- 27 Mar, 2018 14 commits
-
-
Yan Markman authored
Use relaxed I/O on the hot path. This achieves significant performance improvements. On a 10G link, this makes a basic iperf TCP test go from an average of 4.5 Gbits/sec to about 9.40 Gbits/sec. Signed-off-by:
Yan Markman <ymarkman@marvell.com> [Maxime: Commit message, cosmetic changes] Signed-off-by:
Maxime Chevallier <maxime.chevallier@bootlin.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linuxDavid S. Miller authored
Saeed Mahameed says: ==================== mlx5-updates-2018-03-22 (Misc updates) This series includes misc updates for mlx5 core and netdev dirver, Highlights: From Inbar, three patches to add support for PFC stall prevention statistics and enable/disable through new ethtool tunable, as requested from previous submission. From Moshe, four patches, added more drop counters: - drop counter for netdev steering miss - drop counter for when VF logical link is down - drop counter for when netdev logical link is down. From Or, three patches to support vlan push/pop offload via tc HW action, for newer HW (Connectx-5 and onward) via HW steering flow actions rather than the emulated path for the older HW brands. And five more misc small trivial patches. ==================== Signed-off-by: -
Intiyaz Basha authored
Napi is checking Tx queue status and waking the Tx queue if required. Same operation is being done while freeing every Tx buffer. So removed the duplicate operation of checking Tx queue status from the Tx buffer free functions. Signed-off-by:
Intiyaz Basha <intiyaz.basha@cavium.com> Signed-off-by:
Felix Manlunas <felix.manlunas@cavium.com> Signed-off-by:
-
Joe Perches authored
Remove local ADBG macro and use netdev_dbg/pr_debug Miscellanea: o Remove unnecessary debug message after allocation failure as there already is a dump_stack() on the failure paths o Leave the allocation failure message on snmp6_alloc_dev as there is one code path that does not do a dump_stack() Signed-off-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Lucas Bates authored
If tdc is executing test cases inside a namespace, only the first command in a compound statement will be executed inside the namespace by tdc. As a result, the subsequent commands are not executed inside the namespace and the test will fail. Example: for i in {x..y}; do args="foo"; done && tc actions add $args The namespace execution feature will prepend 'ip netns exec' to the command: ip netns exec tcut for i in {x..y}; do args="foo"; done && \ tc actions add $args So the actual tc command is not parsed by the shell as being part of the namespace execution. Enclosing these compound statements inside a bash invocation with proper escape characters resolves the problem by creating a subshell inside the namespace. Signed-off-by:Lucas Bates <lucasb@mojatatu.com> Signed-off-by:
-
Wei Yongjun authored
Fixes the following sparse warning: net/tipc/node.c:336:18: warning: symbol 'tipc_node_create' was not declared. Should it be static? Signed-off-by:
Wei Yongjun <weiyongjun1@huawei.com> Acked-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
-
Wei Yongjun authored
Release alloced resource before return from the error handling case in tipc_udp_enable(), otherwise will cause memory leak. Fixes: 52dfae5c ("tipc: obtain node identity from interface by default") Signed-off-by:
-
Wei Yongjun authored
Fixes the following sparse warning: drivers/net/ethernet/aquantia/atlantic/hw_atl/hw_atl_utils.c:508:5: warning: symbol 'hw_atl_utils_mpi_set_speed' was not declared. Should it be static? Signed-off-by:
-
David S. Miller authored
Maxime Chevallier says: ==================== net: mvpp2: Remove unnecessary dynamic allocs Some utility functions in mvpp2 make use of dynamic alloc to exchange temporary objects representing Parser Entries (which are generic filtering entries in the PPv2 controller). These objects are small (44 bytes each), we can use the stack to exchange them. Some previous discussion on this topic showed that the mvpp2_prs_hw_read, which initializes a struct mvpp2_prs_entry based on one of its fields, can easily lead to erroneous code if we don't zero-out the struct beforehand : https://lkml.org/lkml/2018/3/21/739 To fix this, I propose to rename mvpp2_prs_hw_read into mvpp2_prs_init_from_hw, make it zero-out the struct and take the index as a parameter. That's what's done in the first patch of the series. The second patch is the V3 of ("net: mvpp2: Don't use dynamic allocs for local variables"), making use of mvpp2_prs_init_from_hw and taking previous comments into account. ==================== Signed-off-by:
-
Maxime Chevallier authored
