- 17 Sep, 2019 1 commit
-
-
David S. Miller authored
Pull in bug fixes from 'net' tree for the merge window. Signed-off-by:David S. Miller <davem@davemloft.net>
-
- 16 Sep, 2019 39 commits
-
-
David S. Miller authored
Ido Schimmel says: ==================== mlxsw: spectrum_buffers: Add the ability to query the CPU port's shared buffer Shalom says: While debugging packet loss towards the CPU, it is useful to be able to query the CPU port's shared buffer quotas and occupancy. Patch #1 prevents changing the CPU port's threshold and binding. Patch #2 registers the CPU port with devlink. Patch #3 adds the ability to query the CPU port's shared buffer quotas and occupancy. v3: Patch #2: * Remove unnecessary wrapping v2: Patch #1: * s/0/MLXSW_PORT_CPU_PORT/ * Assign "mlxsw_sp->ports[MLXSW_PORT_CPU_PORT]" at the end of mlxsw_sp_cpu_port_create() to avoid NULL assignment on error path * Add common functions for mlxsw_core_port_init/fini() Patch #2: * Move "changing CPU port's threshold and binding" check to a separate patch ==================== Signed-off-by: -
Shalom Toledo authored
While debugging packet loss towards the CPU, it is useful to be able to query the CPU port's shared buffer quotas and occupancy. Since the CPU port has no ingress buffers, all the shared buffers ingress information will be cleared. Signed-off-by:
Shalom Toledo <shalomt@mellanox.com> Acked-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
Ido Schimmel <idosch@mellanox.com> Signed-off-by:
-
Shalom Toledo authored
Register CPU port with devlink. Signed-off-by:
-
Shalom Toledo authored
Next patch is going to register the CPU port with devlink, but only so that the CPU port's shared buffer configuration and occupancy could be queried. Prevent changing CPU port's shared buffer threshold and binding configuration. Signed-off-by:
-
David S. Miller authored
Arthur Kiyanovski says: ==================== net: ena: implement adaptive interrupt moderation using dim In this patchset we replace our adaptive interrupt moderation implementation with the dim library implementation. The dim library showed great improvement in throughput, latency and CPU usage in different scenarios on ARM CPUs. This patchset also includes a few bug fixes to the parts of the old implementation of adaptive interrupt moderation that were left. Changes from V1 patchset: Removed stray empty lines from patches 01/11, 09/11. ==================== Signed-off-by: -
Arthur Kiyanovski authored
ena_dev->intr_moder_rx/tx_interval save the intervals received from the user after dividing them by ena_dev->intr_delay_resolution. Therefore when intr_delay_resolution changes, the code needs to first mutiply intr_moder_rx/tx_interval by the previous intr_delay_resolution to get the value originally given by the user, and only then divide it by the new intr_delay_resolution. Current code does not first multiply intr_moder_rx/tx_interval by the old intr_delay_resolution. This commit fixes it. Also initialize ena_dev->intr_delay_resolution to be 1. Signed-off-by:
Arthur Kiyanovski <akiyano@amazon.com> Signed-off-by:
-
Arthur Kiyanovski authored
Nonadaptive interrupt moderation intervals are assigned the value set by the user in ethtool -C divided by ena_dev->intr_delay_resolution. Therefore when the user tries to get the nonadaptive interrupt moderation intervals with ethtool -c the code needs to multiply the saved value by ena_dev->intr_delay_resolution. The current code erroneously divides instead of multiplying in ethtool -c. This patch fixes this. Signed-off-by:
-
Arthur Kiyanovski authored
Current implementation always updates the interrupt register with the smoothed_interval of the rx_ring. However this should be done only in case of adaptive interrupt moderation. If non-adaptive interrupt moderation is used, the non-adaptive interrupt moderation interval should be used. This commit fixes that. Signed-off-by:
-
Arthur Kiyanovski authored
Remove previous implementation of adaptive rx interrupt moderation from ena_com files. Signed-off-by:
-
Arthur Kiyanovski authored
Deleted unused 4 fields from struct ena_adapter and their only user ena_restore_ethtool_params(). Signed-off-by:
-
Arthur Kiyanovski authored
