- 27 May, 2020 40 commits
-
-
Johannes Berg authored
As discussed with Mathy almost two years ago in http://lore.kernel.org/r/20180806224857.14853-1-Mathy.Vanhoef@cs.kuleuven.be we should let userspace process SA-QUERY frames if it wants to, so that it can handle OCV (operating channel validation) which mac80211 doesn't know how to. Evidently I had been expecting Mathy to (re)send such a patch, but he never did, perhaps expecting me to do it after our discussion. In any case, this came up now with OCV getting more attention, so move the code around as discussed there to let userspace handle it, and do it properly. Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Link: https://lore.kernel.org/r/20200526103131.1f9cf7e5b6db.Iae5b42b09ad2b1cbcbe13492002c43f0d1d51dfc@changeidSigned-off-by:
-
Markus Theil authored
This adds the necessary capabilities in nl80211 to allow drivers to assign a cookie to control port TX frames (returned via extack in the netlink ACK message of the command) and then later report the frame's status. Signed-off-by:
Markus Theil <markus.theil@tu-ilmenau.de> Link: https://lore.kernel.org/r/20200508144202.7678-2-markus.theil@tu-ilmenau.de [use extack cookie instead of explicit message, recombine patches] Signed-off-by:
-
Gustavo A. R. Silva authored
The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member[1][2], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being inadvertently introduced[3] to the codebase from now on. Also, notice that, dynamic memory allocations won't be affected by this change: "Flexible array members have incomplete type, and so the sizeof operator may not be applied. As a quirk of the original implementation of zero-length arrays, sizeof evaluates to zero."[1] sizeof(flexible-array-member) triggers a warning because flexible array members have incomplete type[1]. There are some instances of code in which the sizeof operator is being incorrectly/erroneously applied to zero-length arrays and the result is zero. Such instances may be hiding some bugs. So, this work (flexible-array member conversions) will also help to get completely rid of those sorts of issues. This issue was found with the help of Coccinelle. [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://github.com/KSPP/linux/issues/21 [3] commit 76497732 ("cxgb3/l2t: Fix undefined behaviour") Signed-off-by:Gustavo A. R. Silva <gustavoars@kernel.org> Link: https://lore.kernel.org/r/20200507185907.GA15102@embeddedorSigned-off-by:
-
Gustavo A. R. Silva authored
The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member[1][2], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being inadvertently introduced[3] to the codebase from now on. Also, notice that, dynamic memory allocations won't be affected by this change: "Flexible array members have incomplete type, and so the sizeof operator may not be applied. As a quirk of the original implementation of zero-length arrays, sizeof evaluates to zero."[1] sizeof(flexible-array-member) triggers a warning because flexible array members have incomplete type[1]. There are some instances of code in which the sizeof operator is being incorrectly/erroneously applied to zero-length arrays and the result is zero. Such instances may be hiding some bugs. So, this work (flexible-array member conversions) will also help to get completely rid of those sorts of issues. This issue was found with the help of Coccinelle. [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://github.com/KSPP/linux/issues/21 [3] commit 76497732 ("cxgb3/l2t: Fix undefined behaviour") Signed-off-by: -
Tamizh Chelvam authored
Add IEEE80211_HE_VHT_MAX_AMPDU_FACTOR and IEEE80211_HE_HT_MAX_AMPDU_FACTOR as per spec to use for peer max ampdu factor. Signed-off-by:
Tamizh Chelvam <tamizhr@codeaurora.org> Link: https://lore.kernel.org/r/1588611900-21185-1-git-send-email-tamizhr@codeaurora.orgSigned-off-by:
-
