aboutsummaryrefslogtreecommitdiff
path: root/drivers/net/ethernet/intel
AgeCommit message (Collapse)AuthorFilesLines
2018-09-25Merge branch '40GbE' of ↵David S. Miller6-216/+412
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue Jeff Kirsher says: ==================== 40GbE Intel Wired LAN Driver Updates 2018-09-25 This series contains updates to i40e and xsk. Mariusz fixes an issue where the VF link state was not being updated properly when the PF is down or up. Also cleaned up the promiscuous configuration during a VF reset. Patryk simplifies the code a bit to use the variables for PF and HW that are declared, rather than using the VSI pointers. Cleaned up the message length parameter to several virtchnl functions, since it was not being used (or needed). Harshitha fixes two potential race conditions when trying to change VF settings by creating a helper function to validate that the VF is enabled and that the VSI is set up. Sergey corrects a double "link down" message by putting in a check for whether or not the link is up or going down. Björn addresses an AF_XDP zero-copy issue that buffers passed from userspace to the kernel was leaked when the hardware descriptor ring was torn down. A zero-copy capable driver picks buffers off the fill ring and places them on the hardware receive ring to be completed at a later point when DMA is complete. Similar on the transmit side; The driver picks buffers off the transmit ring and places them on the transmit hardware ring. In the typical flow, the receive buffer will be placed onto an receive ring (completed to the user), and the transmit buffer will be placed on the completion ring to notify the user that the transfer is done. However, if the driver needs to tear down the hardware rings for some reason (interface goes down, reconfiguration and such), the userspace buffers cannot be leaked. They have to be reused or completed back to userspace. ==================== Signed-off-by: David S. Miller <[email protected]>
2018-09-25i40e: disallow changing the number of descriptors when AF_XDP is onBjörn Töpel3-0/+30
When an AF_XDP UMEM is attached to any of the Rx rings, we disallow a user to change the number of descriptors via e.g. "ethtool -G IFNAME". Otherwise, the size of the stash/reuse queue can grow unbounded, which would result in OOM or leaking userspace buffers. Signed-off-by: Björn Töpel <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25i40e: clean zero-copy XDP Rx ring on shutdown/resetBjörn Töpel3-9/+96
Outstanding Rx descriptors are temporarily stored on a stash/reuse queue. When/if the HW rings comes up again, entries from the stash are used to re-populate the ring. The latter required some restructuring of the allocation scheme for the AF_XDP zero-copy implementation. There is now a fast, and a slow allocation. The "fast allocation" is used from the fast-path and obtains free buffers from the fill ring and the internal recycle mechanism. The "slow allocation" is only used in ring setup, and obtains buffers from the fill ring and the stash (if any). Signed-off-by: Björn Töpel <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25i40e: clean zero-copy XDP Tx ring on shutdown/resetBjörn Töpel3-6/+43
When the zero-copy enabled XDP Tx ring is torn down, due to configuration changes, outstanding frames on the hardware descriptor ring are queued on the completion ring. The completion ring has a back-pressure mechanism that will guarantee that there is sufficient space on the ring. Signed-off-by: Björn Töpel <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25i40e: Remove unused msglen parameter from virtchnl functionsPatryk Małek1-59/+37
msglen parameter seems to be unused in several virtchnl function. This patch removes it from signatures of those functions. Signed-off-by: Patryk Małek <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25i40e: fix double 'NIC Link is Down' messagesSergey Nemov1-1/+4
When isup is false meaning that interface is going to shut down set new speed to 0 to avoid double 'NIC Link is Down' messages. Signed-off-by: Sergey Nemov <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25i40e: add a helper function to validate a VF based on the vf idHarshitha Ramamurthy1-21/+41
When we are trying to change VF settings, it is possible for 2 race conditions to happen. One, when the VF is created but not yet enabled. Second, the VF is enabled but the VSI is still not created or not yet re-created in the VF reset flow. This patch introduces a helper function to validate that the VF is enabled and that the VSI is set up. This patch also calls this function from other functions which could get into these race conditions. While we are poking around here, remove unnecessary parenthesis that checkpatch was complaining about. Signed-off-by: Harshitha Ramamurthy <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25i40e: use declared variables for pf and hwPatryk Małek1-5/+4