1. Out of the fields {per_napi_bytes, per_napi_packets} in struct ena_ring, only rx_ring->per_napi_packets are used to determine if napi did work for dim. This commit removes all other uses of these fields. 2. Remove ena_ring->moder_tbl_idx, which is not used by dim. 3. Remove all calls to ena_com_destroy_interrupt_moderation(), since all it did was to destroy the interrupt moderation table, which is removed as part of removing old interrupt moderation code. Signed-off-by: -
Arthur Kiyanovski authored
Remove code duplication in: ena_com_update_nonadaptive_moderation_interval_tx() ena_com_update_nonadaptive_moderation_interval_rx() functions. Signed-off-by:
-
Arthur Kiyanovski authored
Add driver_supported_features to host_host info which is a new API used to communicate to the device which features are supported by the driver. Add the interrupt_moderation bit to host_info->driver_supported_features and enable it to signal the device that this driver supports interrupt moderation properly. Reserved bits are for features implemented in the future Signed-off-by:
-
Arthur Kiyanovski authored
1. Remove old adaptive interrupt moderation code from set/get_coalesce() 2. Add ena_update_rx_rings_intr_moderation() function for updating nonadaptive interrupt moderation intervals similarly to ena_update_tx_rings_intr_moderation(). 3. Remove checks of multiple unsupported received interrupt coalescing parameters. This makes code cleaner and cancels the need to update it every time a new coalescing parameter is invented. Signed-off-by:
-
Arthur Kiyanovski authored
Use the dim library for the rx adaptive interrupt moderation implementation Signed-off-by:
-
Arthur Kiyanovski authored
Add intr_moder_rx_interval to struct ena_com_dev and use it as the location where the interrupt moderation rx interval is saved, instead of the interrupt moderation table. This is done as a first step before removing the old interrupt moderation code. Signed-off-by:
-
David S. Miller authored
Alexandru Ardelean says: ==================== ethtool: implement Energy Detect Powerdown support via phy-tunable This changeset proposes a new control for PHY tunable to control Energy Detect Power Down. The `phy_tunable_id` has been named `ETHTOOL_PHY_EDPD` since it looks like this feature is common across other PHYs (like EEE), and defining `ETHTOOL_PHY_ENERGY_DETECT_POWER_DOWN` seems too long. The way EDPD works, is that the RX block is put to a lower power mode, except for link-pulse detection circuits. The TX block is also put to low power mode, but the PHY wakes-up periodically to send link pulses, to avoid lock-ups in case the other side is also in EDPD mode. Currently, there are 2 PHY drivers that look like they could use this new PHY tunable feature: the `adin` && `micrel` PHYs. This series updates only the `adin` PHY driver to support this new feature, as this chip has been tested. A change for `micrel` can be proposed after a discussion of the PHY-tunable API is resolved. ==================== Signed-off-by: -
Alexandru Ardelean authored
This driver becomes the first user of the kernel's `ETHTOOL_PHY_EDPD` phy-tunable feature. EDPD is also enabled by default on PHY config_init, but can be disabled via the phy-tunable control. When enabling EDPD, it's also a good idea (for the ADIN PHYs) to enable TX periodic pulses, so that in case the other PHY is also on EDPD mode, there is no lock-up situation where both sides are waiting for the other to transmit. Via the phy-tunable control, TX pulses can be disabled if specifying 0 `tx-interval` via ethtool. The ADIN PHY supports only fixed 1 second intervals; they cannot be configured. That is why the acceptable values are 1, ETHTOOL_PHY_EDPD_DFLT_TX_MSECS and ETHTOOL_PHY_EDPD_NO_TX (which disables TX pulses). Reviewed-by:
Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
Alexandru Ardelean <alexandru.ardelean@analog.com> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Signed-off-by:
-
Alexandru Ardelean authored
