- 22 5月, 2020 1 次提交
-
-
由 Tony Nguyen 提交于
Create a boost TCAM entry for each tunnel port in order to get a tunnel PTYPE. Update netdev feature flags and implement the appropriate logic to get and set values for hardware offloads. Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: NHenry Tieman <henry.w.tieman@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 27 3月, 2020 1 次提交
-
-
由 Jacob Keller 提交于
Add a devlink region for exposing the device's Non Volatime Memory flash contents. Support the recently added .snapshot operation, enabling userspace to request a snapshot of the NVM contents via DEVLINK_CMD_REGION_NEW. Signed-off-by: NJacob Keller <jacob.e.keller@intel.com> Reviewed-by: NJiri Pirko <jiri@mellanox.com> Signed-off-by: NDavid S. Miller <davem@davemloft.net>
-
- 21 3月, 2020 1 次提交
-
-
由 Jacob Keller 提交于
Begin implementing support for the devlink interface with the ice driver. The pf structure is currently memory managed through devres, via a devm_alloc. To mimic this behavior, after allocating the devlink pointer, use devm_add_action to add a teardown action for releasing the devlink memory on exit. The ice hardware is a multi-function PCIe device. Thus, each physical function will get its own devlink instance. This means that each function will be treated independently, with its own parameters and configuration. This is done because the ice driver loads a separate instance for each function. Due to this, the implementation does not enable devlink to manage device-wide resources or configuration, as each physical function will be treated independently. This is done for simplicity, as managing a devlink instance across multiple driver instances would significantly increase the complexity for minimal gain. Signed-off-by: NJacob Keller <jacob.e.keller@intel.com> Reviewed-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 11 3月, 2020 4 次提交
-
-
由 Lukasz Czapnik 提交于
Currently the PF's mailbox receive queue is only 512 entries. This fine, but considering that all VF's mailbox send queues funnel into the PF's single mailbox receive queue, let's increase it to the maximum size. This will help prevent any possible bottleneck/slowdown occurring from the PF's mailbox receive queue being full. Signed-off-by: NLukasz Czapnik <lukasz.czapnik@intel.com> Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brett Creeley 提交于
Currently, if there are bare-metal VFs passing traffic and the ice driver is removed, there is a possibility of VFs triggering a Tx timeout right before iavf_remove(). This is causing iavf_close() to not be called because there is a check in the beginning of iavf_remove() that bails out early if (adapter->state < IAVF_DOWN_PENDING). This makes it so some resources do not get cleaned up. Specifically, free_irq() is never called for data interrupts, which results in the following line of code to trigger: pci_disable_msix() free_msi_irqs() ... BUG_ON(irq_has_action(entry->irq + i)); ... To prevent the Tx timeout from occurring on the VF during driver unload for ice and the iavf there are a few changes that are needed. [1] Don't disable all active VF Tx/Rx queues prior to calling pci_disable_sriov. [2] Call ice_free_vfs() before disabling the service task. [3] Disable VF resets when the ice driver is being unloaded by setting the pf->state flag __ICE_VF_RESETS_DISABLED. Changing [1] and [2] allow each VF driver's remove flow to successfully send VIRTCHNL requests, which includes queue disable. This prevents unexpected Tx timeouts because the PF driver is no longer forcefully disabling queues. Due to [1] and [2] there is a possibility that the PF driver will get a VFLR or reset request over VIRTCHNL from a VF during PF driver unload. Prevent that by doing [3]. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brett Creeley 提交于
Currently when the device runs out of MSI-X interrupts a cryptic and unhelpful message is printed. This will cause confusion when hitting this case. Fix this by clearing up the error message for both SR-IOV and non SR-IOV use cases. Also, make a few minor changes to increase clarity of variables. 1. Change per VF MSI-X and queue pair variables in the PF structure. 2. Use ICE_NONQ_VECS_VF when determining pf->num_msix_per_vf instead of the magic number "1". This vector is reserved for the OICR. All of the resource tracking functions were moved to avoid adding any forward declaration function prototypes. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Mitch Williams 提交于