Thomas Pedersen authored
These are found in IEEE-802.11ah-2016. Signed-off-by:
Thomas Pedersen <thomas@adapt-ip.com> Link: https://lore.kernel.org/r/20200430172554.18383-5-thomas@adapt-ip.comSigned-off-by:
-
Thomas Pedersen authored
If the driver advertises NL80211_EXT_FEATURE_SCAN_FREQ_KHZ userspace can omit NL80211_ATTR_SCAN_FREQUENCIES in favor of an NL80211_ATTR_SCAN_FREQ_KHZ. To get scan results in KHz userspace must also set the NL80211_SCAN_FLAG_FREQ_KHZ. This lets nl80211 remain compatible with older userspaces while not requring and sending redundant (and potentially incorrect) scan frequency sets. Signed-off-by:
-
Thomas Pedersen authored
cfg80211 recently gained the ability to understand a frequency offset component in KHz. Expose this in nl80211 through the new attributes NL80211_ATTR_WIPHY_FREQ_OFFSET, NL80211_FREQUENCY_ATTR_OFFSET, NL80211_ATTR_CENTER_FREQ1_OFFSET, and NL80211_BSS_FREQUENCY_OFFSET. These add support to send and receive a KHz offset component with the following NL80211 commands: - NL80211_CMD_FRAME - NL80211_CMD_GET_SCAN - NL80211_CMD_AUTHENTICATE - NL80211_CMD_ASSOCIATE - NL80211_CMD_CONNECT Along with any other command which takes a chandef, ie: - NL80211_CMD_SET_CHANNEL - NL80211_CMD_SET_WIPHY - NL80211_CMD_START_AP - NL80211_CMD_RADAR_DETECT - NL80211_CMD_NOTIFY_RADAR - NL80211_CMD_CHANNEL_SWITCH - NL80211_JOIN_IBSS - NL80211_CMD_REMAIN_ON_CHANNEL - NL80211_CMD_JOIN_OCB - NL80211_CMD_JOIN_MESH - NL80211_CMD_TDLS_CHANNEL_SWITCH If the driver advertises a band containing channels with frequency offset, it must also verify support for frequency offset channels in its cfg80211 ops, or return an error. Signed-off-by:
-
Thomas Pedersen authored
Drivers may wish to report the RX frequency in units of KHz. Provide cfg80211_rx_mgmt_khz() and wrap it with cfg80211_rx_mgmt() so exisiting drivers which can't report KHz anyway don't need to change. Add a similar wrapper for cfg80211_report_obss_beacon() so the frequency units stay somewhat consistent. This doesn't actually change the nl80211 API yet. Signed-off-by:
-
Sergey Matyukevich authored
Current rule for applying TID configuration for specific peer looks overly complicated. No need to reject new TID configuration when override flag is specified. Another call with the same TID configuration, but without override flag, allows to apply new configuration anyway. Use the same approach as for the 'all peers' case: if override flag is specified, then reset existing TID configuration and immediately apply a new one. Signed-off-by:
Sergey Matyukevich <sergey.matyukevich.os@quantenna.com> Link: https://lore.kernel.org/r/20200424112905.26770-5-sergey.matyukevich.os@quantenna.comSigned-off-by:
-
Sergey Matyukevich authored
This patch adds support to control per TID MSDU aggregation using the NL80211_TID_CONFIG_ATTR_AMSDU_CTRL attribute. Signed-off-by:
-
Sergey Matyukevich authored
Fix all variable names from 'tid' to 'tids' to avoid confusion. Now this is not TID number, but TID mask. Signed-off-by:
-
Sergey Matyukevich authored
TIDs mask type is u64 in wiphy settings and nl80211 processing, see: - wiphy TIDs mask sizes in tid_config_support structure - prepare driver command in parse_tid_conf Use the same type for TIDs mask in cfg80211_tid_cfg. Signed-off-by:
-
Oleksij Rempel authored
Add support for Atheros 100Base-T PHYs. The only difference seems to be the ability to test 2 pairs instead of 4 and the lack of 1000Base-T specific register. Only the ATH9331 was tested with this patch. Signed-off-by:
