- 22 Mar, 2020 18 commits
-
-
David S. Miller authored
Jian Yang says: ==================== selftests: expand txtimestamp with new features Current txtimestamp selftest issues requests with no delay, or fixed 50 usec delay. Nsec granularity is useful to measure fine-grained latency. A configurable delay is useful to simulate the case with cold cachelines. This patchset adds new flags and features to the txtimestamp selftest, including: - Printing in nsec (-N) - Polling interval (-b, -S) - Using epoll (-E, -e) - Printing statistics - Running individual tests in txtimestamp.sh ==================== Signed-off-by:David S. Miller <davem@davemloft.net>
-
Jian Yang authored
Statistics on timestamps is useful to quantify average and tail latency. Print timestamp statistics in count/avg/min/max format. Signed-off-by:
Jian Yang <jianyang@google.com> Acked-by:
Willem de Bruijn <willemb@google.com> Acked-by:
Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by:
-
Jian Yang authored
Add the following new flags: -e: use level-triggered epoll() instead of poll(). -E: use event-triggered epoll() instead of poll(). Signed-off-by:
-
Jian Yang authored
A longer sleep duration between sendmsg()s makes more cachelines to be evicted and results in higher latency. Making the duration configurable. Add the following new flags: -S: Configurable sleep duration. -b: Busy loop instead of poll(). Remove the following flag: -D: No delay between packets: subsumed by -S. Signed-off-by:
-
Jian Yang authored
Txtimestamp reports latencies in uses resolution, while nsec is needed in cases such as measuring latencies on localhost. Add the following new flag: -N: print timestamps and durations in nsec (instead of usec) Signed-off-by:
-
Jian Yang authored
The wrapper script txtimestamp.sh executes a pre-defined list of testcases sequentially without configuration options available. Add an option (-r/--run) to setup the test namespace and pass remaining arguments to txtimestamp binary. The script still runs all tests when no argument is passed. Signed-off-by:
-
David S. Miller authored
Jakub Sitnicki says: ==================== net/tls: Annotate lockless access to sk_prot We have recently noticed that there is a case of lockless read/write to sk->sk_prot [0]. sockmap code on psock tear-down writes to sk->sk_prot, while holding sk_callback_lock. Concurrently, tcp can access it. Usually to read out the sk_prot pointer and invoke one of the ops, sk->sk_prot->handler(). The lockless write (lockless in regard to concurrent reads) happens on the following paths: tcp_bpf_{recvmsg|sendmsg} / sock_map_unref sk_psock_put sk_psock_drop sk_psock_restore_proto WRITE_ONCE(sk->sk_prot, proto) To prevent load/store tearing [1], and to make tooling aware of intentional shared access [2], we need to annotate sites that access sk_prot with READ_ONCE/WRITE_ONCE. This series kicks off the effort to do it. Starting with net/tls. [0] https://lore.kernel.org/bpf/a6bf279e-a998-84ab-4371-cd6c1ccbca5d@gmail.com/ [1] https://lwn.net/Articles/793253/ [2] https://github.com/google/ktsan/wiki/READ_ONCE-and-WRITE_ONCE ==================== Signed-off-by: -
Jakub Sitnicki authored
sockmap performs lockless writes to sk->sk_prot on the following paths: tcp_bpf_{recvmsg|sendmsg} / sock_map_unref sk_psock_put sk_psock_drop sk_psock_restore_proto WRITE_ONCE(sk->sk_prot, proto) To prevent load/store tearing [1], and to make tooling aware of intentional shared access [2], we need to annotate other sites that access sk_prot with READ_ONCE/WRITE_ONCE macros. Change done with Coccinelle with following semantic patch: @@ expression E; identifier I; struct sock *sk; identifier sk_prot =~ "^sk_prot$"; @@ ( E = -sk->sk_prot +READ_ONCE(sk->sk_prot) | -sk->sk_prot = E +WRITE_ONCE(sk->sk_prot, E) | -sk->sk_prot +READ_ONCE(sk->sk_prot) ->I ) Signed-off-by:Jakub Sitnicki <jakub@cloudflare.com> Signed-off-by:
-
Jakub Sitnicki authored
