Complete the transition of all WQ types to use fragmented
order-0 coherent memory instead of high-order allocations.
CQ-WQ already uses order-0.
Here we do the same for cyclic and linked-list WQs.
This allows the driver to load cleanly on systems with a highly
fragmented coherent memory.
Performance tests:
ConnectX-5 100Gbps, CPU: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
Packet rate of 64B packets, single transmit ring, size 8K.
No degradation is sensed.
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Avoid using the kernel's irq_descriptor and return IRQ vector affinity
directly from the driver.
This fixes the following build break when CONFIG_SMP=n
include/linux/mlx5/driver.h: In function ‘mlx5_get_vector_affinity_hint’:
include/linux/mlx5/driver.h:1299:13: error:
‘struct irq_desc’ has no member named ‘affinity_hint’
Fixes: 6082d9c9c9 ("net/mlx5: Fix mlx5_get_vector_affinity function")
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
CC: Randy Dunlap <rdunlap@infradead.org>
CC: Guenter Roeck <linux@roeck-us.net>
CC: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Israel Rukshin <israelr@mellanox.com>
Reported-by: kbuild test robot <lkp@intel.com>
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adding the vector offset when calling to mlx5_vector2eqn() is wrong.
This is because mlx5_vector2eqn() checks if EQ index is equal to vector number
and the fact that the internal completion vectors that mlx5 allocates
don't get an EQ index.
The second problem here is that using effective_affinity_mask gives the same
CPU for different vectors.
This leads to unmapped queues when calling it from blk_mq_rdma_map_queues().
This doesn't happen when using affinity_hint mask.
Fixes: 2572cf57d7 ("mlx5: fix mlx5_get_vector_affinity to start from completion vector 0")
Fixes: 05e0cc84e0 ("net/mlx5: Fix get vector affinity helper function")
Signed-off-by: Israel Rukshin <israelr@mellanox.com>
Reviewed-by: Max Gurtovoy <maxg@mellanox.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Add two new parameters: max_burst_sz and typical_pkt_size (both
in bytes) to rate limit configurations.
max_burst_sz: The device will schedule bursts of packets for an
SQ connected to this rate, smaller than or equal to this value.
Value 0x0 indicates packet bursts will be limited to the device
defaults. This field should be used if bursts of packets must be
strictly kept under a certain value.
typical_pkt_size: When the rate limit is intended for a stream of
similar packets, stating the typical packet size can improve the
accuracy of the rate limiter. The expected packet size will be
the same for all SQs associated with the same rate limit index.
Ethernet driver is updated according to this change, but these two
parameters will be kept as 0 due to lacking of proper way to get the
configurations from user space which requires to change
ndo_set_tx_maxrate interface.
Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Due to bug fixes found by the syzkaller bot and taken into the for-rc
branch after development for the 4.17 merge window had already started
being taken into the for-next branch, there were fairly non-trivial
merge issues that would need to be resolved between the for-rc branch
and the for-next branch. This merge resolves those conflicts and
provides a unified base upon which ongoing development for 4.17 can
be based.
Conflicts:
drivers/infiniband/hw/mlx5/main.c - Commit 42cea83f95
(IB/mlx5: Fix cleanup order on unload) added to for-rc and
commit b5ca15ad7e (IB/mlx5: Add proper representors support)
add as part of the devel cycle both needed to modify the
init/de-init functions used by mlx5. To support the new
representors, the new functions added by the cleanup patch
needed to be made non-static, and the init/de-init list
added by the representors patch needed to be modified to
match the init/de-init list changes made by the cleanup
patch.
Updates:
drivers/infiniband/hw/mlx5/mlx5_ib.h - Update function
prototypes added by representors patch to reflect new function
names as changed by cleanup patch
drivers/infiniband/hw/mlx5/ib_rep.c - Update init/de-init
stage list to match new order from cleanup patch
Signed-off-by: Doug Ledford <dledford@redhat.com>
This patch validates user provided input to prevent integer overflow due
to integer manipulation in the mlx5_ib_create_srq function.
