mellanox ConnectX 10GbE NIC
mlx4_core: Multiple port type support
mlx4_core: Ethernet MAC/VLAN management
mlx4_core: Get ethernet MTU and default address from firmware
mlx4_core: Support multiple pre-reserved QP regions
Update NetEffect maintainer emails to Intel emails
RDMA/cxgb3: Remove cmid reference on tid allocation failures
IB/mad: Use krealloc() to resize snoop table
IPoIB: Always initialize poll_timer to avoid crash on unload
IB/ehca: Don't allow creating UC QP with SRQ
mlx4_core: Add QP range reservation support
RDMA/ucma: Test ucma_alloc_multicast() return against NULL, not with IS_ERR()
Since the ehca device driver does not support dynamic memory add and
remove operations, the driver must explicitly reject such requests in
order to prevent unpredictable behaviors related to existing memory
regions that cover all of memory being used by InfiniBand protocols in
the kernel.
The solution (for now at least) is to add a memory notifier to the
ehca device driver and if a request for dynamic memory add or remove
comes in, ehca will always reject it. The user can add or remove
memory by hot-removing the ehca adapter, performing the memory
operation, and then hot-adding the ehca adapter back.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Because ehca adapters can differ in the maximum number of QPs and CQs
we have to save the maximum number of these ressources per adapter and
not globally per ehca driver. This fix introduces 2 new members to the
shca structure to store the maximum value for QPs and CQs per adapter.
The module parameters are now used as initial values for those
variables. If a user selects an invalid number of CQs or QPs we don't
print an error any longer, instead we will inform the user with a
warning and set the values to the respective maximum supported by the
HW.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Child devices were created without any offload features set, fix this by
moving the code that computes the features into generic function which is
now called through non-child and child device creation.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
-- v1 has a bug where the 'result' flag in ipoib_vlan_add may be used uninitialized
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add a get_rx_csum method. Remove the driver's own get_tso method, as
the ethtool kernel code uses the default one if nothing is provided.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Multi-protocol adapters support different port types. Each consumer
of mlx4_core queries for supported port types; in particular mlx4_ib
can no longer assume that all physical ports belong to it. Port type
is configured through a sysfs interface. When the type of a port is
changed, all mlx4 interfaces are unregistered, and then registered
again with the new port types.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Tejun's commit 7b595756ec made sysfs
attribute->owner unnecessary. But the field was left in the structure to
ease the merge. It's been over a year since that change and it is now
time to start killing attribute->owner along with its users - one arch at
a time!
This patch is attempt #1 to get rid of attribute->owner only for
CONFIG_X86_64 or CONFIG_X86_32 . We will deal with other arches later on
as and when possible - avr32 will be the next since that is something I
can test. Compile (make allyesconfig / make allmodconfig / custom config)
and boot tested.
akpm: the idea is that we put the declaration of sttribute.owner inside
`#ifndef CONFIG_X86'. But that proved to be too ambitious for now because
new usages kept on turning up in subsystem trees.
[akpm: remove the ifdef for now]
Signed-off-by: Parag Warudkar <parag.lkml@gmail.com>
Cc: Greg KH <greg@kroah.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Tejun Heo <htejun@gmail.com>
Cc: Len Brown <lenb@kernel.org>
Cc: Jens Axboe <jens.axboe@oracle.com>
Cc: Jean Delvare <khali@linux-fr.org>
Cc: Roland Dreier <rolandd@cisco.com>
Cc: David Brownell <david-b@pacbell.net>
Cc: Alessandro Zummo <a.zummo@towertech.it>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now that device_create() has been audited, rename things back to the
original call to be sane.
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The error path in iwch_connect() can fail to drop the cmid reference,
which will cause the process to hang when destroying the cmid.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use krealloc() instead of kmalloc() followed by memcpy() when resizing
the MAD module's snoop table.
Also put parentheses around the new table size to avoid calculating
the wrong size to allocate, which fixes a bug pointed out by Haven
Hash <haven.hash@isilon.com>.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This is a much better version of a previous patch to make the parser
tables constant. Rather than changing the typedef, we put the "const" in
all the various places where its required, allowing the __initconst
exception for nfsroot which was the cause of the previous trouble.
This was posted for review some time ago and I believe its been in -mm
since then.
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Alexander Viro <aviro@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
I had this in my patchset to add target reset support, but
it got dropped due to patching conflicts. This initial patch
just renames the function and users. We are actually just
dropping the session, and so this does not have anything to do
with the host exactly. It does for software iscsi because
we allocate a host per session, but for cxgb3i this makes no
sense.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
If the driver knows when hardware is removed like with cxgb3i,
bnx2i, qla4xxx and iser then we will want to remove the sessions/devices
that are bound to that device before removing the host.
cxgb3i and in the future bnx2i will remove the host and that will
remove all the sessions on the hba. iser can call iscsi_kill_session
when it gets an event that indicates that a hca is removed.
And when qla4xxx is hooked in to the lib (it is only hooked into
the class right now) it can call iscsi remove host like the
partial offload card drivers.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
ipoib_ib_dev_stop() does del_timer_sync(&priv->poll_timer), but if a
P_key for an interface is not found, poll_timer is not initialized, so
this leads to a crash or hang. Fix this by moving where poll_timer is
initialized to ipoib_ib_dev_init(), which is always called.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1172>.
Debugged-by: Yosef Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch prevents a UC QP to be created attached to an SRQ, since
current firmware does not support this feature.
Signed-off-by: Michael Faath <micfaath@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
To allow allocating an aligned range of consecutive QP numbers, add an
interface to reserve an aligned range of QP numbers and have the QP
allocation function always take a QP number.
This will be used for RSS support in the mlx4_en Ethernet driver and
also potentially by IPoIB RSS support.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
In case of error, the function ucma_alloc_multicast() returns a NULL
pointer, but never returns an ERR pointer. So after a call to this
function, an IS_ERR test should be replaced by a NULL test.
The semantic match that finds this problem is as follows:
(http://www.emn.fr/x-info/coccinelle/)
// <smpl>
@match bad_is_err_test@
expression x, E;
@@
x = ucma_alloc_multicast(...)
... when != x = E
IS_ERR(x)
// </smpl>
Signed-off-by: Julien Brunel <brunel@diku.dk>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Referencing cm_node after it is freed via rem_ref_cm_node() causes a
slab corruption. There is no need to set cm_node->cm_id to NULL in
mini_cm_close().