Oleksij Rempel <o.rempel@pengutronix.de> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Andrew Lunn says: ==================== Raw PHY TDR data Some ethernet PHYs allow access to raw TDR data in addition to summary diagnostics information. Add support for retrieving this data via netlink ethtool. The basic structure in the core is the same as for normal phy diagnostics, the PHY driver simply uses different helpers to fill the netlink message with different data. There is a graphical tool under development, as well a ethtool(1) which can dump the data as text and JSON. A patched ethtool(1) can be found in https://github.com/lunn/ethtool.git feature/cable-test-v5 Thanks for Chris Healy for lots of testing. v2: See the individual patches but: Pass distances in centimeters, not meters Allow the PHY to round distances to what it supports and report what it actually used along with the results. Make the Marvell PHY use steps a multiple of 0.805 meters, its native step size. v3: Move the TDR configuration into a structure Add a range check on step Use NL_SET_ERR_MSG_ATTR() when appropriate Move TDR configuration into a nest Document attributes in the request Unsquash the last two patches ==================== Signed-off-by:
-
Andrew Lunn authored
When performing a TDR measurement for a short distance, the pulse width should be low, to help differentiate between the outgoing pulse and any reflection. For longer distances, the pulse should be wider, to help with attenuation. Signed-off-by:
-
Andrew Lunn authored
Getting the TDR data requires a large number of MDIO bus transactions. The number can however be reduced if the page is only changed once. Add the needed locking to allow this, and make use of unlocked read/write methods where needed. Signed-off-by:
-
Andrew Lunn authored
Allow the user to configure where on the cable the TDR data should be retrieved, in terms of first and last sample, and the step between samples. Also add the ability to ask for TDR data for just one pair. If this configuration is not provided, it defaults to 1-150m at 1m intervals for all pairs. Signed-off-by:
-
Andrew Lunn authored
The Marvell PHYs can measure the amplitude of the returned signal for a given distance. Implement this option of the cable test infrastructure. When reporting the step, convert the distance into cm. Signed-off-by:
-
Andrew Lunn authored
Add helpers for returning raw TDR helpers in netlink messages. Signed-off-by:
-
Andrew Lunn authored
Add the generic parts of the code used to trigger a cable test and return raw TDR data. Any PHY driver which support this must implement the new driver op. Signed-off-by:
-
Andrew Lunn authored
Some Ethernet PHYs can return the raw time domain reflectromatry data. Add the attributes to allow this data to be requested and returned via netlink ethtool. Signed-off-by:
-
Armin Wolf authored
Fixed a ton of minor checkpatch errors/warnings and remove version printing at module init/when device is found and use MODULE_VERSION instead. Also modifying the RTL8029 PCI string to include the compatible RTL8029AS nic. The only mayor issue remaining is the missing SPDX tag, but since the exact version of the GPL is not stated anywhere inside the file, its impossible to add such a tag at the moment. But maybe it is possible, since 8390.h states Donald Becker's 8390 drivers are licensed under GPL 2.2 only (= GPL-2.0-only ?). The kernel module containing this patch compiles and runs without problems on a RTL8029AS-based NE2000 clone card with kernel 5.7.0-rc6. Signed-off-by:
Armin Wolf <W_Armin@gmx.de> Signed-off-by:
-
Alexander Sverdlin authored
Ignore loopback-originatig packets soon enough and don't try to process L2 header where it doesn't exist. The very similar br_handle_frame() in bridge code performs exactly the same check. This is an example of such ICMPv6 packet: skb len=96 headroom=40 headlen=96 tailroom=56 mac=(40,0) net=(40,40) trans=80 shinfo(txflags=0 nr_frags=0 gso(size=0 type=0 segs=0)) csum(0xae2e9a2f ip_summed=1 complete_sw=0 valid=0 level=0) hash(0xc97ebd88 sw=1 l4=1) proto=0x86dd pkttype=5 iif=24 dev name=etha01.212 feat=0x0x0000000040005000 skb headroom: 00000000: 00 7c 86 52 84 88 ff ff 00 00 00 00 00 00 08 00 skb headroom: 00000010: 45 00 00 9e 5d 5c 40 00 40 11 33 33 00 00 00 01 