In order to slightly simplify the code use the variables for pf and hw that are declared in i40e_set_mac function. Signed-off-by: Patryk Małek <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25i40e: Unset promiscuous settings on VF resetMariusz Stachura1-110/+157
This patch cleans up promiscuous configuration when a VF reset occurs. Previously the promiscuous mode settings were still there after the VF driver removal. Signed-off-by: Mariusz Stachura <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25i40e: Fix VF's link state notificationMariusz Stachura1-5/+0
This resolves an issue where the VF link state was not being updated when the PF is down or up, and the VF link state would always show that it is running. Signed-off-by: Mariusz Stachura <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-25Merge ra.kernel.org:/pub/scm/linux/kernel/git/davem/netDavid S. Miller9-182/+0
Version bump conflict in batman-adv, take what's in net-next. iavf conflict, adjustment of netdev_ops in net-next conflicting with poll controller method removal in net. Signed-off-by: David S. Miller <[email protected]>
2018-09-23i40evf: remove ndo_poll_controllerEric Dumazet1-26/+0
As diagnosed by Song Liu, ndo_poll_controller() can be very dangerous on loaded hosts, since the cpu calling ndo_poll_controller() might steal all NAPI contexts (for all RX/TX queues of the NIC). This capture can last for unlimited amount of time, since one cpu is generally not able to drain all the queues under load. i40evf uses NAPI for TX completions, so we better let core networking stack call the napi->poll() to avoid the capture. Signed-off-by: Eric Dumazet <[email protected]> Cc: Jeff Kirsher <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-09-23ice: remove ndo_poll_controllerEric Dumazet1-27/+0
As diagnosed by Song Liu, ndo_poll_controller() can be very dangerous on loaded hosts, since the cpu calling ndo_poll_controller() might steal all NAPI contexts (for all RX/TX queues of the NIC). This capture can last for unlimited amount of time, since one cpu is generally not able to drain all the queues under load. ice uses NAPI for TX completions, so we better let core networking stack call the napi->poll() to avoid the capture. Signed-off-by: Eric Dumazet <[email protected]> Cc: Jeff Kirsher <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-09-23igb: remove ndo_poll_controllerEric Dumazet1-30/+0
As diagnosed by Song Liu, ndo_poll_controller() can be very dangerous on loaded hosts, since the cpu calling ndo_poll_controller() might steal all NAPI contexts (for all RX/TX queues of the NIC). This capture can last for unlimited amount of time, since one cpu is generally not able to drain all the queues under load. igb uses NAPI for TX completions, so we better let core networking stack call the napi->poll() to avoid the capture. Signed-off-by: Eric Dumazet <[email protected]> Cc: Jeff Kirsher <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-09-23ixgb: remove ndo_poll_controllerEric Dumazet1-25/+0
As diagnosed by Song Liu, ndo_poll_controller() can be very dangerous on loaded hosts, since the cpu calling ndo_poll_controller() might steal all NAPI contexts (for all RX/TX queues of the NIC). This capture can last for unlimited amount of time, since one cpu is generally not able to drain all the queues under load. ixgb uses NAPI for TX completions, so we better let core networking stack call the napi->poll() to avoid the capture. This also removes a problematic use of disable_irq() in a context it is forbidden, as explained in commit af3e0fcf7887 ("8139too: Use disable_irq_nosync() in rtl8139_poll_controller()") Signed-off-by: Eric Dumazet <[email protected]> Cc: Jeff Kirsher <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-09-23fm10k: remove ndo_poll_controllerEric Dumazet3-28/+0
As diagnosed by Song Liu, ndo_poll_controller() can be very dangerous on loaded hosts, since the cpu calling ndo_poll_controller() might steal all NAPI contexts (for all RX/TX queues of the NIC). This capture lasts for unlimited amount of time, since one cpu is generally not able to drain all the queues under load. fm10k uses NAPI for TX completions, so we better let core networking stack call the napi->poll() to avoid the capture. Signed-off-by: Eric Dumazet <[email protected]> Cc: Jeff Kirsher <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-09-23ixgbevf: remove ndo_poll_controllerEric Dumazet1-21/+0