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Set RLKEY bit in the HW context for kernel QPs so that kernel QPs can
use the reserved L_Key for memory reference.
Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fix routed RDMA connections to destinations where the next hop is not
the final destination. Use neigh_*() to properly locate neighbor.
Signed-off-by: Bob Sharp <bsharp@neteffect.com>
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Mask off a critical error after 100 critical error interrupts to
keep the system "sane".
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Mask off MAC interrupts on netdev_stop to prevent spurious MAC interrupts
on unload/reload of iw_nes.
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fill in firmware version for ethtool_drvinfo.
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use timer value set via ethtool intead of #defines.
Signed-off-by: John Lacombe <jlacombe@neteffect.com>
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use correct define for max TSO fragments.
Signed-off-by: Bob Sharp <bsharp@neteffect.com>
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Clear MDC bits before setting them to a new value. Adjust MDC value
for 10G.
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add a module parameter wqm_quanta. It controls the number of segments
transmitted at a time.
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Change permission to 0644 so root can set mpa_version, disable_mpa_crc,
send_first, and nes_drv_opt at runtime.
Signed-off-by: Sweta Bhatt <sweta.bhatt@einfochips.com>
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When running ibv_devinfo, the active_mtu returned is garbage. This is
due to the field not being populated in the query_port function in the
driver. The patch below populates the active_mtu field with a MTU of
2k. It also zeros the struct, so that any new additions to it will
return 0.
Signed-off-by: Jon Mason <jon@opengridcomputing.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add support for NetEffect 4 port 1G HP blade card. The mapping
between physical port and MAC is different from the standup card.
Signed-off-by: Chien Tung <ctung@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
commit 110cf374 ("infiniband: make cm_device use a struct device and
not a kobject.") introduced a memory leak, since it deleted
cm_release_dev_obj(), which was where cm_dev was freed. Fix this by
freeing the leaked structure after calling device_unregister().
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Currently, IPoIB is an LLTX driver that uses its own IRQ-disabling
tx_lock. Not only do we want to get rid of LLTX, this actually causes
problems because of the skb_orphan() done with this tx_lock held: some
skb destructors expect to be run with interrupts enabled.
The simplest fix for this is to get rid of the driver-private tx_lock
and stop using LLTX. We kill off priv->tx_lock and use
netif_tx_lock[_bh]() instead; the patch to do this is a tiny bit
tricky because we need to update places that take priv->lock inside
the tx_lock to disable IRQs, rather than relying on tx_lock having
already disabled IRQs.
Also, there are a couple of places where we need to disable BHs to
make sure we have a consistent context to call netif_tx_lock() (since
we no longer can use _irqsave() variants), and we also have to change
ipoib_send_comp_handler() to call drain_tx_cq() through a timer rather
than directly, because ipoib_send_comp_handler() runs in interrupt
context and drain_tx_cq() must run in BH context so it can call
netif_tx_lock().
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Back in prehistoric (pre-git!) days, the kernel's MSI-X support did
request_mem_region() on a device's MSI-X tables, which meant that a
driver that enabled MSI-X couldn't use pci_request_regions() (since
that would clash with the PCI layer's MSI-X request).
However, that was removed (by me!) years ago, so mthca can just use
pci_request_regions() and pci_release_regions() instead of its own
much more complicated code that avoids requesting the MSI-X tables.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Handle the case where posting a send is requested when the link is
down. This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1117>.
Signed-off-by: Yannick Cote <yannick.cote@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit ee1e2c82 ("IPoIB: Refresh paths instead of flushing them on SM
change events") changed how paths are flushed on an SM event. This
change introduces a problem if the path record query triggered by
fails, causing path->ah to become NULL. A later successful path query
will then trigger WARN_ON() in path_rec_completion(), and crash
because path->ah has already been freed, so the ipoib_put_ah() inside
the lock in path_rec_completion() may actually drop the last reference
(contrary to the comment that claims this is safe).
Fix this by updating path->ah and freeing old_ah only when the path
record query is successful. This prevents the neighbour AH and that
path AH from getting out of sync.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1194>
Reported-by: Rabah Salem <ravah@mellanox.com>
Debugged-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
A break after a return serves no purpose, remove it.
Signed-off-by: Julia Lawall <julia@diku.dk>
Reviewed-by: Richard Genoud <richard.genoud@gmail.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This fixes the problem of incoming BMA responses being dropped due to
a bad "is response" check. Fix the test to use the ib_response_mad()
predicate, which correctly handles BMA MADs.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=988>.
Signed-off-by: Michael Brooks <michael.brooks@qlogic.com>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The code to set the source LID in the sent LRH was not setting the low
bits if LMC != 0 for RC/UC QPs.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When a QP goes into error state, it is required that CQ entries with a
flush error status are delivered to the application for any
outstanding work requests. eHCA does not do this in hardware, so this
patch adds software flush CQE generation to the ehca driver.
Whenever a QP gets into error state, it is added to the QP error list
of its respective CQ. If the error QP list of a CQ is not empty,
poll_cq() generates flush CQEs before polling the actual CQ.
Signed-off-by: Alexander Schmidt <alexs@linux.vnet.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IPoIB: Fix deadlock on RTNL between bcast join comp and ipoib_stop()
RDMA/nes: Fix client side QP destroy
IB/mlx4: Fix up fast register page list format
mlx4_core: Set RAE and init mtt_sz field in FRMR MPT entries
Taking rtnl_lock in ipoib_mcast_join_complete() causes a deadlock with
ipoib_stop(). We avoid it by scheduling the piece of code that takes
the lock on ipoib_workqueue instead of executing it directly. This
works because we only flush the ipoib_workqueue with the RTNL not held.
The deadlock happens because ipoib_stop() calls ipoib_ib_dev_down()
which calls ipoib_mcast_dev_flush(), which calls ipoib_mcast_free(),
which calls ipoib_mcast_leave(). The latter calls
ib_sa_free_multicast(), and this waits until the multicast completion
handler finishes. This handler is ipoib_mcast_join_complete(), which
waits for the rtnl_lock(), which was already taken by ipoib_stop().
This bug was introduced in commit a77a57a1 ("IPoIB: Fix deadlock on
RTNL in ipoib_stop()").
Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Fix QP not being destroyed properly on the client, which leads to
userspace programs hanging on exit. This is a missing chunk from the
connection management rewrite in commit 6492cdf3 ("RDMA/nes: CM
connection setup/teardown rework").
