- 10 Mar, 2020 40 commits
-
-
Jakub Kicinski authored
Set ethtool_ops->supported_coalesce_params to let the core reject unsupported coalescing parameters. This driver only rejected some of the unsupported parameters. Signed-off-by:
Jakub Kicinski <kuba@kernel.org> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Jakub Kicinski authored
Set ethtool_ops->supported_coalesce_params to let the core reject unsupported coalescing parameters. Signed-off-by:
Sameeh Jubran <sameehj@amazon.com> Acked-by:
-
Heiner Kallweit authored
Let netdev_stats_to_stats64() do the copy work for us. Signed-off-by:
Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by:
-
Heiner Kallweit authored
Clearing opts2 belongs to preparing the descriptor for DMA engine use. Therefore move it into rtl8169_mark_to_asic(). Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queueDavid S. Miller authored
Jeff Kirsher says: ==================== 100GbE Intel Wired LAN Driver Updates 2020-03-10 This series contains updates to ice and iavf drivers. Cleaned up unnecessary parenthesis, which was pointed out by Sergei Shtylyov. Mitch updates the iavf and ice drivers to expand the limitation on the number of queues that the driver can support to account for the newer 800-series capabilities. Brett cleans up the error messages for both SR-IOV and non SR-IOV use cases. Fixed the logic when the ice driver is removed and a bare-metal VF is passing traffic, which was causing a transmit hang on the VF. Updated the ice driver to display "Link detected" field via ethtool, when the driver is in safe mode. Updated ice driver to properly set VLAN pruning when transmit anti-spoof is off. Avinash fixed a corner case in DCB, when switching from IEEE to CEE mode, the DCBX mode does not get properly updated. Dave updates the logic when switching from software DCB to firmware DCB to renegotiate DCBX to ensure the firmware agent has up to date information about the DCB settings of the link partner. Lukasz increases the PF's mailbox receive queue size to the maximum to prevent potential bottleneck or slow down occurring from the PF's mailbox receive queue being full. Bruce updates the ice driver to use strscpy() instead of strlcpy(). Cleaned up variable names that were not very descriptive with names that had more meaning. Anirudh replaces the use of ENOTSUPP with EOPNOTSUPP in the ice driver. Jake fixed up a function header comment to properly reflect the variable size and use. v2: Dropped patch 5 of the original series, where Tony added tunnel offload support. Based on community feedback, the patch needed changes, so giving Tony additional time to work on those changes and not hold up the remaining changes in the series. ==================== Acked-by: -
DENG Qingfang authored
The inner pair of parentheses should be around the variable x Fixes: 37feab60 ("net: dsa: mt7530: add support for port mirroring") Signed-off-by:
DENG Qingfang <dqfext@gmail.com> Signed-off-by:
-
George McCollister authored
Simplify ksz_common.c by using delayed_work instead of a combination of timer and work. Signed-off-by:
George McCollister <george.mccollister@gmail.com> Signed-off-by:
-
David S. Miller authored
Jiri Pirko says: ==================== flow_offload: follow-ups to HW stats type patchset This patchset includes couple of patches in reaction to the discussions to the original HW stats patchset. The first patch is a fix, the other two patches are basically cosmetics. ==================== Acked-by:
Edward Cree <ecree@solarflare.com> Reviewed-by:
-
Jiri Pirko authored
The intention of this helper was to allow driver to specify one type that it supports, so not only "any" value would pass. So make the API more strict and allow driver to pass only 1 bit that is going to be checked. Signed-off-by:
Jiri Pirko <jiri@resnulli.us> Signed-off-by:
-
Jiri Pirko authored
Put the values into enum and add an enum to define the bits. Suggested-by:
-
Jiri Pirko authored
Change the check to see if the passed allowed type bit is enabled. Fixes: 319a1d19 ("flow_offload: check for basic action hw stats type") Signed-off-by:
-
David S. Miller authored
