Pull networking fixes from David Miller:1) Use after free in __dev_map_entry_free(), from Eric Dumazet.
1) Use after free in __dev_map_entry_free(), from Eric Dumazet.
2) Fix TCP retransmission timestamps on passive Fast Open, from Yuchung
Cheng.
3) Orphan NFC, we'll take the patches directly into my tree. From
Johannes Berg.
4) We can't recycle cloned TCP skbs, from Eric Dumazet.
5) Some flow dissector bpf test fixes, from Stanislav Fomichev.
6) Fix RCU marking and warnings in rhashtable, from Herbert Xu.
7) Fix some potential fib6 leaks, from Eric Dumazet.
8) Fix a _decode_session4 uninitialized memory read bug fix that got
lost in a merge. From Florian Westphal.
9) Fix ipv6 source address routing wrt. exception route entries, from
Wei Wang.
10) The netdev_xmit_more() conversion was not done %100 properly in mlx5
driver, fix from Tariq Toukan.
11) Clean up botched merge on netfilter kselftest, from Florian
Westphal.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (74 commits)
of_net: fix of_get_mac_address retval if compiled without CONFIG_OF
net: fix kernel-doc warnings for socket.c
net: Treat sock->sk_drops as an unsigned int when printing
kselftests: netfilter: fix leftover net/net-next merge conflict
mlxsw: core: Prevent reading unsupported slave address from SFP EEPROM
mlxsw: core: Prevent QSFP module initialization for old hardware
vsock/virtio: Initialize core virtio vsock before registering the driver
net/mlx5e: Fix possible modify header actions memory leak
net/mlx5e: Fix no rewrite fields with the same match
net/mlx5e: Additional check for flow destination comparison
net/mlx5e: Add missing ethtool driver info for representors
net/mlx5e: Fix number of vports for ingress ACL configuration
net/mlx5e: Fix ethtool rxfh commands when CONFIG_MLX5_EN_RXNFC is disabled
net/mlx5e: Fix wrong xmit_more application
net/mlx5: Fix peer pf disable hca command
net/mlx5: E-Switch, Correct type to u16 for vport_num and int for vport_index
net/mlx5: Add meaningful return codes to status_to_err function
net/mlx5: Imply MLXFW in mlx5_core
Revert "tipc: fix modprobe tipc failed after switch order of device registration"
vsock/virtio: free packets during the socket release
...
To avoid any ambiguity between vport index and vport number,
rename functions that had vport, to vport_num or vport_index appropriately.
vport_num is u16 hence change mlx5_eswitch_index_to_vport_num() return
type to u16.
vport_index is an int in vport array. Hence change input type of vport
index in mlx5_eswitch_index_to_vport_num() to int.
Correct multiple eswitch representor interfaces use type u16 of
rep->vport as type int vport_index.
Send vport FW commands with correct eswitch u16 vport_num instead
host int vport_index.
Fixes: 5ae5162066 ("net/mlx5: E-Switch, Assign a different position for uplink rep and vport")
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Vu Pham <vuhuong@mellanox.com>
Reviewed-by: Bodong Wang <bodong@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Pull more rdma updates from Jason Gunthorpe:
"This is being sent to get a fix for the gcc 9.1 build warnings, and
I've also pulled in some bug fix patches that were posted in the last
two weeks.
- Avoid the gcc 9.1 warning about overflowing a union member
- Fix the wrong callback type for a single response netlink to doit
- Bug fixes from more usage of the mlx5 devx interface"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
net/mlx5: Set completion EQs as shared resources
IB/mlx5: Verify DEVX general object type correctly
RDMA/core: Change system parameters callback from dumpit to doit
RDMA: Directly cast the sockaddr union to sockaddr
Mark completion EQs as shared resources so that they can be used by CQs
with uid != 0.
Fixes: 7efce3691d ("IB/mlx5: Add obj create and destroy functionality")
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Pull rdma updates from Jason Gunthorpe:
"This has been a smaller cycle than normal. One new driver was
accepted, which is unusual, and at least one more driver remains in
review on the list.
Summary:
- Driver fixes for hns, hfi1, nes, rxe, i40iw, mlx5, cxgb4,
vmw_pvrdma
- Many patches from MatthewW converting radix tree and IDR users to
use xarray
- Introduction of tracepoints to the MAD layer
- Build large SGLs at the start for DMA mapping and get the driver to
split them
- Generally clean SGL handling code throughout the subsystem
- Support for restricting RDMA devices to net namespaces for
containers
- Progress to remove object allocation boilerplate code from drivers
- Change in how the mlx5 driver shows representor ports linked to VFs
- mlx5 uapi feature to access the on chip SW ICM memory
- Add a new driver for 'EFA'. This is HW that supports user space
packet processing through QPs in Amazon's cloud"
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (186 commits)
RDMA/ipoib: Allow user space differentiate between valid dev_port
IB/core, ipoib: Do not overreact to SM LID change event
RDMA/device: Don't fire uevent before device is fully initialized
lib/scatterlist: Remove leftover from sg_page_iter comment
RDMA/efa: Add driver to Kconfig/Makefile
RDMA/efa: Add the efa module
RDMA/efa: Add EFA verbs implementation
RDMA/efa: Add common command handlers
RDMA/efa: Implement functions that submit and complete admin commands
RDMA/efa: Add the ABI definitions
RDMA/efa: Add the com service API definitions
RDMA/efa: Add the efa_com.h file
RDMA/efa: Add the efa.h header file
RDMA/efa: Add EFA device definitions
RDMA: Add EFA related definitions
RDMA/umem: Remove hugetlb flag
RDMA/bnxt_re: Use core helpers to get aligned DMA address
RDMA/i40iw: Use core helpers to get aligned DMA address within a supported page size
RDMA/verbs: Add a DMA iterator to return aligned contiguous memory blocks
RDMA/umem: Add API to find best driver supported page size in an MR
...