Signed-off-by: Faisal Latif <flatif@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Byte swap the addresses in the page list for fast register work requests
to big endian to match what the HCA expectx. Also, the addresses must
have the "present" bit set so that the HCA knows it can access them.
Otherwise the HCA will fault the first time it accesses the memory
region.
Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Initialize the L_Key and R_Key for memory regions returned from
mlx4_ib_alloc_fast_reg_mr(). Otherwise callers just get garbage for
the memory keys and can't do anything useful with these MRs.
Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch lets the files using linux/version.h match the files that
#include it.
Signed-off-by: Adrian Bunk <bunk@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit c8c2afe3 ("IPoIB: Use rtnl lock/unlock when changing device
flags") added a call to rtnl_lock() in ipoib_mcast_join_task(), which
is run from the ipoib_workqueue. However, ipoib_stop() (which is run
inside rtnl_lock()) flushes this workqueue, which leads to a deadlock
if the join task is pending.
Fix this by simply not flushing the workqueue from ipoib_stop(). It
turns out that we really don't care about workqueue tasks running
during or after ipoib_stop(), as long as we make sure to flush the
workqueue before unregistering a netdev.
This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1114>.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The check for max physical address was incorrect, thus limiting the
range of allowed physical addresses.
Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If a UD QP has some work requests queued to be sent by the DMA engine
followed by a local loopback work request, we have to wait for the
previous work requests to finish or the completion for the local
loopback work request would be generated out of order. The problem
was that the work request queue pointer was already updated so that
the request would not be processed when the DMA queue drained.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Under rare circumstances, the ehca hardware might erroneously generate
two CQEs for the same WQE, which is not compliant to the IB spec and
will cause unpredictable errors like memory being freed twice. To
avoid this problem, the driver needs to detect the second CQE and
discard it.
For this purpose, introduce an array holding as many elements as the
SQ of the QP, called sq_map. Each sq_map entry stores a "reported"
flag for one WQE in the SQ. When a work request is posted to the SQ,
the respective "reported" flag is set to zero. After the arrival of a
CQE, the flag is set to 1, which allows to detect the occurence of a
second CQE.
The mapping between WQE / CQE and the corresponding sq_map element is
implemented by replacing the lowest 16 Bits of the wr_id with the
index in the queue map. The original 16 Bits are stored in the sq_map
entry and are restored when the CQE is passed to the application.
Signed-off-by: Alexander Schmidt <alexs@linux.vnet.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The idr_find() function may fail when trying to get the QP that is
associated with a CQE, e.g. when a QP has been destroyed between the
generation of a CQE and the poll request for it. In consequence, the
return value of idr_find() must be checked and the CQE must be
discarded when the QP cannot be found.
Signed-off-by: Alexander Schmidt <alexs@linux.vnet.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When the ehca driver detects an invalid opcode in a CQE, it currently
passes the CQE to the application and returns with success. This patch
changes the CQE handling to discard CQEs with invalid opcodes and to
continue reading the next CQE from the CQ.
Signed-off-by: Alexander Schmidt <alexs@linux.vnet.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Rename the "poll_cq_one_read_cqe" goto label to what it actually does,
namely "repoll".
Signed-off-by: Alexander Schmidt <alexs@linux.vnet.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Since the introduction of the port auto-detect mode for ehca, calls to
modify_qp() may be cached in the device driver when the ports are not
activated yet. When a modify_qp() call is cached, the qp state remains
untouched until the port is activated, which will leave the qp in the
reset state. In the reset state, however, it is not allowed to post SQ
WQEs, which confuses applications like ib_mad.
The solution for this problem is to immediately set the qp state as
requested by modify_qp(), even when the call is cached.
Signed-off-by: Alexander Schmidt <alexs@linux.vnet.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
There are users that are running UDP applications that require a large
receive queue size in order to get good performance. To prevent
allocation failures for rx_rings when using non-SRQ mode and large
recv_queue_size (1K or larger), use vmalloc() instead of kcalloc() to
alocate rx_rings.
Signed-off-by: David Wilder <dwilder@us.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
IB/mad: Test ib_create_send_mad() return with IS_ERR(), not == NULL
IB/mlx4: Allow 4K messages for UD QPs
mlx4_core: Add ethernet fields to CQE struct
IB/ipath: Fix printk format warnings
RDMA/cxgb3: Fix deadlock initializing iw_cxgb3 device
RDMA/cxgb3: Fix up MW access rights
RDMA/cxgb3: Fix QP capabilities
RDMA/cma: Remove padding arrays by using struct sockaddr_storage
IB/ipath: Use unsigned long for irq flags
IPoIB/cm: Set correct SG list in ipoib_cm_init_rx_wr()
In case of error, the function ib_create_send_mad() returns an ERR
pointer, but never returns a NULL pointer. So testing the return
value for error should be done with IS_ERR, not by comparing with
NULL.
A simplified version of the semantic patch that makes this change is
as follows:
(http://www.emn.fr/x-info/coccinelle/)
// <smpl>
@correct_null_test@
expression x,E;
statement S1, S2;
@@
x = ib_create_send_mad(...)
<... when != x = E
if (
(
- x@p2 != NULL
+ ! IS_ERR ( x )
|
- x@p2 == NULL
+ IS_ERR( x )
)
)
S1
else S2
...>
? x = E;
// </smpl>
Signed-off-by: Julien Brunel <brunel@diku.dk>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Current code limits the max message size to 2K for UD QPs, while MTU
might be as big as 4K. This patch sets the maximum message size to
4K, which is needed for UD to work correctly on fabrics with a 4K MTU.
Signed-off-by: Alex Naslednikov <xalex@mellanox.co.il>
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add ethernet-related fields to struct mlx4_cqe so that the mlx4_en
ethernet NIC driver can share the same definition.
