IB and iWARP specs both spell out that posting a receive work request
to a queue pair in the RESET state is an invalid operation and required
to fail.  Postpone posting receive work requests until after the
transition to the INIT state.

Fixes: commit dea26209 ("IB/srpt: Change default behavior from using SRQ to using RC")
Signed-off-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

NAPI budget is 64 packets, while maximum polling size for
the send CQ is 16. Let's bring them in sync, so the NAPI
budget will be reused completely.

Cc: Kamal Heib <kamalh@mellanox.com>
Signed-off-by: NErez Shitrit <erezsh@mellanox.com>
Reviewed-by: NAlex Vesker <valex@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Instead of explicit call to poll_cq of the tx ring, use the NAPI mechanism
to handle the completions of each packet that has been sent to the HW.

The next major changes were taken:
 * The driver init completion function in the creation of the send CQ,
   that function triggers the napi scheduling.
 * The driver uses CQ for RX for both modes UD and CM, and CQ for TX
   for CM and UD.

Cc: Kamal Heib <kamalh@mellanox.com>
Signed-off-by: NErez Shitrit <erezsh@mellanox.com>
Reviewed-by: NAlex Vesker <valex@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The first step toward using NAPI in the UD/TX flow is to separate
between two flows, the NAPI and the xmit, meaning no use of shared
variables between both flows.

This patch takes out the tx_outstanding variable that was used in both
flows and instead the driver uses the 2 cyclic ring variables: tx_head
and tx_tail, tx_head used in the xmit flow and tx_tail in the NAPI flow.

Cc: Kamal Heib <kamalh@mellanox.com>
Signed-off-by: NErez Shitrit <erezsh@mellanox.com>
Reviewed-by: NAlex Vesker <valex@mellanox.com>
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

For small networks it is safe to reduce the subnet timeout from
its default value (18 for opensm) to 16. Make the SRP CM timeout
dependent on the subnet timeout such that decreasing the subnet
timeout also causes SRP failover and failback to occur faster.
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: NSagi Grimberg <sagi@grimberg.me>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This is a micro-optimization for the hot path.
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This patch does not change any functionality.
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: NSagi Grimberg <sagi@grimberg.me>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This patch fixes the following kernel crash:

general protection fault: 0000 [#1] PREEMPT SMP
Workqueue: ib_mad2 timeout_sends [ib_core]
Call Trace:
 ib_sa_path_rec_callback+0x1c4/0x1d0 [ib_core]
 send_handler+0xb2/0xd0 [ib_core]
 timeout_sends+0x14d/0x220 [ib_core]
 process_one_work+0x200/0x630
 worker_thread+0x4e/0x3b0
 kthread+0x113/0x150

Fixes: commit aef9ec39 ("IB: Add SCSI RDMA Protocol (SRP) initiator")
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: NSagi Grimberg <sagi@grimberg.me>
Cc: <stable@vger.kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Although in the RC mode more resources are needed that mode has three
advantages over SRQ:
- It works with all RDMA adapters, even those that do not support
  SRQ.
- Posting WRs and polling WCs does not trigger lock contention
  because only one thread at a time accesses a WR or WC queue in
  non-SRQ mode.
- The end-to-end flow control mechanism is used.

>From the IB spec:

    C9-150.2.1: For QPs that are not associated with an SRQ, each HCA
    receive queue shall generate end-to-end flow control credits. If
    a QP is associated with an SRQ, the HCA receive queue shall not
    generate end-to-end flow control credits.

Add new configfs attributes that allow to configure which mode to use
(/sys/kernel/config/target/srpt/$GUID/$GUID/attrib/use_srq). Note:
only the attribute for port 1 is relevant on multi-port adapters.
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

This patch is a micro-optimization for the hot path.
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Additionally, correct the comment about ch->rq_size.
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: NSagi Grimberg <sagi@grimberg.me>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Make srpt_parse_i_port_id() return a negative value if hex2bin()
fails.