Cc: syzkaller <syzkaller@googlegroups.com>
Fixes: e126ba97db ("mlx5: Add driver for Mellanox Connect-IB adapters")
Signed-off-by: Boris Pismenny <borisp@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
In preparation for IB representors, move representors structs to a global
scope, also expose functions needed for registration, unregistration,
eswitch mode and creating a flow rule to direct traffic from SQs to the
right VF.
Signed-off-by: Mark Bloch <markb@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
The current implementation of create CQ requires contiguous
memory, such requirement is problematic once the memory is
fragmented or the system is low in memory, it causes for
failures in dma_zalloc_coherent().
This patch implements new scheme of fragmented CQ to overcome
this issue by introducing new type: 'struct mlx5_frag_buf_ctrl'
to allocate fragmented buffers, rather than contiguous ones.
Base the Completion Queues (CQs) on this new fragmented buffer.
It fixes following crashes:
kworker/29:0: page allocation failure: order:6, mode:0x80d0
CPU: 29 PID: 8374 Comm: kworker/29:0 Tainted: G OE 3.10.0
Workqueue: ib_cm cm_work_handler [ib_cm]
Call Trace:
[<>] dump_stack+0x19/0x1b
[<>] warn_alloc_failed+0x110/0x180
[<>] __alloc_pages_slowpath+0x6b7/0x725
[<>] __alloc_pages_nodemask+0x405/0x420
[<>] dma_generic_alloc_coherent+0x8f/0x140
[<>] x86_swiotlb_alloc_coherent+0x21/0x50
[<>] mlx5_dma_zalloc_coherent_node+0xad/0x110 [mlx5_core]
[<>] ? mlx5_db_alloc_node+0x69/0x1b0 [mlx5_core]
[<>] mlx5_buf_alloc_node+0x3e/0xa0 [mlx5_core]
[<>] mlx5_buf_alloc+0x14/0x20 [mlx5_core]
[<>] create_cq_kernel+0x90/0x1f0 [mlx5_ib]
[<>] mlx5_ib_create_cq+0x3b0/0x4e0 [mlx5_ib]
Signed-off-by: Yonatan Cohen <yonatanc@mellanox.com>
Reviewed-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
EQ structure and API is private to mlx5_core driver only, external
drivers should not have access or the means to manipulate EQ objects.
Remove redundant exports and move API functions out of the linux/mlx5
include directory into the driver's mlx5_core.h private include file.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Gal Pressman <galp@mellanox.com>
Since CQ tree is now per EQ, CQ completion and event forwarding became
specific implementation of EQ logic, this patch moves that logic to eq.c
and makes those functions static.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Gal Pressman <galp@mellanox.com>
Before this patch the driver had one CQ database protected via one
spinlock, this spinlock is meant to synchronize between CQ
adding/removing and CQ IRQ interrupt handling.
On a system with large number of CPUs and on a work load that requires
lots of interrupts, this global spinlock becomes a very nasty hotspot
and introduces a contention between the active cores, which will
significantly hurt performance and becomes a bottleneck that prevents
seamless cpu scaling.
To solve this we simply move the CQ database and its spinlock to be per
EQ (IRQ), thus per core.
Tested with:
system: 2 sockets, 14 cores per socket, hyperthreading, 2x14x2=56 cores
netperf command: ./super_netperf 200 -P 0 -t TCP_RR -H <server> -l 30 -- -r 300,300 -o -s 1M,1M -S 1M,1M
WITHOUT THIS PATCH:
Average: CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle
Average: all 4.32 0.00 36.15 0.09 0.00 34.02 0.00 0.00 0.00 25.41
Samples: 2M of event 'cycles:pp', Event count (approx.): 1554616897271
Overhead Command Shared Object Symbol
+ 14.28% swapper [kernel.vmlinux] [k] intel_idle
+ 12.25% swapper [kernel.vmlinux] [k] queued_spin_lock_slowpath
+ 10.29% netserver [kernel.vmlinux] [k] queued_spin_lock_slowpath
+ 1.32% netserver [kernel.vmlinux] [k] mlx5e_xmit
WITH THIS PATCH:
Average: CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle
Average: all 4.27 0.00 34.31 0.01 0.00 18.71 0.00 0.00 0.00 42.69
Samples: 2M of event 'cycles:pp', Event count (approx.): 1498132937483
Overhead Command Shared Object Symbol
+ 23.33% swapper [kernel.vmlinux] [k] intel_idle
+ 1.69% netserver [kernel.vmlinux] [k] mlx5e_xmit
Tested-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Reviewed-by: Gal Pressman <galp@mellanox.com>
The consumers of this routine expects the affinity map of of vector
index relative to the first completion vector. The upper layers are
not aware of internal/private completion vectors that mlx5 allocates
for its own usage.