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ipath_driver.c:1260: warning: format '%Lx' expects type 'long long unsigned int', but argument 6 has type 'long unsigned int'
ipath_driver.c:1459: warning: format '%Lx' expects type 'long long unsigned int', but argument 4 has type 'u64'
ipath_intr.c:358: warning: format '%Lx' expects type 'long long unsigned int', but argument 3 has type 'u64'
ipath_intr.c:358: warning: format '%Lu' expects type 'long long unsigned int', but argument 6 has type 'u64'
ipath_intr.c:1119: warning: format '%Lx' expects type 'long long unsigned int', but argument 5 has type 'u64'
ipath_intr.c:1119: warning: format '%Lx' expects type 'long long unsigned int', but argument 3 has type 'u64'
ipath_intr.c:1123: warning: format '%Lx' expects type 'long long unsigned int', but argument 3 has type 'u64'
ipath_intr.c:1130: warning: format '%Lx' expects type 'long long unsigned int', but argument 4 has type 'u64'
ipath_iba7220.c:1032: warning: format '%llx' expects type 'long long unsigned int', but argument 4 has type 'u64'
ipath_iba7220.c:1045: warning: format '%llX' expects type 'long long unsigned int', but argument 3 has type 'u64'
ipath_iba7220.c:2506: warning: format '%Lu' expects type 'long long unsigned int', but argument 4 has type 'u64'
Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Running 'ifconfig up' on the cxgb3 interface with iw_cxgb3 loaded
causes a deadlock. The rtnl lock is already held in this path. The
function fw_supports_fastreg() was introduced in 2.6.27 to
conditionally set the IB_DEVICE_MEM_MGT_EXTENSIONS bit iff the
firmware was at 7.0 or greater, and this function also acquires the
rtnl lock and which thus causes a deadlock. Further, if iw_cxgb3 is
loaded _after_ the nic interface is brought up, then the deadlock does
not occur and therefore fw_supports_fastreg() does need to grab the
rtnl lock in that path.
It turns out this code is all useless anyway. The low level driver
will NOT allow the open if the firmware isn't 7.0, so iw_cxgb3 can
always set the MEM_MGT_EXTENSIONS bit. Simplify...
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- MWs don't have local read/write permissions.
- Set the MW_BIND enabled bit if a MR has MW_BIND access.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- Set the stag0 and fastreg capability bits only for kernel qps.
- QP_PRIV flag is no longer used, so don't set it.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
There are a few places where the RDMA CM code handles IPv6 by doing
struct sockaddr addr;
u8 pad[sizeof(struct sockaddr_in6) -
sizeof(struct sockaddr)];
This is fragile and ugly; handle this in a better way with just
struct sockaddr_storage addr;
[ Also roll in patch from Aleksey Senin <alekseys@voltaire.com> to
switch to struct sockaddr_storage and get rid of padding arrays in
struct rdma_addr. ]
Signed-off-by: Roland Dreier <rolandd@cisco.com>
from include/asm-powerpc. This is the result of a
mkdir arch/powerpc/include/asm
git mv include/asm-powerpc/* arch/powerpc/include/asm
Followed by a few documentation/comment fixups and a couple of places
where <asm-powepc/...> was being used explicitly. Of the latter only
one was outside the arch code and it is a driver only built for powerpc.
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Paul Mackerras <paulus@samba.org>
Some module parameters with only one line have the '\n' at the end of the
description. This is not needed nor wanted as after the description the
type (i.e. int) is followed by a newline.
Some modules contain a multi-line description, these are not affected
by this patch.
Signed-off-by: Niels de Vos <niels.devos@wincor-nixdorf.com>
Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: John W. Linville <linville@tuxdriver.com>
Cc: Ed L. Cashin <ecashin@coraid.com>
Cc: Dave Airlie <airlied@linux.ie>
Cc: Roland Dreier <rolandd@cisco.com>
Acked-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Cc: Jeff Garzik <jeff@garzik.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
wr->sg_list should be set to the sge pointer passed in, not
priv->cm.rx_sge.
Reported-by: Hoang-Nam Nguyen <HNGUYEN@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
mlx4: Update/add Mellanox Technologies copyright lines to mlx4 driver files
mlx4_core: Add VLAN tag field to WQE control segment struct
RDMA/nes: CM connection setup/teardown rework
IPoIB: Correct help text for INFINIBAND_IPOIB_DEBUG
IPoIB/cm: Connected mode is no longer EXPERIMENTAL
RDMA/ucm: BKL is not needed for ib_ucm_open()
RDMA/ucma: BKL is not needed for ucma_open()
Add per-device dma_mapping_ops support for CONFIG_X86_64 as POWER
architecture does:
This enables us to cleanly fix the Calgary IOMMU issue that some devices
are not behind the IOMMU (http://lkml.org/lkml/2008/5/8/423).
I think that per-device dma_mapping_ops support would be also helpful for
KVM people to support PCI passthrough but Andi thinks that this makes it
difficult to support the PCI passthrough (see the above thread). So I
CC'ed this to KVM camp. Comments are appreciated.
A pointer to dma_mapping_ops to struct dev_archdata is added. If the
pointer is non NULL, DMA operations in asm/dma-mapping.h use it. If it's
NULL, the system-wide dma_ops pointer is used as before.
If it's useful for KVM people, I plan to implement a mechanism to register
a hook called when a new pci (or dma capable) device is created (it works
with hot plugging). It enables IOMMUs to set up an appropriate
dma_mapping_ops per device.
The major obstacle is that dma_mapping_error doesn't take a pointer to the
device unlike other DMA operations. So x86 can't have dma_mapping_ops per
device. Note all the POWER IOMMUs use the same dma_mapping_error function
so this is not a problem for POWER but x86 IOMMUs use different
dma_mapping_error functions.
The first patch adds the device argument to dma_mapping_error. The patch
is trivial but large since it touches lots of drivers and dma-mapping.h in
all the architecture.
This patch:
dma_mapping_error() doesn't take a pointer to the device unlike other DMA
operations. So we can't have dma_mapping_ops per device.
Note that POWER already has dma_mapping_ops per device but all the POWER
IOMMUs use the same dma_mapping_error function. x86 IOMMUs use device
argument.
[akpm@linux-foundation.org: fix sge]
[akpm@linux-foundation.org: fix svc_rdma]
[akpm@linux-foundation.org: build fix]
[akpm@linux-foundation.org: fix bnx2x]
[akpm@linux-foundation.org: fix s2io]
[akpm@linux-foundation.org: fix pasemi_mac]
[akpm@linux-foundation.org: fix sdhci]
[akpm@linux-foundation.org: build fix]
[akpm@linux-foundation.org: fix sparc]
[akpm@linux-foundation.org: fix ibmvscsi]
Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Muli Ben-Yehuda <muli@il.ibm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Avi Kivity <avi@qumranet.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Update existing Mellanox copyright lines to 2008, and add such lines
to files where they are missing.
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Major rework of CM connection setup/teardown. We had a number of issues
with MPI applications not starting/terminating properly over time.
With these changes we were able to run longer on larger clusters.
* Remove memory allocation from nes_connect() and nes_cm_connect().
* Fix mini_cm_dec_refcnt_listen() when destroying listener.