As diagnosed by Song Liu, ndo_poll_controller() can be very dangerous on loaded hosts, since the cpu calling ndo_poll_controller() might steal all NAPI contexts (for all RX/TX queues of the NIC). This capture can last for unlimited amount of time, since one cpu is generally not able to drain all the queues under load. ixgbevf uses NAPI for TX completions, so we better let core networking stack call the napi->poll() to avoid the capture. Signed-off-by: Eric Dumazet <[email protected]> Cc: Jeff Kirsher <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-09-23ixgbe: remove ndo_poll_controllerEric Dumazet1-25/+0
As diagnosed by Song Liu, ndo_poll_controller() can be very dangerous on loaded hosts, since the cpu calling ndo_poll_controller() might steal all NAPI contexts (for all RX/TX queues of the NIC). This capture can last for unlimited amount of time, since one cpu is generally not able to drain all the queues under load. ixgbe uses NAPI for TX completions, so we better let core networking stack call the napi->poll() to avoid the capture. Reported-by: Song Liu <[email protected]> Signed-off-by: Eric Dumazet <[email protected]> Tested-by: Song Liu <[email protected]> Cc: Jeff Kirsher <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-09-19e1000: remove set but not used variable 'txb2b'YueHaibing1-3/+0
Fixes gcc '-Wunused-but-set-variable' warning: drivers/net/ethernet/intel/e1000/e1000_main.c: In function 'e1000_watchdog': drivers/net/ethernet/intel/e1000/e1000_main.c:2436:9: warning: variable 'txb2b' set but not used [-Wunused-but-set-variable] Signed-off-by: YueHaibing <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-09-18intel-ethernet: use correct module licenseJesse Brandeburg12-13/+13
We recently updated all our SPDX identifiers to correctly indicate our net/ethernet/intel/* drivers were always released and intended to be released under GPL v2, but the MODULE_LICENSE declaration was never updated. Fix the MODULE_LICENSE to be GPL v2, for all our drivers. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: finish renaming files to iavfJesse Brandeburg16-21/+21
This finishes the process of renaming the files that make sense to rename (skipping adminq related files that talk to i40e), and fixes up the build and the #includes so that everything builds nicely. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: rename most of i40e stringsJesse Brandeburg17-1407/+1366
This is the big rename patch, it takes most of the i40e_ and I40E_ strings and renames them to iavf_ and IAVF_. Some of the adminq code, as well as most of the client interface code used by RDMA is left unchanged in order to indicate that the driver is talking to non-internal to iavf code. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: tracing infrastructure renameJesse Brandeburg3-22/+22
Rename the i40e_trace file and fix up all the callers to the new names inside the iavf_trace.h file. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: replace i40e_debug with iavf versionJesse Brandeburg5-23/+23
Change another string (i40e_debug) Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: rename i40e_hw to iavf_hwJesse Brandeburg9-99/+98
Fix up the i40e_hw names to new name, including versions inside other strings. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: rename I40E_ADMINQ_DESCJesse Brandeburg2-11/+11
Take care of some renames containing I40E_ADMINQ_DESC. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: rename device ID definesJesse Brandeburg5-78/+27
Rename the device ID defines to have IAVF in them and remove all the unused defines. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: remove references to old namesJesse Brandeburg9-273/+269
Remove the register name references to I40E_VF* and change to IAVF_VF. Update the descriptor names and defines to the IAVF name. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: move i40evf files to new nameJesse Brandeburg9-11/+11
Simply move the i40evf files to the new name, updating the #includes to track the new names, and updating the Makefile as well. A future patch will remove the i40e references (after the code removal patches later in this series). Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: rename i40e_status to iavf_statusJesse Brandeburg10-132/+119
This is just a rename of an internal variable i40e_status, but it was a pretty big change and so deserved it's own patch. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: rename functions and structs to new nameJesse Brandeburg14-1691/+1665
This basically begins the internal portion of the rename of i40evf to iavf, by renaming many of the functions, structs, variables and defines. Most of the changes were made mechanically, which introduces some alignment issues. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18iavf: diet and reformatJesse Brandeburg8-4014/+6
Remove a bunch of unused code and reformat a few lines. Also remove some now un-necessary files. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-18intel-ethernet: rename i40evf to iavfJesse Brandeburg26-24/+31