Claudiu Manoil says: ==================== enetc: Support extended BD rings at runtime First two patches are just misc code cleanup. The 3rd patch prepares the Rx BD processing code to be extended to processing both normal and extended BDs. The last one adds extended Rx BD support for timestamping without the need of a static config. Finally, the config option FSL_ENETC_HW_TIMESTAMPING can be dropped. Care was taken not to impact non-timestamping usecases. ==================== Signed-off-by: -
Claudiu Manoil authored
Hardware timestamping support (PTP) on Rx requires extended buffer descriptors, double the size of normal Rx descriptors. On the current controller revision only the timestamping offload requires extended Rx descriptors. Since Rx timestamping can be turned on/off at runtime, make Rx ring allocation configurable at runtime too. As a result, the static config option FSL_ENETC_HW_TIMESTAMPING can be dropped and the extended descriptors can be used only when Rx timestamping gets activated. The extension has the same size as the base descriptor, making the descriptor iterators easy to update for the extended case. Signed-off-by:
Claudiu Manoil <claudiu.manoil@nxp.com> Signed-off-by:
-
Claudiu Manoil authored
Improve maintainability of the code iterating the Rx buffer descriptors to prepare it to support iterating extended Rx BD descriptors as well. Don't increment by one the h/w descriptor pointers explicitly, provide an iterator that takes care of the h/w details. Signed-off-by:
-
Claudiu Manoil authored
Refactor the stats len computation code to make it easier to add new stats counters. Signed-off-by:
-
Claudiu Manoil authored
The existence of the DT port node is the first thing checked at probe time, and probing won't reach this point if the node is missing. Signed-off-by:
-
Lukas Wunner authored
When pktgen is used to measure the performance of dev_queue_xmit() packet handling in the core, it is preferable to not hand down packets to a low-level Ethernet driver as it would distort the measurements. Allow using pktgen on the loopback device, thus constraining measurements to core code. Signed-off-by:
Lukas Wunner <lukas@wunner.de> Signed-off-by:
-
Jiri Pirko authored
Instead of manually iterating over entries, use flow_action_for_each helper. Move the helper and wrap it to fit to 80 cols on the way. Signed-off-by:
Pablo Neira Ayuso <pablo@netfilter.org> Acked-by:
-
Jacob Keller authored
The function comment for ice_get_nvm_version indicated that the ver_hi and ver_lo values were 16 bits. In fact, they are only uint8_t values, meaning that they have a maximum size of 8 bits. Fix the comment to match the correct size. Signed-off-by:
Jacob Keller <jacob.e.keller@intel.com> Tested-by:
Andrew Bowers <andrewx.bowers@intel.com> Signed-off-by:
Jeff Kirsher <jeffrey.t.kirsher@intel.com>
-
Bruce Allan authored
The variable name 'type' is not very descriptive. Replace instances of those with a variable name that is more descriptive or replace it if not needed. Signed-off-by:
Bruce Allan <bruce.w.allan@intel.com> Signed-off-by:
Tony Nguyen <anthony.l.nguyen@intel.com> Tested-by:
-
Anirudh Venkataramanan authored
Using ENOTSUPP almost always results in some bizarre error message to be printed in userspace. This is likely because ENOTSUPP was defined for the NFS protocol (as per a comment in include/linux/errno.h). Use EOPNOTSUPP instead. Signed-off-by:
Anirudh Venkataramanan <anirudh.venkataramanan@intel.com> Signed-off-by:
-
Tony Nguyen authored
Commit ed5a3f66 ("ice: Removing hung_queue variable to use txqueue function parameter") began utilizing the txqueue variable over the hung_queue variable. hung_queue was an int where txqueue is an unsigned int. Update the format specifiers to reflect the new type. Fixes: ed5a3f66 ("ice: Removing hung_queue variable to use txqueue function parameter") Signed-off-by:
-
Bruce Allan authored
checkpatch complains "CHECK:DEPRECATED_API: Deprecated use of 'strlcpy', prefer 'stracpy or strscpy' instead"; use strscpy. Signed-off-by:
-
Lukasz Czapnik authored
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:
Lukasz Czapnik <lukasz.czapnik@intel.com> Signed-off-by:
Brett Creeley <brett.creeley@intel.com> Signed-off-by:
-
Brett Creeley authored
VLAN pruning is not always being set correctly due to a previous change that set Tx antispoof off. ice_vsi_is_vlan_pruning_ena() currently checks for both Tx antispoof and Rx pruning. The expectation for this function is to only check Rx pruning so fix the check. Fixes: cd6d6b83 ("ice: Fix VF spoofchk") Signed-off-by:
-
Dave Ertman authored
When switching from SW DCB to FW DCB it is necessary to renegotiate DCBx so that the FW agent can have up to date information about the DCB settings of the link partner. Perform an autoneg restart on the link when activating FW DCB. Signed-off-by:
Dave Ertman <david.m.ertman@intel.com> Tested-by:
-
Avinash JD authored
While testing DCB for a corner case in which mode is switched from IEEE to CEE and pfc_ena bitmask unchanged then DCBX mode doesn't get updated. This is happening because the function ice_dcb_get_mode() is called in a "no change detected block" instead of "change detected block". Signed-off-by:
Avinash JD <avinash.dayanand@intel.com> Signed-off-by:
Scott Register <scottx.register@intel.com> Tested-by:
-
Brett Creeley authored
Currently the "Link detected" field is not shown when the device goes into safe mode. This is because the safe mode Ethtool ops does not set the get_link function. Fix this by setting the safe mode Ethtool op get_link function. Signed-off-by:
-
Brett Creeley authored
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:
-
Brett Creeley authored
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:
-
Mitch Williams authored
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:
Mitch Williams <mitch.a.williams@intel.com> Signed-off-by:
-
Mitch Williams authored
Previous devices could only allocate 4 MSI-X vectors per VF so there was a limitation of 4 queues. 800-series hardware can allocate more than 4 MSI-X vectors, so expand the limitation on the number of queues that the driver can support to account for these capabilities. Fix ethtool channel operations to accommodate this change and adjust the reporting of max number of queues to what is given to us by the PF. Since we're not requesting queues above this value, just trigger reset to activate the queues, which we already own. Finally, fix a test condition that would display an incorrect error message. Signed-off-by:
-
Jeff Kirsher authored
Sergei Shtylyov pointed out that two instances of parenthesis are not needed, so remove them. Suggested-by:
Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by:
-
David S. Miller authored
Jose Abreu says: ==================== net: Add support for Synopsys DesignWare XPCS This adds support for Synopsys DesignWare XPCS in net subsystem and integrates it into stmmac. At 1/8, we start by removing the limitation of stmmac selftests that needed a PHY to pass all the tests. Then at 2/8 we use some helpers in stmmac so that some code can be simplified. At 3/8, we fallback to dev_fwnode() so that PCI based setups wich may not have CONFIG_OF can still use FW node. At 4/8, we adapt stmmac to the new PHYLINK changes as suggested by Russell King. We proceed by doing changes in PHYLINK in order to support XPCS: At 5/8 we add some missing speeds that USXGMII supports and at 6/8 we check if Autoneg is supported after initial parameters are validated. Support for XPCS is finally introduced at 7/8, along with the usage of it in stmmac driver at 8/8. ==================== Signed-off-by: -
Jose Abreu authored
Adds all the necessary logic so that stmmac can be used with Synopsys DesignWare XPCS. Signed-off-by:
Jose Abreu <Jose.Abreu@synopsys.com> Signed-off-by:
-
Jose Abreu authored
Synopsys DesignWare XPCS is an MMD that can manage link status, auto-negotiation, link training, ... In this commit we add basic support for XPCS using USXGMII interface and Clause 73 Auto-negotiation. This is highly tied with PHYLINK and can't be used without it. A given ethernet driver can use the provided callbacks to add the support for XPCS. Signed-off-by:
-
Jose Abreu authored
We may have cases where MAC or PCS do not support Autoneg. Check if it is supported after validate callback is called. Signed-off-by:
-
Jose Abreu authored
USXGMII also supports these missing backplane speeds. Signed-off-by:
-
Jose Abreu authored
Convert the stmmac ethernet driver to use the finalised link parameters in mac_link_up() rather than the parameters in mac_config(). Suggested-by:
Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by:
-
Jose Abreu authored
When CONFIG_OF is not enabled, of_fwnode_handle() will return NULL, even though we can have a FW handle from a given device. Fallback to dev_fwnode() helper if needed. Signed-off-by:
-