The `phy_tunable_id` has been named `ETHTOOL_PHY_EDPD` since it looks like this feature is common across other PHYs (like EEE), and defining `ETHTOOL_PHY_ENERGY_DETECT_POWER_DOWN` seems too long. The way EDPD works, is that the RX block is put to a lower power mode, except for link-pulse detection circuits. The TX block is also put to low power mode, but the PHY wakes-up periodically to send link pulses, to avoid lock-ups in case the other side is also in EDPD mode. Currently, there are 2 PHY drivers that look like they could use this new PHY tunable feature: the `adin` && `micrel` PHYs. The ADIN's datasheet mentions that TX pulses are at intervals of 1 second default each, and they can be disabled. For the Micrel KSZ9031 PHY, the datasheet does not mention whether they can be disabled, but mentions that they can modified. The way this change is structured, is similar to the PHY tunable downshift control: * a `ETHTOOL_PHY_EDPD_DFLT_TX_MSECS` value is exposed to cover a default TX interval; some PHYs could specify a certain value that makes sense * `ETHTOOL_PHY_EDPD_NO_TX` would disable TX when EDPD is enabled * `ETHTOOL_PHY_EDPD_DISABLE` will disable EDPD As noted by the `ETHTOOL_PHY_EDPD_DFLT_TX_MSECS` the interval unit is 1 millisecond, which should cover a reasonable range of intervals: - from 1 millisecond, which does not sound like much of a power-saver - to ~65 seconds which is quite a lot to wait for a link to come up when plugging a cable Reviewed-by:
-
Dongli Zhang authored
When skb_shinfo(skb) is not able to cache extra fragment (that is, skb_shinfo(skb)->nr_frags >= MAX_SKB_FRAGS), xennet_fill_frags() assumes the sk_buff_head list is already empty. As a result, cons is increased only by 1 and returns to error handling path in xennet_poll(). However, if the sk_buff_head list is not empty, queue->rx.rsp_cons may be set incorrectly. That is, queue->rx.rsp_cons would point to the rx ring buffer entries whose queue->rx_skbs[i] and queue->grant_rx_ref[i] are already cleared to NULL. This leads to NULL pointer access in the next iteration to process rx ring buffer entries. Below is how xennet_poll() does error handling. All remaining entries in tmpq are accounted to queue->rx.rsp_cons without assuming how many outstanding skbs are remained in the list. 985 static int xennet_poll(struct napi_struct *napi, int budget) ... ... 1032 if (unlikely(xennet_set_skb_gso(skb, gso))) { 1033 __skb_queue_head(&tmpq, skb); 1034 queue->rx.rsp_cons += skb_queue_len(&tmpq); 1035 goto err; 1036 } It is better to always have the error handling in the same way. Fixes: ad4f15dc ("xen/netfront: don't bug in case of too many frags") Signed-off-by:Dongli Zhang <dongli.zhang@oracle.com> Signed-off-by:
-
Markus Elfring authored
The dev_kfree_skb() function performs also input parameter validation. Thus the test around the shown calls is not needed. This issue was detected by using the Coccinelle software. Signed-off-by:
Markus Elfring <elfring@users.sourceforge.net> Signed-off-by:
Julian Wiedmann <jwi@linux.ibm.com> Signed-off-by:
-
Sameeh Jubran authored
There is a race condition that can occur when calling ena_down(). The ena_clean_tx_irq() - which is a part of the napi handler - function might wake up the tx queue when the queue is supposed to be down (during recovery or changing the size of the queues for example) This causes the ena_start_xmit() function to trigger and possibly try to access the destroyed queues. The race is illustrated below: Flow A: Flow B(napi handler) ena_down() netif_carrier_off() netif_tx_disable() ena_clean_tx_irq() netif_tx_wake_queue() ena_napi_disable_all() ena_destroy_all_io_queues() After these flows the tx queue is active and ena_start_xmit() accesses the destroyed queue which leads to a kernel panic. fixes: 1738cd3e (net: ena: Add a driver for Amazon Elastic Network Adapters (ENA)) Signed-off-by:Sameeh Jubran <sameehj@amazon.com> Signed-off-by:
-
David S. Miller authored
Ido Schimmel says: ==================== drop_monitor: Better sanitize notified packets When working in 'packet' mode, drop monitor generates a notification with a potentially truncated payload of the dropped packet. The payload is copied from the MAC header, but I forgot to check that the MAC header was set, so do it now. Patch #1 sets the offsets to the various protocol layers in netdevsim, so that it will continue to work after the MAC header check is added to drop monitor in patch #2. ==================== Signed-off-by: -
Ido Schimmel authored
When working in 'packet' mode, drop monitor generates a notification with a potentially truncated payload of the dropped packet. The payload is copied from the MAC header, but I forgot to check that the MAC header was set, so do it now. Fixes: ca30707d ("drop_monitor: Add packet alert mode") Fixes: 5e58109b ("drop_monitor: Add support for packet alert mode for hardware drops") Acked-by:
-
Ido Schimmel authored
The driver periodically generates "trapped" UDP packets that it then passes on to devlink. Set the offsets to the various protocol layers. This is a prerequisite to the next patch, where drop monitor is taught to check that the offset to the MAC header was set. Acked-by:
-
David S. Miller authored
Vladimir Oltean says: ==================== tc-taprio offload for SJA1105 DSA This is the third attempt to submit the tc-taprio offload model for inclusion in the networking tree. The sja1105 switch driver will provide the first implementation of the offload. Only the bare minimum is added: - The offload model and a DSA pass-through - The hardware implementation - The interaction with the netdev queues in the tagger code - Documentation What has been removed from previous attempts is support for PTP-as-clocksource in sja1105, as well as configuring the traffic class for management traffic. These will be added as soon as the offload model is settled. ==================== Signed-off-by: -
Vladimir Oltean authored
While not an exhaustive usage tutorial, this describes the details needed to build more complex scenarios. Signed-off-by:
Vladimir Oltean <olteanv@gmail.com> Signed-off-by:
-
Vladimir Oltean authored
This qdisc offload is the closest thing to what the SJA1105 supports in hardware for time-based egress shaping. The switch core really is built around SAE AS6802/TTEthernet (a TTTech standard) but can be made to operate similarly to IEEE 802.1Qbv with some constraints: - The gate control list is a global list for all ports. There are 8 execution threads that iterate through this global list in parallel. I don't know why 8, there are only 4 front-panel ports. - Care must be taken by the user to make sure that two execution threads never get to execute a GCL entry simultaneously. I created a O(n^4) checker for this hardware limitation, prior to accepting a taprio offload configuration as valid. - The spec says that if a GCL entry's interval is shorter than the frame length, you shouldn't send it (and end up in head-of-line blocking). Well, this switch does anyway. - The switch has no concept of ADMIN and OPER configurations. Because it's so simple, the TAS settings are loaded through the static config tables interface, so there isn't even place for any discussion about 'graceful switchover between ADMIN and OPER'. You just reset the switch and upload a new OPER config. - The switch accepts multiple time sources for the gate events. Right now I am using the standalone clock source as opposed to PTP. So the base time parameter doesn't really do much. Support for the PTP clock source will be added in a future series. Signed-off-by:
-
Vladimir Oltean authored
This is a preparation patch for the tc-taprio offload (and potentially for other future offloads such as tc-mqprio). Instead of looking directly at skb->priority during xmit, let's get the netdev queue and the queue-to-traffic-class mapping, and put the resulting traffic class into the dsa_8021q PCP field. The switch is configured with a 1-to-1 PCP-to-ingress-queue-to-egress-queue mapping (see vlan_pmap in sja1105_main.c), so the effect is that we can inject into a front-panel's egress traffic class through VLAN tagging from Linux, completely transparently. Unfortunately the switch doesn't look at the VLAN PCP in the case of management traffic to/from the CPU (link-local frames at 01-80-C2-xx-xx-xx or 01-1B-19-xx-xx-xx) so we can't alter the transmission queue of this type of traffic on a frame-by-frame basis. It is only selected through the "hostprio" setting which ATM is harcoded in the driver to 7. Signed-off-by:
-
Vladimir Oltean authored
In order to support tc-taprio offload, the TTEthernet egress scheduling core registers must be made visible through the static interface. Signed-off-by:
-
Vladimir Oltean authored
DSA currently handles shared block filters (for the classifier-action qdisc) in the core due to what I believe are simply pragmatic reasons - hiding the complexity from drivers and offerring a simple API for port mirroring. Extend the dsa_slave_setup_tc function by passing all other qdisc offloads to the driver layer, where the driver may choose what it implements and how. DSA is simply a pass-through in this case. Signed-off-by:
Kurt Kanzenbach <kurt@linutronix.de> Reviewed-by:
Ilias Apalodimas <ilias.apalodimas@linaro.org> Signed-off-by:
-
Vinicius Costa Gomes authored
This allows taprio to offload the schedule enforcement to capable network cards, resulting in more precise windows and less CPU usage. The gate mask acts on traffic classes (groups of queues of same priority), as specified in IEEE 802.1Q-2018, and following the existing taprio and mqprio semantics. It is up to the driver to perform conversion between tc and individual netdev queues if for some reason it needs to make that distinction. Full offload is requested from the network interface by specifying "flags 2" in the tc qdisc creation command, which in turn corresponds to the TCA_TAPRIO_ATTR_FLAG_FULL_OFFLOAD bit. The important detail here is the clockid which is implicitly /dev/ptpN for full offload, and hence not configurable. A reference counting API is added to support the use case where Ethernet drivers need to keep the taprio offload structure locally (i.e. they are a multi-port switch driver, and configuring a port depends on the settings of other ports as well). The refcount_t variable is kept in a private structure (__tc_taprio_qopt_offload) and not exposed to drivers. In the future, the private structure might also be expanded with a backpointer to taprio_sched *q, to implement the notification system described in the patch (of when admin became oper, or an error occurred, etc, so the offload can be monitored with 'tc qdisc show'). Signed-off-by:
Vinicius Costa Gomes <vinicius.gomes@intel.com> Signed-off-by:
Voon Weifeng <weifeng.voon@intel.com> Signed-off-by:
-
Russell King authored
Update the phylink documentation to make it clear that phylink is designed to be used on the MAC facing side of the link, rather than between a SFP and PHY. Signed-off-by:
Russell King <rmk+kernel@armlinux.org.uk> Reviewed-by:
-
David S. Miller authored
Michael Chan says: ==================== bnxt_en: error recovery follow-up patches. A follow-up patchset for the recently added health and error recovery feature. The first fix is to prevent .ndo_set_rx_mode() from proceeding when reset is in progress. The 2nd fix is for the firmware coredump command. The 3rd and 4th patches update the error recovery process slightly to add a state that polls and waits for the firmware to be down. ==================== Signed-off-by: -
Vasundhara Volam authored
This new state is required when firmware indicates that the error recovery process requires polling for firmware state to be completely down before initiating reset. For example, firmware may take some time to collect the crash dump before it is down and ready to be reset. Signed-off-by:
Vasundhara Volam <vasundhara-v.volam@broadcom.com> Signed-off-by:
Michael Chan <michael.chan@broadcom.com> Signed-off-by:
-
Michael Chan authored
Some error recovery updates to the spec., among other minor changes. Signed-off-by:
-
Vasundhara Volam authored
Firmware coredump messages take much longer than standard messages, so increase the timeout accordingly. Fixes: 6c5657d0 ("bnxt_en: Add support for ethtool get dump.") Signed-off-by:
-
Michael Chan authored
Check the BNXT_STATE_OPEN flag instead of netif_running() in bnxt_set_rx_mode(). If the driver is going through any reset, such as firmware reset or even TX timeout, it may not be ready to set the RX mode and may crash. The new rx mode settings will be picked up when the device is opened again later. Fixes: 230d1f0d ("bnxt_en: Handle firmware reset.") Signed-off-by:
-
Thomas Higdon authored
Neal Cardwell mentioned that snd_wnd would be useful for diagnosing TCP performance problems -- > (1) Usually when we're diagnosing TCP performance problems, we do so > from the sender, since the sender makes most of the > performance-critical decisions (cwnd, pacing, TSO size, TSQ, etc). > From the sender-side the thing that would be most useful is to see > tp->snd_wnd, the receive window that the receiver has advertised to > the sender. This serves the purpose of adding an additional __u32 to avoid the would-be hole caused by the addition of the tcpi_rcvi_ooopack field. Signed-off-by:
Thomas Higdon <tph@fb.com> Acked-by:
Yuchung Cheng <ycheng@google.com> Acked-by:
Neal Cardwell <ncardwell@google.com> Acked-by:
Soheil Hassas Yeganeh <soheil@google.com> Signed-off-by:
-