skb headroom: 00000020: 02 40 43 80 00 00 86 dd skb linear: 00000000: 60 09 88 bd 00 38 3a ff fe 80 00 00 00 00 00 00 skb linear: 00000010: 00 40 43 ff fe 80 00 00 ff 02 00 00 00 00 00 00 skb linear: 00000020: 00 00 00 00 00 00 00 01 86 00 61 00 40 00 00 2d skb linear: 00000030: 00 00 00 00 00 00 00 00 03 04 40 e0 00 00 01 2c skb linear: 00000040: 00 00 00 78 00 00 00 00 fd 5f 42 68 23 87 a8 81 skb linear: 00000050: 00 00 00 00 00 00 00 00 01 01 02 40 43 80 00 00 skb tailroom: 00000000: ... skb tailroom: 00000010: ... skb tailroom: 00000020: ... skb tailroom: 00000030: ... Call Trace, how it happens exactly: ... macvlan_handle_frame+0x321/0x425 [macvlan] ? macvlan_forward_source+0x110/0x110 [macvlan] __netif_receive_skb_core+0x545/0xda0 ? enqueue_task_fair+0xe5/0x8e0 ? __netif_receive_skb_one_core+0x36/0x70 __netif_receive_skb_one_core+0x36/0x70 process_backlog+0x97/0x140 net_rx_action+0x1eb/0x350 ? __hrtimer_run_queues+0x136/0x2e0 __do_softirq+0xe3/0x383 do_softirq_own_stack+0x2a/0x40 </IRQ> do_softirq.part.4+0x4e/0x50 netif_rx_ni+0x60/0xd0 dev_loopback_xmit+0x83/0xf0 ip6_finish_output2+0x575/0x590 [ipv6] ? ip6_cork_release.isra.1+0x64/0x90 [ipv6] ? __ip6_make_skb+0x38d/0x680 [ipv6] ? ip6_output+0x6c/0x140 [ipv6] ip6_output+0x6c/0x140 [ipv6] ip6_send_skb+0x1e/0x60 [ipv6] rawv6_sendmsg+0xc4b/0xe10 [ipv6] ? proc_put_long+0xd0/0xd0 ? rw_copy_check_uvector+0x4e/0x110 ? sock_sendmsg+0x36/0x40 sock_sendmsg+0x36/0x40 ___sys_sendmsg+0x2b6/0x2d0 ? proc_dointvec+0x23/0x30 ? addrconf_sysctl_forward+0x8d/0x250 [ipv6] ? dev_forward_change+0x130/0x130 [ipv6] ? _raw_spin_unlock+0x12/0x30 ? proc_sys_call_handler.isra.14+0x9f/0x110 ? __call_rcu+0x213/0x510 ? get_max_files+0x10/0x10 ? trace_hardirqs_on+0x2c/0xe0 ? __sys_sendmsg+0x63/0xa0 __sys_sendmsg+0x63/0xa0 do_syscall_64+0x6c/0x1e0 entry_SYSCALL_64_after_hwframe+0x49/0xbe Signed-off-by:
Alexander Sverdlin <alexander.sverdlin@nokia.com> Signed-off-by:
-
David S. Miller authored
Ido Schimmel says: ==================== mlxsw: Various trap changes - part 2 This patch set contains another set of small changes in mlxsw trap configuration. It is the last set before exposing control traps (e.g., IGMP query, ARP request) via devlink-trap. Tested with existing devlink-trap selftests. Please see individual patches for a detailed changelog. ==================== Signed-off-by: -
Ido Schimmel authored
The device has a trap for IPv6 packets that need be routed and have a unicast link-local destination IP (i.e., fe80::/10). This allows mlxsw to ignore link-local routes, as the packets will be trapped to the CPU in any case. However, since link-local routes are not programmed, it is possible for routed packets to hit the default route which might also be programmed to trap packets. This means that packets with a link-local destination IP might be trapped for the wrong reason. To overcome this, allow programming link-local prefix routes (usually one fe80::/64 per-table), so that the packets will be forwarded until reaching the link-local trap. Signed-off-by:
Ido Schimmel <idosch@mellanox.com> Reviewed-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
-
Ido Schimmel authored
Bidirectional Forwarding Detection (BFD) provides "low-overhead, short-duration detection of failures in the path between adjacent forwarding engines" (RFC 5880). This is accomplished by exchanging BFD packets between the two forwarding engines. Up until now these packets were trapped via the general local delivery (i.e., IP2ME) trap which also traps a lot of other packets that are not as time-sensitive as BFD packets. Expose dedicated traps for BFD packets so that user space could configure a dedicated policer for them. Signed-off-by:
-
Ido Schimmel authored
IPv6 packets that need to be forwarded and have a link-local source IP are dropped by the kernel and an ICMPv6 "Destination unreachable" is sent to the sending host. As such, change the trap group of such packets so that they do not interfere with IPv6 management packets. In the future this trap will be exposed as an exception via devlink-trap. Signed-off-by:
-
Ido Schimmel authored
Routed IP packets with the Router Alert option need to be trapped to the CPU as they might need to be locally delivered to raw sockets with the IP_ROUTER_ALERT / IPV6_ROUTER_ALERT socket option. Move them to the same group with other packets that might need to be trapped following route lookup. Signed-off-by:
-
Ido Schimmel authored
After the previous patch the split is no longer necessary and all the trap groups can be moved under the same enum. Signed-off-by:
-
Ido Schimmel authored
As explained in commit e6125230 ("mlxsw: spectrum_trap: Introduce dummy group with thin policer"), the purpose of the "thin" policer is to pass as less packets as possible to the CPU. The identifier of this policer is currently set according to the maximum number of used trap groups, but this is fragile: On Spectrum-1 the maximum number of policers is less than the maximum number of trap groups, which might result in an invalid policer identifier in case the number of used trap groups grows beyond the policer limit. Solve this by dynamically allocating the policer identifier. Signed-off-by:
-
Ido Schimmel authored
The number of Spectrum trap groups is not infinite, but two identifiers are occupied by SwitchX-2 specific trap groups. Free these identifiers by moving them out of the main enum. Signed-off-by:
-
Ido Schimmel authored
To align with recent recommended values. Will be configurable by future patches. Signed-off-by:
-
Ido Schimmel authored
Packets with an IPv6 link-local destination (i.e., fe80::/10) should not be forwarded and are therefore trapped to the CPU for local delivery. Since these packets are trapped for the same logical reason as packets hitting local routes, associate both traps with the same group. Signed-off-by:
-
Ido Schimmel authored
When a packet enters the device it is classified to a filtering identifier (FID) based on the ingress port and VLAN. The FID miss trap is used to trap packets for which a FID could not be found. In mlxsw this trap should only be triggered when a port is enslaved to an OVS bridge and a matching ACL rule could not be found, so as to trigger learning. These packets are therefore completely unrelated to packets hitting local routes and should be in a different group. Move them. Signed-off-by:
-
Ido Schimmel authored
Group these various IPv6 packets (e.g., router solicitations, router advertisement) together and subject them to the same policer. Signed-off-by:
-
Ido Schimmel authored
The IPv6 Neighbour Discovery (ND) group will be used for various IPv6 packets, not all of which fall under the definition of ND, so rename it to "IPV6" which is more appropriate. Signed-off-by:
-
Ido Schimmel authored
Trap groups that use the same policer settings can share the same switch case. Signed-off-by:
-
Ido Schimmel authored
Packets that are trapped via tc's trap action are currently subject to the same policer as packets hitting local routes. The latter are critical to the correct functioning of the control plane, while the former are mainly used for traffic inspection. Split the ACL trap to a separate group with its own policer. Use a higher priority for these traps than for traps using mirror action (e.g., ARP, IGMP). Otherwise, packets matching both traps will not be forwarded in hardware (because of trap action) and also not forwarded in software because they will be marked with 'offload_fwd_mark'. Signed-off-by:
-
Florian Westphal authored
We can try to coalesce skbs we take from the subflows rx queue with the tail of the mptcp rx queue. If successful, the skb head can be discarded early. We can also free the skb extensions, we do not access them after this. Signed-off-by:
Florian Westphal <fw@strlen.de> Signed-off-by:
-