* Remove unnecessary code from schedule_nes_timer() and nes_cm_timer_tick().
* Functionalize mini_cm_recv_pkt() and process_packet().
* Clean up cm_node->ref_count usage.
* Reuse skbs if available.
Signed-off-by: Faisal Latif <flatif@neteffect.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The help text for INFINIBAND_IPOIB_DEBUG refers to "ipoib_debugfs,"
which no longer exists. Correct this to talk about the files under
debugfs that are really created.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Connected mode is now tested and used by lots of people. No need to
hide it under CONFIG_EXPERIMENTAL.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
MAINTAINERS: Remove Glenn Streiff from NetEffect entry
mlx4_core: Improve error message when not enough UAR pages are available
IB/mlx4: Add support for memory management extensions and local DMA L_Key
IB/mthca: Keep free count for MTT buddy allocator
mlx4_core: Keep free count for MTT buddy allocator
mlx4_code: Add missing FW status return code
IB/mlx4: Rename struct mlx4_lso_seg to mlx4_wqe_lso_seg
mlx4_core: Add module parameter to enable QoS support
RDMA/iwcm: Remove IB_ACCESS_LOCAL_WRITE from remote QP attributes
IPoIB: Include err code in trace message for ib_sa_path_rec_get() failures
IB/sa_query: Check if sm_ah is NULL in ib_sa_remove_one()
IB/ehca: Release mutex in error path of alloc_small_queue_page()
IB/ehca: Use default value for Local CA ACK Delay if FW returns 0
IB/ehca: Filter PATH_MIG events if QP was never armed
IB/iser: Add support for RDMA_CM_EVENT_ADDR_CHANGE event
RDMA/cma: Add RDMA_CM_EVENT_TIMEWAIT_EXIT event
RDMA/cma: Add RDMA_CM_EVENT_ADDR_CHANGE event
* 'cpus4096-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (31 commits)
NR_CPUS: Replace NR_CPUS in speedstep-centrino.c
cpumask: Provide a generic set of CPUMASK_ALLOC macros, FIXUP
NR_CPUS: Replace NR_CPUS in cpufreq userspace routines
NR_CPUS: Replace per_cpu(..., smp_processor_id()) with __get_cpu_var
NR_CPUS: Replace NR_CPUS in arch/x86/kernel/genapic_flat_64.c
NR_CPUS: Replace NR_CPUS in arch/x86/kernel/genx2apic_uv_x.c
NR_CPUS: Replace NR_CPUS in arch/x86/kernel/cpu/proc.c
NR_CPUS: Replace NR_CPUS in arch/x86/kernel/cpu/mcheck/mce_64.c
cpumask: Optimize cpumask_of_cpu in lib/smp_processor_id.c, fix
cpumask: Use optimized CPUMASK_ALLOC macros in the centrino_target
cpumask: Provide a generic set of CPUMASK_ALLOC macros
cpumask: Optimize cpumask_of_cpu in lib/smp_processor_id.c
cpumask: Optimize cpumask_of_cpu in kernel/time/tick-common.c
cpumask: Optimize cpumask_of_cpu in drivers/misc/sgi-xp/xpc_main.c
cpumask: Optimize cpumask_of_cpu in arch/x86/kernel/ldt.c
cpumask: Optimize cpumask_of_cpu in arch/x86/kernel/io_apic_64.c
cpumask: Replace cpumask_of_cpu with cpumask_of_cpu_ptr
Revert "cpumask: introduce new APIs"
cpumask: make for_each_cpu_mask a bit smaller
net: Pass reference to cpumask variable in net/sunrpc/svc.c
...
Fix up trivial conflicts in drivers/cpufreq/cpufreq.c manually
Add support for the following operations to mlx4 when device firmware
supports them:
- Send with invalidate and local invalidate send queue work requests;
- Allocate/free fast register MRs;
- Allocate/free fast register MR page lists;
- Fast register MR send queue work requests;
- Local DMA L_Key.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
MTT entries are allocated with a buddy allocator, which just keeps
bitmaps for each level of the buddy table. However, all free space
starts out at the highest order, and small allocations start scanning
from the lowest order. When the lowest order tables have no free
space, this can lead to scanning potentially millions of bits before
finding a free entry at a higher order.
We can avoid this by just keeping a count of how many free entries
each order has, and skipping the bitmap scan when an order is
completely empty. This provides a nice performance boost for a
negligible increase in memory usage.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Remove IB_ACCESS_LOCAL_WRITE from qp.qp_access_flags because this
attribute is only used to set remote permissions.
Signed-off-by: Dotan Barak <dotanba@gmail.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Print the return code of ib_sa_path_rec_get() if it fails to help
debug errors.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
If update_sm_ah() fails, it leaves the port's sm_ah as NULL. Then if
the device or module is removed, ib_sa_remove_one() will dereference a
NULL pointer when it calls kref_put(). Fix this by testing if sm_ah
is NULL before dropping the reference.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The pd->lock mutex is released on a successful return, so it should be
released on an error return as well.
The semantic patch that makes this change is as follows:
(http://www.emn.fr/x-info/coccinelle/)
// <smpl>
@@
expression l;
@@
mutex_lock(l);
... when != mutex_unlock(l)
when any
when strict
(
if (...) { ... when != mutex_unlock(l)
+ mutex_unlock(l);
return ...;
}
|
mutex_unlock(l);
)
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Some firmware versions report a Local CA ACK Delay of 0. In that
case, return a more sensible default value of 12 (-> 16 msec) instead.
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Certain firmware versions sometimes cause spurious PATH_MIG events to
occur during QP creation. Filter these events by making sure PATH_MIG
events are only handed down when they actually make sense (i.e. when
the QP has been armed at least once).
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Enhance iser to act upon notification on network stack changes that
make its RDMA connection unaligned with the link used by the stack for
the <src,dst> IPs used to establish the connection.
When RDMA_CM_EVENT_ADDR_CHANGE arrives, just disconnect the
connection, assuming that the user space iscsid daemon will reconnect,
and the new connection will be aligned with the IP stack.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Consumers that want to re-use their QPs in new connections need to
know when the QP has exited the timewait state. Report the timewait
event through the rdma_cm.
Signed-off-by: Amir Vadai <amirv@mellanox.co.il>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add an RDMA_CM_EVENT_ADDR_CHANGE event can be used by rdma-cm
consumers that wish to have their RDMA sessions always use the same
links (eg <hca/port>) as the IP stack does. In the current code, this
does not happen when bonding is used and fail-over happened but the IB
link used by an already existing session is operating fine.
