- Dec 20, 2018
-
-
Eli Britstein authored
Change the order to first route IPv4/6 and return if error. Only after successful route continue to allocate an encap header, with no functional change. Signed-off-by:
Eli Britstein <elibr@mellanox.com> Reviewed-by:
Roi Dayan <roid@mellanox.com> Signed-off-by:
Saeed Mahameed <saeedm@mellanox.com>
-
Eli Britstein authored
In tunnel encap we prepare the encap header for IPv4/6 cases, in two separate functions. For ETH header generation the code is almost duplicated. Move the ETH header generation code from IPv4/6 functions to a helper function, with no functional change. Signed-off-by:
-
Eli Britstein authored
Currently we don't support nor fail attempts to offload encap flows routed to vlan device on the underlay network. We wrongly consider a vlan underlay device to be on the same e-switch b/c the switchdev ID is retrieved recursively. Add explicit check for that and fail such attempts. Also align to a more strict check for the ingress and the underlay devices to practically be on the same eswitch. Fixes: ce99f6b9 ('net/mlx5e: Support SRIOV TC encapsulation offloads for IPv6 tunnels') Fixes: 3e621b19 ('net/mlx5e: Support TC encapsulation offloads with upper devices') Signed-off-by:
-
Eli Britstein authored
For tunnel we determine the output devs for IPv4/6 cases, in two separate functions, with a duplicated code. Move that code from IPv4/6 functions to a helper function, with no functional change. Signed-off-by:
-
Eli Britstein authored
We use the switchdev parent HW id helper to identify if the mirred device shares the same ASIC/port with the ingress device. This can get us wrong in the presence of upper devices such as vlan or bridge set over the HW devices (VF or uplink representors), b/c the switchdev ID is retrieved recursively. To fail offload attempts in such cases, we condition the check on the egress device to have not only the same switchdev ID but also the relevant mlx5 netdev ops. Fixes: 03a9d11e ('net/mlx5e: Add TC drop and mirred/redirect action parsing for SRIOV offloads') Signed-off-by:
Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by:
-
Or Gerlitz authored
There are cases (e.g tunneling with vlan on underlay and potentially more) where this makes sense, so allow that. Signed-off-by:
-
Gavi Teitz authored
The default size for the VF representors' SQ was too small to handle high packet rates. Doubling the size from 64 to 128 drastically improves the packet rate under stress (by about 50%), whereas increasing the size beyond 128 has not shown to make any further difference. The impact of the SQ size was measured with UDP traffic, in the following topology: TG <-> PF <-> TC forwarding <-> VF representor <-> VF in VM over a single core processing bi-directional traffic, with the following results: SQ size of 64: SQ size of 128: Packet rate for 64B UDP packets: 860 [Kpps] 1280 [Kpps] Packet rate for 114B VxLan encapsulated UDP packets: 320 [Kpps] 500 [Kpps] Signed-off-by:Gavi Teitz <gavi@mellanox.com> Reviewed-by:
-
Miroslav Lichvar authored
Read the system time right before and immediately after reading the low register of the internal timer. This adds support for the PTP_SYS_OFFSET_EXTENDED ioctl. Cc: Richard Cochran <richardcochran@gmail.com> Cc: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by:
Miroslav Lichvar <mlichvar@redhat.com> Signed-off-by:
-
Miroslav Lichvar authored
The timecounter needs to be updated at least once in half of the cyclecounter interval to prevent timecounter_cyc2time() interpreting a new timestamp as an old value and causing a backward jump. This would be an issue if the timecounter multiplier was so small that the update interval would not be limited by the 64-bit overflow in multiplication. Shorten the calculated interval to make sure the timecounter is updated in time even when the system clock is slowed down by up to 10%, the multiplier is increased by up to 10%, and the scheduled overflow check is late by 15%. Cc: Richard Cochran <richardcochran@gmail.com> Cc: Ariel Levkovich <lariel@mellanox.com> Cc: Saeed Mahameed <saeedm@mellanox.com> Signed-off-by:
-
David Ahern authored
Add NDA_PROTOCOL to nda_policy and use the policy for attribute parsing and validation for adding neighbors and in dump requests. Remove the now duplicate checks on nla_len. Signed-off-by:
David Ahern <dsahern@gmail.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Peng Li says: ==================== net: hns3: code optimizations & bugfixes for HNS3 driver This patchset includes bugfixes and code optimizations for the HNS3 ethernet controller driver ==================== Signed-off-by: -
Peng Li authored
This patch removes the redundant variable initialization, as driver will devm_kzalloc to set value to hdev soon. Signed-off-by:
Peng Li <lipeng321@huawei.com> Signed-off-by:
-
Peng Li authored
Driver gets rss information from the last descriptor of the packet. When driver handle the rss type, ring->next_to_clean indicates the first descriptor of next packet. This patch fix the descriptor index with "ring->next_to_clean - 1". Fixes: 232fc64b ("net: hns3: Add HW RSS hash information to RX skb") Signed-off-by:
-
Jian Shen authored
When user disables flow director, all the rules will be disabled. But when reset happens, it will restore all the rules again. It's not reasonable. This patch fixes it by add flow director status check before restore fules. Fixes: 6871af29 ("net: hns3: Add reset handle for flow director") Fixes: c17852a8 ("net: hns3: Add support for enable/disable flow director") Signed-off-by:
Jian Shen <shenjian15@huawei.com> Signed-off-by:
-
Jian Shen authored
When add flow director fule for vf, the vf id is used as array subscript before valid checking, which may cause memory overflow. Fixes: dd74f815 ("net: hns3: Add support for rule add/delete for flow director") Signed-off-by:
-
Huazhong Tan authored
While doing DOWN operation, the driver will reclaim the memory which has already used for TX. If the hardware is processing this memory, it will cause a RCB error to the hardware. According the hardware's description, the driver should reset the tqp before reclaim the memory during DOWN. Signed-off-by:
Huazhong Tan <tanhuazhong@huawei.com> Signed-off-by:
-
Jian Shen authored
Each pf supports max 64 vectors and 128 tqps. For 2p/4p core scenario, there may be more than 64 cpus online. So the result of min_t(u16, num_Online_cpus(), tqp_num) may be more than 64. This patch adds check for the vector number. Fixes: dd38c726 ("net: hns3: fix for coalesce configuration lost during reset") Signed-off-by:
-
Peng Li authored
udelay() in driver may always occupancy processor. If there is only one cpu in system, the VF driver may initialize fail when insmod PF and VF driver in the same system. This patch use msleep() to free cpu when VF wait PF message. Signed-off-by:
-
Jian Shen authored
In original codes, default tc value is set to the max tc. It's more reasonable to close tc by changing default tc value to 1. Users can enable it with lldp tool when they want to use tc. Signed-off-by:
-
Jian Shen authored
When triggering nic down, there is a time window between bringing down the protocol stack and stopping the work task. If the net is up in the time window, it may bring up the protocol stack again. This patch fixes it by stop the work task at the beginning of hns3_nic_net_stop(). To keep symmetrical, start the work task at the end of hns3_nic_net_open(). Signed-off-by:
-
David Ahern authored
pneigh_lookup uses kmalloc versus kzalloc when new entries are allocated. Given that the newly added protocol field needs to be initialized. Fixes: df9b0e30 ("neighbor: Add protocol attribute") Signed-off-by:
-
Peter Oskolkov authored
This patch refactors reuseport_add_any selftest a bit: - makes it more modular (eliminates several copy/pasted blocks); - skips DCCP tests if DCCP is not supported V2: added "Signed-off-by" tag. Signed-off-by:
Peter Oskolkov <posk@google.com> Signed-off-by:
-
Andrew Lunn authored
The 6320 family of switches uses the same watchdog registers as the 6390. Signed-off-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Vivien Didelot <vivien.didelot@gmail.com> Signed-off-by:
-
David S. Miller authored
Biao Huang says: ==================== add ethernet binding and modify ethernet driver for mt2712 changes in v3: resend this series base on the latest net-next tree. changes in v2 as comments from Sean: 1. fix typo. 2. use capital letters for RMII/MII/RGMII in driver and bindings. v1: This new series is the result of discussion in: http://lkml.org/lkml/2018/12/13/1007 http://lkml.org/lkml/2018/12/14/53 1. ethernet binding file move to this series. 2. remove fine tune property in device tree 3. remove fine tune flow in ethernet driver 4. set rgmii timing according to the value in device tree, and don't care whether phy insert internal delay or not. ==================== Signed-off-by:
-
Biao Huang authored
1. remove fine-tune property and related setting to simplify the timing adjustment flow. 2. set timing value according to the value from device tree, and will not care whether PHY insert internal delay. Signed-off-by:
Biao Huang <biao.huang@mediatek.com> Signed-off-by:
-
Biao Huang authored
remove fine-tune property in device tree, modify the corresponding description in dt-binding. Signed-off-by:
-
David S. Miller authored
Roopa Prabhu says: ==================== neigh get support This series adds support for neigh get similar to route and recently added fdb get. v2: fix key len check. and some other fixes ==================== Signed-off-by: -
Roopa Prabhu authored
Signed-off-by:
Roopa Prabhu <roopa@cumulusnetworks.com> Reviewed-by:
David Ahern <dsa@cumulusnetworks.com> Signed-off-by:
-
Roopa Prabhu authored
this patch registers neigh doit handler. The doit handler returns a neigh entry given dst and dev. This is similar to route and fdb doit (get) handlers. Also moves nda_policy declaration from rtnetlink.c to neighbour.c Signed-off-by:
-
David S. Miller authored
Ido Schimmel says: ==================== mlxsw: Make driver more robust In recent months we fixed several bugs in the driver that could have been avoided by re-evaluating some of the involved code paths and by introducing relevant and comprehensive test cases. This patchset tries to do that by introducing a set of small and mostly non-functional changes in addition to a new test. I have further improvements in mind, but they can be done in a different set. Patch #1 makes sure we correctly sanitize upper devices of a VLAN interface. Patch #2 removes an unexpected behavior from the driver, in which routes configured on a VLAN interface will cease being offloaded after certain operations. Patch #3 is a small cleanup. Patch #4 simplifies the driver by removing reference counting from VLAN entries configured on a port. Patches #5-#6 simplify linking/unlinking from a bridge, especially when LAG and VLAN devices are involved. They make both operations symmetric even when ports are unlinked from a bridged LAG device. Patch #7-#9 make router interface (RIF) deletion more robust by removing reliance on device chain to indicate whether a NETDEV_DOWN event in the inet{,6}addr notification chains should be processed. This is due to the fact that IP addresses can be flushed from a netdev after it was unlinked from its lower device. Patch #10 adds a new test to for valid and invalid configurations over mlxsw ports. Some of the test cases are derived from recent fixes. I expect that more test cases will be added over time. ==================== Signed-off-by: -
Ido Schimmel authored
Add a new test that is focused on rtnetlink configuration. Its purpose is to test valid and invalid (as deemed by mlxsw) configurations and make sure that they succeed / fail without producing a trace. Some of the test cases are derived from recent fixes in order to make sure that the fixed bugs are not introduced again. Signed-off-by:
Ido Schimmel <idosch@mellanox.com> Reviewed-by:
Petr Machata <petrm@mellanox.com> Signed-off-by:
-
Ido Schimmel authored
Previous patches tried to make RIF deletion more robust and avoid use-after-free situations. As another precaution, hold a reference on a RIF's netdev and release it when the RIF is deleted. Signed-off-by:
-
Ido Schimmel authored
In the past we had multiple instances where RIFs were not properly deleted. One of the reasons for leaking a RIF was that at the time when IP addresses were flushed from the respective netdev (prompting the destruction of the RIF), the netdev was no longer a mlxsw upper. This caused the inet{,6}addr notification blocks to ignore the NETDEV_DOWN event and leak the RIF. Instead of checking whether the netdev is our upper when an IP address is removed, we can instead check if the netdev has a RIF configured. To look up a RIF we need to access mlxsw private data, so the patch stores the notification blocks inside a mlxsw struct. This then allows us to use container_of() and extract the required private data. Signed-off-by: -
Ido Schimmel authored
Next patch is going to make RIF deletion more robust by removing reliance on fragile mlxsw_sp_lower_get(). This is because a netdev is not necessarily our upper anymore when its IP addresses are flushed. The inet{,6}addr notification blocks are going to resolve 'struct mlxsw_sp' using container_of(), but the functions they call still use mlxsw_sp_lower_get(). As a preparation for the next patch, propagate 'struct mlxsw_sp' down to the functions called from the notification blocks and remove reliance on mlxsw_sp_lower_get(). Signed-off-by: -