Unlike the XL710 series, 800-series hardware can allocate more than 4 MSI-X vectors per VF. This patch enables that functionality. We dynamically allocate vectors and queues depending on how many VFs are enabled. Allocating the maximum number of VFs replicates XL710 behavior with 4 queues and 4 vectors. But allocating a smaller number of VFs will give you 16 queues and 16 vectors. Signed-off-by: NMitch Williams <mitch.a.williams@intel.com> Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 20 2月, 2020 1 次提交
-
-
由 Paul Greenwalt 提交于
Update the PF VFs MDD event message to rate limit once per second and report the total number Rx|Tx event count. Add support to print pending MDD events that occur during the rate limit. The use of net_ratelimit did not allow for per VF Rx|Tx granularity. Additional PF MDD log messages are guarded by netif_msg_[rx|tx]_err(). Since VF RX MDD events disable the queue, add ethtool private flag mdd-auto-reset-vf to configure VF reset to re-enable the queue. Disable anti-spoof detection interrupt to prevent spurious events during a function reset. To avoid race condition do not make PF MDD register reads conditional on global MDD result. Signed-off-by: NPaul Greenwalt <paul.greenwalt@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 04 1月, 2020 3 次提交
-
-
由 Krzysztof Kazimierczak 提交于
In ice_xsk_umem(), variable qid which is later used as an array index, is not validated for a possible boundary exceedance. Because of that, a calling function might receive an invalid address, which causes general protection fault when dereferenced. To address this, add a boundary check to see if qid is greater than the size of a UMEM array. Also, don't let user change vsi->num_xsk_umems just by trying to setup a second UMEM if its value is already set up (i.e. UMEM region has already been allocated for this VSI). While at it, make sure that ring->zca.free pointer is always zeroed out if there is no UMEM on a specified ring. Signed-off-by: NKrzysztof Kazimierczak <krzysztof.kazimierczak@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brett Creeley 提交于
We can't have more than one default VSI so prevent another VSI from overwriting the current dflt_vsi. This was achieved by adding the following functions: ice_is_dflt_vsi_in_use() - Used to check if the default VSI is already being used. ice_is_vsi_dflt_vsi() - Used to check if VSI passed in is in fact the default VSI. ice_set_dflt_vsi() - Used to set the default VSI via a switch rule ice_clear_dflt_vsi() - Used to clear the default VSI via a switch rule. Also, there was no need to introduce any locking because all mailbox events and synchronization of switch filters for the PF happen in the service task. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brett Creeley 提交于
There are many things wrong with the function ice_set_vf_spoofchk(). 1. The VSI being modified is the PF VSI, not the VF VSI. 2. We are enabling Rx VLAN pruning instead of Tx VLAN anti-spoof. 3. The spoofchk setting for each VF is not initialized correctly or re-initialized correctly on reset. To fix [1] we need to make sure we are modifying the VF VSI. This is done by using the vf->lan_vsi_idx to index into the PF's VSI array. To fix [2] replace setting Rx VLAN pruning in ice_set_vf_spoofchk() with setting Tx VLAN anti-spoof. To Fix [3] we need to make sure the initial VSI settings match what is done in ice_set_vf_spoofchk() for spoofchk=on. Also make sure this also works for VF reset. This was done by modifying ice_vsi_init() to account for the current spoofchk state of the VF VSI. Because of these changes, Tx VLAN anti-spoof needs to be removed from ice_cfg_vlan_pruning(). This is okay for the VF because this is now controlled from the admin enabling/disabling spoofchk. For the PF, Tx VLAN anti-spoof should not be set. This change requires us to call ice_set_vf_spoofchk() when configuring promiscuous mode for the VF which requires ice_set_vf_spoofchk() to move in order to prevent a forward declaration prototype. Also, add VLAN 0 by default when allocating a VF since the PF is unaware if the guest OS is running the 8021q module. Without this, MDD events will trigger on untagged traffic because spoofcheck is enabled by default. Due to this change, ignore add/delete messages for VLAN 0 from VIRTCHNL since this is added/deleted during VF initialization/teardown respectively and should not be modified. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 23 11月, 2019 2 次提交
-
-
由 Henry Tieman 提交于
Add code to query and set the number of channels on the primary VSI for a PF. This is accessed from the 'ethtool -l' and 'ethtool -L' commands, respectively. Though the ice driver supports asymmetric queues report an IRQ vector that has both Rx and Tx queues attached and is counted as a 'combined' channel. Signed-off-by: NHenry Tieman <henry.w.tieman@intel.com> Co-developed-by: NMaciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by: NMaciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brett Creeley 提交于