When the state of rep was introduced, it was also designed to prevent
duplicate unloading of the same rep. Considering the following two
flows when an eswitch manager is at switchdev mode with n VF reps loaded.
+--------------------------------------+--------------------------------+
| cpu-0 | cpu-1 |
| -------- | -------- |
| mlx5_ib_remove | mlx5_eswitch_disable_sriov |
| mlx5_ib_unregister_vport_reps | esw_offloads_cleanup |
| mlx5_eswitch_unregister_vport_reps | esw_offloads_unload_all_reps |
| __unload_reps_all_vport | __unload_reps_all_vport |
+--------------------------------------+--------------------------------+
These two flows will try to unload the same rep. Per original design,
once one flow unloads the rep, the state moves to REGISTERED. The 2nd
flow will no longer needs to do the unload and bails out. However, as
read and write of the state is not atomic, when 1st flow is doing the
unload, the state is still LOADED, 2nd flow is able to do the same
unload action. Kernel crash will happen.
To solve this, driver should do atomic test-and-set for the state. So
that only one flow can change the rep state from LOADED to REGISTERED,
and proceed to do the actual unloading.
Since the state is changing to atomic type, all other read/write should
be atomic action as well.
Fixes: f121e0ea95 (net/mlx5: E-Switch, Add state to eswitch vport representors)
Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Reviewed-by: Vu Pham <vuhuong@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
mlx5_query_nic_vport_vlans() is not used anymore. Hence remove it.
This patch doesn't change any functionality.
Signed-off-by: Bodong Wang <bodong@mellanox.com>
Reviewed-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Add the support to read additional EEPROM information from high pages.
Information for modules such as SFF-8436 and SFF-8636:
1) Application select table
2) User writable EEPROM
3) Thresholds and alarms
Signed-off-by: Erez Alfasi <ereza@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
This merge commit includes some misc shared code updates from mlx5-next branch needed
for net-next.
1) From Aya: Enable general events on all physical link types and
restrict general event handling of subtype DELAY_DROP_TIMEOUT in mlx5 rdma
driver to ethernet links only as it was intended.
2) From Eli: Introduce low level bits for prio tag mode
3) From Maor: Low level steering updates to support RDMA RX flow
steering and enables RoCE loopback traffic when switchdev is enabled.
4) From Vu and Parav: Two small mlx5 core cleanups
5) From Yevgeny add HW definitions of geneve offloads
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
The cited commit broke the offsets of hca cap struct, fix it.
While at it, cleanup a white space introduced by the same commit.
Fixes: b169e64a24 ("net/mlx5: Geneve, Add flow table capabilities for Geneve decap with TLV options")
Reported-by: Qian Cai <cai@lca.pw>
Cc: Yevgeny Kliteynik <kliteyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Introduce specification for Geneve decap flow with encapsulation options
and allow creation of rules that are matching on Geneve TLV options.
Reviewed-by: Oz Shlomo <ozsh@mellanox.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Introduce support for Geneve flow specification and allow
the creation of rules that are matching on basic Geneve
protocol fields: VNI, OAM bit, protocol type, options length.
Reviewed-by: Oz Shlomo <ozsh@mellanox.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
When in switchdev mode, we would like to treat loopback RoCE
traffic (on eswitch manager) as RDMA and not as regular
Ethernet traffic
In order to enable it we add flow steering rule that forward RoCE
loopback traffic to the HW RoCE filter (by adding allow rule).
In addition we add RoCE address in GID index 0, which will be
set in the RoCE loopback packet.
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Acked-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Flow table supports three types of miss action:
1. Default miss action - go to default miss table according to table.
2. Go to specific table.
3. Switch domain - go to the root table of an alternative steering
table domain.
New table miss action was added - switch_domain.
The next domain for RDMA_RX namespace is the NIC RX domain.
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Add new flow steering namespace - MLX5_FLOW_NAMESPACE_RDMA_RX.
Flow steering rules in this namespace are used to filter
RDMA traffic.
Signed-off-by: Maor Gottlieb <maorg@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Currently mlx5 core stores copy of the PCI device name in a
mlx5_priv structure and uses pr_warn, pr_err helpers.