Use the netevent notification for sensing that a change has happened
in the IP stack, then scan the rdma-cm ID list to see if there is an
ID that is "misaligned" with respect to the IP stack, and deliver
RDMA_CM_EVENT_ADDR_CHANGE for this ID. The consumer can act on the
event or just ignore it.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This object really should be a struct device, or at least contain a
pointer to a struct device, as it is trying to create a separate device
tree outside of the main device tree. This patch fixes this problem.
It is needed for the class core rework that is being done in the driver
core.
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This pointer really is a struct ib_device, not a struct device, so name
it properly to help prevent confusion.
This makes the followon patch in this series much smaller and easier to
understand as well.
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Acked-by: Sean Hefty <sean.hefty@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
device_create() is race-prone, so use the race-free
device_create_drvdata() instead as device_create() is going away.
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6: (102 commits)
[SCSI] scsi_dh: fix kconfig related build errors
[SCSI] sym53c8xx: Fix bogus sym_que_entry re-implementation of container_of
[SCSI] scsi_cmnd.h: remove double inclusion of linux/blkdev.h
[SCSI] make struct scsi_{host,target}_type static
[SCSI] fix locking in host use of blk_plug_device()
[SCSI] zfcp: Cleanup external header file
[SCSI] zfcp: Cleanup code in zfcp_erp.c
[SCSI] zfcp: zfcp_fsf cleanup.
[SCSI] zfcp: consolidate sysfs things into one file.
[SCSI] zfcp: Cleanup of code in zfcp_aux.c
[SCSI] zfcp: Cleanup of code in zfcp_scsi.c
[SCSI] zfcp: Move status accessors from zfcp to SCSI include file.
[SCSI] zfcp: Small QDIO cleanups
[SCSI] zfcp: Adapter reopen for large number of unsolicited status
[SCSI] zfcp: Fix error checking for ELS ADISC requests
[SCSI] zfcp: wait until adapter is finished with ERP during auto-port
[SCSI] ibmvfc: IBM Power Virtual Fibre Channel Adapter Client Driver
[SCSI] sg: Add target reset support
[SCSI] lib: Add support for the T10 (SCSI) Data Integrity Field CRC
[SCSI] sd: Move scsi_disk() accessor function to sd.h
...
Now that we have a specific lock to protect the network
device unicast and multicast lists, remove extraneous
grabs of the TX lock in cases where the code only needs
address list protection.
Signed-off-by: David S. Miller <davem@davemloft.net>
Add netif_addr_{lock,unlock}{,_bh}() helpers.
Use them to protect operations that operate on or read
the network device unicast and multicast address lists.
Also use them in cases where the code simply wants to
block calls into the driver's ->set_rx_mode() and
->set_multicast_list() methods.
Signed-off-by: David S. Miller <davem@davemloft.net>
Current code uses kmalloc() and then just does a bitwise OR operation on
qp->flags in create_qp_common(), which means that qp->flags may
potentially have some unintended bits set. This patch uses kzalloc()
and avoids further explicit clearing of structure members, which also
shrinks the code:
add/remove: 0/0 grow/shrink: 0/1 up/down: 0/-65 (-65)
function old new delta
create_qp_common 2024 1959 -65
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The RDMA CM has some logic in place to make sure that callbacks on a
given CM ID are delivered to the consumer in a serialized manner.
Specifically it has code to protect against a device removal racing
with a running callback function.
This patch simplifies this logic by using a mutex per ID instead of a
wait queue and atomic variable. This means that cma_disable_remove()
now is more properly named to cma_disable_callback(), and
cma_enable_remove() can now be removed because it just would become a
trivial wrapper around mutex_unlock().
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Keep a pointer to the local (src) netdevice in struct rdma_dev_addr,
and copy it in as part of rdma_copy_addr(). Use rdma_translate_ip()
in cma_new_conn_id() to reduce some code duplication and also make
sure the src_dev member gets set.
In a high-availability configuration the netdevice pointer can be used
by the RDMA CM to align RDMA sessions to use the same links as the IP
stack does under fail-over and route change cases.
Signed-off-by: Or Gerlitz <ogerlitz@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Handling the zero STag in receive work request requires some extra
logic in the driver:
- Only set the QP_PRIV bit for kernel mode QPs.
- Add a zero STag build function for recv wrs. The uP needs a PBL
allocated and passed down in the recv WR so it can construct a HW
PBL for the zero STag S/G entries. Note: we need to place a few
restrictions on zero STag usage because of this:
1) all SGEs in a recv WR must either be zero STag or not. No mixing.
2) an individual SGE length cannot exceed 128MB for a zero-stag SGE.
This should be OK since it's not really practical to allocate
such a large chunk of pinned contiguous DMA mapped memory.
- Add an optimized non-zero-STag recv wr format for kernel users.
This is needed to optimize both zero and non-zero STag cracking in
the recv path for kernel users.
- Remove the iwch_ prefix from the static build functions.
- Bump required FW version.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
- Change the IB_DEVICE_ZERO_STAG flag to the transport-neutral name
IB_DEVICE_LOCAL_DMA_LKEY, which is used by iWARP RNICs to indicate 0
STag support and IB HCAs to indicate reserved L_Key support.
- Add a u32 local_dma_lkey member to struct ib_device. Drivers fill
this in with the appropriate local DMA L_Key (if they support it).
- Fix up the drivers using this flag.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The MLX transport requires two extra gather entries for sends (one for
the header and one for the checksum at the end, as the comment says).
However the code checked that max_recv_sge was not too big, instead of
checking max_send_sge as it should have. Fix the code to check the
correct condition.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Exactly when the catastrophic error polling timer function runs is not
important, so use round_jiffies() to save unnecessary wakeups.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Since we use del_timer_sync() anyway, there's no need for an
additional flag to tell the timer not to rearm.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Increase IPoIB ring sizes to twice their original sizes (RX: 128->256,
TX: 64->128) to act as a shock absorber for high traffic peaks. With
the current settings, we have seen cases that there are many calls to
netif_stop_queue(), which causes degradation in throughput. Also,
larger receive buffer sizes help IPoIB in CM mode to avoid experiencing
RNR NAK conditions due to insufficient receive buffers at the SRQ.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Since IPoIB connected mode does not NETIF_F_SG, we only have one DMA
mapping per send, so we don't need a mapping[] array. Define a new
struct with a single u64 mapping member and use it for the CM tx_ring.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The IB spe. for SubnGet(NodeInfo) and query HCA says that the vendor
ID field should be the IEEE OUI assigned to the vendor. The ipath
driver was returning the PCI vendor ID instead. This will affect
applications which call ibv_query_device(). The old value was
0x001fc1 or 0x001077, the new value is 0x001175.