We use &pf->dev->pdev all over the code. Add a simple macro to do this for us. When multiple de-references like this are being done add a local struct device variable. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 09 11月, 2019 2 次提交
-
-
由 Dave Ertman 提交于
Implement interface layer for the DCBNL subsystem. These are the functions to support the callbacks defined in the dcbnl_rtnl_ops struct. These callbacks are going to be used to interface with the DCB settings of the device. Implementation of dcb_nl set functions and supporting SW DCB functions. Signed-off-by: NDave Ertman <david.m.ertman@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Anirudh Venkataramanan 提交于
DCB configuration flow needs to disable and enable only the PF (main) VSI, so use ice_ena_vsi and ice_dis_vsi. To avoid the use of ifdef to control the staticness of these functions, move them to ice_lib.c. Also replace the allocate and copy of old_cfg to kmemdup() in ice_pf_dcb_cfg(). Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 05 11月, 2019 4 次提交
-
-
由 Maciej Fijalkowski 提交于
Add an ethtool "legacy-rx" priv flag for toggling the Rx path. This control knob will be mainly used for build_skb usage as well as buffer size/MTU manipulation. In preparation for adding build_skb support in a way that it takes care of how we set the values of max_frame and rx_buf_len fields of struct ice_vsi. Specifically, in this patch mentioned fields are set to values that will allow us to provide headroom and tailroom in-place. This can be mostly broken down onto following: - for legacy-rx "on" ethtool control knob, old behaviour is kept; - for standard 1500 MTU size configure the buffer of size 1536, as network stack is expecting the NET_SKB_PAD to be provided and NET_IP_ALIGN can have a non-zero value (these can be typically equal to 32 and 2, respectively); - for larger MTUs go with max_frame set to 9k and configure the 3k buffer in case when PAGE_SIZE of underlying arch is less than 8k; 3k buffer is implying the need for order 1 page, so that our page recycling scheme can still be applied; With that said, substitute the hardcoded ICE_RXBUF_2048 and PAGE_SIZE values in DMA API that we're making use of with rx_ring->rx_buf_len and ice_rx_pg_size(rx_ring). The latter is an introduced helper for determining the page size based on its order (which was figured out via ice_rx_pg_order). Last but not least, take care of truesize calculation. In the followup patch the headroom/tailroom computation logic will be introduced. This change aligns the buffer and frame configuration with other Intel drivers, most importantly with iavf. Signed-off-by: NMaciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Krzysztof Kazimierczak 提交于
Add zero copy AF_XDP support. This patch adds zero copy support for Tx and Rx; code for zero copy is added to ice_xsk.h and ice_xsk.c. For Tx, implement ndo_xsk_wakeup. As with other drivers, reuse existing XDP Tx queues for this task, since XDP_REDIRECT guarantees mutual exclusion between different NAPI contexts based on CPU ID. In turn, a netdev can XDP_REDIRECT to another netdev with a different NAPI context, since the operation is bound to a specific core and each core has its own hardware ring. For Rx, allocate frames as MEM_TYPE_ZERO_COPY on queues that AF_XDP is enabled. Signed-off-by: NKrzysztof Kazimierczak <krzysztof.kazimierczak@intel.com> Co-developed-by: NMaciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by: NMaciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Maciej Fijalkowski 提交于
Add support for XDP. Implement ndo_bpf and ndo_xdp_xmit. Upon load of an XDP program, allocate additional Tx rings for dedicated XDP use. The following actions are supported: XDP_TX, XDP_DROP, XDP_REDIRECT, XDP_PASS, and XDP_ABORTED. Signed-off-by: NMaciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Anirudh Venkataramanan 提交于
Remove a few uses of kernel configuration flags from ice_lib.c by introducing a new source file ice_base.c. Also move corresponding function prototypes from ice_lib.h to ice_base.h and include ice_base.h where required. Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 13 9月, 2019 3 次提交
-
-
由 Tony Nguyen 提交于