Fixes: commit a42d985b ("ib_srpt: Initial SRP Target merge for v3.3-rc1")
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Avoid that gcc 7 reports the following warning when building with W=1:

warning: this statement may fall through [-Wimplicit-fallthrough=]
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Acked-by: NSagi Grimberg <sagi@grimberg.me>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

To support passing child interfaces to the lower device a new
rdma_netdev function was used, set_id. This will allow us to
attach the PKEY index lower device resources such as TIS/QP.
For devices that do not support offloads in IPoIB same logic
will be used, setting the PKEY index to priv struct.
Signed-off-by: NAlex Vesker <valex@mellanox.com>
Reviewed-by: NErez Shitrit <erezsh@mellanox.com>

When ndo_open and ndo_stop are called RTNL lock should be held.
In this specific case ipoib_ib_dev_open calls the offloaded ndo_open
which re-sets the number of TX queue assuming RTNL lock is held.
Since RTNL lock is not held, RTNL assert will fail.
Signed-off-by: NAlex Vesker <valex@mellanox.com>

Instead of making every caller convert the second argument of
sa_path_set_slid() and sa_path_set_dlid() to big endian format,
make these two functions accept LIDs in CPU endian format.
This patch does not change any functionality.
Signed-off-by: NBart Van Assche <bart.vanassche@wdc.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Dasaratharaman Chandramouli <dasaratharaman.chandramouli@intel.com>
Cc: Don Hiatt <don.hiatt@intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

In preparation for unconditionally passing the struct timer_list pointer to
all timer callbacks, switch to using the new timer_setup() and from_timer()
to pass the timer pointer explicitly.

Cc: Doug Ledford <dledford@redhat.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Cc: Leon Romanovsky <leon@kernel.org>
Cc: Alex Vesker <valex@mellanox.com>
Cc: Erez Shitrit <erezsh@mellanox.com>
Cc: Zhu Yanjun <yanjun.zhu@oracle.com>
Cc: Dasaratharaman Chandramouli <dasaratharaman.chandramouli@intel.com>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Yuval Shaia <yuval.shaia@oracle.com>
Cc: linux-rdma@vger.kernel.org
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: NKees Cook <keescook@chromium.org>
Reviewed-by: NLeon Romanovsky <leonro@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Vendor error print should be consistent across protocols to avoid any
confusion.
This patch corrects that.
Suggested-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: NAjaykumar Hotchandani <ajaykumar.hotchandani@oracle.com>
Reviewed-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Acked-by: NWengang Wang <wen.gang.wang@oracle.com>
Reviewed-by: NYuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: NYuval Shaia <yuval.shaia@oracle.com>
Acked-by: NSagi Grimberg <sagi@grimberg.me>
Acked-by: NRoi Dayan <roid@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Add protocol specific routing control information in the encapsulation
header as per the configuration.
Reviewed-by: NSudeep Dutt <sudeep.dutt@intel.com>
Signed-off-by: NNiranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
Signed-off-by: NScott Franco <safranco@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Update eth_link_status and operating status information to
represent the overall status of the virtual Ethernet switch port.
Reviewed-by: NSudeep Dutt <sudeep.dutt@intel.com>
Signed-off-by: NNiranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Clear the MAC table digest when the MAC table is freed.
Reviewed-by: NNiranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
Signed-off-by: NScott Franco <safranco@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Do not include EM specified MAC address in total MACs of the
UC MAC list.
Reviewed-by: NSudeep Dutt <sudeep.dutt@intel.com>
Signed-off-by: NNiranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Ensure MAC address is reset while deleting the vesw port.
Reviewed-by: NSudeep Dutt <sudeep.dutt@intel.com>
Signed-off-by: NNiranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Set power on default value of 1500 for eth_mtu.
Reviewed-by: NSudeep Dutt <sudeep.dutt@intel.com>
Signed-off-by: NNiranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Per pcp mtu fields are not used, mark them as reserved.
Reviewed-by: NSudeep Dutt <sudeep.dutt@intel.com>
Signed-off-by: NNiranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

commit 9a9b8112 will cause core to fail UD QP from being destroyed
on ipoib unload, therefore cause resources leakage.
On pkey change event above patch modifies mgid before calling underlying
driver to detach it from QP. Drivers' detach_mcast() will fail to find
modified mgid it was never given to attach in a first place.
Core qp->usecnt will never go down, so ib_destroy_qp() will fail.