Get rid of the copy of this name; instead store the parent device
pointer that contains name as well as dma specific parameters.
This also allows to use kernel's well defined dev_warn, dev_err, dev_dbg
device specific print routines.
This is also a preparation patch to access non PCI parent device in
future.
Signed-off-by: Parav Pandit <parav@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Current ConnectX HW is unable to perform VLAN pop in TX path and VLAN
push on RX path. To workaround that limitation untagged packets will be
tagged with VLAN ID 0x000 (priority tag) and pop/push actions will be
replaced by VLAN re-write actions (which are supported by the HW).
Introduce prio tag mode as a pre-step to controlling the workaround
behavior.
Signed-off-by: Eli Britstein <elibr@mellanox.com>
Reviewed-by: Oz Shlomo <ozsh@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Ariel Levkovich says:
====================
The series exposes the ICM address of the receive transport
interface (TIR) of Raw Packet and RSS QPs to the user since they are
required to properly create and insert steering rules that direct flows to
these QPs.
====================
For dependencies this branch is based on mlx5-next from
git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux
* branch 'mlx5_tir_icm':
IB/mlx5: Expose TIR ICM address to user space
net/mlx5: Introduce new TIR creation core API
net/mlx5: Expose TIR ICM address in command outbox
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Introducing new TIR creation core API which allows caller
to receive back from the call the full command outbox.
This comes as a preparation for the next patch that will
retrieve the TIR ICM address from the command outbox.
Signed-off-by: Ariel Levkovich <lariel@mellanox.com>
Reviewed-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Adding the TIR ICM address to the create_tir command outbox
through which the device reports the ICM address of the newly
created TIR.
The TIR address can be used for direct attachment to a steering
rule in SW managed steering mode.
Signed-off-by: Ariel Levkovich <lariel@mellanox.com>
Reviewed-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Add SW ICM related fields to the device memory capabilities
structure and sw ownership capability in flow table properties.
The currently supported SW ICM types are steering and header modify
and the changes exposes the device memory capabilities for each
of these two types.
SW ICM memory can be allocated by SW and then be accessed by RDMA
operations for direct management of the HW packet handling tables.
Signed-off-by: Ariel Levkovich <lariel@mellanox.com>
Reviewed-by: Eli Cohen <eli@mellanox.com>
Reviewed-by: Mark Bloch <markb@mellanox.com>
Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Jason Gunthorpe says:
====================
Upon review it turns out there are some long standing problems in BAR
mapping area:
* BAR pages intended for read-only can be switched to writable via mprotect.
* Missing use of rdma_user_mmap_io for the mlx5 clock BAR page.
* Disassociate causes SIGBUS when touching the pages.
* CPU pages are being mapped through to the process via remap_pfn_range
instead of the more appropriate vm_insert_page, causing weird behaviors
during disassociation.
This series adds the missing VM_* flag manipulation, adds faulting a zero
page for disassociation and revises the CPU page mappings to use
vm_insert_page.
====================
For dependencies this branch is based on for-rc from
git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma.git
* branch 'rdma_mmap':
RDMA: Remove rdma_user_mmap_page
RDMA/mlx5: Use get_zeroed_page() for clock_info
RDMA/ucontext: Fix regression with disassociate
RDMA/mlx5: Use rdma_user_map_io for mapping BAR pages
RDMA/mlx5: Do not allow the user to write to the clock page
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Upon high packet rate with multiple CPUs TX workloads, much of the HCA's
resources are spent on prefetching TX descriptors, thus affecting
transmission rates.
This patch comes to mitigate this problem by moving some workload to the
CPU and reducing the HW data prefetch overhead for small packets (<= 256B).
When forwarding packets with XDP, a packet that is smaller
than a certain size (set to ~256 bytes) would be sent inline within
its WQE TX descrptor (mem-copied), when the hardware tx queue is congested
beyond a pre-defined water-mark.
This is added to better utilize the HW resources (which now makes
one less packet data prefetch) and allow better scalability, on the
account of CPU usage (which now 'memcpy's the packet into the WQE).
To load balance between HW and CPU and get max packet rate, we use
watermarks to detect how much the HW is congested and move the work
loads back and forth between HW and CPU.
Performance:
Tested packet rate for UDP 64Byte multi-stream
over two dual port ConnectX-5 100Gbps NICs.
CPU: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz
* Tested with hyper-threading disabled
XDP_TX:
| | before | after | |
| 24 rings | 51Mpps | 116Mpps | +126% |
| 1 ring | 12Mpps | 12Mpps | same |
XDP_REDIRECT:
** Below is the transmit rate, not the redirection rate
which might be larger, and is not affected by this patch.
| | before | after | |
| 32 rings | 64Mpps | 92Mpps | +43% |
| 1 ring | 6.4Mpps | 6.4Mpps | same |
As we can see, feature significantly improves scaling, without
hurting single ring performance.
Signed-off-by: Shay Agroskin <shayag@mellanox.com>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