Attempt to request an optional device-specific DDP package file (one with the PCIe Device Serial Number in its name so that different DDP package files can be used on different devices). If the optional package file exists, download it to the device. If not, download the default package file. Log an appropriate message based on whether or not a DDP package file exists and the return code from the attempt to download it to the device. If the download fails and there is not already a package file on the device, go into "Safe Mode" where some features are not supported. Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Lukasz Czapnik 提交于
The FW build id is currently being displayed as an int which doesn't make sense. Instead display FW build id as a hex value. Also add other useful information to the output such as NVM version, API patch info, and FW build hash. Signed-off-by: NLukasz Czapnik <lukasz.czapnik@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Paul M Stillwell Jr 提交于
The driver is required to send a version to the firmware to indicate that the driver is up. If the driver doesn't do this the firmware doesn't behave properly. Signed-off-by: NPaul M Stillwell Jr <paul.m.stillwell.jr@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 05 9月, 2019 4 次提交
-
-
由 Jesse Brandeburg 提交于
The driver should start out with a reasonable number of descriptors that can prevent drops due to a CPU being in a power management state. Change the default number of descriptors to 2048. The user can always change the value at runtime. Transmit descriptor counts are not modified because they don't need to change due to the speed of the interface, or for power managed CPUs, but the code is simplified to a fixed value for the transmit default. Signed-off-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Anirudh Venkataramanan 提交于
Remove q_left_tx and q_left_rx from the PF struct as these can be obtained by calling ice_get_avail_txq_count and ice_get_avail_rxq_count respectively. The function ice_determine_q_usage is only setting num_lan_tx and num_lan_rx in the PF structure, and these are later assigned to vsi->alloc_txq and vsi->alloc_rxq respectively. This is an unnecessary indirection, so remove ice_determine_q_usage and just assign values for vsi->alloc_txq and vsi->alloc_rxq in ice_vsi_set_num_qs and use these to set num_lan_tx and num_lan_rx respectively. Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Jesse Brandeburg 提交于
The driver has supported a transmit work limit that was configurable from ethtool for a long time, but there are no good use cases for having it be a variable that can be changed at run time. In addition, this variable was noted to be causing performance overhead due to cache misses. Just remove the variable and let the code use a constant so that the functionality is maintained (a limit on the number of transmits that will be cleaned in any one call to the clean routines) without the cache miss. Removes code, removes a variable, removes testing surface. Yay. Signed-off-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Anirudh Venkataramanan 提交于
There are multiple places where we currently use ice_find_vsi_by_type to get the PF (a.k.a. main) VSI. The PF VSI by definition is always the first element in the pf->vsi array (i.e. pf->vsi[0]). So instead add and use a new helper function ice_get_main_vsi, which just returns pf->vsi[0]. Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 27 8月, 2019 1 次提交
-
-
由 Anirudh Venkataramanan 提交于
The total number of queues available on the device is divided between multiple physical functions (PF) in the firmware and provided to the driver when it gets function capabilities from the firmware. Thus each PF knows how many Tx/Rx queues it has. These queues are then doled out to different VSIs (for LAN traffic, SR-IOV VF traffic, etc.) To track usage of these queues at the PF level, the driver uses two bitmaps avail_txqs and avail_rxqs. At the VSI level (i.e. struct ice_vsi instances) the driver uses two arrays txq_map and rxq_map, to track ownership of VSIs' queues in avail_txqs and avail_rxqs respectively. The aforementioned bitmaps and arrays should be allocated dynamically, because the number of queues supported by a PF is only available once function capabilities have been queried. The current static allocation consumes way more memory than required. This patch removes the DECLARE_BITMAP for avail_txqs and avail_rxqs and instead uses bitmap_zalloc to allocate the bitmaps during init. Similarly txq_map and rxq_map are now allocated in ice_vsi_alloc_arrays. As a result ICE_MAX_TXQS and ICE_MAX_RXQS defines are no longer needed. Also as txq_map and rxq_map are now allocated and freed, some code reordering was required in ice_vsi_rebuild for correct functioning. Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 24 8月, 2019 2 次提交
-
-
由 Bruce Allan 提交于