IPoIB driver actually does take care of new broadcast mgid based on new
pkey by destroying an old mcast object in ipoib_mcast_dev_flush())
....
	if (priv->broadcast) {
		rb_erase(&priv->broadcast->rb_node, &priv->multicast_tree);
		list_add_tail(&priv->broadcast->list, &remove_list);
		priv->broadcast = NULL;
	}
...

then in restarted ipoib_macst_join_task() creating a new broadcast mcast
object, sending join request and on completion tells the driver to attach
to reinitialized QP:
...
if (!priv->broadcast) {
...
	broadcast = ipoib_mcast_alloc(dev, 0);
...
	memcpy(broadcast->mcmember.mgid.raw, priv->dev->broadcast + 4,
	       sizeof (union ib_gid));
	priv->broadcast = broadcast;
...

Fixes: 9a9b8112 ("IB/ipoib: Update broadcast object if PKey value was changed in index 0")
Cc: stable@vger.kernel.org
Reviewed-by: NMike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NAlex Estrin <alex.estrin@intel.com>
Signed-off-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NFeras Daoud <ferasda@mellanox.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Call ipoib_remove_one when one of the IPoIB ports fails to initialize in
order not to leave the module in unstable state.
Signed-off-by: NYuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

No reason to have dependency on PCI for the entire infiniband stack so
move it to KConfig of only the drivers that actually using PCI.
Signed-off-by: NYuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Call free_rdma_netdev instead of free_netdev each time we want to
release a netdevice. This call is also relevant for future freeing
of offloaded child interfaces.

This patch also adds a missing call for free netdevice when releasing
a parent interface that has child interfaces using ipoib_remove_one.

Fixes: cd565b4b ('IB/IPoIB: Support acceleration options callbacks')
Signed-off-by: NAlex Vesker <valex@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

A possible ABBA lock can happen with RTNL and vlan_rwsem.
For example:

Flow A: Device Flush
	__ipoib_ib_dev_flush
	down_read(vlan_rwsem) 			// Lock A
	ipoib_flush_ah
	flush_workqueue(priv->wq) 		// Wait for completion
	A work on shared WQ (Mcast carrier)
		ipoib_mcast_carrier_on_task
		while (!rtnl_trylock())         // Wait for lock B

Flow B: Sysfs PKEY delete
	ipoib_vlan_delete
	lock(RTNL) 				// Lock B
	down_write(vlan_rwsem) 			// Wait for lock A

This can happen with PKEY creates as well. The solution is to release
the RTNL lock in sysfs functions in case it is not possible to lock
VLAN RW semaphore and reset the SYS call.

Fixes: 69956d83 ("IB/ipoib: Sync between remove_one to sysfs calls that use rtnl_lock")
Signed-off-by: NShalom Lagziel <shaloml@mellanox.com>
Signed-off-by: NAlex Vesker <valex@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The ib_mr->length represents the length of the MR in bytes as per
the IBTA spec 1.3 section 11.2.10.3 (REGISTER PHYSICAL MEMORY REGION).

Currently ib_mr->length field is defined as only 32-bits field.
This might result into truncation and failed WRs of consumers who
registers more than 4GB bytes memory regions and whose WRs accessing
such MRs.

This patch makes the length 64-bit to avoid such truncation.

Cc: Sagi Grimberg <sagi@grimberg.me>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Faisal Latif <faisal.latif@intel.com>
Fixes: 4c67e2bf ("IB/core: Introduce new fast registration API")
Signed-off-by: NIlya Lesokhin <ilyal@mellanox.com>
Signed-off-by: NParav Pandit <parav@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

IPoIB doesn't support transport/rnr retry schemes as per
RFC so those errors are expected. No need to flood the
log files with them.
Tested-by: NMichael Nowak <michael.nowak@oracle.com>
Tested-by: NRafael Alejandro Peralez <rafael.peralez@oracle.com>
Tested-by: NLiwen Huang <liwen.huang@oracle.com>
Tested-by: NHong Liu <hong.x.liu@oracle.com>
Reviewed-by: NMukesh Kacker <mukesh.kacker@oracle.com>
Reported-by: NRajiv Raja <rajiv.raja@oracle.com>
Signed-off-by: NSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: NYuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Adds support for ioctl callback in the RDMA netdevs to allow
supporting functions not handled by the generic interface code.
Signed-off-by: NFeras Daoud <ferasda@mellanox.com>
Signed-off-by: NEitan Rabin <rabin@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Reviewed-by: NYuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

In order to avoid deadlock between sysfs functions (like create/delete
child) and remove_one (both of them are using the sysfs lock and
rtnl_lock) the driver will use a state mutex for sync.