The vendor ID doesn't appear to be exported via /sys so that should
reduce possible compatibility issues. I'm only aware of Open MPI as a
major application which depends on this change, and they have made
necessary adjustments.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
When the driver sets the MTU of the net device outside of its
change_mtu method, it should make use of dev_set_mtu() instead of
directly setting the mtu field of struct netdevice. Otherwise
functions registered to be called upon MTU change will not get called
(this is done through call_netdevice_notifiers() in dev_set_mtu()).
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Use of this lock is required to synchronize changes to the netdvice's
data structs. Also move the call to ipoib_flush_paths() after the
modification of the netdevice flags in set_mode().
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ipoib_mcast_detach() does nothing except call ib_detach_mcast(), so just
use the core API in the one place that does a multicast group detach.
add/remove: 0/1 grow/shrink: 0/1 up/down: 0/-105 (-105)
function old new delta
ipoib_mcast_leave 357 319 -38
ipoib_mcast_detach 67 - -67
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The current code will set the Q_Key for any join of a non-sendonly
multicast group. The operation involves a modify QP operation, which
is fairly heavyweight, and is only really required after the join of
the broadcast group. Fix this by adding a parameter to ipoib_mcast_attach()
to control when the Q_Key is set.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
No need for a mutex around calls to ib_attach_mcast/ib_detach_mcast
since these operations are synchronized at the HW driver layer.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The IPOIB_MCAST_STARTED flag is not used at all since commit b3e2749b
("IPoIB: Don't drop multicast sends when they can be queued"), so
remove it.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Every caller of nes_post_cqp_request() passed it NES_CQP_REQUEST_RING_DOORBELL,
so just remove that parameter and always ring the doorbell.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Faisal Latif <flatif@neteffect.com>
cxgb3 does not currently report the page size capabilities, and
incorrectly reports them internally.
This version changes the bit-shifting to a static value (per Steve's
request).
Signed-off-by: Jon Mason <jon@opengridcomputing.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The patch tries to solve the problem of device going down and paths being
flushed on an SM change event. The method is to mark the paths as candidates for
refresh (by setting the new valid flag to 0), and wait for an ARP
probe a new path record query.
The solution requires a different and less intrusive handling of SM
change event. For that, the second argument of the flush function
changes its meaning from a boolean flag to a level. In most cases, SM
failover doesn't cause LID change so traffic won't stop. In the rare
cases of LID change, the remote host (the one that hadn't changed its
LID) will lose connectivity until paths are refreshed. This is no
worse than the current state. In fact, preventing the device from
going down saves packets that otherwise would be lost.
Signed-off-by: Moni Levy <monil@voltaire.com>
Signed-off-by: Moni Shoua <monis@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This gives ehca an autogenerated modalias and therefore enables automatic loading.
Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add "ipoib_use_lro" module parameter to enable LRO and an
"ipoib_lro_max_aggr" module parameter to set the max number of packets
to be aggregated. Make LRO controllable and LRO statistics accessible
through ethtool.
Signed-off-by: Vladimir Sokolovsky <vlad@mellanox.co.il>
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Set IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK for IPoIB's UD QPs if
supported by the underlying device. This creates an improvement of up
to 39% in bandwidth when sending multicast packets with IPoIB, and an
improvment of 12% in cpu usage.
Signed-off-by: Ron Livne <ronli@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Add support for handling the IB_QP_CREATE_MULTICAST_BLOCK_LOOPBACK
flag by using the per-multicast group loopback blocking feature of
mlx4 hardware.
Signed-off-by: Ron Livne <ronli@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- Add a new rdma ctl command called RDMA_GET_MIB to the cxgb3 low
level driver to obtain the protocol mib from the rnic hardware.
- Add new iw_cxgb3 provider method to get the MIB from the low level
driver.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch adds a sysfs attribute group called "proto_stats" under
/sys/class/infiniband/$device/ and populates this group with protocol
statistics if they exist for a given device. Currently, only iWARP
stats are defined, but the code is designed to allow InfiniBand
protocol stats if they become available. These stats are per-device
and more importantly -not- per port.
Details:
- Add union rdma_protocol_stats in ib_verbs.h. This union allows
defining transport-specific stats. Currently only iwarp stats are
defined.
- Add struct iw_protocol_stats to define the current set of iwarp
protocol stats.
- Add new ib_device method called get_proto_stats() to return protocol
statistics.
- Add logic in core/sysfs.c to create iwarp protocol stats attributes
if the device is an RNIC and has a get_proto_stats() method.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
For devices that don't support SRQs, ipoib_cm_post_receive_nonsrq() is
called from both ipoib_cm_handle_rx_wc() and ipoib_cm_nonsrq_init_rx(),
and these two callers are not synchronized against each other.
However, ipoib_cm_post_receive_nonsrq() always reuses the same receive
work request and scatter list structures, so multiple callers can end
up stepping on each other, which leads to posting garbled work
requests.
Fix this by having the caller pass in the ib_recv_wr and ib_sge
structures to use, and allocating new local structures in
ipoib_cm_nonsrq_init_rx().
Based on a patch by Pradeep Satyanarayana <pradeep@us.ibm.com> and
David Wilder <dwilder@us.ibm.com>, with debugging help from Hoang-Nam
Nguyen <hnguyen@de.ibm.com>.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The members struct iwch_rnic_attributes.vendor_id and .vendor_part_id
are write-only, so we might as well get rid of them.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Acked-by: Steve Wise <swise@opengridcomputing.com>
- set fw_ver
- set hw_ver
- set max_qp_wr to something reasonable
- set max_cqe to something reasonable
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
During corner case testing, we noticed that some versions of ehca do
not properly transition to interrupt done in special load situations.
This can be resolved by periodically triggering EOI through H_EOI, if
EQEs are pending.
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit 65adfa91 ("IB/mlx4: Fix RESET to RESET and RESET to ERROR
transitions") added some extra code to handle a QP state transition
from RESET to ERROR. However, the latest 1.2.1 version of the IB spec
has clarified that this transition is actually not allowed, so we can
remove this extra code again.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
Commit b18aad71 ("IB/mthca: Fix RESET to ERROR transition") added some
extra code to handle a QP state transition from RESET to ERROR.