Hence, return the affinity map of vector index relative to the first
completion vector.
Fixes: 05e0cc84e0 ("net/mlx5: Fix get vector affinity helper function")
Reported-by: Logan Gunthorpe <logang@deltatee.com>
Tested-by: Max Gurtovoy <maxg@mellanox.com>
Reviewed-by: Max Gurtovoy <maxg@mellanox.com>
Cc: <stable@vger.kernel.org> # v4.15
Signed-off-by: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Doug Ledford <dledford@redhat.com>
To resolve conflicts in:
drivers/infiniband/hw/mlx5/main.c
drivers/infiniband/hw/mlx5/qp.c
From patches merged into the -rc cycle. The conflict resolution matches
what linux-next has been carrying.
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Adds a new page to mlx5 core containing clock info data that allows
user level applications to translate between cqe timestamp to
nanoseconds. The information stored into this page is represented
through mlx5_ib_clock_info.
In order to synchronize between kernel and user space a sequence
number is incremented at the beginning and end of each update.
An odd number means the data is being updated while an even means
the access was already done. To guarantee that the data structure
was accessed atomically user will:
repeat:
seq1 = <read sequence>
goto <repeate> while odd
<read data structure>
seq2 = <read sequence>
if seq1 != seq2 goto repeat
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Reviewed-by: Jason Gunthorpe <jgg@mellanox.com>
Reviewed-by: Alex Vesker <valex@mellanox.com>
Signed-off-by: Feras Daoud <ferasda@mellanox.com>
Signed-off-by: Eitan Rabin <rabin@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Doug Ledford <dledford@redhat.com>
mlx5_get_vector_affinity used to call pci_irq_get_affinity and after
reverting the patch that sets the device affinity via PCI_IRQ_AFFINITY
API, calling pci_irq_get_affinity becomes useless and it breaks RDMA
mlx5 users. To fix this, this patch provides an alternative way to
retrieve IRQ vector affinity using legacy IRQ API, following
smp_affinity read procfs implementation.
Fixes: 231243c827 ("Revert mlx5: move affinity hints assignments to generic code")
Fixes: a435393aca ("mlx5: move affinity hints assignments to generic code")
Cc: Sagi Grimberg <sagi@grimberg.me>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
When in dual port mode setting a RoCE GID for any port flows through the
master ports mlx5_core_dev. Provide an interface to set the port when
sending this command.
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
When mlx5_ib_add is called determine if the mlx5 core device being
added is capable of dual port RoCE operation. If it is, determine
whether it is a master device or a slave device using the
num_vhca_ports and affiliate_nic_vport_criteria capabilities.
If the device is a slave, attempt to find a master device to affiliate it
with. Devices that can be affiliated will share a system image guid. If
none are found place it on a list of unaffiliated ports. If a master is
found bind the port to it by configuring the port affiliation in the NIC
vport context.
Similarly when mlx5_ib_remove is called determine the port type. If it's
a slave port, unaffiliate it from the master device, otherwise just
remove it from the unaffiliated port list.
The IB device is registered as a multiport device, even if a 2nd port is
not available for affiliation. When the 2nd port is affiliated later the
GID cache must be refreshed in order to get the default GIDs for the 2nd
port in the cache. Export roce_rescan_device to provide a mechanism to
refresh the cache after a new port is bound.
In a multiport configuration all IB object (QP, MR, PD, etc) related
commands should flow through the master mlx5_core_dev, other commands
must be sent to the slave port mlx5_core_mdev, an interface is provide
to get the correct mdev for non IB object commands.