Apart from being a "tremendous" win when it comes to generated machine code (see bloat-o-meter output for x86-64 below) this mainly prepares ground for annotating access to sk_prot with READ_ONCE, so that we don't pepper the code with access annotations and needlessly repeat loads. add/remove: 0/0 grow/shrink: 0/1 up/down: 0/-46 (-46) Function old new delta tls_init 851 805 -46 Total: Before=21063, After=21017, chg -0.22% Signed-off-by:
-
Jakub Sitnicki authored
The helper that builds kTLS proto ops doesn't need to and should not modify the base proto ops. Annotate the parameter as read-only. Signed-off-by:
-
David S. Miller authored
Shannon Nelson says: ==================== ionic error recovery fixes These are a few little patches to make error recovery a little more safe and successful. ==================== Signed-off-by: -
Shannon Nelson authored
Make sure the queue structs exist before trying to tear them down to make for safer error recovery. Fixes: 0f3154e6 ("ionic: Add Tx and Rx handling") Signed-off-by:
Shannon Nelson <snelson@pensando.io> Signed-off-by:
-
Shannon Nelson authored
Add a little more cleanup when tearing down the queues. Fixes: 1d062b7b ("ionic: Add basic adminq support") Signed-off-by:
-
Shannon Nelson authored
Don't worry if the rx filter add firmware request fails on EEXIST, at least we know the filter is there. Same for the delete request, at least we know it isn't there. Fixes: 2a654540 ("ionic: Add Rx filter and rx_mode ndo support") Signed-off-by:
-
Shannon Nelson authored
Don't save the lif->dentry until we know we have a good value. Fixes: 1a58e196 ("ionic: Add basic lif support") Signed-off-by:
-
Shannon Nelson authored
It is possible (but unlikely) that FW was busy and missed a heartbeat check but is still alive and will process the pending request, so don't clean the dev_cmd in this case. This occasionally occurs when working with a card that is supporting many devices and is trying to shut them all down at once, but still wants to see that last LIF disable request. Fixes: 97ca4865 ("ionic: add heartbeat check") Signed-off-by:
-
Shannon Nelson authored
Short circuit the cleanup if we get a timeout error from ionic_qcq_disable() so as to not have to wait too long on shutdown when we already know the FW is not responding. Fixes: 0f3154e6 ("ionic: Add Tx and Rx handling") Signed-off-by:
-
Alex Elder authored
Don't assume the receive buffer size is a power-of-2 number of pages. Instead, define the receive buffer size independently, and then compute the page order from that size when needed. This fixes a build problem that arises when the ARM64_PAGE_SHIFT config option is set to have a page size greater than 4KB. The problem was identified by Linux Kernel Functional Testing. The IPA code basically assumed the page size to be 4KB. A larger page size caused the receive buffer size to become correspondingly larger (32KB or 128KB for ARM64_16K_PAGES and ARM64_64K_PAGES, respectively). The receive buffer size is used to compute an "aggregation byte limit" value that gets programmed into the hardware, and the large page sizes caused that limit value to be too big to fit in a 5 bit field. This triggered a BUILD_BUG_ON() call in ipa_endpoint_validate_build(). This fix causes a lot of receive buffer memory to be wasted if system is configured for page size greater than 4KB. But such a misguided configuration will now build successfully. Reported-by:
Naresh Kamboju <naresh.kamboju@linaro.org> Signed-off-by:
Alex Elder <elder@linaro.org> Reviewed-by:
Bjorn Andersson <bjorn.andersson@linaro.org> Signed-off-by:
-
- 20 Mar, 2020 22 commits
-
-
David S. Miller authored
Merge tag 'mac80211-next-for-net-next-2020-03-20' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211-next Johannes Berg says: ==================== Another set of changes: * HE ranging (fine timing measurement) API support * hwsim gets virtio support, for use with wmediumd, to be able to simulate with multiple machines * eapol-over-nl80211 improvements to exclude preauth * IBSS reset support, to recover connections from userspace * and various others. ==================== Signed-off-by:
-
Vladimir Oltean authored
SJA1105 switches R and S have one SerDes port with an 802.3z quasi-compatible PCS, hardwired on port 4. The other ports are still MII/RMII/RGMII. The PCS performs rate adaptation to lower link speeds; the MAC on this port is hardwired at gigabit. Only full duplex is supported. The SGMII port can be configured as part of the static config tables, as well as through a dedicated SPI address region for its pseudo-clause-22 registers. However it looks like the static configuration is not able to change some out-of-reset values (like the value of MII_BMCR), so at the end of the day, having code for it is utterly pointless. We are just going to use the pseudo-C22 interface. Because the PCS gets reset when the switch resets, we have to add even more restoration logic to sja1105_static_config_reload, otherwise the SGMII port breaks after operations such as enabling PTP timestamping which require a switch reset. >From PHYLINK perspective, the switch supports *only* SGMII (it doesn't support 1000Base-X). It also doesn't expose access to the raw config word for in-band AN in registers MII_ADV/MII_LPA. It is able to work in the following modes: - Forced speed - SGMII in-band AN slave (speed received from PHY) - SGMII in-band AN master (acting as a PHY) The latter mode is not supported by this patch. It is even unclear to me how that would be described. There is some code for it left in the patch, but 'an_master' is always passed as false. Signed-off-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Reviewed-by:
Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by:
-
David S. Miller authored
Nikolay Aleksandrov says: ==================== net: bridge: vlan options: nest the tunnel options After a discussion with Roopa about the new tunnel vlan option, she suggested that we'll be adding more tunnel options and attributes, so it'd be better to have them all grouped together under one main vlan entry tunnel attribute instead of making them all main attributes. Since the tunnel code was added in this net-next cycle and still hasn't been released we can easily nest the BRIDGE_VLANDB_ENTRY_TUNNEL_ID attribute in BRIDGE_VLANDB_ENTRY_TUNNEL_INFO and allow for any new tunnel attributes to be added there. In addition one positive side-effect is that we can remove the outside vlan info flag which controlled the operation (setlink/dellink) and move it under a new nested attribute so user-space can specify it explicitly. Thus the vlan tunnel format becomes: [BRIDGE_VLANDB_ENTRY] [BRIDGE_VLANDB_ENTRY_TUNNEL_INFO] [BRIDGE_VLANDB_TINFO_ID] [BRIDGE_VLANDB_TINFO_CMD] ... ==================== Signed-off-by: -
Nikolay Aleksandrov authored
Now that we have a nested tunnel info attribute we can add a separate one for the tunnel command and require it explicitly from user-space. It must be one of RTM_SETLINK/DELLINK. Only RTM_SETLINK requires a valid tunnel id, DELLINK just removes it if it was set before. This allows us to have all tunnel attributes and control in one place, thus removing the need for an outside vlan info flag. Signed-off-by:
Nikolay Aleksandrov <nikolay@cumulusnetworks.com> Signed-off-by:
-
Nikolay Aleksandrov authored
While discussing the new API, Roopa mentioned that we'll be adding more tunnel attributes and options in the future, so it's better to make it a nested attribute, since this is still in net-next we can easily change it and nest the tunnel id attribute under BRIDGE_VLANDB_ENTRY_TUNNEL_INFO. The new format is: [BRIDGE_VLANDB_ENTRY] [BRIDGE_VLANDB_ENTRY_TUNNEL_INFO] [BRIDGE_VLANDB_TINFO_ID] Any new tunnel attributes can be nested under BRIDGE_VLANDB_ENTRY_TUNNEL_INFO. Suggested-by:Roopa Prabhu <roopa@cumulusnetworks.com> Signed-off-by:
-
Yan-Hsuan Chuang authored
Some of the drivers are not using channel context, but let the stack to control/switch channels instead. For such cases, driver can still remain on channel because the mac80211 stack actually supports it. The stack will check if the driver is using chan_ctx and has ops->remain_on_channel been hooked. Otherwise it will start its ROC work to remain on channel. So, even if the driver is not using chan_ctx, the driver is still capable of doing remain on channel. Signed-off-by:
Yan-Hsuan Chuang <yhchuang@realtek.com> Link: https://lore.kernel.org/r/20200312074337.16198-1-yhchuang@realtek.comSigned-off-by:
Johannes Berg <johannes.berg@intel.com>
-
Johannes Berg authored
The long if chain of interface types is hard to read, especially now with the additional condition after it. Use a switch statement to clarify this code. Signed-off-by:
-
Veerendranath Jakkam authored
Drivers that trigger roaming need to know the lifetime of the configured PMKSA for deciding whether to trigger the full or PMKSA cache based authentication. The configured PMKSA is invalid after the PMK lifetime has expired and must not be used after that and the STA needs to disassociate if the PMK expires. Hence the STA is expected to refresh the PMK with a full authentication before this happens (e.g., when reassociating to a new BSS the next time or by performing EAPOL reauthentication depending on the AKM) to avoid unnecessary disconnection. The PMK reauthentication threshold is the percentage of the PMK lifetime value and indicates to the driver to trigger a full authentication roam (without PMKSA caching) after the reauthentication threshold time, but before the PMK timer has expired. Authentication methods like SAE need to be able to generate a new PMKSA entry without having to force a disconnection after this threshold timeout. If no roaming occurs between the reauthentication threshold time and PMK lifetime expiration, disassociation is still forced. The new attributes for providing these values correspond to the dot11 MIB variables dot11RSNAConfigPMKLifetime and dot11RSNAConfigPMKReauthThreshold. This type of functionality is already available in cases where user space component is in control of roaming. This commit extends that same capability into cases where parts or all of this functionality is offloaded to the driver. Signed-off-by:
Veerendranath Jakkam <vjakkam@codeaurora.org> Signed-off-by:
Jouni Malinen <jouni@codeaurora.org> Link: https://lore.kernel.org/r/20200312235903.18462-1-jouni@codeaurora.orgSigned-off-by:
-
Seevalamuthu Mariappan authored
Use perCPU pointers to get rx_stats in sta_set_sinfo when RSS is enabled Signed-off-by:
Seevalamuthu Mariappan <seevalam@codeaurora.org> Link: https://lore.kernel.org/r/1584526555-25960-1-git-send-email-seevalam@codeaurora.orgSigned-off-by:
-
Taehee Yoo authored
->ndo_get_iflink() is useful for finding lower interface. Test commands: ip link add dummy0 type dummy ip link add vw1 link dummy0 type virt_wifi ip link show vw1 Before: 9: vw1: <BROADCAST,MULTICAST> ... After: 9: vw1@dummy0: <BROADCAST,MULTICAST> ... Signed-off-by:Taehee Yoo <ap420073@gmail.com> Link: https://lore.kernel.org/r/20200305090636.28221-1-ap420073@gmail.comSigned-off-by:
-
Nicolas Cavallari authored
Set the NL80211_EXT_FEATURE_DEL_IBSS_STA if the interface support IBSS mode, so that stations can be reset from user space. mac80211 already deletes stations by itself, so mac80211 drivers must already support this. This has been successfully tested with ath9k. Signed-off-by:
Nicolas Cavallari <nicolas.cavallari@green-communications.fr> Link: https://lore.kernel.org/r/20200305135754.12094-2-cavallar@lri.frSigned-off-by:
-
Nicolas Cavallari authored