Some helper functions that search for given entries in the TCAM filter on PPv2 controller make use of dynamically alloced temporary variables, allocated with GFP_KERNEL. These functions can be called in atomic context, and dynamic alloc is not really needed in these cases anyways. This commit gets rid of dynamic allocs and use stack allocation in the following functions, and where they're used : - mvpp2_prs_flow_find - mvpp2_prs_vlan_find - mvpp2_prs_double_vlan_find - mvpp2_prs_mac_da_range_find For all these functions, instead of returning an temporary object representing the TCAM entry, we simply return the TCAM id that matches the requested entry. Signed-off-by:
-
Maxime Chevallier authored
The mvpp2_prs_hw_read function uses the 'index' field of the struct mvpp2_prs_entry to initialize the rest of the fields. This makes it unclear from a caller's perspective, who needs to manipulate a struct that is not entirely initialized. This commit makes it an init function for prs_entry, by passing it the index as a parameter. The function now zeroes the entry, and sets the index field before doing all other init from HW. The function is renamed 'mvpp2_prs_init_from_hw' to make that clear. Signed-off-by:
-
Colin Ian King authored
The call to nla_nest_start calls nla_put which can lead to a NULL return so it's possible for attr to become NULL and we can potentially get a NULL pointer dereference on attr. Fix this by checking for a NULL return. Detected by CoverityScan, CID#1466125 ("Dereference null return") Fixes: 955dc68c ("net/ncsi: Add generic netlink family") Signed-off-by:Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Xin Long authored
After Commit dae399d7 ("sctp: hold transport instead of assoc when lookup assoc in rx path"), it put transport instead of asoc in sctp_has_association. Variable 'asoc' is not used any more. So this patch is to remove it, while at it, it also changes the return type of sctp_has_association to bool, and does the same for it's caller sctp_endpoint_is_peeled_off. Signed-off-by:
Xin Long <lucien.xin@gmail.com> Acked-by:
Neil Horman <nhorman@tuxdriver.com> Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queueDavid S. Miller authored
Jeff Kirsher says: ==================== 40GbE Intel Wired LAN Driver Updates 2018-03-26 This series contains updates to i40e only. Jake provides several patches which remove the need for cmpxchg64(), starting with moving I40E_FLAG_[UDP]_FILTER_SYNC from pf->flags to pf->state since they are modified during run time possibly when the RTNL lock is not held so they should be a state bits and not flags. Moved additional "flags" which should be state fields, into pf->state. Ensure we hold the RTNL lock for the entire sequence of preparing for reset and when resuming, which will protect the flags related to interrupt scheme under RTNL lock so that their modification is properly threaded. Finally, cleanup the use of cmpxchg64() since it is no longer needed. Cleaned up the holes in the feature flags created my moving some flags to the state field. Björn Töpel adds XDP_REDIRECT support as well as tweaking the page counting for XDP_REDIRECT so that it will function properly. ==================== Signed-off-by:
-
- 26 Mar, 2018 26 commits
-
-
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queueDavid S. Miller authored
Jeff Kirsher says: ==================== 100GbE Intel Wired LAN Driver Updates 2018-03-26 This patch series adds the ice driver, which will support the Intel(R) E800 Series of network devices. This is the first phase in the release of this driver where we implement basic transmit and receive. The idea behind the multi-phase release is to aid in code review as well as testing. Subsequent phases will implement advanced features (like SR-IOV, tunnelling, flow director, QoS, etc.) that build upon the previous phase(s). Each phase will be submitted as a patch series. ==================== Signed-off-by: -
Björn Töpel authored
The driver now acts upon the XDP_REDIRECT return action. Two new ndos are implemented, ndo_xdp_xmit and ndo_xdp_flush. XDP_REDIRECT action enables XDP program to redirect frames to other netdevs. Signed-off-by:
Björn Töpel <bjorn.topel@intel.com> Tested-by:
Andrew Bowers <andrewx.bowers@intel.com> Signed-off-by:
Jeff Kirsher <jeffrey.t.kirsher@intel.com>
-
Björn Töpel authored