Users expect ethtool statistics to be updated on-demand when invoking 'ethtool -S <iface>' instead of providing a snapshot of statistics taken once a second (the frequency of the watchdog task where stats are currently updated). Update stats every time 'ethtool -S <iface>' is run. Also, fix an indentation style issue and an unnecessary local variable initialization in ice_get_ethtool_stats() discovered while investigating the subject issue. Signed-off-by: NBruce Allan <bruce.w.allan@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Dave Ertman 提交于
The current flag name of "enable-fw-lldp" is a bit cumbersome. Change priv-flag name to "fw-lldp-agent" with a value of on or off. This is more straight-forward in meaning. Signed-off-by: NDave Ertman <david.m.ertman@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 21 8月, 2019 3 次提交
-
-
由 Akeem G Abodunrin 提交于
In order to use some of the VF resources definition in the SR-IOV specific virtchnl header file, this patch moves applicable code to ice_virtchnl_pf.h file accordingly... and they should have been defined in the destination file originally. Signed-off-by: NAkeem G Abodunrin <akeem.g.abodunrin@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brett Creeley 提交于
Currently we use the ICE_MBXQ_LEN for both the Mailbox send and receive queues that are used to communicate with VFs. This is fine for the send queue because the PF driver will lock the queue for every single send, but for the Mailbox receive queue every VF is posting to its Mailbox send queue and the hardware is then handing the message to the PF on its Mailbox receive queue. This becomes a problem with many VFs because it seems to overburden the Mailbox receive queue on the PF. Fix this by increasing the Mailbox receive queue for the PF to 512 entries. The number 512 was determined based on the number of VFs supported by the device. We can have a total of 256 VFs so in the worst case this allows the VFs to put 2 messages in the PFs Mailbox receive queue at the same time. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Akeem G Abodunrin 提交于
This patch restructures how VFs are configured, and resources allocated. Instead of freeing resources that were never allocated, and resetting empty VFs that have never been created - the new flow will just allocate resources for number of requested VFs based on the availability. During VFs initialization process, global interrupt is disabled, and rearmed after getting MSIX vectors for VFs. This allows immediate mailbox communications, instead of delaying it till later and VFs. PF communications resulted to using polling instead of actual interrupt. The issue manifested when creating higher number of VFs (128 VFs) per PF. Signed-off-by: NAkeem G Abodunrin <akeem.g.abodunrin@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 01 8月, 2019 2 次提交
-
-
由 Brett Creeley 提交于
This flag is not needed and is called every time we re-enable interrupts in the hotpath so remove it. Also remove ice_vsi_req_irq() because it was a wrapper function for ice_vsi_req_irq_msix() whose sole purpose was checking the ICE_FLAG_MSIX_ENA flag. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Tony Nguyen 提交于
The firmware reports an error when trying to configure a port with no media. Instead of always configuring the port, check for media before attempting to configure it. In the absence of media, turn off link and poll for media to become available before re-enabling link. Move ice_force_phys_link_state() up to avoid forward declaration. Signed-off-by: NTony Nguyen <anthony.l.nguyen@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 31 5月, 2019 1 次提交
-
-
由 Preethi Banala 提交于
Change minimum number of descriptor count from 32 to 64. This is to have a feature parity with previous Intel NIC drivers. Signed-off-by: NPreethi Banala <preethi.banala@intel.com> Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 29 5月, 2019 3 次提交
-
-
由 Brett Creeley 提交于
Currently we have two MSI-x (IRQ) trackers, one for OS requested MSI-x entries (sw_irq_tracker) and one for hardware MSI-x vectors (hw_irq_tracker). Generally the sw_irq_tracker has less entries than the hw_irq_tracker because the hw_irq_tracker has entries equal to the max allowed MSI-x per PF and the sw_irq_tracker is mainly the minimum (non SR-IOV portion of the vectors, kernel granted IRQs). All of the non SR-IOV portions of the driver (i.e. LAN queues, RDMA queues, OICR, etc.) take at least one of each type of tracker resource. SR-IOV only grabs entries from the hw_irq_tracker. There are a few issues with this approach that can be seen when doing any kind of device reconfiguration (i.e. ethtool -L, SR-IOV, etc.). One of them being, any time the driver