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
When multiple RoCE ports are supported registration for events on
multiple netdevs is required. Refactor the event registration and
handling to support multiple ports.
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
There are two potential problems with the existing implementation.
1. Enable and disable can race after the atomic operations.
2. If a command fails the refcount is left in an inconsistent state.
Introduce a lock and perform error checking.
Fixes: a6f7d2aff6 ("net/mlx5: Add support for multiple RoCE enable")
Signed-off-by: Daniel Jurgens <danielj@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Add a missing command interface to work with a DCT. It includes: creating,
destroying and get events for.
Signed-off-by: Moni Shoua <monis@mellanox.com>
Reviewed-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Pull rdma fixes from Jason Gunthorpe:
"This is the next batch of for-rc patches from RDMA. It includes the
fix for the ipoib regression I mentioned last time, and the result of
a fairly major debugging effort to get iser working reliably on cxgb4
hardware - it turns out the cxgb4 driver was not handling QP error
flushing properly causing iser to fail.
- cxgb4 fix for an iser testing failure as debugged by Steve and
Sagi. The problem was a driver bug in the handling of shutting down
a QP.
- Various vmw_pvrdma fixes for bogus WARN_ON, missed resource free on
error unwind and a use after free bug
- Improper congestion counter values on mlx5 when link aggregation is
enabled
- ipoib lockdep regression introduced in this merge window
- hfi1 regression supporting the device in a VM introduced in a
recent patch
- Typo that breaks future uAPI compatibility in the verbs core
- More SELinux related oops fixing
- Fix an oops during error unwind in mlx5"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
IB/mlx5: Fix mlx5_ib_alloc_mr error flow
IB/core: Verify that QP is security enabled in create and destroy
IB/uverbs: Fix command checking as part of ib_uverbs_ex_modify_qp()
IB/mlx5: Serialize access to the VMA list
IB/hfi: Only read capability registers if the capability exists
IB/ipoib: Fix lockdep issue found on ipoib_ib_dev_heavy_flush
IB/mlx5: Fix congestion counters in LAG mode
RDMA/vmw_pvrdma: Avoid use after free due to QP/CQ/SRQ destroy
RDMA/vmw_pvrdma: Use refcount_dec_and_test to avoid warning
RDMA/vmw_pvrdma: Call ib_umem_release on destroy QP path
iw_cxgb4: when flushing, complete all wrs in a chain
iw_cxgb4: reflect the original WR opcode in drain cqes
iw_cxgb4: Only validate the MSN for successful completions
This patch extends the alloc context flow to be prepared for working
with dynamic UAR allocations.
Currently upon alloc context there is some fix size of UARs that are
allocated (named 'static allocation') and there is no option to user
application to ask for more or control which UAR will be used by which
QP.
In this patch the driver prepares its data structures to manage both the
static and the dynamic allocations and let the user driver knows about
the max value of dynamic blue-flame registers that are allowed.
Downstream patches from this series will enable the dynamic allocation
and the association as part of QP creation.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Congestion counters are counted and queried per physical function.
When working in LAG mode, CNP packets can be sent or received on both
of the functions, thus congestion counters should be aggregated from
the two physical functions.
Fixes: e1f24a79f4 ("IB/mlx5: Support congestion related counters")
Signed-off-by: Majd Dibbiny <majd@mellanox.com>
Reviewed-by: Aviv Heller <avivh@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
When mlx5_stop_eqs fails to destroy any of the eqs it returns with an error.
In such failure flow the function will return without
releasing all EQs irqs and then pci_free_irq_vectors will fail.
Fix by only warn on destroy EQ failure and continue to release other
EQs and their irqs.
It fixes the following kernel trace:
kernel: kernel BUG at drivers/pci/msi.c:352!
...
...
kernel: Call Trace:
kernel: pci_disable_msix+0xd3/0x100
kernel: pci_free_irq_vectors+0xe/0x20
kernel: mlx5_load_one.isra.17+0x9f5/0xec0 [mlx5_core]
Fixes: e126ba97db ("mlx5: Add driver for Mellanox Connect-IB adapters")
Signed-off-by: Moshe Shemesh <moshe@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