This commit tweaks the page counting for XDP_REDIRECT to function properly. XDP_REDIRECT support will be added in a future commit. The current page counting scheme assumes that the reference count cannot decrease until the received frame is sent to the upper layers of the networking stack. This assumption does not hold for the XDP_REDIRECT action, since a page (pointed out by xdp_buff) can have its reference count decreased via the xdp_do_redirect call. To work around that, we now start off by a large page count and then don't allow a refcount less than two. Signed-off-by:
-
Jacob Keller authored
Remove the gaps created by the recent refactor of various feature flags that have moved to the state field. Use only a u32 now that we have fewer than 32 flags in the field. Signed-off-by:
Jacob Keller <jacob.e.keller@intel.com> Tested-by:
-
Jacob Keller authored
Now that the only places which modify flags are either (a) during initialization prior to creating a netdevice, or (b) while holding the rtnl lock, we no longer need the cmpxchg64 call in i40e_set_priv_flags. Signed-off-by:
-
Jacob Keller authored
When we suspend and resume, we need to clear and re-enable the interrupt scheme. This was previously not done while holding the RTNL lock, which could be problematic, because we are actually destroying and re-creating queues. Hold the RTNL lock for the entire sequence of preparing for reset, and when resuming. This additionally protects the flags related to interrupt scheme under RTNL lock so that their modification is properly threaded. This is part of a larger effort to remove the need for cmpxchg64 in i40e_set_priv_flags(). Signed-off-by:
-
Jacob Keller authored
The iWarp client flags are all potentially changed when the RTNL lock is not held, so they should not be part of the pf->flags variable. Instead, move them into the state field so that we can use atomic bit operations. This is part of a larger effort to remove cmpxchg64 in i40e_set_priv_flags() Signed-off-by:
-
Jacob Keller authored
This flag is modified outside of the RTNL lock and thus should not be part of the pf->flags variable. Use a state bit instead, so that we can use atomic bit operations. This is part of a larger effort to remove cmpxchg64 in i40e_set_priv_flags() Signed-off-by:
-
Aviv Heller authored
We support outer VLAN offload in driver and HW regardless of whether an encapsulation is present in the next headers. Exposing this in hw_enc_features will allow us to offload outer VLANs in cases where encapsulation protocols like VXLAN and IPsec are used. Signed-off-by:
Aviv Heller <avivh@mellanox.com> Signed-off-by:
Saeed Mahameed <saeedm@mellanox.com>
-
Gal Pressman authored
Add a new macro to prevent copy-pasting the same code for each new feature. Signed-off-by:
Gal Pressman <galp@mellanox.com> Signed-off-by:
-
Gal Pressman authored
The function is used in en_main.c only, we can make it static and remove its declaration from en.h Signed-off-by:
-
Gal Pressman authored
ethtool core code makes sure data isn't NULL before calling get_ethtool_stats, testing it again in the driver is redundant. Signed-off-by:
-
Leon Romanovsky authored
The system with CONFIG_UBSAN enabled on produces the following error during driver initialization. The reason to it that max_reg_cmds can be larger enough to cause to "1 << max_reg_cmds" overflow the unsigned long. ================================================================================ UBSAN: Undefined behaviour in drivers/net/ethernet/mellanox/mlx5/core/cmd.c:1805:42 signed integer overflow: -2147483648 - 1 cannot be represented in type 'int' CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.15.0-rc2-00032-g06cda2358d9b-dirty #724 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5-0-ge51488c-20140602_164612-nilsson.home.kraxel.org 04/01/2014 Call Trace: dump_stack+0xe9/0x18f ? dma_virt_alloc+0x81/0x81 ubsan_epilogue+0xe/0x4e handle_overflow+0x187/0x20c mlx5_cmd_init+0x73a/0x12b0 mlx5_load_one+0x1c3d/0x1d30 init_one+0xd02/0xf10 pci_device_probe+0x26c/0x3b0 driver_probe_device+0x622/0xb40 __driver_attach+0x175/0x1b0 bus_for_each_dev+0xef/0x190 bus_add_driver+0x2db/0x490 driver_register+0x16b/0x1e0 __pci_register_driver+0x177/0x1b0 init+0x6d/0x92 do_one_initcall+0x15b/0x270 kernel_init_freeable+0x2d8/0x3d0 kernel_init+0x14/0x190 ret_from_fork+0x24/0x30 ================================================================================ Signed-off-by:
Leon Romanovsky <leonro@mellanox.com> Signed-off-by:
-
Or Gerlitz authored
Currently, we are emulating the offload of vlan push/pop actions using global setup as done by commit f5f82476 ("net/mlx5: E-Switch, Support VLAN actions in the offloads mode"). With newer NICs, we can apply a flow action for that matter, do that while keeping the emulated path for the older HW brands. Signed-off-by:
Or Gerlitz <ogerlitz@mellanox.com> Reviewed-by:
Mark Bloch <markb@mellanox.com> Signed-off-by:
-
Or Gerlitz authored
Newer NICs (ConnectX-5 and onward) can apply vlan pop or push as an action taking place during flow steering. Add the core bits for that. Signed-off-by:
-
Or Gerlitz authored
Align the checks for modify header and encap actions with the rest of the code. Signed-off-by:
-
Moshe Shemesh authored
Added the following packets drop counter: Rx interface down dropped packets - counts packets which were received while the ETH interface was down. This counter will be shown on ethtool as a new counter called rx_if_down_packets. The implementation allocates a q_counter for drop rq which gets all the received traffic while the interface is down. Signed-off-by:
Moshe Shemesh <moshe@mellanox.com> Signed-off-by:
-
Moshe Shemesh authored
Added the following packets dropped while vport down statistics: Rx dropped while vport down - counts packets which were steered by e-switch to a vport, but dropped since the vport was down. This counter will be shown on ip link tool as part of the vport rx_dropped counter. Tx dropped while vport down - counts packets which were transmitted by a vport, but dropped due to vport logical link down. This counter will be shown on ip link tool as part of the vport tx_dropped counter. The counters are read from FW by command QUERY_VNIC_ENV. Signed-off-by:
-
Moshe Shemesh authored
Added the following packets drop counter: Rx steering missed dropped packets - counts packets which were dropped due to miss on NIC rx steering rules. This counter will be shown on ethtool as a new counter called rx_steer_missed_packets. Signed-off-by:
-
Moshe Shemesh authored
Add support for new FW command QUERY_VNIC_ENV. The command is used by the driver to query vnic diagnostic statistics from FW. Signed-off-by:
-
Inbar Karmy authored
Implement set/get functions to configure PFC stall prevention timeout by tunables api through ethtool. By default the stall prevention timeout is configured to 8 sec. Timeout range is: 80-8000 msec. Enabling stall prevention with the auto timeout will set the timeout to 100 msec. Signed-off-by:
Inbar Karmy <inbark@mellanox.com> Signed-off-by:
-
Inbar Karmy authored
In the event where the device unexpectedly becomes unresponsive for a long period of time, flow control mechanism may propagate pause frames which will cause congestion spreading to the entire network. To prevent this scenario, when the device is stalled for a period longer than a pre-configured timeout, flow control mechanisms are automatically disabled. This patch adds support for the ETHTOOL_PFC_STALL_PREVENTION as a tunable. This API provides support for configuring flow control storm prevention timeout (msec). Signed-off-by:
-
Jacob Keller authored
The two Flow Directory auto disable flags are used at run time to mark when the flow director features needed to be disabled. Thus the flags could change even when the RTNL lock is not held. They also have some code constructions which really should be test_and_set or test_and_clear using atomic bit operations. Create new state fields to mark this, and stop including them in pf->flags. This is part of a larger effort to remove the need for cmpxchg64 in i40e_set_priv_flags(). Signed-off-by:
-
Jacob Keller authored
This flag is modified during run time, possibly even when the RTNL lock is not held. Additionally it has a few places which should be using test_and_set or test_and_clear atomic bit operations. Create a new state bit __I40E_UDP_SYNC_PENDING and use it instead of the ole I40E_FLAG_UDP_FILTER_SYNC flag. This is part of a larger effort to remove the need for using cmpxchg64 in i40e_set_priv_flags. Signed-off-by:
-
Inbar Karmy authored
Add the needed capability bit and counters to device spec description. Expose the following two counters in ethtool: tx_pause_storm_warning_events: when the device is stalled for a period longer than a pre-configured watermark, the counter increase, allowing the debug utility an insight into current device status. tx_pause_storm_error_events: when the device is stalled for a period longer than a pre-configured timeout, the pause transmission is disabled, and the counter increase. Signed-off-by:
-
Jacob Keller authored
The I40E_FLAG_FILTER_SYNC flag is modified during run time possibly when the RTNL lock is not held. Thus, it should not be part of pf->flags, and instead should be using atomic bit operations in the pf->state field. Create a __I40E_MACVLAN_SYNC_PENDING state bit, and use it instead of the old I40E_FLAG_FILTER_SYNC flag. This is part of a larger effort to remove the need for cmpxchg64 in i40e_set_priv_flags(). Signed-off-by:
-