creates an ice_q_vector and associates it to a LAN queue pair it will grab and use one entry from the hw_irq_tracker and one from the sw_irq_tracker. If the indices on these does not match it will cause a Tx timeout, which will cause a reset and then the indices will match up again and traffic will resume. The mismatched indices come from the trackers not being the same size and/or the search_hint in the two trackers not being equal. Another reason for the refactor is the co-existence of features with SR-IOV. If SR-IOV is enabled and the interrupts are taken from the end of the sw_irq_tracker then other features can no longer use this space because the hardware has now given the remaining interrupts to SR-IOV. This patch reworks how we track MSI-x vectors by removing the hw_irq_tracker completely and instead MSI-x resources needed for SR-IOV are determined all at once instead of per VF. This can be done because when creating VFs we know how many are wanted and how many MSI-x vectors each VF needs. This also allows us to start using MSI-x resources from the end of the PF's allowed MSI-x vectors so we are less likely to use entries needed for other features (i.e. RDMA, L2 Offload, etc). This patch also reworks the ice_res_tracker structure by removing the search_hint and adding a new member - "end". Instead of having a search_hint we will always search from 0. The new member, "end", will be used to manipulate the end of the ice_res_tracker (specifically sw_irq_tracker) during runtime based on MSI-x vectors needed by SR-IOV. In the normal case, the end of ice_res_tracker will be equal to the ice_res_tracker's num_entries. The sriov_base_vector member was added to the PF structure. It is used to represent the starting MSI-x index of all the needed MSI-x vectors for all SR-IOV VFs. Depending on how many MSI-x are needed, SR-IOV may have to take resources from the sw_irq_tracker. This is done by setting the sw_irq_tracker->end equal to the pf->sriov_base_vector. When all SR-IOV VFs are removed then the sw_irq_tracker->end is reset back to sw_irq_tracker->num_entries. The sriov_base_vector, along with the VF's number of MSI-x (pf->num_vf_msix), vf_id, and the base MSI-x index on the PF (pf->hw.func_caps.common_cap.msix_vector_first_id), is used to calculate the first HW absolute MSI-x index for each VF, which is used to write to the VPINT_ALLOC[_PCI] and GLINT_VECT2FUNC registers to program the VFs MSI-x PCI configuration bits. Also, the sriov_base_vector is used along with VF's num_vf_msix, vf_id, and q_vector->v_idx to determine the MSI-x register index (used for writing to GLINT_DYN_CTL) within the PF's space. Interrupt changes removed any references to hw_base_vector, hw_oicr_idx, and hw_irq_tracker. Only sw_base_vector, sw_oicr_idx, and sw_irq_tracker variables remain. Change all of these by removing the "sw_" prefix to help avoid confusion with these variables and their use. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Anirudh Venkataramanan 提交于
This patch adds a handler for ethtool selftest. Selftest includes testing link, interrupts, eeprom, registers and packet loopback. Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Brett Creeley 提交于
Currently we set the default number of Rx descriptors per queue to the system's page size divided by the number of bytes per descriptor. For 4K page size systems this is resulting in 128 Rx descriptors per queue. This is causing more dropped packets than desired in the default configuration. Fix this by setting the minimum default Rx descriptor count per queue to 512. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
- 24 5月, 2019 2 次提交
-
-
由 Brett Creeley 提交于
Currently the driver is calling ice_napi_del() and then unregister_netdev(). The call to unregister_netdev() will result in a call to ice_stop() and then ice_vsi_close(). This is where we call napi_disable() for all the MSI-X vectors. This flow is reversed so make the changes to ensure napi_disable() happens prior to napi_del(). Before calling napi_del() and free_netdev() make sure unregister_netdev() was called. This is done by making sure the __ICE_DOWN bit is set in the vsi->state for the interested VSI. Signed-off-by: NBrett Creeley <brett.creeley@intel.com> Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-
由 Jesse Brandeburg 提交于
We can use bit fields to store boolean values and when the bit fields are next to each other, the compiler will combine them (as long as the size holds enough). Signed-off-by: NJesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com> Tested-by: NAndrew Bowers <andrewx.bowers@intel.com> Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
-