Sometimes, userspace is able to detect that a peer silently lost its state (like, if the peer reboots). wpa_supplicant does this for IBSS-RSN by registering for auth/deauth frames, but when it detects this, it is only able to remove the encryption keys of the peer and close its port. However, the kernel also hold other state about the station, such as BA sessions, probe response parameters and the like. They also need to be resetted correctly. This patch adds the NL80211_EXT_FEATURE_DEL_IBSS_STA feature flag indicating the driver accepts deleting stations in IBSS mode, which should send a deauth and reset the state of the station, just like in mesh point mode. Signed-off-by:
-
Johannes Berg authored
We use the parsing CRC for checking if the beacon changed, and if the WLAN_EID_EXT_HE_OPERATION extended element changes we need to track it so we can react to that. Include it in the CRC calculation. Link: https://lore.kernel.org/r/20200131111300.891737-22-luca@coelho.fiSigned-off-by:
-
Shaul Triebitz authored
Add API for telling whether the driver supports protected TWT. The protected_twt capability in the RSNXE will be based on this. Signed-off-by:
Shaul Triebitz <shaul.triebitz@intel.com> Signed-off-by:
Luca Coelho <luciano.coelho@intel.com> Link: https://lore.kernel.org/r/20200131111300.891737-23-luca@coelho.fiSigned-off-by:
-
Shaul Triebitz authored
In AP mode, set htc_trig_based_pkt_ext and frame_time_rts_th for driver use. Signed-off-by:
-
Shaul Triebitz authored
Pass the AP's HE operation element to the driver. Signed-off-by:
-
Avraham Stern authored
Add support for requesting that the ranging measurement will use the trigger-based / non trigger-based flow instead of the EDCA based flow. Signed-off-by:
Avraham Stern <avraham.stern@intel.com> Signed-off-by:
-
Johannes Berg authored
In beacon protection, don't leave skb->next/prev pointing to the on-stack list, even if that's actually harmless since we don't use them again afterwards. While at it, check that the SKB on the list is still the same, as that's required here. If not, the encryption (protection) code is buggy. Fixes: 0a3a8436 ("mac80211: Beacon protection using the new BIGTK (AP)") Signed-off-by:
-
Qiujun Huang authored
The structure member added at some point, but the kernel-doc was not updated. Signed-off-by:
Qiujun Huang <hqjagain@gmail.com> Link: https://lore.kernel.org/r/20200312144424.3023-1-hqjagain@gmail.comSigned-off-by:
-
Markus Theil authored
This patch adds support for disabling pre-auth rx over the nl80211 control port for mac80211. Signed-off-by:
Markus Theil <markus.theil@tu-ilmenau.de> Link: https://lore.kernel.org/r/20200312091055.54257-3-markus.theil@tu-ilmenau.de [fix indentation slightly, squash feature enablement] Signed-off-by:
-
Markus Theil authored
If the nl80211 control port is used before this patch, pre-auth frames (0x88c7) are send to userspace uncoditionally. While this enables userspace to only use nl80211 on the station side, it is not always useful for APs. Furthermore, pre-auth frames are ordinary data frames and not related to the control port. Therefore it should for example be possible for pre-auth frames to be bridged onto a wired network on AP side without touching userspace. For backwards compatibility to code already using pre-auth over nl80211, this patch adds a feature flag to disable this behavior, while it remains enabled by default. An additional ext. feature flag is added to detect this from userspace. Thanks to Jouni for pointing out, that pre-auth frames should be handled as ordinary data frames. Signed-off-by:
-
Erel Geron authored
This allows communication with external entities. It also required fixing up the netlink policy, since NLA_UNSPEC attributes are no longer accepted. Signed-off-by:
Erel Geron <erelx.geron@intel.com> [port to backports, inline the ID, use 29 as the ID as requested, drop != NULL checks, reduce ifdefs] Link: https://lore.kernel.org/r/20200305143212.c6e4c87d225b.I7ce60bf143e863dcdf0fb8040aab7168ba549b99@changeidSigned-off-by:
-