That will fix traces as the following:
schedule+0x3e/0x90
kernfs_drain+0x75/0xf0
? wait_woken+0x90/0x90
__kernfs_remove+0x12e/0x1c0
kernfs_remove+0x25/0x40
sysfs_remove_dir+0x57/0x90
kobject_del+0x22/0x60
device_del+0x195/0x230
 pm_runtime_set_memalloc_noio+0xac/0xf0
netdev_unregister_kobject+0x71/0x80
rollback_registered_many+0x205/0x2f0
rollback_registered+0x31/0x40
unregister_netdevice_queue+0x58/0xb0
unregister_netdev+0x20/0x30
ipoib_remove_one+0xb7/0x240 [ib_ipoib]
ib_unregister_device+0xbc/0x1b0 [ib_core]
ib_unregister_mad_agent+0x29/0x30 [ib_core]
mlx4_ib_remove+0x67/0x280 [mlx4_ib]
INFO: task echo:24082 blocked for more than 120 seconds.
Tainted: G           OE   4.1.12-37.5.1.el6uek.x86_64 #2
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this
message.
Call Trace:
schedule+0x3e/0x90
schedule_preempt_disabled+0xe/0x10
__mutex_lock_slowpath+0x95/0x110
? _rcu_barrier+0x177/0x220
mutex_lock+0x23/0x40
rtnl_lock+0x15/0x20
netdev_run_todo+0x81/0x1f0
rtnl_unlock+0xe/0x10
ipoib_vlan_delete+0x12f/0x1c0 [ib_ipoib]
delete_child+0x69/0x80 [ib_ipoib]
dev_attr_store+0x20/0x30
sysfs_kf_write+0x41/0x50
Signed-off-by: NErez Shitrit <erezsh@mellanox.com>
Reviewed-by: NAlex Vesker <valex@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

The functions ib_register_event_handler() and
ib_unregister_event_handler() always returned success and they can't fail.

Let's convert those functions to be void, remove redundant checks and
cleanup tons of goto statements.
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

There is a need to forward FW version to user space
application through RDMA netlink. In order to make it safe, there
is need to declare nla_policy and limit the size of FW string.

The new define IB_FW_VERSION_NAME_MAX will limit the size of
FW version string. That define was chosen to be equal to
ETHTOOL_FWVERS_LEN, because many drivers anyway are limited
by that value indirectly.

The introduction of this define allows us to remove the string size
from get_fw_str function signature.
Signed-off-by: NLeon Romanovsky <leonro@mellanox.com>

IPoIB contains local_lid field which is 16 bits in
length, increase it to 32 bits.
Signed-off-by: NDasaratharaman Chandramouli <dasaratharaman.chandramouli@intel.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Signed-off-by: NDon Hiatt <don.hiatt@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

srpt contains lid and sm_lid fields which are 16 bits in
length, increase them to 32 bits.
Signed-off-by: NDasaratharaman Chandramouli <dasaratharaman.chandramouli@intel.com>
Reviewed-by: NIra Weiny <ira.weiny@intel.com>
Signed-off-by: NDon Hiatt <don.hiatt@intel.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

All modules in drivers/infiniband defined and used MODULE_VERSION, which
was pointless because the kernel version describes their state more accurate
then those arbitrary numbers.
Signed-off-by: NLeon Romanovsky <leon@kernel.org>
Acked-by: NSagi Grimbrg <sagi@grimberg.me>
Reviewed-by: NSagi Grimberg <sagi@grimbeg.me>
Acked-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Reviewed-by: NDennis Dalessandro <dennis.dalessandro@intel.com>
Acked-by: NSelvin Xavier <selvin.xavier@broadcom.com>
Acked-by: NRam Amrani <Ram.Amrani@cavium.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Acked-by: NAdit Ranadive <aditr@vmware.com>
Signed-off-by: NDoug Ledford <dledford@redhat.com>

Modify QP can fail and it can be acceptable, like when moving from RST to
ERR state, all the rest are not acceptable and a message to the log
should be printed.

The current code prints on all failures and many messages like:
"Failed to modify QP to ERROR state" appear, even when supported by the
state machine of the QP object.
Signed-off-by: NErez Shitrit <erezsh@mellanox.com>
Signed-off-by: NLeon Romanovsky <leon@kernel.org>