Rename the Intel Ethernet Adaptive Virtual Function driver (i40evf) to a new name (iavf) that is more consistent with the ongoing maintenance of the driver as the universal VF driver for multiple product lines. This first patch fixes up the directory names and the .ko name, intentionally ignoring the function names inside the driver for now. Basically this is the simplest patch that gets the rename done and will be followed by other patches that rename the internal functions. This patch also addresses a couple of string/name issues and updates the Copyright year. Also, made sure to add a MODULE_ALIAS to the old name. Signed-off-by: Jesse Brandeburg <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-09-08i40e(vf): remove i40e_ethtool_stats.h header fileJacob Keller4-444/+436
Essentially reverts commit 8fd75c58a09a ("i40e: move ethtool stats boiler plate code to i40e_ethtool_stats.h", 2018-08-30), and additionally moves the similar code in i40evf into i40evf_ethtool.c. The code was intially moved from i40e_ethtool.c into i40e_ethtool_stats.h as a way of better logically organizing the code. This has two problems. First, we can't have an inline function with variadic arguments on all platforms. Second, it gave the appearance that we had plans to share code between the i40e and i40evf drivers, due to having a near copy of the contents in the i40evf/i40e_ethtool_stats.h file. Patches which actually attempt to combine or share code between the i40e and i40evf drivers have not materialized, and are likely a ways off. Rather than fixing the one function which causes build issues, just move this code back into the i40e_ethtool.c and i40evf_ethtool.c files. Note that we also change these functions back from static inlines to just statics, since they're no longer in a header file. We can revisit this if/when work is done to actually attempt to share code between drivers. Alternatively, this stats code could be made more generic so that it can be shared across drivers as part of ethtool kernel work. Signed-off-by: Jacob Keller <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2018-08-31Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-nextDavid S. Miller8-93/+1385
Daniel Borkmann says: ==================== pull-request: bpf-next 2018-09-01 The following pull-request contains BPF updates for your *net-next* tree. The main changes are: 1) Add AF_XDP zero-copy support for i40e driver (!), from Björn and Magnus. 2) BPF verifier improvements by giving each register its own liveness chain which allows to simplify and getting rid of skip_callee() logic, from Edward. 3) Add bpf fs pretty print support for percpu arraymap, percpu hashmap and percpu lru hashmap. Also add generic percpu formatted print on bpftool so the same can be dumped there, from Yonghong. 4) Add bpf_{set,get}sockopt() helper support for TCP_SAVE_SYN and TCP_SAVED_SYN options to allow reflection of tos/tclass from received SYN packet, from Nikita. 5) Misc improvements to the BPF sockmap test cases in terms of cgroup v2 interaction and removal of incorrect shutdown() calls, from John. 6) Few cleanups in xdp_umem_assign_dev() and xdpsock samples, from Prashant. ==================== Signed-off-by: David S. Miller <[email protected]>
2018-09-01xsk: i40e: get rid of useless struct xdp_umem_propsMagnus Karlsson1-2/+2
This commit gets rid of the structure xdp_umem_props. It was there to be able to break a dependency at one point, but this is no longer needed. The values in the struct are instead stored directly in the xdp_umem structure. This simplifies the xsk code as well as af_xdp zero-copy drivers and as a bonus gets rid of one internal header file. The i40e driver is also adapted to the new interface in this commit. Signed-off-by: Magnus Karlsson <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]>
2018-09-01i40e: fix possible compiler warning in xsk TX pathMagnus Karlsson1-4/+2
With certain gcc versions, it was possible to get the warning "'tx_desc' may be used uninitialized in this function" for the i40e_xmit_zc. This was not possible, however this commit simplifies the code path so that this warning is no longer emitted. Signed-off-by: Magnus Karlsson <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]>
2018-08-30i40e: Prevent deleting MAC address from VF when set by PFPatryk Małek1-0/+10
To prevent VF from deleting MAC address that was assigned by the PF we need to check for that scenario when we try to delete a MAC address from a VF. Signed-off-by: Patryk Małek <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40evf: cancel workqueue sync for adminq when a VF is removedLihong Yang1-0/+2