Ido Schimmel authored
When a LAG device or a VLAN device on top of it is enslaved to a bridge, the driver propagates the CHANGEUPPER event to the LAG's slaves. This causes each physical port to increase the reference count of the internal representation of the bridge port by calling mlxsw_sp_port_bridge_join(). However, when a port is removed from a LAG, the corresponding leave() function is not called and the reference count is not decremented. This leads to ugly hacks such as mlxsw_sp_bridge_port_should_destroy() that try to understand if the bridge port should be destroyed even when its reference count is not 0. Instead, make sure that when a port is unlinked from a LAG it would see the same events as if the LAG (or its uppers) were unlinked from a bridge. The above is achieved by walking the LAG's uppers when a port is unlinked and calling mlxsw_sp_port_bridge_leave() for each upper that is enslaved to a bridge. Signed-off-by:
-
Ido Schimmel authored
Commit b3529af6 ("spectrum: Reference count VLAN entries") started reference counting port-VLAN entries in a similar fashion to the 8021q driver. However, this is not actually needed and only complicates things. Instead, the driver should forbid the creation of a VLAN on a port if this VLAN already exists. This would also solve the issue fixed by the mentioned commit. Therefore, remove the get()/put() API and use create()/destroy() instead. One place that needs special attention is VLAN addition in a VLAN-aware bridge via switchdev operations. In case the VLAN flags (e.g., 'pvid') are toggled, then the VLAN entry already exists. To prevent the driver from wrongly returning EEXIST, the driver is changed to check in the prepare phase whether the entry already exists and only returns an error in case it is not associated with the correct bridge port. Signed-off-by:
-
Ido Schimmel authored
In commit 993107fe ("mlxsw: spectrum_switchdev: Fix VLAN device deletion via ioctl") I fixed a bug caused by the fact that the driver views differently the deletion of a VLAN device when it is deleted via an ioctl and netlink. Instead of relying on a specific order of events (device being unregistered vs. VLAN filter being updated), simply make sure that the driver performs the necessary cleanup when the VLAN device is unlinked, which always happens before the other two events. Signed-off-by:
-
Ido Schimmel authored
This function is no longer used. Remove it. Signed-off-by:
-
Ido Schimmel authored
Currently, when a RIF is constructed on top of a FID, the RIF increments the FID's reference count and the RIF is destroyed when the FID's reference count drops to 1. This effectively means that when no local ports are member in the FID, the FID is destroyed regardless if the router port is a member in the FID or not. The above can lead to the unexpected behavior in which routes using a VLAN interface as their nexthop device are no longer offloaded after the last local port leaves the corresponding VLAN (FID). Example: # ip -4 route show dev br0.10 192.0.2.0/24 proto kernel scope link src 192.0.2.1 offload # bridge vlan del vid 10 dev swp3 # ip -4 route show dev br0.10 192.0.2.0/24 proto kernel scope link src 192.0.2.1 After the patch, the route is offloaded before and after the VLAN is removed from local port 'swp3', as the RIF corresponding to 'br0.10' continues to exists. In order to remove RIFs' reliance on the underlying FID's reference count, we need to add a reference count to sub-port RIFs, which are RIFs that correspond to physical ports and their uppers (e.g., LAG devices). In this case, each {Port, VID} ('struct mlxsw_sp_port_vlan') needs to hold a reference on the RIF. For example: bond0.10 | bond0 | +-------+ | | swp1 swp2 Both {Port 1, VID 10} and {Port 2, VID 10} will hold a reference on the RIF corresponding to 'bond0.10'. When the last reference is dropped, the RIF will be destroyed. Signed-off-by: -
Ido Schimmel authored
Currently, only VRF and macvlan uppers are supported on top of VLAN device configured over a bridge, so make sure the driver forbids other uppers. Note that enslavement to a VRF is handled earlier in the notification block, so there is no need to check for a VRF upper here. Signed-off-by:
-