However, the latest 1.2.1 version of the IB spec has clarified that
this transition is actually not allowed, so we can remove this extra
code again.
Signed-off-by: Roland Dreier <rolandd@cisco.com>
I was reviewing the QP state transition diagram in the IB 1.2.1 spec
and the code for qp_state_table[], and noticed that the code allows a
QP to be modified from IB_QPS_RESET to IB_QPS_ERR whereas the notes
for figure 124 (pg 457) specifically says that this transition isn't
allowed. This is a clarification from earlier versions of the IB
spec, which were ambiguous in this area and suggested that the RESET
to ERR transition was allowed.
Fix up the qp_state_table[] to make RESET->ERR not allowed.
Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ConnectX HCAs support the IB_MGMT_CLASS_CONG_MGMT management class, so
process MADs of this class through the MAD_IFC firmware command.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
ConnectX returns the max message size it supports through the
QUERY_DEV_CAP firmware command. When modifying a QP to RTR, the max
message size for the QP must be specified. This value must not exceed
the value declared through QUERY_DEV_CAP. The current code ignores
the max allowed size and unconditionally sets the value to 2^31. This
patch sets all QPs to the max value allowed as returned from firmware.
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
- set IB_DEVICE_MEM_MGT_EXTENSIONS capability bit if fw supports it.
- set max_fast_reg_page_list_len device attribute.
- add iwch_alloc_fast_reg_mr function.
- add iwch_alloc_fastreg_pbl
- add iwch_free_fastreg_pbl
- adjust the WQ depth for kernel mode work queues to account for
fastreg possibly taking 2 WR slots.
- add fastreg_mr work request support.
- add local_inv work request support.
- add send_with_inv and send_with_se_inv work request support.
- removed useless duplicate enums/defines for TPT/MW/MR stuff.
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
This patch adds support for the IB "base memory management extension"
(BMME) and the equivalent iWARP operations (which the iWARP verbs
mandates all devices must implement). The new operations are:
- Allocate an ib_mr for use in fast register work requests.
- Allocate/free a physical buffer lists for use in fast register work
requests. This allows device drivers to allocate this memory as
needed for use in posting send requests (eg via dma_alloc_coherent).
- New send queue work requests:
* send with remote invalidate
* fast register memory region
* local invalidate memory region
* RDMA read with invalidate local memory region (iWARP only)
Consumer interface details:
- A new device capability flag IB_DEVICE_MEM_MGT_EXTENSIONS is added
to indicate device support for these features.
- New send work request opcodes IB_WR_FAST_REG_MR, IB_WR_LOCAL_INV,
IB_WR_RDMA_READ_WITH_INV are added.
- A new consumer API function, ib_alloc_mr() is added to allocate
fast register memory regions.
- New consumer API functions, ib_alloc_fast_reg_page_list() and
ib_free_fast_reg_page_list() are added to allocate and free
device-specific memory for fast registration page lists.
- A new consumer API function, ib_update_fast_reg_key(), is added to
allow the key portion of the R_Key and L_Key of a fast registration
MR to be updated. Consumers call this if desired before posting
a IB_WR_FAST_REG_MR work request.
Consumers can use this as follows:
- MR is allocated with ib_alloc_mr().
- Page list memory is allocated with ib_alloc_fast_reg_page_list().
- MR R_Key/L_Key "key" field is updated with ib_update_fast_reg_key().
- MR made VALID and bound to a specific page list via
ib_post_send(IB_WR_FAST_REG_MR)
- MR made INVALID via ib_post_send(IB_WR_LOCAL_INV),
ib_post_send(IB_WR_RDMA_READ_WITH_INV) or an incoming send with
invalidate operation.
- MR is deallocated with ib_dereg_mr()
- page lists dealloced via ib_free_fast_reg_page_list().
Applications can allocate a fast register MR once, and then can
repeatedly bind the MR to different physical block lists (PBLs) via
posting work requests to a send queue (SQ). For each outstanding
MR-to-PBL binding in the SQ pipe, a fast_reg_page_list needs to be
allocated (the fast_reg_page_list is owned by the low-level driver
from the consumer posting a work request until the request completes).
Thus pipelining can be achieved while still allowing device-specific
page_list processing.
The 32-bit fast register memory key/STag is composed of a 24-bit index
and an 8-bit key. The application can change the key each time it
fast registers thus allowing more control over the peer's use of the
key/STag (ie it can effectively be changed each time the rkey is
rebound to a page list).
Signed-off-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
The connected mode implementation in the IPoIB driver has a large
overhead in the way SKBs are handled in the receive flow. It usually
allocates an SKB with as big as was used in the currently received SKB
and moves unused fragments from the old SKB to the new one. This
involves a loop on all the remaining fragments and incurs overhead on
the CPU. This patch, for small SKBs, allocates an SKB just large
enough to contain the received data and copies to it the data from the
received SKB. The newly allocated SKB is passed to the stack and the
old SKB is reposted.
When running netperf, UDP small messages, without this pach I get:
UDP UNIDIRECTIONAL SEND TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to
14.4.3.178 (14.4.3.178) port 0 AF_INET
Socket Message Elapsed Messages
Size Size Time Okay Errors Throughput
bytes bytes secs # # 10^6bits/sec
114688 128 10.00 5142034 0 526.31
114688 10.00 1130489 115.71
With this patch I get both send and receive at ~315 mbps.
The reason that send performance actually slows down is as follows:
When using this patch, the overhead of the CPU for handling RX packets
is dramatically reduced. As a result, we do not experience RNR NAK
messages from the receiver which cause the connection to be closed and
reopened again; when the patch is not used, the receiver cannot handle
the packets fast enough so there is less time to post new buffers and
hence the mentioned RNR NACKs. So what happens is that the
application *thinks* it posted a certain number of packets for
transmission but these packets are flushed and do not really get
transmitted. Since the connection gets opened and closed many times,
each time netperf gets the CPU time that otherwise would have been
given to IPoIB to actually transmit the packets. This can be verified
when looking at the port counters -- the output of ifconfig and the
oputput of netperf (this is for the case without the patch):
tx packets
==========
port counter: 1,543,996
ifconfig: 1,581,426
netperf: 5,142,034
rx packets
==========
netperf 1,1304,089
Signed-off-by: Eli Cohen <eli@mellanox.co.il>