If a VF is being removed, there is no need to continue with the workqueue sync for the adminq task, thus cancel it. Without this call, when VFs are created and removed right away, there might be a chance for the driver to crash with events stuck in the adminq. Signed-off-by: Lihong Yang <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40e: hold the rtnl lock on clearing interrupt schemePatryk Małek1-0/+8
Hold the rtnl lock when we're clearing interrupt scheme in i40e_shutdown and in i40e_remove. Signed-off-by: Patryk Małek <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40evf: Don't enable vlan stripping when rx offload is turned onPatryk Małek1-5/+6
With current implementation of i40evf_set_features when user sets any offload via ethtool we set I40EVF_FLAG_AQ_ENABLE_VLAN_STRIPPING as a required aq which triggers driver to call i40evf_enable_vlan_stripping. This shouldn't take place. This patches fixes it by setting the flag only when VLAN offload is turned on. Signed-off-by: Patryk Małek <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40e: Check and correct speed values for link on openJan Sokolowski1-3/+24
If our card has been put in an unstable state due to other drivers interacting with it, speed settings might be incorrect. If incorrect, forcefully reset them on open to known default values. Signed-off-by: Jan Sokolowski <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40e: report correct statistics when XDP is enabledBjörn Töpel1-11/+13
When XDP is enabled, the driver will report incorrect statistics. Received frames will reported as transmitted frames. This commits fixes the i40e implementation of ndo_get_stats64 (struct net_device_ops), so that iproute2 will report correct statistics (e.g. when running "ip -stats link show dev eth0") even when XDP is enabled. Reported-by: Jesper Dangaard Brouer <[email protected]> Fixes: 74608d17fe29 ("i40e: add support for XDP_TX action") Signed-off-by: Björn Töpel <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40e: static analysis report from communityMartyna Szapar1-1/+1
Static analysis tools report a problem from original driver submission. Removing unnecessary check in condition. Signed-off-by: Martyna Szapar <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40evf: set IFF_UNICAST_FLT flag for the VFLihong Yang1-0/+2
Set IFF_UNICAST_FLT flag for the VF to prevent it from entering promiscuous mode when macvlan is added to the VF. Signed-off-by: Lihong Yang <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40e: use correct length for strncpyMitch Williams1-1/+2
Caught by GCC 8. When we provide a length for strncpy, we should not include the terminating null. So we must tell it one less than the size of the destination buffer. Signed-off-by: Mitch Williams <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40evf: Validate the number of queues a PF sendsPaul M Stillwell Jr1-0/+32
A PF can send any number of queues to the VF and the VF may not be able to support that many. Check to see that the number of queues is less than or equal to the max number of queues the VF can have. Signed-off-by: Paul M Stillwell Jr <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40evf: Change a VF mac without reloading the VF driverPaweł Jabłoński2-5/+14
Add possibility to change a VF mac address from host side without reloading the VF driver on the guest side. Without this patch it is not possible to change the VF mac because executing i40evf_virtchnl_completion function with VIRTCHNL_OP_GET_VF_RESOURCES opcode resets the VF mac address to previous value. Signed-off-by: Paweł Jabłoński <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40evf: update ethtool stats code and use helper functionsJacob Keller2-60/+298
Fix a bug in the way we handled VF queues, by always showing stats for the maximum number of queues, even if they aren't allocated. It is not safe to change the number of strings reported to ethtool, as grabbing statistics occurs over multiple ethtool ops for which the rtnl_lock() cannot be held the entire time. Avoid this by always reporting queue stats for the maximum number of queues in the netdevice. Share some of the helper functionality for adding stats with the PF code in i40e_ethtool_stats.h This should reduce the chance of potential future bugs, and make adding new statistics easier. Note for the queue stats, unlike the PF driver we do not keep an array of queue pointers, but an array of queues, so care must be taken to avoid accessing queue memory that hasn't yet been allocated. Signed-off-by: Jacob Keller <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>
2018-08-30i40e: move ethtool stats boiler plate code to i40e_ethtool_stats.hJacob Keller2-181/+222
Move the boiler plate structures and helper functions we recently added into their own header file, so that the complete collection is located together. Signed-off-by: Jacob Keller <[email protected]> Tested-by: Andrew Bowers <[email protected]> Signed-off-by: Jeff Kirsher <[email protected]>